Nuke

USER GUIDE
VERSION 11.3v5
Nuke™ 11.3v5 Online Help. Copyright © 2019 The Foundry Visionmongers Ltd. All Rights Reserved. Use of this document and the Nuke
software is subject to an End User License Agreement (the "EULA"), the terms of which are incorporated herein by reference. This
document and the Nuke software may be used or copied only in accordance with the terms of the EULA. This document, the Nuke
software and all intellectual property rights relating thereto are and shall remain the sole property of The Foundry Visionmongers Ltd.
("The Foundry") and/or The Foundry's licensors.
The EULA is available here: End User License Agreement (EULA)
The Foundry assumes no responsibility or liability for any errors or inaccuracies that may appear in this document and this document is
subject to change without notice. The content of this document is furnished for informational use only.
Except as permitted by the EULA, no part of this document may be reproduced, stored in a retrieval system or transmitted, in any form
or by any means, electronic, mechanical, recording or otherwise, without the prior written permission of The Foundry. To the extent that
the EULA authorizes the making of copies of this Reference Guide, such copies shall be reproduced with all copyright, trademark and
other proprietary rights notices included herein. The EULA expressly prohibits any action that could adversely affect the property rights
of The Foundry and/or The Foundry's licensors, including, but not limited to, the removal of the following (or any other copyright,
trademark or other proprietary rights notice included herein):
Nuke™ compositing software © 2019 The Foundry Visionmongers Ltd. All Rights Reserved.
Nuke™ is a trademark of The Foundry Visionmongers Ltd.
Digital Domain ® is a registered trademark of Digital Domain, Inc.
Primatte™ keyer tool © 1997-2019 Photron USA, Inc. All Rights Reserved.
Primatte™ is a trademark of IMAGICA Corp.
Primatte™ patent is held by IMAGICA Corp.
In addition to those names set forth on this page, the names of other actual companies and products mentioned in this User Guide
(including, but not limited to, those set forth below) may be the trademarks or service marks, or registered trademarks or service marks,
of their respective owners in the United States and/or other countries. No association with any company or product is intended or
inferred by the mention of its name in this User Guide.
ACADEMY AWARD ® is a registered service mark of the Academy of Motion Picture Arts and Sciences.
Linux ® is a registered trademark of Linus Torvalds.
Windows ® is the registered trademark of Microsoft Corporation.
Mac, Mac OS X, macOS, Sierra, High Sierra, Mojave, Shake, Final Cut Pro, and QuickTime are trademarks of Apple, Inc., registered in the
U.S. and other countries.
Adobe ® and Photoshop ® are either registered trademarks or trademarks of Adobe Systems Incorporated in the United States and/or
other countries.
Maya ® is a registered trademark of Autodesk, Inc., in the USA and other countries.
Houdini ® is a registered trademark of Side Effects Software, Inc.
Boujou is a trademark of 2d3 Ltd.
3D-Equalizer is a trademark of Science.D.Visions.
RenderMan ® is a registered trademark of Pixar.
Cineon™ is a trademark of Eastman Kodak Company.
Thank you to Diogo Girondi for providing icons for the Nuke user interface and Tim Baier for proofreading.
The Foundry Visionmongers Ltd.
5 Golden Square
London
W1F 9HT
UK
Rev: Wednesday, July 24, 2019

Contents
Getting Started 80
Meet the Nuke Product Family 81
Nuke Products 81
About Nuke 82
About NukeX 83
About Nuke Assist 84
Using Gizmos, Groups, Precomps, Knobs, and Python 86
About Nuke Non-commercial 87
Key Concepts 88
Understanding the Workflow 88
Working with Multiple Image Formats 89
Channel Operations 90
8-, 16-, and 32-Bit Image Processing 91
Compositing in 3D 91
Render Farms and Frame Servers 92
Installation and Licensing 93
Operating Systems 93
Windows 94
System Requirements 94
Minimum Hardware Requirements 94
Requirements for GPU Acceleration 95

NVIDIA 95
AMD 96
Multi-GPU Processing 96
Installing on Windows 97
Installing with the User Interface (UI) 97
Installing from the Command Line 98
Launching on Windows 99
Nuke Analytics 99
Nuke Non-commercial 100
Licensing on Windows 101
Obtaining Licenses 101
Installing Licenses 102

To Install a License from Disk 103
To Install an Activation Key or License Text 103
To Install a Floating License 104
Further Reading 105
Licensing Nuke Non-commercial on Windows 106
Uninstalling Apps on Windows 108
Mac OS X and macOS 109

NVIDIA 110
AMD 111
Installing on Mac 112
Installing with the User Interface (UI) 112
Installing from the Terminal 112
Launching on Mac 114
Nuke Analytics 114

Licensing on Mac 116

Further Reading 120
Licensing Nuke Non-commercial on Mac 121
Uninstalling Apps on Mac 123
Linux 124

NVIDIA 125
AMD 126
Installing on Linux 127
Installing from the Terminal 127
Installing Remotely from the Terminal 128
Launching on Linux 129
Nuke Analytics 129
Licensing on Linux 131

Further Reading 135

Licensing Nuke Non-commercial on Linux 136
Uninstalling Apps on Linux 138
Nuke Studio Environments 139
The Compositing Environment 139
The Timeline Environment 140
How the Panels Link 141
Panel Focus and Keyboard Shortcuts 142
Default Workspaces 143
Customizing Workspaces 144

Customizing Panes 144
Moving the Toolbar 145
Adding Tabs 145
Soloing Tabs 145
Floating Windows 145
Maximizing Windows 146
Hiding Tab Names and Controls 146
Saving and Loading Workspaces 146
Setting the Startup Workspace 146
Setting Preferences 147

Changing Preferences 147
Saving Preferences 148
To save your preferences 148
Resetting Preferences 148
Using the Compositing Environment 150
Toolbar, Menu Bar, and Content Menus 150
Using the Toolbar 152
Working with Nodes 154
Adding Nodes 154
Selecting Nodes 155
Replacing Nodes 156
Renaming Nodes 157

Editing Nodes 158
Cloning Nodes 159
Disabling and Deleting Nodes 160
Connecting Nodes 160
Node Indicators 163
Displaying Node Information 164
Customizing the Node Display 165
Creating Node Tool Sets 167
Using the Tab Menu 168
Adding Nodes by Sub-string 168
Adding and Removing Favorites 170
Managing Weighting 171
Navigating Inside the Node Graph 173
Panning 173
Zooming, Fitting in the Node Graph 174
Bookmarking the Pan and Zoom Level 175
Bookmarking Nodes 175
Searching for Nodes 176
Cleaning up the Node Graph 177
Properties Panels 178
Managing the Properties Panel 178
Properties Panel Controls 179
Displaying Controls 182
Using Input Fields 183
Using the Color Controls 185
Using the Color Swatch 185
Using the In-Panel Color Picker 186
Using the Floating Color Picker Window 187

Using the Channel Selector 189
Color Slider Functions 190
Customizing a Node’s Defaults with Node Presets 192

To Create a Node Preset: 192
Loading and Deleting a Preset 192
Animating Parameters 193
Working with Animated Parameters 193
Animated Parameters in the Dope Sheet and the Curve Editor 195
Using the Dope Sheet 196

Using the Dope Sheet Interface 196
Viewing Keys in the Dope Sheet 197
The Dope Sheet and Time Nodes 198
Multiple Branches of Nodes in the Dope Sheet 199
Selecting Keyframes 199
Copying and Pasting 200
Deleting and Adding Keyframes 200
Editing Clips in the Dope Sheet 200
Synchronizing the Frame Range 204
Using the Curve Editor 204
Editing Curves 207
Compositing Viewers 215
Adding Viewer Nodes 216
Connecting Viewer Nodes 216
Toggling Views 217
Panning and Zooming the Viewer Window 217

Panning the Frame 217
Re-centering the Frame 217
Zooming In and Out 217
Restoring the Zoom to 100% 218
Showing / Hiding Floating Viewers 218
Showing / Hiding Viewer Toolbars 218
Locking the Viewer Zoom Level 218

Viewing Overscan in the Viewer 219
Using the Viewer Controls 220
Compositing Viewer Controls 220
Jumping to a Specific Frame 222
Synchronizing Viewer Playback 223
Pausing the Display 223
Displaying a Single Channel 223
Layer and Channel Dropdown Menus 224
Displaying Alpha Mattes 225
Viewer Info Bar 226
Adjust Display Gain and Gamma 227
2D / 3D Toggle and Camera Controls 228

Selection Modes 228
Monitor Output Toggle 230
Controlling Zoom and Resolution 231
Using the Zoom Menu 231
Proxy Mode 231
Lowering the Display Resolution of Individual Viewers 232
Pixel Aspect Ratio 232
Full-frame processing 233
Region of Interest (ROI) 234
Viewer Overlays and Input Processes 236
Guides and Masks 236
Adding Custom Guides and Masks 240
Using the Viewer Composite Display Modes 240
Input Process and Viewer Process Controls 242
Using the File Browser 247
To Use the Navigation Controls 248

Path Name Field 248
To Use the Path Name Field: 248
To Use the Filter Dropdown Menu and Sequences Checkbox 248
To Preview Files in the File Browser 249
To Select Multiple Files with the File Browser 250
To Open Incomplete Sequences in Separate Read Nodes 250
Undoing and Redoing 251
Progress Bars 252
Handling Errors 253
Using Nuke Studio's Timeline Environment 254
Achieving Real-time Playback 254

Windows and Linux Certified Hardware 254
About Clips and Shots 257
Source Clips 257
Shots 258
A Source Clip Opened in the Timeline View 258
Clip and Shot Properties 259
Setting Source Clip Ranges 260
Using Relative Paths to Media 261
Ingesting Media 263
Using Drag-and-Drop 265
Using the File Browser 266
To Use the Navigation Controls 267
Path Name Field 267
To Use the Filter Dropdown Menu and Sequences Checkbox 268
To Preview Files in the File Browser 268

To Select Multiple Files with the File Browser 270
Sorting and Searching Media 272
Sorting the Project Panel 272
Searching for Media 274
Setting Poster Frames 275
Color-coding Source Clips and Shots 277
Setting Default Colors 277
Setting Colors by Selection 278
Setting Colors by File Type 279
Reconnecting and Refreshing Clips 281
Localizing Media 282
Setting Localization Preferences 282
Managing Localization 285
Updating On Demand Clips 288
Clearing Localized Files 289
Using the Timeline Viewer 291
Deleting Media 292
Timeline Playback Tools 293
In and Out Points 294
Playback Controls 296
Timeline Viewer Tools 298
Using In and Out Markers 302
Using In and Out Markers 304
Working with Colorspaces 307
To apply colorspace changes to shots: 308
Previewing on a Broadcast Monitor 309

To preview output on an external broadcast video monitor: 312
Using Scopes 314
Histogram 314
Waveform 316
Vector 318
About Anamorphic Media 320
About QuickTime Media 321
About RED Media 323
To modify the RED Rocket options: 323
Compositing with Nuke 325
Organization of the Section 325
Managing Scripts 327
Setting Up Your Script 327

Name, Time Span, and Frame Rate 327
Setting the Default Project Directory 328
Full-Size Formats 329

To Set Up a Full-Size Format 329
Proxy Mode 330

Proxy Format and Proxy Scale 330
To Set Up Proxy Formats 331
To Set Up a Proxy Scale 332
Read Nodes and Proxy Files 332

To Define Which File (Full-Res or Proxy) is Used in the Proxy Mode: 333
Write Nodes and Proxy Files 333
Using Proxy Mode to Enlarge a Script 333

Using Small Proxy Files 334
Using Large Proxy Files 334
Rendering the Scaled-up Output 334
To Toggle Between Full Resolution and Proxy Mode 334
Loading Image Sequences 335

Importing Image Sequences 335
To Import an Image Sequence from an External File Browser 337
Notes on Importing QuickTime Files 337
Notes on Importing AVI Files 338
Notes on Importing OpenEXR Files 338
Notes on Importing PSD Files 339
Naming Conventions 340
Changing the Relation Between the Current Frame and the Frame
Read In 340
Using Expressions 340
Specifying a Start Frame for a Clip 341
Offsetting All Frames by a Constant Value 341
Reformatting Image Sequences 343
To Insert a Reformat Node: 343
Image Caching 344
The Cache Directory 346
Defining the Settings for Caching 346
Clearing the Disk Cache 346

To Empty the Disk Cache 347
Using the DiskCache Node 348
To Cache Images Upstream 348
Localizing Files for Better Performance 350
Setting Localization Preferences 350
Managing Localization 353
Updating On Demand Clips 355
Clearing Localized Files 356
Saving Scripts and Recovering Back-Ups 358
Saving Scripts 358

To Save a Script 358
Automatic Back-Up of Scripts 359
To Define Autosave Options for a Script 359
To Turn off Automatic Back-Up 359
Recovering Back-Ups 359
Closing Scripts 361
Loading Scripts 362
Defining Frame Ranges 363
Reformatting Elements 364
Using the Reformat Node 364
Converting Images to a Desired Image Format 364

To Create a New Output Format: 365
To Edit a Format: 365
To Delete a Format: 366
To Apply a Format: 366
Creating Thumbnails 366
Scaling Image Sequences 367
Cropping Elements 368
Adjusting the Bounding Box 370
Resizing the Bounding Box 370

Copying a Bounding Box from One Input to Another 371
Adding a Black Outside Edge to the Bounding Box 372
Bounding Box Warnings 372
Channels 375
Introduction 375
Quick Start 376
Understanding Channels and Layers 377
Channels 377
Layers 377
Creating Channels and Layers 379

To Create a New Layer and/or Channel 379
Calling Channels 381
Viewing Channels in the Viewer 382
Selecting Input Channels 383

To Select a Single Input Channel 383
To Select Multiple Input Channels 383
Selecting Masks 383

To Select a Channel for Use as a Matte from the Mask Input 384
To Select a Channel for Use as a Matte from the Main Input 385
Linking Channels Using the Link Menu 386
Tracing Channels 387
Renaming Channels 388
Removing Channels and Layers 389
To Remove a Layer or a Channel Within a Layer 389
Swapping Channels 390
Channel Outputs 391
Assigning Constants 392
Creating Swap Layers 392
Merging Images 393
Layering Images Together with the Merge Node 393

To Layer Images with the Merge Node 393
Merge Operations 397
Generating Contact Sheets 408
To Generate a Contact Sheet 408
Copying a Rectangle from one Image to Another 411
To Copy a Rectangle from One Image to Another 412

Removing Noise with Denoise 414
Quick Start 414
Connecting Denoise 416
Analyzing and Removing Noise 417
Reviewing the Results 419
Fine Tuning 421
Keying with ChromaKeyer 424
Quick Key 424
Picking the Screen Color 426
Adjusting Screen Gain and Balance 427
Improving Mattes 429
Adjusting the Gain, White Point, and Black Point 429
Despilling and Color Replacement 432
Custom Despilling 432
Color Replacement 432
Multi-Pass Keying 434
Keying with Keylight 435
Quick Key 435
Basic Keying 437
Picking the Screen Color 437
Screen Matte 437
Viewing the Key 438
Keying More 439

Advanced Keying 440
Under the Hood 440
View 440
Status 441
View 443
Status 443
Screen Color 446
Background Pixel 446
Edge Pixel 447
Foreground pixel 447
Biasing 448
The Bias Colors in Everyday Use 449

Picking a bias color 449
Why are there two bias colors? 450
Clip Black and White 452
Screen Gain 453
Screen Balance 454
PreBlur and Tuning 455
Tuning 455
Screen Processing 456
Two-stage keying 456
Clip Rollback 456
Dilate 457
Softness 457
Despot 458
Mattes 459
Inside and Outside Masks 460
Source Alpha 462
Color Replacement 463
Inside mask 463
Edges 463
InM component 464
OutM component 464
Keying with Primatte 466
Quick Start 466
Connecting the Primatte Node 467
Primatte Basic Operation Tutorial 469
Auto-Compute 469
Smart Select BG Color 471
Clean BG Noise 473
Clean FG Noise 474
Spill Removal - Method #1 476
Sampling Tools 479
The Spill Sampling Tools 479
The Matte Sampling Tools 479
The Detail Sampling Tools 480
Replacing Spill 481

Primatte Controls 483
Initialize Section 483
Primatte Viewer Tools 485
Degrain Section 487
Degrain type 487
Degrain Tools Tutorial 487
Actions Section 491
Smart Select BG Color 491
Clean BG Noise 491
Clean FG Noise 491
Matte Sponge 491
Make FG Trans 492
Restore Detail 492
Spill Sponge 492
Spill (-) 492
Spill (+) 492
Matte (-) 493
Matte (+) 493
Detail (-) 493
Detail (+) 493
Fine Tuning Sliders 494
3D Sample 494
Simple Select BG Color 494
Current Color Chip 494
Adjust Lighting 494
Hybrid Render 495
Adjust Lighting 496

Hybrid Matte 497
Fine Tuning 498
Spill or L-poly slider (Spill Removal) 498
Transparency or M-poly slider (Adjust Transparency) 498
Detail or S-poly slider (Add/Restore Lost Detail) 498
Spill Process Section 500
Replace with 500
Replace color slider 500
Defocus slider 500
Output Section 501
The Primatte Algorithm 504
Explanation of How Primatte Works 505
Explanation of How Primatte RT+ works 512
Explanation of How Primatte RT works 512
Contact Details 514
Main Office 514
Primatte Office 514
Proprietary Notices 514
Keying with Ultimatte 515
Ultimatte Quick Start 515
Connecting the Ultimatte Node 516
To use the Ultimatte inputs 516
Sampling the Screen Color 517
Adjusting the Controls on Ultimatte Tab 517
Using Overlay Tools and Screen Correct 518

Adjusting overlay controls 519
Adjusting the screen correct controls 520
Adjusting the Density of the Matte 521
Adjusting Density controls 521
Adjusting Spill Controls 522
Retaining Shadows and Removing Noise 524
Adjusting Shadows controls 524
Adjusting Cleanup controls 524
Adjusting Color Controls 526
Adjusting Film Controls 527
Choosing an Output Mode 528
Using RotoPaint 529
Roto or RotoPaint? 529
RotoPaint Quick Start 529
Connecting the RotoPaint Node 531
To Connect the RotoPaint Node 531
Working with the Toolbars 532
Working with the Stroke/Shape List 534
To Use the Stroke/Shape List 534
Drawing Paint Strokes 537
Using the Brush tool 539
To Use the Brush Tool 539
Using the Eraser Tool 540

To Use the Eraser Tool 540
Using the Clone Tool 541
To Use the Clone Tool 541
Using the Reveal Tool 543
To Use the Reveal Tool 543
Using the Blur Tool 545
To Use the Blur Tool 545
Using the Sharpen Tool 546
To Use the Sharpen Tool 546
Using the Smear Tool 547
To Use the Smear Tool 547
Using the Dodge Tool 548
To Use the Dodge Tool 548
Using the Burn Tool 549
To Use the Burn Tool 549
Drawing Shapes 550
Using the Bezier and Cusped Bezier Tools 552
To Use the Bezier or Cusped Bezier Tools 552
Using the B-Spline Tool 554
To Use the B-Spline Tool 554
Using the Ellipse, Rectangle, and Cusped

Rectangle Tools 556
To Use the Ellipse, Rectangle, and Cusped Rectangle Tools 556
Using the Open Spline Tool 558

To Use the Open Spline Tool 558
Setting Default RotoPaint Tools and Settings 561
To Set a Tool as Your Default Tool: 561
To Set Your Default Tool Properties: 562
Selecting the Output Format and Channels 564
Selecting Existing Strokes/Shapes for Editing 566
To Adjust Display Properties for Selected Shapes/Strokes 566
To Select an Entire Stroke/Shape 567
To Select a Spline 567
To Select Only Points on a Stroke/Shape 568
To Select Feather Points 568
To Select Points Using a Transform Handle 568
Editing Existing Stroke/Shape Attributes 569
Editing Color 569
Editing Opacity 569
Selecting a Source for Your Stroke/Shape 570
Editing Blending Mode 570
Transforming Strokes/Shapes/Groups 572
To Transform a Stroke/Shape Using a Transform Handle Jack: 573
To Transform Points Using a Transform Box: 573
To Transform Onion Skin Source 573
Adjusting Mask Controls 575
Editing Shape-Specific Attributes 576
Adding and Removing Feather 576
Editing Stroke-Specific Attributes 577
Selecting a Source Image 577

Editing Brush Type 577
Editing Brush Size 577
Editing Brush Spacing 578
Editing Brush Hardness 578
Adjusting Write On 579
Editing Clone or Reveal Attributes 580
Editing Existing Stroke/Shape Timing 581
Editing Existing Stroke/Shape Stack Order 582
Editing Existing Stroke/Shape Splines 583
To Add a Point to a Stroke/Shape 583
To Move a Point 583
To Move Several Points Together 583
To Delete a Point 584
To Delete an Entire Stroke/Shape 584
Copying, Cutting, and Pasting Stroke Attributes 585
To Copy Stroke/Shape Attributes 585
To Cut Stroke/Shape Attributes 585
To Paste Stroke/Shape Attributes 586
Point Cusping, Smoothing, Expressions, and

Links 587
To Cusp or Smooth Points 587
To Add Expressions to Points 587
To Link Other Controls to a Point’s Position 588
Animating Strokes/Shapes 589
To animate a stroke/shape using autokey 589
To animate strokes/shapes manually 589

Viewing Spline Keyframes 591
To View Keyframes on the Timeline 591
Deleting or Rippling Keyframes 592
To Delete a Keyframe 592
To Delete All Keyframes for a Stroke/Shape 592
To Ripple Keyframes 592
Copying, Cutting, and Pasting Animations 594
To Copy Animations 594
To Cut Animations 594
To Paste Animations 595
Adding Motion Blur 597
Adding Blur to Shapes 597
Adding Global Motion Blur 599
Viewing Points in the Curve Editor and the Dope

Sheet 601
Copying, Pasting, and Cutting Stroke Positions 602
To Copy Stroke/Shape Positions 602
To Cut Stroke/Shape Positions 602
To Paste Stroke/Shape Positions 603
Copying, Pasting, and Cutting Point Positions 604
To Copy Point Values 604
Pasting Point Positions 605
To Paste Point Values 605
Cutting Point Positions 606
To Cut Point Values 606

RotoPaint and Stereoscopic Projects 607
Selecting the View to Draw On 607
Selecting the View to Clone From 607
Reproducing Strokes/Shapes in Other Views 608
Editing Strokes/Shapes in One View Only 609
To Edit Stroke/Shape Points 609
To Edit Stroke/Shape Attributes 609
Applying a Stereo Offset 611
Where Are the Bezier and Paint Nodes? 612
Tracking and Stabilizing 613
Before Tracking 613
Quick Start 613
Connecting the Tracker Node 614
Adding Track Anchors 615
Positioning Track Anchors 615
Tracking Preferences and Viewer Tools 618
Automatic vs. Keyframe Tracking 620
Automatic Tracking 621
Calculating Auto-Tracks 621
Troubleshooting Auto-Tracks 622
Dealing with Occlusions 623
Keyframe Tracking 625
Calculating Keyframe Tracks 625
Importing Tracking Data 626

Troubleshooting Keyframe Tracks 628
Dealing with Occlusions 629
Applying Tracking Data 631
Applying Tracking Data Using Tracker’s Controls 631

Stabilizing Elements 631
Match-moving Elements 632
Removing or Adding Jitter 633
Applying Tracking Data Using Linking Expressions 634

Creating Linking Expressions 634
To Link Animated Parameters with a Tracker Node 635
Transforming Masks with Tracking Data 635
Using the CornerPin2D Node 637

To Use the CornerPin2D Node 638
Transforming Elements 640
Transforming in 2D 641
Using the 2D Transformation Overlay 641
Translating Elements 642

Using the Transform Node 642
Using the Position Node 643
Rotating Elements 643
Scaling Elements 644
Skewing Elements 645
To Invert a Transform Effect 645
Choosing a Filtering Algorithm 648
How Nodes Concatenate 653
Applying Core Transformations in 2.5D 656
Using the 3D Transformation Handles 656
Adding a Card3D Node 657
Translating Elements 657

To Translate an Element Using the Card3D Node 657
Rotating Elements 657
Scaling Elements 658
Skewing Elements 658

To Skew an Element Using the Card3D Node 658
Specifying the Order of Operations 659

To Choose the Operation Order for Scales, Rotations, and Translations 659
To Choose the Operation Order for Rotations 659
Choosing a Filtering Algorithm 659
To Add Motion Blur 660
To Add Motion Blur to an Image Rendered in a Third-party

Application 661
To Create Motion Blur with the VectorBlur Node 662
Replicating the Input Image Across the Output 664
To Use the Tile Node 664
Warping Images 666
Quick Start 666
Warping 667
Warping Images Using the GridWarp Node 667

To Warp an Image Using the GridWarp Node 667
Warping an Image Using the SplineWarp Node 675

To Warp an Image Using the SplineWarp Node 675
Using Boundary Shapes and Pins to Limit Warp 683
Morphing 686
Morphing One Image into Another Using the GridWarp Node 687
Morphing One Image into Another Using the SplineWarp Node 689
Using the cookie-cutter to create traveling masks 691
Transforming and Animating Warps and Morphs 694
Transforming Warps 694

Using the Tracker Node 694
Using the PlanarTracker Node 694
Using Expressions with GridWarp 695
Animating Warps 696

To Animate a Warp Using the GridWarp Node 696
To Animate a Warp Using the SplineWarp Node 696
Linking Transforms Using the SplineWarp Node 697
Temporal Operations 698
Quick Start 698
Distorting Time 700
Simple Retiming 701
To Retime All Frames in a Clip 701
To Retime a Range of Frames in a Clip 701
Interpolation 702
Frame-Blending 702
To Insert a FrameBlend Node 703
OFlow Retiming 705
To Retime with OFlow 705
Refining the Results 706
Warping Clips 709
To Warp a Clip 710
Global Frame Range and Speed 711
Applying a TimeBlur Filter 712
Applying Motion Blur to a Clip 712
Editing Clips 713
Offsetting and Slipping Clips 713

To Offset a Clip Using the TimeOffset Node 713
To Slip a Clip Using the TimeClip Node 714
Cutting Clips 716
To Cut a Clip 716
Splicing Clips 716

To Splice Slips 716
Freeze Framing Clips 718

To Output a Freeze Frame 718
To Output a Frame Interval 718
Working with Color 719
Using Scopes 720
Histogram 720
Using the Histogram Node 722

To Define Tonal Range with the Histogram Node 722
Waveform 723
Vectorscope 724
Using the Pixel Analyzer 727
Analyzing Pixel Selections 728
Analyzing Full Frames 730
Applying Analysis Data 731
Making Tonal Adjustments 733
Sampling White and Black Points 733

To Define Tonal Range with the Grade Node 733
Making Basic Corrections 734
Using ColorCorrect Sliders 735

To Adjust Contrast, Gain, Gamma or Offset with the ColorCorrect Node 735
Adjusting Black Levels with the Toe Node 737
Using ColorLookup Curves 737

To Make Basic Corrections with the ColorLookup Node 738
Making Hue, Saturation, and Value Adjustments 740
Correcting HSV 741

To make HSV corrections with the HSVTool node 741
Correcting Hue Only 742

Suppressing spill 743
Correcting Saturation Only 744

To make saturation corrections with the Saturation node 744
Masking Color Corrections 745
To Mask a Color Correction 745
Applying Grain 747
Using Synthetic Grain 747

To add synthetic grain with the Grain node 747
Using Practical Grain 748

To Create Film Stock Sequences 748
To Add Scanned Grain with the ScannedGrain Node 749
To Mix the Grain and Backplate 750
Applying Mathematical Operations 751
Clamping Channel Values 751
Offsetting Channel Values 752
Inverting Channel Values 752
Multiplying Channel Values 753
Applying Expressions to Channel Values 753
Transforming the Color Space 755
Overriding the Default Cineon Conversion 756
To Override the Default Cineon Conversions Using Log2Lin 756
To Convert Between Logarithmic and Linear Color Spaces 756
To Create a 3D LUT in Log Color Space (for example Cineon) 757
Making Other Colorspace Conversions 758
To Convert an Element in Nuke’s Native Color Space into Another

Color Space 758
Converting Color Spaces with the OCIOColorSpace Node 758
Changing the Viewer Color Space 760

Adding Context Viewer Processes 760
Filtering and Spatial Effects 762
Applying Convolves 762
Using the Convolve Node 762
Simulating Depth-of-Field Blurring 765
Quick Start 766
Connecting ZDefocus 766
Adjusting Blur Settings 768
Adjusting the Shape of Out-of-Focus Highlights 772

Using a Predefined Disc Image 772
Using a Predefined Bladed Image 773
Using a Custom Filter Image 775
Enhancing Out-of-Focus Highlights 776
Masking Blur Effects 777
Creating Effects 779
Quick Start 779
Background Reflections on Foreground Elements 780
Using the LightWrap Node 781
Creating Star Filter Effects on Image Highlights 783
Using the Glint Node 783
Creating Text Overlays 786
Preparing a Text Overlay 786
Entering Text 788
Example Variables and Entities 788
Fonts and Font Properties 793
Selecting a Font 794
Adjusting Font Size and Spacing 794

Justifying Fonts 796
Updating the Font Cache 799
Transforming Text 800
Transforming Selections 800
Transforming Groups 801
Animating Transforms 802
Animating Selections 802
Animating Groups 802
Changing the Text Color 803
Masking Regions of the Viewer 804
Adding Shadows to Text 805
Adding a Drop Shadow 808
Analyzing and Matching Clips 811
Introduction 811
Cropping Black Edges 813
Analyzing the Intensity of a Frame Sequence 814
Removing Flicker 815
Analyzing Exposure Differences 816
Tracking the Brightest and Darkest Pixels 817
3D Compositing 818
Overview 818
Setting Up a Scene 820

The Scene Node 820
To Add a Scene Node 820
The ScanlineRender Node 821

To Add a ScanlineRender Node 821
The Camera Node 821
Using the 3D Viewer 823
Switching to the 3D Viewer 823
To Navigate in the 3D Viewer 824
To Change the 3D Viewer Display Properties 824
To Look Through a Camera 825
To Lock the 3D Camera View 826
To Use the Interactive 3D Camera View Mode 826
3D Scene Geometry 827
Using Built-in Primitive Geometry 828
Working with Cards 828

To Add a Card Object: 828
Deforming Card Objects 829
Working with Cubes 831

To Add a Cube 831
To Translate a Cube’s Sides 831
Working with Cylinders 832

To Add a Cylinder 832
Adjusting Geometric Resolution 832
Working with Spheres 833

To Add a Sphere 833
Adjusting Geometric Resolution 834
Importing Geometry and Point Clouds from Other

Applications 835
Importing Geometry from OBJ Files 835

To import an OBJ object 835
Importing Geometry and Point Clouds from FBX Files 836

Importing Meshes from FBX Files 836
Importing Point Clouds from FBX Files 837
Importing Geometry and Point Clouds from Alembic Files 838
To Import Meshes and Point Clouds from an Alembic File 838
Creating Point Clouds and Geometry from Scratch 841
Creating a Position Pass Using the

DepthToPosition Node 842
To Create a Position Pass using DepthToPosition 842
Creating a Dense Point Cloud Using the

PositionToPoints Node 844
To Create a Point Cloud Using PositionToPoints 844
Creating a Point Cloud Using the DepthToPoints

Node 846
To Create a Point Cloud Using DepthToPoints 846
Object Display Properties 848
To Edit an Object’s Display Attributes 848
3D Selection Tools 850
Selection Modes 850
Selecting Nodes 851
Selecting Vertices on a 3D Object 851

To select vertices in the Viewer 852
To Save or Restore Selected Vertices Using the GeoSelect Node 852
Selecting Faces on a 3D Object 853
Selecting 3D Objects 854
Matching Position, Orientation, and Size to 3D Selection 854
Merging Objects 855
To Merge Your 3D Objects 855
Modifying Object Shapes 856
Modifying Objects Using the EditGeo Node 856

To Modify Objects Using EditGeo 856
To Animate Edits 858
Modifying Objects Using Lookup Curves 859
To Modify Objects Using Lookup Curves 860
Modifying Objects Using a Power Function 861
To Modify Objects Using a Power Function 861
Modifying Objects Using an Image - Method 1 863
To Modify Objects Using an Image 863
Modifying Objects Using an Image - Method 2 865
Connecting the Displacement Node 865
Adjusting the Displacement Controls 865
Adjusting Displacement Controls for Rendering 866
Modifying Objects Using a Perlin Noise Function 868
To Modify Objects Using a Perlin Noise Function 868
Modifying Objects Using a Distortion Function 870
To Modify Objects Using a Distortion Function 870
Modifying Objects Using a Trilinear Interpolation 872
To Modify Objects Using a Trilinear Interpolation 872
Materials and Textures 873
Object Material Properties 874
Applying a Material Using the ApplyMaterial Node 875
Adjusting the Diffuse Color 877
Adjusting Specular Highlights 878
Simulating Materials That Emit Light 879
Adjusting Diffuse, Specular, and Emission Using a Single Node 879
Simulating Smooth, Regular Surfaces 880

Rendering a Wireframe Overlay on Your Geometry 881
Merging Two Shader Nodes 885
To Merge Two Shaders 885
Merging a Material with the Objects Behind 887
To Merge a Material with the Objects Behind 887
Replacing Material Channels with a Constant

Color 891
Projecting Textures onto Objects 893
Projecting Textures with the UVProject Node 893
Projecting Textures with the Project3D Node 894
Importing UDIM Patches 895
Lighting 898
Working with Lights 899
Inserting Direct Lights 899
Inserting Point Lights 899
Inserting Spot Lights 900
Inserting Environment Lights 901
Inserting Direct Lights, Point Lights, or Spot Lights 903

Using the Look Input 903
Importing Lights from FBX Files 907
Casting Shadows 909
To Cast Shadows: 910
Adjusting Shadows When Using ScanlineRender 911
Adjusting Shadows When Using PrmanRender 914
Manipulating Object Normals 916
To Manipulate Object Normals 916

Relighting a 2D Image Using 3D Lights 917
To Relight a 2D Image Using Relight 917
Cameras 919
Working with Cameras 920
To Add a Camera 920
To Look Through a Camera 920
To Lock the 3D Camera View 921
To Use the Interactive 3D Camera View Mode 921
To Edit a Camera’s Lens Characteristics 921
Projection Cameras 926
First a Little Math... 926
Setting Up the Projection Camera Script 927

To Add a Projection Camera 927
To View a 3D Scene over a 2D Background Image 928
Importing Cameras from FBX Files 930
Importing Cameras from an FBX File 930
Importing Cameras from Alembic Files 932
To Import Cameras from an Alembic File: 932
Importing Cameras from Boujou 934
To Import a Camera from Boujou: 934
Locating a 3D Point from an Animated Camera 935
Connecting the PointsTo3D Node 935
Setting Reference Frames 935
Calculating the 3D Position 936
Transforming Geometry, Cameras, and Lights 938

Using the Transform Handles 939
To Move an Object with the Transform Handles 939
To Rotate an Object with the Transform Handles 939
To Scale an Object with the Transform Handles 940
Transforming from the Node Properties Panel 941
To Set Transformation Options 941
To Transform an Object from Its Panel 941
Transformations and the Pivot Point 942
To Move the Pivot Point 942
Parenting to Axis Objects 943
To Add an Axis Object 943
Using the TransformGeo Node 946
To Have One 3D Object Always Face Another 946
Applying Tracks to an Object 949
To apply a channel file to an object 949
Importing Transforms from FBX Files 950
Importing Transforms from an FBX File 950
Importing Transforms from Alembic Files 952
To Import Transforms from an Alembic file: 952
Adding Motion Blur to the 3D Scene 954
Adding Motion Blur Using a Renderer 955
Adding Motion Blur Using VectorBlur 958
To use VectorBlur with ScanlineRender 958
To use VectorBlur with RayRender 960

To use VectorBlur with MotionBlur3D 962
To use VectorBlur with Third-Party Motion Vectors 964
Exporting Geometry, Cameras, Lights, Axes, or

Point Clouds 965
Rendering a 3D Scene 966
Choosing a Render Node 966
Rendering Out a Scene 966
To Add Motion Blur to the 3D Scene 967
Stereoscopic Scripts 968
Quick Start 968
Setting Up Views for the Script 970
Creating Views Automatically 970
Creating and Managing Views Manually 970
Loading Multi-View Images 974
To Read Images in 974
Combining Views from Different Files into a Single Output 975
Displaying Views in the Viewer 977
To Display a Particular View 977
To Display Two Views Next to Each Other 978
To Display a Blend Between Two Views 978
Selecting Which Views Display Changes 980
Splitting Views Off 980
To Show Separate Values for Each View 981
To Unsplit Views 982
Selecting the View to Process When Using the RotoPaint Node 982
Performing Different Actions on Different Views 983

To Extract a View for Processing 983
Reproducing Changes Made to One View 985
Reproducing X and Y Values 985
Swapping Views 987
To Rearrange Views 987
Converting Images into Anaglyph 988
To Convert Your Images into Anaglyph 988
Changing Convergence 990
To Change the Convergence Point of a Stereo Image 991
Using the Same Convergence Point Throughout a Sequence 992
Previewing Stereoscopic Images 994
Previewing Using the Viewer Stereo Modes 994
Displaying OpenGL Stereo in Comp Viewers 994

Enabling OpenGL Stereo Output 995
Windows 995
Linux 996
Switching to OpenGL Stereo Output 996
Flipbooking Stereo Sequences 997
Rendering Stereoscopic Images 999
Rendering EXR Files 999
Rendering Other File Formats 999
Deep Compositing 1001
About Deep Compositing 1001
Quick Start 1001
Reading in Deep Footage 1003
Importing DTEX Files 1004
Importing Scanline OpenEXR Files 1004

Creating Deep Data 1005
Converting a 2D Image Sequence to a Deep Frame Using Input

Frames 1005
Converting a 2D Image to a Deep Image 1006
Recoloring Depth Data 1006
Using ScanlineRender to Generate Deep Data 1007
Merging Deep Images 1010
Creating Holdouts 1012
Creating a Flattened Holdout with the DeepHoldout Node 1012
Creating a Deep Holdout with the DeepMerge Node 1013
Creating 2D and 3D Elements from Deep Images 1015
Creating a 2D Image from a Deep Image 1015
Creating a Point Cloud from a Deep Image 1015
Modifying Deep Data 1017
Color Correcting Deep Images 1017
Adjusting the Effect of Deep Color Correction 1017
Modifying Your Deep Images with Expressions 1017
Cropping, Reformatting, and Transforming Deep

Images 1019
Cropping Deep Images 1019
Reformatting Deep Images 1019
Transforming Deep Samples 1020
Sampling Deep Images 1021
Writing Deep Data 1022
Working with File Metadata 1023
Metadata 1023
Viewing Metadata 1025
Comparing Metadata Between Inputs 1026
Modifying Metadata 1027
To Add Metadata 1027
To Edit Metadata 1027
To Remove Metadata 1028
To Edit the List of Actions in the ModifyMetaData Properties 1029
Copying and Filtering Metadata Between Inputs 1030
Adding a Time Code to Metadata 1031
Rendering Metadata 1032
Accessing Metadata Using Tcl Expressions 1033
Accessing Metadata Using Python 1034
Audio in Nuke 1035
Quick Start 1035
Reading Audio Files into the Node Graph 1036
Creating and Editing Audio Curves 1038
Creating a Keyframe Curve 1038
Modifying the Audio Curve in the Curve Editor and Dope Sheet 1038
Flipbooking the Audio Track 1040
Previews and Rendering 1041
About Rendering in Nuke 1041
Quick Start 1041
Previewing Output 1043

Previewing in a Nuke Viewer 1044
To Enable the ROI Render Feature 1044
To Edit the Position or Size of Current ROI 1044
To Disable the ROI Render Feature 1044
Flipbooking Sequences 1045
Setting the Viewer Cache Location and Size 1045
To Use HieroPlayer as Nuke's Flipbook 1046
Flipbooking a Sequence 1046
Capturing the Viewer 1048
Previewing on an External Broadcast Video

Monitor 1051
To Preview Output on an External Broadcast Video Monitor 1054
To Disable Viewing on an External Broadcast Video Monitor 1055
Rendering Output 1056
Render Resolution and Format 1057
Output (Write) Nodes 1058
To Render a Single Write Node 1058
To Render Selected or All Write Nodes in the Script 1061
Selecting Which Views to Render 1062
Notes on Rendering QuickTime Files 1063

Advanced QuickTimes Options 1064
Notes on Rendering OpenEXR Files 1067
Rendering Using the Frame Server 1068
Frame Server Preferences 1069
Using the Frame Server on External Machines 1070
Configuring the Frame Server on External Machines 1070
Frame Server Logs 1072

File Name Conventions for Rendered Images 1074
Writing Versions of Rendered Images 1074
Changing the Numbering of Rendered Frames 1075
Using Expressions 1075
Specifying a Frame Number for the First Frame in the Clip 1076
Offsetting All Frame Numbers by a Constant Value 1076
Using a Write Node to Read in the Rendered

Image 1078
To Use a Write Node to Read in the Rendered Image 1078
What is the Hash Value? 1079
Render Farms 1080
Organizing Scripts 1081
Displaying Script Information 1082
Using Visual Diagnostics 1083
Enabling Visual Diagnostics 1083
Filtering Profile Data 1087
Exporting and Importing Profile Data 1090
Exporting Profile Data 1090
Importing Profile Data 1091
Using Performance Timing 1092
Enabling Performance Timings 1092
Outputting Performance Timings Onscreen 1092
Writing Performance Timings to File 1094
File Name Search and Replace 1095
To Search for a File Name or File Path and Replace It 1095

Grouping Nodes in the Node Graph 1096
Grouping Nodes with the Backdrop Node 1096

To Group Nodes with a Backdrop Node 1096
Grouping Nodes with the Group Node 1098

To Group Nodes with a Group Node 1098
To View the Nodes Nested Inside a Group Node 1099
To Ungroup Nodes 1099
Adding Notes to the Node Graph 1101
To Add a Note to the Node Graph 1101
Using the Precomp Node 1102
Creating Precomp Nodes 1103
To Create a Precomp Node 1103
About the Hash Value 1104
Precomp Nodes and Project Settings 1104
Using a Precomp Node to Speed-up Rendering 1106
To Render a Precomp Node 1106
Precomp Revisions 1108
To View and Edit a Precomp Script 1108
To Reload a Revised Precomp Script 1108
Collaborative Workflow Example 1109
To Enable a Collaborative Workflow 1109
Using the LiveGroup Node 1110
Collaborative Scripts 1112
Collaborative Workflow Examples 1113
Creating a Master Script 1113
Combining Existing Scripts 1115
Creating LiveGroups 1116

Importing an Existing Script into a LiveGroup 1116
Adding Nodes to a LiveGroup 1117
Viewing Nodes Inside a LiveGroup 1118
Referencing a Master Script Using LiveInputs 1118
Managing LiveGroup Controls 1120
Picking and Adding Controls 1120
Expression Linking Controls 1121
Editing and Publishing LiveGroups 1123
Editing LiveGroups 1123
Versioning LiveGroup Scripts 1124
Publishing LiveGroups 1125
Overriding LiveGroup Controls 1127
Configuring Nuke 1130
What Is a Terminal and How Do I Use One? 1131
Command Line Operations 1132
On Windows 1132
On Mac 1132
On Linux 1133
Using command line Flags 1134
General syntax 1140
Examples 1140
Using [argv 0] 1141
Using Python to convert TIFFs to JPEGs 1142
Environment Variables 1143
Setting Environment Variables 1143

On Windows 1143
On Mac 1143
On Linux 1144
To Check if an Environment Variable Exists 1145
From Inside Nuke 1145
In the Windows Environment 1145
In the Mac or Linux Environment 1145
To Display a List of Set Environment Variables 1145

On Windows 1145
On Mac or Linux 1146
Nuke Environment Variables 1147
Loading Gizmos, NDK Plug-ins, and Python and

Tcl Scripts 1154
Defining the Nuke Plug-in Path 1156
Loading OFX Plug-ins 1157
Defining Common Favorite Directories 1158
To Define a Common Set of Favorite Directories 1158

Example 1 1158
Example 2 1159
Handling File Paths Cross Platform 1161
Setting Default Values for Controls 1162
Defining Custom Menus and Toolbars 1163
To Add a Toolbar 1164

Example 1 1165
Example 2 1165
To Define a Menu or Toolbar Option 1165

Example 1 1167
Example 2 1167
Example 3 1168
Defining Common Image Formats 1169
Defining an Image Format 1169

Example 1169
Creating and Accessing Gizmos 1170
Creating Gizmos 1170
Managing Gizmo Controls 1172
Adding Knobs Using Drag-and-Drop 1172

Editing Knobs 1175
Organizing Knobs 1177
Adding Tabs 1178
Hiding and Removing Knobs 1179
Adding Knobs Manually 1181
To Pick Existing Controls 1181

To Create New Controls: 1183
To Delete Controls: 1185
Accessing Gizmos in Nuke 1192
Sourcing Custom Plug-ins and Generic Tcl

Scripts 1194
Nuke Custom Plug-ins 1194

To Source a Custom Plug-in 1194
Sourcing Tcl Procedure 1194

To Source a Generic Tcl Procedure 1194
Template Scripts 1195
Creating and Using a Template Script 1195
Defining Common Preferences 1196
Defining Default Preferences 1196
Deleting (and Resetting) the Preferences 1197
Altering a Script’s Lookup Tables (LUTs) 1198
Managing Nuke's Native LUTs 1199
To Display LUT Curves 1199
To Create a New LUT 1199
To Edit LUTs 1200

To Reset the LUT Curves Back to Their Initial Default Shapes 1200
To Remove LUTs 1200
Selecting the LUT to Use 1202
Default LUT Settings 1202
To Change the Default LUT Settings 1203
Example Cases 1204
Working in Video Colorspace 1204
Linear Data in 16-Bit Files 1204
Cineon Displays 1204
Color Management 1204
OCIO Color Management 1205
Creating Custom Viewer Processes 1207
Using a Gizmo as a Custom Viewer Process 1209
To Register a LUT in the Project Settings as a Viewer Process 1209
To Create a Viewer Process Gizmo 1210
To Register a Custom Viewer Process 1211
Applying Custom Viewer Processes to Images 1213
To Apply Your Custom Viewer Process to Images Displayed in a

Viewer 1213
To View the Controls of the Currently Active Viewer Process 1213
Expressions 1214
Quick Start 1214
Linking Expressions 1215
Referencing Values from Another Parameter 1216
Linking Channels and Formats Using Expressions 1218
Adding Mathematical Functions to Expressions 1219

Converting Expressions Between Scripting
Languages 1227
The Script Editor and Python 1228
Quick Start 1228
Using the Script Editor 1229
Opening the Script Editor 1229
Input and Output Panes 1229
Entering a Statement 1229
Moving Through and Clearing the Script History 1231
Clearing the Output Pane 1231
Automating Procedures 1232
Saving Statements in a Python Module 1232
Opening a Python Script in the Script Editor 1232
Importing and Executing a Python Script 1232
Nuke as a Python Module 1234
Getting Help 1235
More Documentation 1235
Viewing More Examples 1235
Using the Help Statement 1235
Python on the Web 1236
Timeline Editing in Nuke Studio 1237
Using Tags 1238
Using Quick Tags 1238
Tagging Using the Viewer 1240
Tagging Shots 1241

Adding Notes to Tags 1243
Removing Tags 1244
Creating Custom Tags 1245
Sharing Custom Tags 1246
Filtering and Flagging Media Using Tags 1246
Viewing Metadata 1248
Source Clip and Shot Metadata 1248
Filtering and Flagging Media Using Metadata 1249
Conforming Sequences 1250
Timeline Environment Project Settings 1251
Color Management Settings 1252
Adding OCIO Configurations 1254
Importing Sequences 1256
Notes on AAF Sequences 1259
Conforming Sequences 1261
Conforming Using a Browser 1262
Conforming with Pre-ingested Media 1266
About the Media Spreadsheet 1268
Sorting and Custom Columns 1268
Spreadsheet Controls 1269
Adjusting Timecodes 1271
Renaming Shots on the Timeline 1274

Saving and Loading Projects 1276
Autosaved Projects 1277
Managing Timelines 1278
Adding Tracks to the Timeline 1281
Adding Clips to the Timeline 1284
Audio and the Timeline 1287
WAV Shots 1290
Displaying Audio Waveforms 1291
Audio Scrubbing 1292
Synchronizing Audio and Video 1293
PulseAudio on Linux 1294
Stopping PulseAudio 1294
Restarting PulseAudio 1295
Using Reference Media 1296
Comparing Media 1297
Caching Frames in the Playback Cache 1299
Caching Frames in the Disk Cache 1301
Caching Sequence Ranges 1303
Caching Selected Shot Ranges 1305
Caching In/Out Ranges 1307

Clearing Cached Frames 1309
Clearing All Cached Frames 1309
Clearing Sequences 1309
Clearing Selected Shot Ranges 1311
Clearing In/Out Ranges 1312
Viewing Multi-Format Timelines 1314
Setting the Sequence Format 1315
Reformatting Shots 1316
Refreshing and Replacing Shots 1318
Setting Soft Trims 1319
Enabling and Disabling Shots 1321
Adding Transitions 1322
Invalid Transitions 1324
Retiming Shots 1325
Using Freeze Frames 1329
Blending Tracks on the Timeline 1330
Adding New Blend Tracks 1330
Converting Tracks to Blend Tracks 1331
Masking Blended Tracks 1332
Stereoscopic and Multi-View Projects 1334
Creating Views in a Project 1336
Creating and Managing Views 1336
Importing Source Clips 1338

Importing Single-View Clips 1338
File Name Variables 1338
Directory Name Variables 1340
Importing Multi-View Clips 1342

Multi-View QuickTime Files 1342
Displaying Views in the Viewer 1343
Displaying a Particular View 1343
Displaying Two Views Next to Each Other 1343
Displaying a Blend Between Two Views 1344
Displaying OpenGL Stereo in Timeline Viewers 1344

Enabling OpenGL Stereo Output 1344
Windows 1344
Linux 1346
Switching to OpenGL Stereo Output 1346
Displaying Views in the Timeline 1348
Splitting Views to Separate Tracks 1348

Displaying Two Views Next to Each Other 1350
Displaying a Blend Between Two Views 1350
Applying Changes to Selected Views 1352
Splitting Views to Tracks 1352
Splitting Views in the Properties Panel 1353
Showing Separate Values for Each View 1354
Unsplitting Views 1355
Soft Effects 1356
Available Soft Effects 1356
Adding Sequence-Level Soft Effects 1359

Using the Spreadsheet View 1360
Adding Shot-Level Soft Effects 1362
Soft Effect Controls 1362
Editing Sequence-Level Soft Effects 1364

Moving 1364
Copying 1364
Cloning 1364
Copying as Clone 1365
Decloning 1365
Deleting 1365
Editing Shot-Level Soft Effects 1366

Enabling and Disabling Soft Effects 1366
Create Comp 1367
Edit Export Settings 1367
Edit Project Settings 1368
Creating and Editing a Comp 1369
Nuke Comp Colors 1371
Creating Multiple Shot Comps 1372
Create Comp Special 1374
Enabling and Disabling a Nuke Comp 1376
Saving a New Version and Updating the Comp 1376
Stereo and Multi-View Comps 1376
Annotations 1380
Workflow 1380
The Annotations Menu 1381
Adding Annotations 1381
Enabling and Disabling Annotations 1384
Editing Sequence-Level Annotations 1384
Editing Shot-Level Annotations 1385
Viewing Annotations in the Compositing Environment 1386

Viewing the Annotation Node Group 1386
Re-Exporting Annotations from the Timeline 1387
Timeline Editing Tools 1389
Using the Multi Tool 1392
Using the Move/Trim Tool 1394
Moving Shots 1394

To Move Shots Using the Spreadsheet View: 1396
Trimming Shots 1397

To Trim Shots Using the Spreadsheet View: 1399
Using the Selection Tools 1402
Using the Slip Clip Tool 1405
Slipping Using the Spreadsheet View 1406
Using the Slide Clip Tool 1408
Using the Roll Edit Tool 1410
Using the Retime Clip Tool 1412
Using the Razor and Join Tools 1414
Copying Cuts Between Tracks 1415
Insert, Overwrite, and 3-Point Editing 1417
Inserting Clips 1417
Overwrite Edits 1418
3-Point Editing 1419
Versions and Snapshots 1423
Using Versions 1424
Versions in Bins 1426
Versions in Sequences 1428
Using Snapshots 1429
Creating Snapshots 1430
Restoring Snapshots 1432

Exporting from Nuke Studio 1433
Introduction to the Export Dialog 1434
Using Local and Project Presets 1436
Using the Shot Template 1436
Custom Shot Templates 1439

Multi-format Exports 1439
Adding Burn-in Text to Exports 1442
Adding Additional Nodes During Export 1444
Using the Frame Server on External Machines 1446
Configuring the Frame Server on External Machines 1446
Frame Server Logs 1448
Exporting Sequences and Shots 1450
Nuke Project File Settings 1451
Nuke Write Node Settings 1453
Tracks, Range, and Handles Settings 1456
Building VFX Tracks and Comp Clips 1458
Building Tracks From Export Structure 1458
Building Comp Clips From Export Tags 1460
Building Render Placeholders From Export Tags 1461
Exporting Multi-View Source Clips 1463
Transcoding 1465
Transcoding a Sequence 1465
Tracks and Range Settings 1467

Transcoding a Sequence as Shots 1469
Transcoding Timeline Selections 1471
Transcoding from the Bin View 1472
Ad Hoc Exports 1473
Exporting EDLs and XMLs 1474
Using the Audio Exporter 1476
Exporting Audio from Sequences 1476
Exporting Audio from Shots 1477
Exporting Audio from Source Clips 1479
Using the Copy Exporter 1481
Using the SymLink Generator 1483
Advanced Compositing with NukeX and Nuke

Studio 1485
Retiming and Motion Blur 1487
VectorGenerator 1488
Quick Start 1489
Connecting VectorGenerator 1490
Viewing and Refining the Results 1491
Kronos 1493
Example Input Speed Calculation 1494
Quick Start 1495

Connecting Kronos 1496
Retiming a Sequence 1498
Retiming a Sequence Using Output Speed 1498
Retiming a Sequence Using Input Speed 1499
Retiming a Sequence Using Frame 1500
MotionBlur 1506
Quick Start 1507
Connecting MotionBlur 1508
Adjusting MotionBlur Controls 1509
Working with Lens Distortion 1511
Quick Start 1511
Estimating Lens Distortion Using a Grid 1513
Adjusting Grid Detection Parameters 1518
Feature Detection 1518
Link Detection 1519
Removing Lens Distortion from an Image 1521
Estimating and Removing Lens Distortion Using

Lines 1524
Applying Lens Distortion to an Image 1529
Working with STMaps 1530

Tracking with PlanarTracker 1532
Tracking a Plane 1533
Drawing a Plane to Track 1533
Tracking the Footage 1534
Tracker Menu Options in the Viewer 1536
Track Motion Controls 1536
Tracking Controls 1537
Clear Tracking Data Controls 1537
Planar Tracker Surface Controls 1538
Reference Frame Controls 1538
Keyframe Controls 1538
Layer Dropdown 1539
CornerPin Dropdown 1539
Tracking Tab Menu Options 1540

Settings 1540
Export 1541
Correction 1541
Reference Frame 1541
Reusing a Track Result 1543
Reusing a Tracked Plane 1543
Reusing a Single Track 1543
Placing an Image on the Planar Surface 1544
Camera Tracking 1548
Introduction 1548
Quick Start 1548
Connecting the CameraTracker Node 1550
Masking Out Regions of the Image 1551
Working with Multi-View Scripts 1553

Setting Camera Parameters 1554
Tracking in Sequence Mode 1556
Viewing Track Data 1559
Reviewing AutoTracks Curves 1559
Troubleshooting Sequence Tracks 1561
Pre-Tracking Checks 1561
Reviewing and Refining Tracking Data 1561
Re-tracking Using Tracking Settings 1563
Extending Existing Camera Tracks 1565
Retracking Partial Frame Ranges 1566
Tracking in Stills Mode 1567
Still Photography Guidelines 1568
General Guidelines 1568
Shooting Near-flat Scenes 1568
Shooting 3D Objects 1568
Shooting Interiors 1569
Tracking Still Frames 1570
Selecting the Frames to Track 1570
Adjusting the Features to Track 1571
Tracking the Selected Frames 1573
Viewing Reference Frames and Track Data 1574
Disconnected Frame Sets 1576
Troubleshooting Still Tracks 1577

Pre-tracking Checks 1577
Post-Tracking Refinements 1577
Retracking Partial Frame Ranges 1578
Working with User Tracks 1579
Adding and Positioning User Tracks 1580
User Tracking Methods 1582
Auto User Tracks 1582
Manual User Tracks 1583
Extracting User Tracks 1584
Importing a Tracker 1584
Tracking Assists 1585
Tracking a Scene Manually 1586
Linking Still Reference Frames 1587
Assigning 3D Survey Points 1588
Solving the Camera Position 1590
Viewing Solve Data 1591
Reviewing Solved User Tracks 1593
Previewing Matchmove Quality 1593

Method 1 1594
Method 2 1595
Troubleshooting Solves 1597
Using Curve Thresholds to Delete Tracks 1597
Refining a Solve 1599
Updating Solves with Extended Tracking Data 1600

Adjusting the Scene 1601
Setting the Ground Plane and Axes 1602
Setting Axes Individually 1603
Setting the Origin Within a Scene 1603
Transforming the Scene Manually 1605
Transforming Using Channel Files 1605
Transforming Using Snap To Position 1606
Using Scene Constraints 1607
Using Solve Data 1609
Creating Camera Nodes 1610
Creating a Camera Rig 1610
Creating a Camera Set 1611
Creating Scenes 1613
Creating a 3D Scene with ScanlineRender and LensDistortion 1613
Creating Point Clouds 1615
Creating Cards 1616
Combining Solves 1618
Placing Objects in the Scene 1619
Accounting for Lens Distortion 1620
Distorting Your CG Elements to Match Your 2D Footage 1620
Undistorting Your 2D Footage to Match Your CG Elements (Using

LensDistortion) 1620
Undistorting Your 2D Footage to Match Your CG Elements

(Using CameraTracker) 1621
About the Lens Model 1622
Using MatchGrade 1624
Extracting a Baked-In Grade 1625
Analyzing Reference Frames 1626
Auto-Analyzing Per Frame 1629
Matching a Grade Between Two Different Clips 1631
Using the Smart Vector Toolset 1636
Quick Start 1636
Generating Motion Vectors 1637
Masking Foreground and Background Areas 1638
Adding Paint to the Source 1640
Adding an Image to the Source 1642
Applying Motion Vectors to the Source 1643
Warping the Vectors 1643
Improving Warps 1646
Warping Images Using VectorCornerPin 1648
Warping Multiple Reference Frames 1653
Generating Depth Maps 1655
Quick Start 1655
Connecting DepthGenerator 1656
Selecting What to Output 1658
Analyzing Depth 1660

Using the Results 1665
Creating Dense Point Clouds 1666
Quick Start 1666
Connecting the PointCloudGenerator Node 1667
Masking Out Regions of the Image 1668
Setting Keyframes in a Sequence 1669
Setting Keyframes Automatically 1669
Setting Keyframes Manually 1669
Tracking a Dense Point Cloud 1670
Filtering Your Point Cloud 1672
Removing Rejected Points 1673
Grouping, Labeling, and Baking Points 1675
Creating a Mesh from a Point Cloud 1677
Adding Texture to a Mesh 1678
Using the PoissonMesh Node 1679
Adding Texture to the PoissonMesh 1680
Using ModelBuilder 1681
Prerequisites 1681
Quick Start 1682
Connecting the ModelBuilder Node 1683

Creating Shapes 1684
Editing Shapes’ Display Characteristics 1687
Positioning Shapes 1689
Editing Shapes 1691
Editing Vertices 1693
Editing Edges and Edge Loops 1695
Editing Faces 1699
Editing Objects 1703
Setting the Initial Transform Action Center 1705
Applying Textures 1708
Projecting Textures onto Your Shapes 1710
UV Unwrapping 1712
Quick Start 1713

Creating Seams on Your Model 1714
To Create Seams 1714
Unwrapping Your Model and Previewing Its UVs 1715
Editing UVs 1717
Applying Textures 1718

Method 1 1718
Method 2 1720
Exporting Shapes to Separate Geometry Nodes 1723
To Export Shapes to Separate Geometry Nodes: 1723
Creating 3D Particles 1724
Quick Start 1724

Connecting Particle Nodes 1725
Emitting Particles 1726
Spawning Particles with ParticleSpawn 1731
Adjusting the Speed and Direction of Particles 1732
Applying Gravity to Particles 1732
Aligning Particles 1732
Controlling Particle Speeds 1732
Attracting Particles to a Specific Point 1733
Modifying the Particles’ Movement 1734
Bouncing Particles off Objects 1734
Adding Drag to Particles 1735
Adding Turbulence Motion on Particles 1735
Adding Spiral Motion to Particles 1736
Adding a Wind Effect to Particles 1737
Adjusting Controls Common to Several Particle

Nodes 1738
Particle Rendering Controls 1738
Particle Transform Controls 1738
Condition and Region Controls 1738

Conditions tab 1738
Region tab 1739
Adjusting Particle Properties Using Curves 1740
Adjusting Particles Using Expressions 1741
Particle Expression Functions 1742
Adjusting Particle Simulation Settings 1746
Merging Particle Streams 1747

Controlling Particles by Channel 1748
Caching Particles 1751
PrmanRender 1753
Setting Up RenderMan Pro Server and PrmanRender 1753
Using The PrmanRender Node 1754
Render Quality 1755
Shadows, Reflections, Refractions and Depth of

Field 1756
Motion Blur Parameters 1757
Shader Parameters 1758
RIB Parameters 1759
Using the ModifyRIB Node 1760
Using the Reflection Node 1761
Using the Refraction Node 1762
Using F_DeFlicker2 1763
Introduction 1763
Quick Start 1764
Parameters 1765
Using F_ReGrain 1766
Introduction 1766
Quick Start 1767

Adding Sampled Grain 1767
Using Pre-Sampled, Standard Grain Types 1768
Response 1769
Checking the Result 1770
Proxy Resolutions 1771
Parameters 1772
Color Space in FurnaceCore Plug-ins 1775
Color Space in F_ReGrain 1775
Using F_WireRemoval 1776
Introduction 1776
Background 1777
Clean Plates 1777
FurnaceCore 1777
Reconstruction Methods 1778
Tracker 1779
Quick Start 1780
Positioning the On-Screen Wire Tool 1781
Tracking 1782
User Keyframes and Track Keyframes 1783
Indicators on the On-Screen Wire Tool 1784
Tracker Controls 1785

Parameters 1787
Global Motion Estimation 1790
Introduction 1790
What is Global Motion Estimation? 1791
Limitations of GME 1791
Global Motion Estimation Effects 1792
Using Them 1792
Controls 1794
Parameters That Affect Analysis 1794
Parameters That Affect Rendering 1795
Widgets 1796
Using F_Align 1797
Introduction 1797
Quick Start 1799
Analyzing with the Analyse button 1799
Analyzing On The Fly 1800
Parameters 1801
Using F_RigRemoval 1803
Introduction 1803
Quick Start 1804
Occlusions 1806
Parameters 1808
Using F_Steadiness 1810

Introduction 1810
Quick Start 1812
Smoothing Out Camera Motion 1812
Locking To A Frame 1812
Parameters 1814
Using the BlinkScript Node 1817
About BlinkScript 1817
Quick Start 1819
Connecting the BlinkScript Node 1820
Loading, Editing, and Saving Kernels 1821
Loading Kernels 1821
Editing Kernels 1822
Saving Kernels 1824
Setting Kernel Parameters 1825
GPU or CPU? 1825
Specifying the Output Format 1826
Changing Performance Settings 1826
Publishing and Protecting Your Kernels 1827
Limitations and Known Issues 1829
RGBA Only 1829
GPU 1829
Crashing and Infinite Loops 1829
Known Issues 1829

OpenCL on OS X 10.8 (Mountain Lion) 1829
Written Tutorials 1830

Introduction 1830
The Projects 1831
Installing the Project Files 1832
To Create the Tutorial Directory (Windows) 1832
To Create the Tutorial Directory (Mac or Linux) 1832
To Download and Install the Project Files 1832
Tutorial 1: Compositing Basics 1833
Introduction 1833
Starting Nuke 1834
To Launch Under Windows 1834
To Launch Under Mac 1834
To Launch Under Linux 1834
Using the Toolbar 1836
Using the Menus 1837
Customizing Your Workspace 1839
Saving Files and File Backup 1841
Defining File/Saving Options 1841
Recovering Back-Up Files 1842
Turning off Automatic Back-Up 1842
Setting Up the Project 1843
To Set Up Your Project 1843
Working with Nodes 1845
Inserting Nodes 1845
Connection Tips 1849

Importing Image Sequences 1852
To Read the Images 1852
Navigating Inside the Windows 1855
Panning Your View 1855
Zooming or Magnifying Your View 1855
Using the Node Graph Overview 1855
Framing the View in the Window 1856
Working with Viewers 1857
Displaying the Images in a Viewer Window 1858
Viewing Multiple Inputs 1860
Reformatting Images 1862
To Conform Images to the Project Format 1862
Using Proxies and “Down-res” 1863
To Activate Proxy Mode 1863
To Activate “Down-Res” 1863
Compositing Images 1865
To Composite Two Nodes 1865
Color-Correcting Images 1868
Masking Effects 1869
To Create and Apply a Bezier Mask 1869
Creating Flipbook Previews 1870
To Generate a Flipbook 1870
Rendering Final Output 1870

To Render the Result of Your Composite 1870
Using the Nuke Frame Number Variable 1872
Image Formats 1872
Rendering with the Active Resolution 1872
Rendering Multiple Channels 1873
Epilogue 1874
Tutorial 2: 2D Point Tracking 1875
Introduction 1875
One-Point, Two-Point, Three-Point, Four 1877
Open the Tutorial Project File 1879
To Open the Project File 1879
Tracking a Single Feature 1880
Setting a Tracking Anchor 1880
Auto-Tracking vs. Keyframe Tracking 1881

Using Auto-Tracking 1881
Using Keyframe Tracking 1883
Editing Track Data 1884
Tracking Obscured Features 1886
To Track a Feature That Moves off Screen 1886
Stabilizing Elements 1888
To Track and Stabilize 1888
Match-Moving Elements 1891
To Match-Move an Element 1891
Epilogue 1893
Tutorial 3: Keying and Mattes 1894
Introduction 1894

Keying with Primatte 1896
To Pull a Key with Primatte (Method 1) 1896
To Pull a Key with Primatte (Method 2) 1897
Image-Based Keying 1901
To Pull a Key with IBK 1901
Rotoscoping 1906
To Draw a Garbage Matte 1906
Epilogue 1911
Tutorial 4: 3D Integration 1912
Introduction 1912
The Basic 3D System 1913
The 3D Viewer 1913
The Geometry or Scene Node 1914
The Camera Node 1914
The ScanlineRender Node 1914
Setting Up a 3D System 1917
To Set Up a 3D Node Tree 1917
To Position Objects in the Scene 1919
Making a Scene 1922
To Set Up a Scene 1922
Merging and Constraining Objects 1925
To Move the Merged Objects Together 1926
To Make Objects ‘Look’ at the Camera 1926

Animating a Scene 1928
To Animate the Camera 1928
Working with Geometry 1931
To Add Primitive Objects to the Scene 1931
Lighting and Surface Properties 1936
To Define Surface Attributes to Objects 1936
To Add Light Objects to a Scene 1937
Epilogue 1939
Appendices 1940
Organization of the Section 1940
Appendix A: Preferences 1941
The Available Preference Settings 1941
General 1941
Project Defaults 1942
Performance 1947
Behaviors 1957
Panels 1962
Appendix B: Keyboard Shortcuts 1979
Conventions 1979
Global 1980
Nuke Studio's Timeline Viewer 1981
Nuke Studio's Timeline 1984
2D Compositing Viewer 1986
3D Compositing Viewer 1991
Node Graph 1992
Project/Tags/Versions Bin 1997
Properties Panel 1998

Curve Editor/Dope Sheet 2003
Script Editor 2004
Roto/RotoPaint 2005
Appendix C: Supported File Formats 2007
Notes 2007
Supported File Formats 2007
Supported Camera Formats 2016
Supported Audio Formats 2018
Appendix D: External Software 2019
Third-Party Contributions 2019
Third-Party Libraries and Fonts 2021
Third-Party Library Versions 2021

Getting Started
An overview of installation and licensing, the default Nuke environments and workspaces, and the
Preferences dialog.
The Nuke Family
Learn about the different products and modes of Nuke as well as some key concepts, everything you
should know before using Nuke products.
Installation & Licensing
Find out how to install and license Nuke on Windows, Mac, and Linux to the point where you have the
application in front of you and are ready to start work.
Comp Environment
Discover the basics of Nuke's compositing environment, designed to streamline day-to-day workflow.
Flexible, efficient and feature packed, this toolset delivers film-grade results, fast.
Timeline Environment
Explore Nuke Studio's timeline environment allowing you to conform, review, edit, and even create
and render compositions from the timeline.
USER GUIDE
80
Meet the Nuke Product Family
These pages introduce the different products and modes of Nuke and provide links to some key
concepts you should know before using Nuke products.
Nuke Products
The Nuke products have different purposes and different levels of functionality.
Nuke Studio is a product that combines the functionality of all the Nuke products listed below, which
are then used as different modes in which to run Nuke Studio. It also includes its own specific
functions that are only applicable to Nuke Studio.
All the Nuke products listed below can also be licensed as separate products, except for Nuke Assist,
which is included with NukeX and Nuke Studio.
The Nuke Products include the following:

• Nuke - Foundry’s award-winning compositing tool. Nuke provides a stable foundation from which
to deliver shot-based VFX work.
• NukeX - includes all of Nuke’s features with the addition of a range of specialized plug-in tools to
eliminate the need to perform certain tasks in third-party software packages. To run it, you need
both a Nuke and a NukeX license, or a Nuke license.
See About NukeX and the Advanced Compositing with NukeX and Nuke Studio section for more
information.
• Nuke Studio- the most powerful application in the Nuke product family, it includes all the
functionality of Nuke and NukeX with the addition of timeline and conform workflows.
• Nuke Assist - a feature of NukeX intended for use as a workstation for artists performing painting,
rotoscoping, and tracking. Two complementary licenses of Nuke Assist are included in every NukeX
maintenance package. Nuke Assist only supports a limited subset of Nuke tools.
See About Nuke Assist for a run down of the functionality provided by Nuke Assist.
• Nuke Non-commercial - a special version of the commercial Nuke application that you can run
without a license, intended for use as an educational tool.
See About Nuke Non-commercial for a run down of the restrictions placed on using the non-
commercial version.
USER GUIDE
81
About Nuke |
About Nuke
When you purchase Nuke, it contains all the Nuke products as different modes. For example, you can
choose to run Nuke in NukeX mode, or Nuke Non-commercial mode.
Nuke modes include the following:
Mode Description
Nuke Studio Includes the Timeline environment and Compositing environment, and has
some additional Nuke Studio specific features, such as the ability to create a
Nuke Comp, add soft effects to the timeline, and the annotations menu. Nuke
Studio combines the tools and abilities of Nuke, NukeX, and Hiero.
For more information about Nuke Studio, see Nuke Studio Environments.
Nuke A single shot compositor (what is referred to as the Compositing environment

in Nuke Studio), offering a Node Graph, Viewer, animation tools, and so on.
NukeX Just like Nuke, a single shot compositor (what is referred to as the
Compositing environment in Nuke Studio), with additional tools such as
CameraTracker, DepthGenerator, FurnaceCore, Kronos, and so on.
See About NukeX for more information.
Nuke Assist Provides basic tools including Bezier, Draw, Roto, and so on. The Nuke Assist
mode is intended for use as a workstation for artists performing painting,
rotoscoping, and tracking.
SeAbout Nuke Assist for more information.
Nuke Non- This mode is intended for personal and educational use. It includes all the
commercial features of the commercial version of Nuke, offering you a chance to explore
and learn the application fully while using it from the comfort of your own
home.
See About Nuke Non-commercial for more information.
USER GUIDE
82
About NukeX |
About NukeX
When using NukeX, you have all the features of Nuke in use, plus the following:
• CameraTracker
• DepthGenerator
• FurnaceCore
• Kronos
• LensDistortion
• MotionBlur
• ModelBuilder
• Particles
• PlanarTracker
• PointCloudGenerator
• PoissonMesh
• PrmanRender
• ProjectionSolver
• VectorGenerator
Nuke and NukeX are fully compatible, with Nuke capable of viewing and rendering - but not editing -
NukeX features.
USER GUIDE
83
About Nuke Assist |
About Nuke Assist

Nuke Assist is licensed as part of a NukeX or Nuke Studio maintenance package and is intended for
use as a workstation for artists performing painting, rotoscoping, and tracking. Nuke Assist doesn’t
support any NukeX-or Nuke Studio-specific features apart from PlanarTracker, and has a limited
subset of Nuke nodes and features available. Nuke Assist does not support any custom plug-ins, only
the following nodes are supported:
Nuke Assist Nodes
Image
Checkerboard ColorBars ColorWheel Constant
Read Viewer
Draw
Bezier Radial Ramp Rectangle
Roto RotoPaint
Note: Bezier is only available through the X menu. Press X in the Node Graph and then
enter Bezier as a Tcl command to add the node.
Time
FrameBlend FrameHold FrameRange TimeEcho
TimeOffset
Channel
Add Copy ChannelMerge Remove
Shuffle ShuffleCopy
Color
Grade Invert OCIO CDLTransform OCIO Colorspace
OCIO Display OCIO FileTransform OCIO LogConvert
USER GUIDE
84
About Nuke Assist |
Nuke Assist Nodes
Filter
Blur
Keyer
Keyer
Merge
AddMix Dissolve KeyMix Merge
Premult Switch Unpremult
Transform
Crop CornerPin PlanarTracker Reformat
Tracker Transform TransformMasked
Views
JoinViews OneView ShuffleView Split and Join
Stereo Anaglyph Stereo MixViews Stereo ReConverge Stereo SideBySide
Metadata
AddTimeCode CompareMetadata CopyMetadata ModifyMetadata
ViewMetadata
Other
Backdrop Dot Group Input
Output PostageStamp StickyNote
Note: You cannot render output using Write nodes in Nuke Assist.
You can load projects created in Nuke's other modes and work as normal, within the constraints of
Nuke Assist. The Viewer renders the output of the node tree whether the components are supported
or not. Any unsupported nodes and plug-ins in the Node Graph are outlined in red and their controls
are grayed out. You cannot modify the output of unsupported nodes and plug-ins.
USER GUIDE
85
About Nuke Assist | Using Gizmos, Groups, Precomps, Knobs, and Python
Tip: For more information on node trees and node-based compositing, see Understanding
the Workflow.
Using Gizmos, Groups, Precomps, Knobs, and Python

Gizmos, Group, and Precomp nodes can be loaded as normal, but if they contain any nodes that are
not supported by Nuke Assist they have a red outline around the node in the Node Graph and the
control panel controls are grayed out.
Note: Nuke Assist allows the creation of custom knobs, but they can't be edited.
Python scripts work as usual for nodes that are supported by Nuke Assist, but any attempt to add
unsupported nodes displays an error message in the Script Editor output section, or terminal (in -t
mode).
For example, executing nuke.createNode('Transform') adds a Transform node to the Node Graph,
but nuke.createNode('Convolve') prints the following error:
# Result:
Traceback (most recent call last):
File "<string>", line 1, in <module>
RuntimeError: Convolve is not available in Nuke Assist
USER GUIDE
86
About Nuke Non-commercial | Using Gizmos, Groups, Precomps, Knobs, and Python
About Nuke Non-commercial

Nuke Non-commercial is a free version of Nuke that runs outside the regular licensing model. Nuke
Non-commercial is meant for personal, educational, and other non-commercial use. It is aimed at
students, industry professionals, and others interested in Nuke. It includes most of the features of the
commercial version of Nuke, offering you a chance to explore and learn the application fully while
using it from the comfort of your own home.
You can run Nuke, NukeX, and Nuke Studio in non-commercial mode using the --nc command line
argument. For example, to launch Nuke Studio in non-commercial mode on Mac, enter:
/Applications/Nuke11.3v5/Nuke11.3v5.app/Contents/MacOS/Nuke11.3v5 --nc --studio
Nuke Non-commercial is a fully functional version of Nuke, but as it's designed for non-commercial
use only, it does differ from the commercial version in some aspects. Here are the main differences:
• Certain nodes are disabled in Nuke Non-commercial, including BlinkScript, GenerateLUT, Primatte,
Ultimatte, and WriteGeo.
• Rendered output is restricted to 1920x1080 HD and the MPEG4 and H.264 formats are disabled.
• command line renders are restricted to encrypted .nknc scripts.
• Frame Server slave rendering is disabled.
• Exporting EDL/XML sequences is disabled.
• Exporting LUTs from MatchGrade is disabled.
• Gizmos, clipboard contents, .hrox project files, and .nk scripts are all encrypted.
• Monitor Output is disabled.
In other respects, Nuke Non-commercial contains all the functionality of the commercial version of
Nuke.
USER GUIDE
87
Key Concepts | Understanding the Workflow
Key Concepts
Nuke products are a resolution-independent compositing system, with extensive channel support,
powerful image manipulation tools, and a rich 3D compositing environment. This section explains
concepts you should know before using Nuke products.
Understanding the Workflow

A Nuke project consists of a network of linked operators called nodes. Nuke's Compositing
environment utilizes a node-based workflow, where you connect a series of nodes to read, process,
and manipulate images. Each node in the script, or comp, performs an operation and contributes to
the output.
You can open a Nuke comp file in a text editor, and a series of sequential commands are displayed,
which are interpreted and executed when you render the output.
USER GUIDE
88
Key Concepts | Working with Multiple Image Formats
In the image above, you see an example of a very simple Nuke script. Two Read nodes reference
media on disk. Effect nodes extract a matte and blur an image. A Merge node set to over, composites
the foreground image (input A) over the background image (input B). Finally, a Write node renders
and outputs the completed composite to disk. There is also a Viewer node, which displays the output
of any node in the script.
Note: Nuke Assist does not support Write nodes or render farms. See Nuke Products for
more information.
Working with Multiple Image Formats

Nuke products support multiple file formats, such as Cineon, TIFF, OpenEXR, HDRI, and RAW camera
data (using the dcraw command line program), and allows you to mix them all within the same
composite. By default, Nuke products convert all imported sequences to their native 32-bit linear RGB
colorspace. You can, however, use the Colorspace node to force one of several color models,
including sRGB, Cineon, rec709, gamma 1.80/2.20, HSV, or HSL. The Log2Lin node lets you convert
between logarithmic and linear colorspace (and vice versa).
USER GUIDE
89
Key Concepts | Channel Operations
Note: Nuke Assist does not support Colorspace or Log2Lin nodes. See Nuke Products for
more information.
There are no restrictions on image resolution - you can freely mix and scale elements of any
resolution within the same script. You can, for example, use a 2k film plate as the background for
video shot in PAL format, and then output the result in HD1080i. Nuke products automatically adjust
their Viewers to accommodate the image you’re viewing.
Channel Operations
In Nuke products, you can assign the output of each node as new channels, and pass them to the
next node in the script. When you need to re-use a particular channel (say, to apply a color correction
to the hair), you simply choose the channel containing the matte from the downstream color-
correction node.
Nuke products support up to 1023 channels of image data. This provides additional benefits when
working with computer-generated (CG) elements, especially when such elements are rendered out in
the OpenEXR format.
For example, your 3D department could render out multiple lighting passes for a particular CG
element (beauty, fill, backlight, reflection, shadow, etc.) as an .exr sequence, which you could then
read into a Nuke script, or comp. You would be able to access all of the render passes stored within
the .exr sequence from any downstream node in your script.
You might choose to color correct only the CG element’s highlights by using the specular pass as a
mask to a particular color correction operator. Such an approach again has the advantage of keeping
the Nuke comp free of unnecessarily complex branching - virtually all render passes and mattes can
be passed through a single pipe in the comp.
USER GUIDE
90
Key Concepts | 8-, 16-, and 32-Bit Image Processing
The Channels chapter explains how to take full advantage of the 1023-channel workflow.
8-, 16-, and 32-Bit Image Processing

Some digital compositing systems, especially those geared for video work, are optimized for
processing exclusively 8-bit elements (that is, images with 256 intensity values per channel). Other
systems allow for the mixing of 8, 16, and 32-bit elements.
For Nuke products, which began as a film effects tool, image quality is paramount. Thus, they support
the processing of exclusively 32-bit-per channel elements (elements with lower bit depths are
converted to 32 bits per channel upon import). Thirty-two bit support allows for a much richer palette
of colors and floating point precision in all script calculations. In practice, this means that Nuke
products carry out every operation - from an increase in gamma to a transform - with much greater
accuracy than a lower-bit-depth system.
Compositing in 3D
Some digital compositing systems support a strictly two-dimensional workflow. Nuke products, by
contrast, offer a robust 3D workspace that lets you create and render complex scenes composed of
polygonal models, cards (planes textured with images), cameras, lights, and textures.
This 3D workspace has countless uses, the simplest of which is generating pan-and-tile scenes. These
are scenes with 2D image planes arranged into a curved shape, and then rendered out through an
animated camera to give the illusion of a seamless environment.
Simple pan-and-tile scene.
The 3D Compositing chapter explains how to make full use of Nuke’s 3D workspace.
USER GUIDE
91
Key Concepts | Render Farms and Frame Servers
Render Farms and Frame Servers

Nuke products support virtually all third-party and proprietary render-queuing software. By
integrating Nuke products with such a system, the render load can be distributed across all the Nuke-
or NukeX-licensed machines on your network, whether Windows, Mac, or Linux-based. See Render
Farms for more information.
In addition, Nuke Studio ships with an internal Frame Server, which also allows you to setup external
slave machines to process renders faster. See Using the Frame Server on External Machines for more
information.
Note: Nuke Assist does not support Write nodes, render farms, or the Frame Server. See
Nuke Products for more information.
USER GUIDE
92
Installation and Licensing
We know the installation and licensing of a new application can be a boring task that you just want to
be done with as soon as possible. To help you with that, this chapter guides you to the point where
you have the application in front of you and are ready to start work.
Operating Systems
To see the installation and licensing instructions for your operating system, go to:
• Windows,
• Mac OS X and macOS, or
• Linux.
Note: For more information on the differences between the Nuke applications, see Nuke
Products.
USER GUIDE
93
Windows | System Requirements
Windows
These pages guide you through installation and licensing to the point where you have the application
in front of you and are ready to start work. After installation, all applications are run from either
desktop icons, the browser, or from the command line using arguments.
System Requirements
Qualified Operating Systems
• Windows 7 64-bit
• Windows 10 64-bit
Note: Other operating systems may work, but have not been fully tested.
Minimum Hardware Requirements

• x86-64 processor, such as Intel Pentium 4 or AMD Athlon, with SSE3 instruction set support (or
newer).
• 5 GB disk space available for caching and temporary files.
• At least 8 GB of RAM.
• Display with at least 1280 x 1024 pixel resolution and 24-bit color.
• Graphics card with at least 512 MB of video memory and driver support for OpenGL 2.0 (minimum
requirement).
• To enable optional GPU acceleration of Viewer processing, you need OpenGL 2.0 with support
for floating point textures and GLSL.
• To enable Nuke to calculate certain nodes using the GPU, there are some additional
requirements. For more information, see Requirements for GPU Acceleration.
• R3D Rocket cards require the Rocket Driver 1.4.19.0 and Firmware 1.1.16.5 or later.
Note: To avoid graphical problems, such as text disappearing in the Viewer and Node
Graph, it is important to keep your graphics card drivers up-to-date. Driver updates can be
USER GUIDE
94
Windows | Requirements for GPU Acceleration
obtained from the websites of the graphics card manufacturers (for example,
www.nvidia.com and support.amd.com).
Note: If you’re using R3D Rocket graphics card, note that using it with Foundry applications
will most likely only be considerably faster when you’re reading in at full resolution. If you’re
reading in at half resolution, for instance, disabling the R3D Rocket card may be faster. This
is because the R3D Rocket graphics card is designed to be fast when reading in multiple
frames at the same time. This is not how Nuke works internally, and therefore reads with the
R3D Rocket card disabled may sometimes be faster when working in lower resolutions (< 4K
widths). Note that the R3D Rocket card always produces better results than Nuke when
downsampling. Also, the R3D Rocket card can only be used by one application at a time, so if
you are viewing multiple Nuke comps at once, you may only be able to use the R3D Rocket
card in one.
Requirements for GPU Acceleration

If you want to enable Nuke to calculate certain nodes using the GPU, there are some additional
requirements.
NVIDIA
• An NVIDIA GPU with compute capability 2.0 (Fermi) or above. A list of the compute capabilities of
NVIDIA GPUs is available at www.nvidia.co.uk/object/cuda_gpus_uk.html
Note: The compute capability is a property of the GPU hardware and can't be altered by a
software update.
• Graphics drivers capable of running CUDA 8.0 & 6.5 or above. These are bundled with the regular
drivers for your NVIDIA GPU. Driver version r361 or above is required.
Go to http://www.nvidia.com/Download/Find.aspx?lang=en-us for more information.
Note: If your computer enters sleep mode, the CUDA drivers cannot recover and you must
restart Nuke to use GPU acceleration.
USER GUIDE
95
Windows | Requirements for GPU Acceleration
Tip: We recommend using the latest graphics drivers, where possible.
AMD
An AMD GPU and recommended driver from the following list:
Card Driver
AMD FirePro W8100 17.Q2.1
AMD Radeon R9 Fury X 17.4.3 - 17.6.2
AMD Radeon RX 480 17.Q2.1
AMD Radeon Pro WX 7100 17.4.3 - 17.6.2
Note: Other AMD GPUs may work, but have not been fully tested.
Multi-GPU Processing
Nuke's GPU support includes an Enable multi-GPU support option. When enabled in the
preferences, GPU processing is shared between the available GPUs for extra processing speed.
Note: Multi-GPU processing is only available for identical GPUs in the same machine. For
example, two NVIDIA GeForce GTX 1080s or two AMD FirePro W9100s.
USER GUIDE
96
Installing on Windows | Installing with the User Interface (UI)
Installing on Windows
The installation bundle installs the entire Nuke family, including Hiero and HieroPlayer, and icons for
the various components appear in your installation folder.
Note: Some modern anti-virus software may wrongly assume that certain files in the Nuke
installer are suspicious. Examples of these files include libnuke-12.0.0.so and geolib-
runtime-prof.so. If you have trouble installing Nuke on your machine, try disabling your
anti-virus software before installation. Don't forget to restart your anti-virus software after
installation.
Installing with the User Interface (UI)

1. Download the correct .zip file from our website at www.foundry.com.
2. Unzip the installer from the .zip file and double-click on the .exe file to install Nuke.
3. Follow the on-screen instructions. By default, Nuke is installed to drive letter:\Program
Files\Nuke11.3v5.
4. The Nuke plug-ins page on our website opens, giving you easy access to a large selection of plug-
ins to use with Nuke.
5. Proceed to Licensing on Windows.
Note: On Windows, if you install Nuke to a network drive to run from multiple computers,
please ensure that the correct Microsoft run time libraries are installed on each machine
that runs Nuke. To do this, run vcredist_x64.exe on each machine. The appropriate one of
these files can be found in the VCRedist subdirectory in the folder where Nuke is installed.
Running Nuke without installing the libraries on your machine may work correctly,
particularly as many systems (such as Windows Vista by default) already have them. If the
libraries are not present, Nuke can still run correctly, but some plug-ins may fail to load with
error messages such as “This application has failed to start because the application
configuration is incorrect. Reinstalling the application may fix this problem.”
Please note that these libraries are set up automatically on the machine that runs the Nuke
installer, so users installing on their local machine do not need to worry about this issue.
USER GUIDE
97
Installing on Windows | Installing from the Command Line
Installing from the Command Line

To install the Nuke bundle from the command line, do the following:
1. Download the correct .exe installation file from our website at www.foundry.com.
2. To open a command prompt window, select Start > All Programs > Accessories > Command
Prompt.
3. Use the cd (change directory) command to move to the directory where you saved the installation
file. For example, if you saved the installation file in C:\Temp, use the following command and
press Return:
cd \Temp
4. To install Nuke, do one of the following:
• To install Nuke to the current directory and display the installation dialog, type the name of the
install file without the file extension and press Return:
Nuke11.3v5-win-x86-release-64
• To install Nuke to a specified directory and display the installation dialog, use the /dir install
option:
Nuke11.3v5-win-x86-release-64 /dir=“C:\Nuke”
• To install Nuke silently so that the installer does not prompt you for anything but displays a
progress bar, enter /silent after the installation command:
Nuke11.3v5-win-x86-release-64 /silent
• To install Nuke silently so that nothing is displayed, enter /verysilent after the installation
command:
Nuke11.3v5-win-x86-release-64 /verysilent
Tip: You can add /MERGETASKS="!desktopicon" to the /verysilent flag to avoid creating
Desktop icons for all the flavors of Nuke.
• You can also use a combination of install options:

Nuke11.3v5-win-x86-release-64 /silent /dir=“C:\Nuke”
Note: By using the silent or verysilent install options, you agree to the terms of the Nuke
End User Licensing Agreement. To see this agreement, please refer to Foundry's End User
License Agreement or run the installer in standard, non-silent mode.
USER GUIDE
98
Launching on Windows | Nuke Analytics
Launching on Windows
To launch the application on Windows, do one of the following:
• Double-click the required icon on the Desktop.
• Navigate to Start > All Programs > The Foundry > Nuke11.3v5 and select the required application.
• Using a command prompt, navigate to the Nuke application directory (by default, \Program
Files\Nuke11.3v5) and enter:
• Nuke11.3.exe --studio to launch Nuke Studio.
• Nuke11.3.exe --nukex to launch NukeX.
• Nuke11.3.exe to launch Nuke.
• Nuke11.3.exe --hiero to launch Hiero.
• Nuke11.3.exe --player to launch HieroPlayer.
• Nuke11.3.exe --nukeassist to launch Nuke Assist.
Note: Nuke Assist licenses are only available as part of the NukeX or Nuke Studio package,
and cannot be purchased separately. For more information, see About Nuke Assist.
Tip: For more information on other command prompt options, such as safe mode, see
Command Line Operations.
If you already have a valid license, the graphical interface appears, and a command line window
opens. If you don't have a license or haven't installed one yet, proceed to Licensing on Windows.
Nuke Analytics
In an effort to further improve quality and reliability, we ask you to allow us to collect usage statistics
from the machines on which you license Nuke, NukeX, Nuke Studio, Hiero, and HieroPlayer. This
usage information also assists our Support team to resolve issues more quickly.
Note: The port number used to communicate with Foundry is 443, the same one used for
uploading crash reports.
USER GUIDE
99
Launching on Windows | Nuke Non-commercial
The first time you start an application, and on every major release, a dialog displays asking for
permission for us to collect this information. You can enable or disable collection at any time in the
Preferences under Behaviors > Startup.
Note: This information is only collected for interactive sessions. Running applications in
terminal mode or under render licenses does not upload data to Foundry.
The following list shows the information we'll collect, if you give us permission to do so:
• Unique session ID • Anonymous user key • Application version string
• Application name • Session start time (GMT) • Session duration (in seconds)
• If the session exited cleanly • Peak memory usage • Model
• Operating system • System OS version • MAC address
• CPU Name • CPU Cores • GPU model name
• Amount of GPU RAM • OpenGL driver version • GPU driver version
• Amount of RAM • Memory speed
Nuke Non-commercial
If you want to try out or learn Nuke, you can run Nuke Non-commercial. This version allows you to
explore most of Nuke’s features, but prevents the commercial use of the application. For more
information, see About Nuke Non-commercial.
To launch the application on Windows, do one of the following:

• Double-click the NukeNC, NukeXNC, or Nuke StudioNC icon on the Desktop.
• Navigate to Start > All Programs > The Foundry > Nuke11.3v5 and select Nuke11.3v5NC,
NukeX11.3v5NC, or NukeStudio11.3v5NC.
• Using a command prompt, navigate to the Nuke application directory and enter:
• Nuke11.3.exe --nc --studio to launch Nuke Studio.
• Nuke11.3.exe --nc --nukex to launch NukeX.
• Nuke11.3.exe --nc to launch Nuke.
If you have already activated Nuke Non-commercial on the current device, the graphical interface
appears, and a command line window opens. If you haven't activated the device yet, proceed to
Licensing Nuke Non-commercial on Windows.
USER GUIDE
100
Licensing on Windows | Obtaining Licenses
Licensing on Windows
The following licensing methods are available:
• Activation Keys - activation keys allow you to activate and generate your actual product license key,
at a later point after purchase, on the machine for which you require the license.
They are provided for both node locked and floating licenses, and generate the appropriate license
type once installed using the product's Licensing dialog or online using the Activate a Product page:
https://www.foundry.com/licensing/activate-product
• Node Locked Licenses - these can be used to license an application on a single machine. They do
not work on different machines and if you need them to, you’ll have to transfer your license.
Node locked licenses, sometimes called uncounted licenses, do not require additional licensing
software to be installed.
• Floating Licenses - also known as counted licenses, enable applications to work on any networked
client machine. The floating license is put on the server and is locked to a unique number on that
server.
Floating licenses on a server requires additional software to be installed on the server to manage
the licenses and give them out to the client stations that want them. This software is called the
Foundry Licensing Tools (FLT) and can be downloaded at no extra cost from our website.
• Subscription Licenses - subscription licensing differs from traditional node locked or floating
licenses in that a single license, or entitlement, is valid on any authorized device up to the
entitlement's maximum number of activations.
For more information on Subscription Licensing, see Licensing Nuke Non-commercial on Windows.
The following instructions run through the basic options for the first two licensing methods, but you
can find a more detailed description in the Foundry Licensing Tools (FLT) User Guide available on our
website: https://www.foundry.com/licensing
Obtaining Licenses
To obtain a license, you'll need your machine's System ID (sometimes called Host ID or rlmhostid).
Just so you know what a System ID number looks like, here’s an example: 000ea641d7a1.
USER GUIDE
101
Licensing on Windows | Installing Licenses
Note: Bear in mind that, for floating licenses, you'll need the System ID of the license
server, not the machines on which you intend to run the application.
There are a number of ways you can find out your machine's System ID:
• Launch the application without a license, click Status, and then scroll down the error report until
you see your System ID.
• Download the Foundry License Utility (FLU) from https://www.foundry.com/licensing and run it.
Your System ID is displayed.
• Download the Foundry Licensing Tools (FLT) free of charge from https://www.foundry.com/licensing
and then run C:\Program Files\TheFoundry\LicensingTools7.0\ Foundry License Utility.exe
When you know your System ID, you can request a license for Foundry products:
• from Foundry's Sales Department at sales@foundry.com
• from the product pages on our website, such as https://www.foundry.com/products/nuke
• by launching the application without a license and selecting:
• Buy Nuke or Buy Hiero - opens a web browser directly to our website to purchase a license.
• Try Nuke or Try Hiero - displays the 15-day trial license download screen. Enter your Foundry
account details or create a new account and follow the on-screen instructions to receive a trial
license.
Note: By default, if you have installed a temporary license, the application displays a dialog
at start-up alerting you to the number of days remaining. If you want to disable this
behavior, you can set the FN_DISABLE_LICENSE_DIALOG environment variable to 1 to
suppress the warning message about imminent license expiration. See Environment
Variables for more information.
Installing Licenses
When you start the application before installing a license, a Licensing dialog displays an error
informing you that no license was available. The installation process is dependent on what type of
license you requested:
• License file - if you requested a license file, typically foundry.lic, this option allows you to browse
to the file location and install it automatically. See To Install a License from Disk for more
information.
USER GUIDE
102
• Activation Key or license text - if you requested an Activation Key or license by email, this option
allows you to paste the key or license text into the Licensing dialog, which then installs the license in
the correct directory. See To Install an Activation Key or License Text for more information.
• A floating license - if you requested a floating license to supply licenses to multiple client
machines, this option allows you to enter the server address that supplies the client licenses.
Note: You must install a floating license and additional software on the license server to use
this option.
See To Install a Floating License for more information.
To Install a License from Disk

1. Save the license file to a known location on disk.
2. Launch the application.
The Licensing dialog displays.
3. Click Install License to display the available license installation options.
4. Click Install from Disk.
5. Browse to the location of the license file.
6. Click Open to install the license automatically in the correct directory.
To Install an Activation Key or License Text

3. Click Activation Key / License Text and then either:
• Enter the Activation Key string in place of Insert Activation Key Here. A license key typically looks
something like this:
nuke-0101-77d3-99bd-a977-93e9-8035
OR
• Copy the license text and paste it over the Copy/Paste license text here string. License text
typically looks something like this:
LICENSE foundry nuke_i 2015.0929 29-sep-2015 uncounted
hostid=000a957bfde5 share=h min_timeout=30 start=29-sep-2015 issued=29-
sep-2015 disable=VM _ck=da32d7372f sig="60P0450MJRP97E3DP
B42C99Y5UAPRMEMGNQ39PG22H4WGH3WFK2KPTXFWJTYR0GYASJBXC0PU8"
4. Click Install.
USER GUIDE
103
The license is automatically installed on your machine in the correct directory.
Note: Activation Keys require an internet connection. If you access the internet through a
proxy server and cannot connect to the activation server, you may get an error dialog
prompting you to either:
Click Use Proxy to enter the proxy server name, port number, username, and password.
This enables the application to connect to the activation server and obtain a license. The
license is then installed automatically, or
Click on the web link in the dialog and use the System ID (also known as hostid) provided to
manually activate and install a license.
To Install a Floating License

If you requested a floating license from Foundry, you will receive your license key (foundry.lic) in an
email or internet download. You should also receive the Foundry License Utility (FLU) application to
help you install the license key on the license server machine. The server manages licenses for the
client machines on your network.
Note: The FLU is also available to download from https://www.foundry.com/licensing
1. Make sure you have saved both the license key (foundry.lic) and the FLU application in the same
directory.
2. Run the FLU application.
The license key automatically appears in the FLU window if the FLU and foundry.lic are in the
same directory.
Tip: If they are not in the same directory, you can either copy and paste the contents of the
license key or drag-and-drop the file into the FLU window.
3. Click Install.
This checks the license file and, provided that the license is valid, installs it into the correct
directory.
4. In order for the floating license to work, you will need to install the Foundry Licensing Tools (FLT)
on the license server machine.
USER GUIDE
104
Licensing on Windows | Further Reading
For more information on how to install floating licenses, refer to the FLT User Guide, which you can
download from our website https://www.foundry.com/licensing
5. Once your license server is up and running, launch the application on the client machine.
6. Click Install License to display the available install methods.
7. Click Use Server and enter the server address in the field provided. The format for the server
name is: <port>@<servername>, for example, 30001@red.
Note: You must perform steps 5 through 7 on each client machine that requires a license
from the server.
Further Reading
There is a lot to learn about licenses, much of which is beyond the scope of this manual. For more
information on licensing, displaying the System ID number, setting up a floating license server,
adding new license keys and managing license usage across a network, you should read the Foundry
Licensing Tools (FLT) User Guide, which can be downloaded from our website,
https://www.foundry.com/licensing
USER GUIDE
105
Licensing Nuke Non-commercial on Windows | Further Reading
Licensing Nuke Non-commercial

on Windows
Subscription licensing differs from traditional node locked or floating licenses in that a single license,
or entitlement, is valid on any authorized device up to the entitlement's maximum number of
activations.
• An Entitlement represents the right to run a Foundry product for a set amount of time on a set
number of devices.
• An Authorized Device is a recognized device, such as a desktop computer, on which entitlements
can be activated.
For example, if an Entitlement for Nuke has five activations, you can use Nuke on five separate
Authorized Devices simultaneously. If you want to activate another device, you have to deactivate an
existing one, but you can activate and deactivate devices as often as you like.
To get started with Nuke Non-commercial, follow these steps:

1. Create a Foundry account using a valid email address on our website,
https://www.foundry.com/user/register
2. Launch Nuke in non-commercial mode as described under Launching on Windows.
A Licensing dialog displays, similar to regular licensing. Nuke Non-commercial is free, but your
entitlement only contains two activations.
3. Click Authorise Device.
4. Enter your account email address and password and then click Authorise Device.
5. A subscription license is created in your home directory:
C:\Users\<username>\FoundryLicensing\<SystemID>
Note: Replace <username> and <SystemID> with the current user and the MAC address of
the device, respectively.
The license looks something like this: c58edf7e-17ab-435b-8d8a-b3a9b347ab11.lic

6. Once the license is installed, click Launch to start using Nuke.
Note: On Windows, there is a known issue with user names containing non-ASCII characters
causing licensing to fail. If a licensing error similar to the following displays:
Unable to create subscription license directory: C:\Users\Zoë
USER GUIDE
106
Licensing Nuke Non-commercial on Windows | Further Reading
Hernández\FoundryLicensing\
Try changing the license directory to an alternate location using the FN_SUBSCRIPTION_
LICENSE_DIR environment variable. See Environment Variables for more information.
7. If you need to deactivate an entitlement or deauthorize a device, navigate to Help > License and,
click:
• Deactivate Nuke to reclaim one of your entitlements,
• Deauthorize Device to reclaim your existing Foundry entitlements on this device and stop
additional ones running, or
• Deauthorize All Devices to reclaim your existing Foundry entitlements on all devices
associated with your account, and stop additional ones running.
USER GUIDE
107
Uninstalling Apps on Windows | Further Reading
Uninstalling Apps on Windows

To uninstall Nuke or Hiero on Windows, there are a few things you need to do:
1. Navigate to Start > All Programs > The Foundry > Nuke11.3v5 and select Uninstall.
The Nuke Uninstall dialog displays.
2. Click Yes to uninstall.
3. Delete, rename, or move your .nuke or .hiero directories, if they exist.
The .nuke and .hiero directories are usually found under the directory pointed to by the HOME
environment variable. If this variable is not set (which is common), the directories are under the
directory specified by the USERPROFILE environment variable, which is generally one of the
following:
drive letter:\Documents and Settings\login name\
drive letter:\Users\login name\
To find out if the HOME and USERPROFILE environment variables are set and where they are
pointing at, enter %HOME% or %USERPROFILE% into the address bar in Windows Explorer. If the
environment variable is set, the folder it’s pointing at is opened. If it’s not set, you get an error.
4. Delete, rename, or move your cached files, which reside in the following directory by default:
~\AppData\Local\Temp\nuke\
Where ~ is equal to %HOME% or %USERPROFILE% as detailed above.
Note: If you specified an alternate directory using the NUKE_TEMP_DIR environment

variable, purge those files as well as the default location. See Nuke Environment Variables
for more information.
USER GUIDE
108
Mac OS X and macOS | System Requirements
Mac OS X and macOS

System Requirements
• macOS 10.13 (High Sierra)

• macOS 10.14 (Mojave)

• x86-64 processor, such as Intel Core 2 Duo or later.
• 5 GB of disk space available for caching and temporary files.
requirement).
USER GUIDE
109
Mac OS X and macOS | Requirements for GPU Acceleration
Note: If you’re using R3D Rocket graphics card, note that using it in Nuke will most likely
only be considerably faster when you’re reading in at full resolution. If you’re reading in at
half resolution, for instance, using Nuke without the R3D Rocket card enabled may be faster.
This is because the R3D Rocket graphics card is designed to be fast when reading in multiple
you are viewing multiple Nuke scripts at once, you may only be able to use the R3D Rocket
card in one.

requirements.
NVIDIA
An NVIDIA GPU with compute capability 2.0 (Fermi) or above. A list of the compute capabilities of
NVIDIA GPUs is available at:
www.nvidia.co.uk/object/cuda_gpus_uk.html.
software update.
Graphics drivers capable of running CUDA 8.0 & 6.5 or above. On Mac, the CUDA driver is separate
from the NVIDIA graphics driver and will need to be installed, if you don't have it already. The
minimum requirement is driver version r361 which can be downloaded from
www.nvidia.com/drivers.
USER GUIDE
110
Mac OS X and macOS | Requirements for GPU Acceleration
AMD
On Mac, AMD GPUs are supported on any Mac Pro running Mac OS X Mavericks (10.9.3 ), mid 2015
MacBook Pros onward, and late 2017 iMac Pros. Bit-wise equality between GPU and CPU holds in
most cases, but for some operations there are limitations to the accuracy possible with this
configuration.
Warning: Although AMD GPUs are enabled on Mac Pros manufactured prior to the late
2013 model, they are not officially supported and used at your own risk.
Note: To ensure you get the best performance from OpenCL GPUs, we recommend
updating Mavericks to 10.9.5, or above for full functionality. However:
If you're running an earlier version of Mac OS X than 10.9.5 and processing images greater
than 4 mega pixels resolution, VectorGenerator, Kronos, and MotionBlur do not support
GPU acceleration.
If you're running an earlier version of Mac OS X than 10.9.4, Kronos and MotionBlur do not
support GPU acceleration.
USER GUIDE
111
Installing on Mac | Installing with the User Interface (UI)
Installing on Mac
installation.
Installing with the User Interface (UI)

1. Download the correct .dmg installation file from our website at www.foundry.com.
2. Double-click on the Nuke11.3v5-mac-x86-release-64.dmg archive to extract the .pkg installer:
A .pkg file is created.
3. Double-click on the .pkg file.
4. Follow the on-screen instructions to install Nuke. By default, Nuke is installed to
/Applications/Nuke11.3v5.
6. Proceed to Launching on Mac.
Installing from the Terminal

1. Download the correct .dmg installation file from our website at www.foundry.com.
2. Launch a Terminal window.
3. To mount the .dmg installation file, use the hdiutil attach command with the directory where you
saved the installation file. For example, if you saved the installation file in Builds/Nuke, use the
following command:
hdiutil attach /Builds/Nuke/Nuke11.3v5-mac-x86-release-64.dmg
4. Enter the following command:
pushd /Volumes/Nuke11.3v5/
This stores the directory path in memory, so it can be returned to later.
5. To install Nuke, use the following command:
USER GUIDE
112
Installing on Mac | Installing from the Terminal
sudo installer -pkg Nuke11.3v5-mac-x86-release-64.pkg -target "/"

You are prompted for a password.
6. Enter the following command:
popd
This changes to the directory stored by the pushd command.
7. Finally, use the following command to eject the mounted disk image:
hdiutil detach /Volumes/Nuke11.3v5
Note: By running a silent install of Nuke, you agree to the terms of the End User Licensing
Agreement. To see this agreement, please refer to Foundry's End User License Agreement or
run the installer in standard, non-silent mode.
USER GUIDE
113
Launching on Mac | Nuke Analytics
Launching on Mac
To launch the application on Mac, do one of the following:
• Open the Nuke application directory (/Applications/Nuke11.3v5/), and double-click the required
icon.
• Using the terminal, navigate to the Nuke application directory
(/Applications/Nuke11.3v5/Nuke11.3v5.app/Contents/MacOS/) and enter:
• ./Nuke11.3v5 --studio to launch Nuke Studio.
• ./Nuke11.3v5 --nukex to launch NukeX.
• ./Nuke11.3v5 to launch Nuke.
• ./Nuke11.3v5 --hiero to launch Hiero.
• ./Nuke11.3v5 --player to launch HieroPlayer.
• ./Nuke11.3v5 --nukeassist to launch Nuke Assist.
Tip: For more information on other terminal options, such as safe mode, see Command
Line Operations.
If you already have a valid license, the graphical interface appears, and a command line window
opens. If you don't have a license or haven't installed one yet, proceed to Licensing on Mac.
Nuke Analytics
from the machines on which you license Nuke, NukeX, Nuke Studio, Hiero, and HieroPlayer. This
usage information also assists our Support team to resolve issues more quickly.
USER GUIDE
114
Launching on Mac | Nuke Non-commercial
Nuke Non-commercial
To launch the application on Mac, do one of the following:

• Double-click the NukeNC, NukeXNC, or Nuke StudioNC dock icon.
• Open the Nuke application directory and double-click the NukeNC, NukeXNC, or Nuke StudioNC
icon.
• Using the terminal, navigate to the Nuke application directory and enter:
• ./Nuke11.3v5 --nc --studio to launch Nuke Studio.
• ./Nuke11.3v5 --nc --nukex to launch NukeX.
• ./Nuke11.3v5 --nc to launch Nuke.
Licensing Nuke Non-commercial on Mac.
USER GUIDE
115
Licensing on Mac | Obtaining Licenses
Licensing on Mac
server.
For more information on Subscription Licensing, see Licensing Nuke Non-commercial on Mac.
Obtaining Licenses
USER GUIDE
116
Licensing on Mac | Installing Licenses
• Download the Foundry License Utility (FLU) from https://www.foundry.com/licensing and run it.
Your System ID is displayed.
and then run /Applications/TheFoundry/LicensingTools7.0/Foundry Licence Utility.app
license.
Note: By default, if you have installed a temporary license, Nuke displays a dialog at start-
up alerting you to the number of days remaining. If you want to disable this behavior, you
can set the FN_DISABLE_LICENSE_DIALOG environment variable to 1 to suppress the
warning message about imminent license expiration. See Environment Variables for more
information.
Installing Licenses
information.
USER GUIDE
117
this option.


nuke-0101-77d3-99bd-a977-93e9-8035
OR
4. Click Install.
USER GUIDE
118

directory.
same directory.
3. Click Install.
directory.
USER GUIDE
119
Licensing on Mac | Further Reading
5. Once your license server is up and running, launch the application on the client machine.
from the server.
Further Reading
information on licensing Nuke, displaying the System ID number, setting up a floating license server,
USER GUIDE
120
Licensing Nuke Non-commercial on Mac | Further Reading

on Mac
activations.
number of devices.
can be activated.

2. Launch Nuke in non-commercial mode as described under Launching on Mac.
/Users/<username>/FoundryLicensing/<SystemID>

click:
USER GUIDE
121
Licensing Nuke Non-commercial on Mac | Further Reading
USER GUIDE
122
Uninstalling Apps on Mac | Further Reading
Uninstalling Apps on Mac

To uninstall Nuke or Hiero on Mac, there are a few things you need to do:
1. Navigate to Applications and delete the Nuke 11.3v5 directory.
The directories are found in your home directory, by default:
/Users/<login name>/.nuke
Note: The directories may be a hidden directory on your machine. To allow your Mac to
display hidden files and directories, type the following command in the Terminal
application, press Return, and then relaunch the Finder application:
defaults write com.apple.finder AppleShowAllFiles YES
/var/tmp/nuke

USER GUIDE
123
Linux | System Requirements
Linux
System Requirements
• CentOS 6 (64-bit)
• CentOS 7 (64-bit)

• x86-64 processor, such as Intel Pentium 4 or AMD Athlon, with SSE3 instruction set support (or
newer).
• 5 GB disk space available for caching and temporary files.
requirement).
USER GUIDE
124
Linux | Requirements for GPU Acceleration
Note: If you’re using R3D Rocket graphics card, note that using it in Nuke will most likely
only be considerably faster when you’re reading in at full resolution. If you’re reading in at
half resolution, for instance, using Nuke without the R3D Rocket card enabled may be faster.
This is because the R3D Rocket graphics card is designed to be fast when reading in multiple
you are viewing multiple Nuke scripts at once, you may only be able to use the R3D Rocket
card in one.

requirements.
NVIDIA
• An NVIDIA GPU with compute capability 2.0 (Fermi) or above. A list of the compute capabilities of
NVIDIA GPUs is available at:
www.nvidia.co.uk/object/cuda_gpus_uk.html
software update.
• Graphics drivers capable of running CUDA 8.0 & 6.5 or above. These are bundled with the regular
drivers for your NVIDIA GPU. Driver version r361 or above is required.
Go to http://www.nvidia.com/Download/Find.aspx?lang=en-us for more information.
USER GUIDE
125
Linux | Requirements for GPU Acceleration
AMD
An AMD GPU and recommended driver from the following list:
Card Driver
AMD Radeon R9 Fury X 17.10
AMD Radeon RX 480 17.Q2.1
AMD Radeon Pro WX 7100 17.10
Note: Other AMD GPUs may work, but have not been fully tested.
USER GUIDE
126
Installing on Linux | Installing from the Terminal
Installing on Linux
installation.
Installing from the Terminal

1. Download the correct .tgz installation file from our website at www.foundry.com.
2. Extract the installer from the .tgz archive with the following terminal command:
tar xvzf Nuke11.3v5-linux-x86-release-64.tgz
This gives you an installer file.
3. Run the installer.
sudo ./Nuke11.3v5-linux-x86-release-64-installer
4. Follow the on-screen instructions. By default, Nuke is installed to /usr/local/Nuke11.3v5
6. If you didn’t add a license key during the installation, do that now. Proceed to Licensing on Linux.
Tip: To install Nuke silently, you can simply unzip the installer file. This creates the properly
formed Nuke directory tree in the current directory.
Note: By installing Nuke silently, you agree to the terms of the End User Licensing
Agreement. To see this agreement, please refer to Foundry's End User License Agreement or
run the installer in standard, non-silent mode.
USER GUIDE
127
Installing on Linux | Installing Remotely from the Terminal
Installing Remotely from the Terminal

If you need to install Nuke on render machines using the terminal, do the following:
1. Download the correct .tgz installation file from our website at www.foundry.com.
2. Extract the installer from the .tgz archive with the following terminal command:
tar xvzf Nuke11.3v5-linux-x86-release-64.tgz
This gives you an installer file.
3. Use the following terminal command to log in to your render machine as root:
ssh root@render_machine
Replace render_machine with the name of your render node.
4. Make a directory to install Nuke to:
mkdir /usr/local/Nuke11.3v5
5. Copy the installer file from the machine that you downloaded it on to your render machine with a
command like:
scp root@download_machine:/tmp/Nuke11.3v5-linux-x86-release-64-installer
root@render_machine:/usr/local/Nuke11.3v5/
Replace download_machine with the name of the machine you downloaded the installer file to, and
render_machine with the name of your render node.
6. Unzip the installer file to unpack its contents into your Nuke directory:
cd /usr/local/Nuke11.3v5
unzip Nuke11.3v5-linux-x86-release-64-installer
7. Repeat steps 3-6 for each render machine.
USER GUIDE
128
Launching on Linux | Nuke Analytics
Launching on Linux
To launch the application on Linux, do one of the following:
• Open the Nuke application directory (by default, /usr/local/Nuke11.3v5) and double-click the
required icon.
• Using a terminal, navigate to the Nuke application directory and enter:
• ./Nuke11.3 --studio to launch Nuke Studio.
• ./Nuke11.3 --nukex to launch NukeX.
• ./Nuke11.3 to launch Nuke.
• ./Nuke11.3 --hiero to launch Hiero.
• ./Nuke11.3 --player to launch HieroPlayer.
• ./Nuke11.3 --nukeassist to launch Nuke Assist.
Tip: For more information on other terminal options, such as safe mode, see Command
Line Operations.
If you already have a valid license, the graphical interface appears. If you don't have a license or
haven't installed one yet, proceed to Licensing on Linux.
Nuke Analytics
from the machines on which you license license Nuke, NukeX, Nuke Studio, Hiero, and HieroPlayer.
This usage information also assists our Support team to resolve issues more quickly.
USER GUIDE
129
Launching on Linux | Nuke Non-commercial
Nuke Non-commercial
To launch the application on Linux, do one of the following:

• Double-click the NukeNC, NukeXNC, or Nuke StudioNC icon on the Desktop.
• Open the Nuke application directory and double-click the NukeNC, NukeXNC, or Nuke StudioNC
icon.
• Using a terminal, navigate to the Nuke application directory and enter:
• ./Nuke11.3 --nc --studio to launch Nuke Studio.
• ./Nuke11.3 --nc --nukex to launch NukeX.
• ./Nuke11.3 --nc to launch Nuke.
Licensing Nuke Non-commercial on Linux.
USER GUIDE
130
Licensing on Linux | Obtaining Licenses
Licensing on Linux
server.
For more information on Subscription Licensing, see Licensing Nuke Non-commercial on Linux.
Obtaining Licenses
USER GUIDE
131
Licensing on Linux | Installing Licenses
and then run the following command in a terminal shell:
/usr/local/foundry/LicensingTools7.0/bin/systemid
license.
Note: By default, if you have installed a temporary license, the application displays a dialog
at start-up alerting you to the number of days remaining. If you want to disable this
behavior, you can set the FN_DISABLE_LICENSE_DIALOG environment variable to 1 to
suppress the warning message about imminent license expiration. See Environment
Variables for more information.
Installing Licenses
information.
USER GUIDE
132
this option.


nuke-0101-77d3-99bd-a977-93e9-8035
OR
4. Click Install.
USER GUIDE
133

directory.
same directory.
3. Click Install.
directory.
USER GUIDE
134
Licensing on Linux | Further Reading
5. Once your license server is up and running, launch Nuke on the client machine.
from the server.
Further Reading
information on licensing Nuke, displaying the System ID number, setting up a floating license server,
USER GUIDE
135
Licensing Nuke Non-commercial on Linux | Further Reading

on Linux
activations.
number of devices.
can be activated.

2. Launch Nuke in non-commercial mode as described under Launching on Linux.
/home/<username>/FoundryLicensing/<SystemID>

click:
USER GUIDE
136
Licensing Nuke Non-commercial on Linux | Further Reading
USER GUIDE
137
Uninstalling Apps on Linux | Further Reading
Uninstalling Apps on Linux

To uninstall Nuke or Hiero on Linux, there are a few things you need to do:
1. Navigate to /usr/local/ and delete the Nuke 11.3v5 directory.
The directories are found in your home directory, by default:
/home/<login name>/.nuke
/var/tmp/nuke

USER GUIDE
138
Nuke Studio Environments
The Compositing Environment

You can use the Compositing environment to perform node-based compositing with a choice of
different VFX tools, manage color grading, review your script, and render out your script. See Using
the Compositing Environment for more information.
By default, there is a Node Graph panel in the lower-left corner, a Viewer panel in the top-left corner,
and a Properties panel on the right.
Node Graph The Node Graph is where you add nodes and build your node tree.
USER GUIDE
139
Nuke Studio Environments | The Timeline Environment
Properties Panel When you add a node to the Node Graph, its properties appear in the
Properties panel on the right.
Viewer To check the result, you can view the output in a Viewer.
The Timeline Environment

The Nuke StudioTimeline environment allows you to conform, create Nuke Comps, add soft effects,
perform timeline-based editing, export your project, and view and edit metadata and properties. See
Using Nuke Studio's Timeline Environment for more information.
Project Tab You can manage all aspects of your projects and bins in the Project tab.
Menu Bar Use the Menu bar to access Nuke Studio's dropdown menus.
Bin View The bin displays the contents of any selected Project tab.
Viewer You can display and review your media in the Viewer.
Timeline Info The timeline info displays the current timeline’s sequence, media, and
USER GUIDE
140
Nuke Studio Environments | How the Panels Link
metadata information.
Editing Tools There is a comprehensive set of editing tools provided in Nuke Studio. See
Timeline Editing Tools for more information.
Spreadsheet Tab Use the Spreadsheet tab to display the contents of the timeline in
spreadsheet form. Note that the spreadsheet and timeline are linked,
mirroring any selections made.
Timeline The timeline displays the current track, including all shots and any effects that
have been added.
How the Panels Link

In Nuke Studio, some panels are linked together so that the linked panel tabs are automatically
displayed at the front of a panel, when one of linked group is selected. This makes it quick and easy
to switch between working in the Timeline and Compositing environments.
The Timeline linked group includes the spreadsheet, the Timeline environment Viewer, and the
timeline. The Compositing linked group includes the Node Graph, the Compositing environment
Viewer, and the node Toolbar.
Timeline Linked Group Compositing Linked Group
For example, when you are in the Compositing environment and you select the a spreadsheet from
the Properties content menu. The spread sheet appears as a new tab in the Properties pane, and
automatically opens the timeline, and the timeline Viewer.
USER GUIDE
141
Nuke Studio Environments | Panel Focus and Keyboard Shortcuts
Note: If more than one tab in the linked group is in the same pane, the most recently
viewed tab is displayed at the front of the pane.
Shared panels take precedence over unshared panels. Shared panels include Curve Editor, Dope
Sheet, Pixel Analyzer, Scopes, and Progress bars. The Properties pane and Script Editor are also
shared panels, but differ slightly in that they show workspace-specific content. For example, the
Properties panels shows soft effect keyframes and Node Graph keyframes depending on the
currently selected workspace.
Panel Focus and Keyboard Shortcuts

Nuke Studio deals with panel focus in two ways: click focus and mouse-over focus. Click focus defines
the main panel where keyboard events are registered and mouse-over focus allows you to
temporarily override that focus. If the mouse-over focus is centered on a panel that doesn't recognize
a particular keyboard event, the event falls back to the click focus panel.
A good example of click focus versus mouse-over focus is between Viewers and the Node Graph:
• If a Compositing Viewer has click focus and you press B, only the blue color channel is displayed in
the Viewer. If the mouse-over focus resides on the Node Graph and you press B, the click focus is
overridden and a Blur node is added to the Node Graph.
• If a Compositing Viewer has click focus and you press , (comma), the gain is reduced in the Viewer.
If the mouse-over focus resides on the Node Graph and you press , (comma), the gain is still
reduced in the Viewer because the Node Graph has no , (comma) equivalent keyboard shortcut.
Some areas of the interface retain focus no matter where the mouse-over focus resides, such as the
message control in the Text node's Properties panel and the Filter field in the Project panel.
Click focus is shown in the interface with a bright orange highlight in the panel name and mouse-over
focus is shown as a muted orange highlight.
USER GUIDE
142
Nuke Studio Environments | Default Workspaces
Default Workspaces
There are six default workspaces in Nuke/NukeX and Nuke Studio. By default, Nuke/NukeX opens in
the Compositing workspace and Nuke Studio opens in the Finishing workspace. To change the
workspace, you can do either of the following:
• Select Workspace from the top menu bar and then select the required workspace.
OR
• Use the keyboard shortcuts to open the required workspace. Press Shift and the required
workspace keyboard shortcut, depending on mode:
Keyboard Shortcut Nuke/NukeX Nuke Studio
Shift+F1 Compositing Conforming
Shift+F2 LargeNodeGraph Editing
USER GUIDE
143
Nuke Studio Environments | Customizing Workspaces
Keyboard Shortcut Nuke/NukeX Nuke Studio
Shift+F3 LargeViewer Reviewing
Shift+F4 Scripting Timeline
Shift+F5 Animation Finishing
Shift+F6 Floating Compositing
Customizing Workspaces
Customizing Panes
You can resize and split panes to make more room for different elements on the screen. To resize a
pane, drag the divider line of the pane into a new location.
You can split a pane by clicking on the content menu button in the top-left corner of the pane,
and then selecting Split Vertical or Split Horizontal from the menu that opens.
Note: Pressing and holding the spacebar brings up the right-click menu for that pane,
where available.
USER GUIDE
144
Nuke Studio Environments | Customizing Workspaces
Moving the Toolbar

You can move the Toolbar into a new position by adding a new panel for it, hiding the panel name
and controls, and resizing the panel. For more information on how to do this, see Adding Tabs, and
Hiding Tab Names and Controls.
Adding Tabs
When you can’t fit more elements into your display, you can use tabs to save space. You can also use
tabs to move the Toolbar into a new location.
You can add a tab by clicking on the content menu button in the top-left corner of the pane, and
then selecting the type of tab you want to add. For example, you can add Node Toolbar, Node
Graph, New Viewer, or Script Editor. The new tab is added on top of the existing tabs.
To move tabs, click on the name of the tab and drag it to a new position inside the same pane or in
another pane.
You can close tabs again by clicking the X in the top-right corner of the tab you want to close.
Note: Closing a linked tab closes all associated tabs. If you hold Alt while closing a linked
tab, it only closes that tab.
Soloing Tabs
You can choose to solo a tab by either right-clicking on the tab name or clicking the content menu,
and then selecting Solo Tab. This automatically closes any other open tabs in the same pane, except
for the one you have chosen to solo.
Floating Windows
You can turn tabs and panes into floating windows and vice versa. To turn a tab or pane into a
floating window, click the content menu button in the top-left corner in the tab or pane you want
to float, and then select Float Tab or Float Pane. You can also float tabs by either clicking
Ctrl/Cmd+click on the tab name, or right-clicking on the tab name and select Float Tab.
USER GUIDE
145
Nuke Studio Environments | Saving and Loading Workspaces
To change a floating window into a tab or pane, click on the tab or pane name in the floating window
and drag it to where you want it to dock. You can close floating windows by clicking the X button in
the top-right corner of the tab or pane.
Maximizing Windows
To make a window fullscreen, first ensure the window you want to make fullscreen is active, and then
press Alt+S. This could be the main application window or a floating Viewer. Making it fullscreen
removes the window borders. You can also maximize tabs and panels by pressing spacebar.
Hiding Tab Names and Controls

You can hide the names and control buttons of tabs, as you may not need them with all panels, such
as the Toolbar panel. To hide the names and controls on tabs, click the content menu button in
the top-left corner of the tab, and disable Show Tabs.
You can show the names and controls on tabs again by moving the cursor over the top of the pane
area until the top edge of the pane highlights, right-click to open the content menu, and select Show
Tabs.
Saving and Loading Workspaces

After you have customized a workspace and you are happy with it, you can save it by selecting
Workspace > Save Workspace... You are then asked to name it. After saving it, your custom
workspace appears in the Workspace dropdown under the existing default workspaces. Select it
from the dropdown to load it.
Workspaces are saved in your .nuke file under Workspaces > Nuke or NukeStudio, depending on
which Nuke mode you're currently using.
Note: The location of the .nuke file varies by platform. See Loading Gizmos, NDK Plug-ins,
and Python and Tcl Scripts for more detailed information.
Setting the Startup Workspace

When you launch Nuke, it opens in the workspace set in the Preferences. You can change the startup
workspace to any other default workspace or a custom workspace, by doing the following:
USER GUIDE
146
Nuke Studio Environments | Setting Preferences
1. Open the Preferences dialog by pressing Shift+S.

2. In the Behaviors section, select Startup.
3. Use the startup workspace dropdown to select the workspace you want Nuke to load on startup.
Setting Preferences
Open the Preferences dialog by pressing Shift+S. The preferences available depend on which mode
Nuke is launched in.
Changing Preferences
The function of each preference is described under Appendix A: Preferences.
When you make a change to a preference, in most cases, the interface registers the change
immediately (for example, an interface element displays in the new color). Some preference changes,
such as Performance > Hardware > default blink device, require you to restart Nuke Studio, for
the changes to take effect.
USER GUIDE
147
Saving Preferences
Nuke stores your preference settings in a file called preferences11.3.nk, which resides in your .nuke
directory. The location of this is dependent on your operating system.
• Linux: /home/login name/.nuke
• Mac: /Users/login name/.nuke
• Windows: drive letter:\Users\login name\.nuke
Note: On Windows, .nuke resides in the directory pointed to by the HOME environment
variable. If this variable is not set (which is common), the .nuke directory is under the folder
specified by the USERPROFILE environment variable.
Each Nuke user can maintain his or her own unique settings. After making a change in the
Preferences dialog, you can simply click OK to save and close your preferences. If you click Cancel,
any changes that you made are not saved.
To save your preferences

Make the desired changes inside the Preferences dialog, then click OK. Nuke writes the new settings
to preferences11.3.nk file, which you can find in the .nuke directory:
• On Windows: The .nuke directory can be found under the directory pointed to by the HOME
environment variable. If this variable is not set (which is common), the .nuke directory is under the
folder specified by the USERPROFILE environment variable - which is generally of the form drive
letter:\Documents and Settings\login name\ (Windows XP) or drive letter:\Users\login name\ (Windows
Vista).
• On Mac: /Users/login name/.nuke
• On Linux: /users/login name/.nuke
Your new preferences remain in effect for the current and all subsequent sessions.
Resetting Preferences
To reset any changes you made simply click Restore Defaults in the bottom-left of the Preferences
dialog. You can reset the preferences to default by deleting the preferences11.3.nk file. After doing
USER GUIDE
148
this, the next time you launch Nuke, it rebuilds the file with the default preferences.
USER GUIDE
149
Using the Compositing
Environment
These pages are designed to help you learn how to use the Compositing environment including
working with nodes, using the Toolbar, and using the Properties panel.
Toolbar, Menu Bar, and Content Menus

The Toolbar is located on the left-hand side of the Viewer in the Compositing environment. It consists
of a number of menu icons. The different nodes are grouped under these icons based on their
functions. You use the Toolbar to add nodes to the Node Graph.
The menu bar is located on top of the Nuke window (in all workspaces). Its menus, such as the File or
Edit menu, let you perform more general actions related to the project or script, the Viewers, or
editing, rather than certain individual nodes.
In addition to the Toolbar and the menu bar, you should also familiarize yourself with the content
menus. They are the gray checkered boxes in the top-left corner of each pane. If you click on the box,
a menu opens as shown in the image below. You can use the options in the menu to customize the
workspace.
Finally, to work faster, you can right-click on the different panels to display a menu with options
related to that particular panel.
USER GUIDE
150
Using the Compositing Environment |
USER GUIDE
151
Using the Toolbar |
Using the Toolbar

The Compositing environment’s Toolbar includes the following icons:
Icon Functions
Image Image Read and Write nodes, built-in Nuke elements, and Viewer
nodes.
Draw Roto shapes, paint tools, film grain, fills, lens flares, sparkles, and
other vector-based image tools.
Time Retiming image sequences.
Channel Channel management.
Color Applying color correction effects.
Filter Applying convolve filters, such as blur, sharpen, edge detect, and
erode.
Keyer Extracting procedural mattes.
Merge Layering background and foreground elements.
Transform Translating, scaling, tracking, and stabilizing elements.
3D 3D compositing nodes and tools.
Particles Creating, spawning, and editing particles.
Deep Creating, merging, and editing deep images.
USER GUIDE
152
Using the Toolbar |
Icon Functions
Views Nodes for working with views and stereoscopic or multi-view material.
Metadata Viewing, editing, and comparing image metadata.
ToolSets Creating, deleting, and managing tool sets.
Other Additional operators for script and Viewer management.
Any installed plug-ins and custom menus that do not have their own
icon.
Note: Nuke Assist does not support all the toolbars available in Nuke. See Nuke Products
To display a tool tip that explains the icon’s function, move the cursor over the icon.
To make selections from the Toolbar, click on an icon and select an option from the menu.
USER GUIDE
153
Working with Nodes | Adding Nodes
Working with Nodes

Nodes are the basic building blocks of any composite. You can create a new compositing script by
inserting and connecting nodes to form a network of operations. These operations concatenate and
allow you to manipulate your images.
Note: Not all nodes are supported in all versions of products. For a run down of supported
nodes, see Nuke Products.
Adding Nodes
You can add nodes using the Toolbar, the Tab menu, or the right-click menu. When you add a node,
Nuke automatically connects it to the currently selected node.
Using the Toolbar
1. Select the existing node that you want the new node to follow by clicking on it.
2. Click an icon on the Toolbar and select a node from the menu that appears. For example, if you
want to add a Blur node, click the Filter icon and select Blur.
Note: You can press the middle-mouse button on a menu icon to repeat the last item used
from that menu. For example, if you first select a Blur node from the Filter menu, you can
then add another Blur node by simply pressing the middle-mouse button on the Filter icon.
Using the Tab Menu
Tip: See Using the Tab Menu for more detailed information.
2. Press the Tab key and start typing the name of the node you want to create.
This opens a prompt displaying a list of matches.
USER GUIDE
154
Working with Nodes | Selecting Nodes
3. To select the node you want to add from the list, you can either click on it, or scroll to it with the
Up and Down arrow keys and press Return.
Note: To add the last node created using this method, simply press Tab and then Return.
Using the Right-Click Menu
To add a node using the right-click menu, do the following:

1. Right-click on an existing node that you want the new node to follow.
2. From the menu that opens, select the node you want to add.
Note: You can also add nodes using keyboard shortcuts. Most menus in the Toolbar include
a note of the relevant keyboard shortcut next to the item in question.
Adding a New Branch of the Node Tree
To add a node in a new branch of the Node Tree, do the following:

2. Hold down Shift and create the node using the Toolbar, Tab menu, or right-click menu. To add a
node in a new branch with the Tab menu, press the Tab key first, then hold down Shift when
selecting the new node.
The node is added after the selected node in a new branch of the node tree.
Selecting Nodes
Nuke offers a number of options for selecting nodes. Selected nodes are highlighted in a color that is
defined in the Preferences dialog. See Appendix A: Preferences for more information.
USER GUIDE
155
Working with Nodes | Replacing Nodes
Type of Selection How to Make the Selection
Select a single To select a single node, simply click on it.

node
Select multiple To select multiple nodes, you can either press Shift while clicking on each
nodes node you want to select, or click and drag in the workspace to draw a marquee
around the nodes you want to select.
Select all Press Ctrl/Cmd while dragging a node. Nuke selects all nodes that feed data to
upstream nodes the selected node. You can also Ctrl/Cmd+Shift+click to select more nodes
without clearing the current selection.
Select all nodes in Click on a node in the Node Graph and select Edit > Select Connected Nodes
a node tree (or press Ctrl/Cmd+Alt+A). This selects all nodes in the node tree, whether they
are upstream or downstream from the current node. Nodes in any other node
trees are not selected.
Select all nodes in Select Edit > Select all (or press Ctrl/Cmd+A).
a script
Select nodes by 1. Choose Edit > Search, or press the forward slash (/).
name A dialog is displayed.
2. Type an alphanumeric string that is included in the names of the nodes you
wish to select. Click OK.
Tip: When typing the above alphanumeric search string, you can use
asterisks (*) and question marks (?) as wild cards. An asterisk stands
for multiple alphanumeric characters. A question mark represents
just one character.
Invert a selection Select Edit > Invert Selection.
Replacing Nodes
To replace a node in Nuke, simply hold down Ctrl/Cmd and create a new node. See below for more
details.
USER GUIDE
156
Working with Nodes | Renaming Nodes
To Replace One Node with Another
1. In the Node Graph, select the node you want to replace by clicking on it.
2. Hold down Ctrl/Cmd and create a new node using the Toolbar, a right-click menu, or the Tab
menu. To replace a node with the Tab menu, press the Tab key first, then hold down Ctrl/Cmd
when selecting the new node.
The new node replaces the selected node in the Node Graph.
Note: Note that you cannot replace nodes in this manner if you are using a keyboard
shortcut (such as B for the Blur node) to create the new node.
Renaming Nodes
There are a couple of ways to rename a node in Nuke.
To rename a node, you can either:

1. Double-click on the node to open its properties panel.
2. In the title field on top of the Properties panel, you should see the current name of the node.
Delete that name and enter a new name in its place,
OR
1. Click on the node in the Node Graph to select it.
2. Press N.
3. Enter a new name for the node in the rename field that appears on top of the node.
USER GUIDE
157
Working with Nodes | Editing Nodes
Editing Nodes
To copy, paste, and perform other editing functions in the Node Graph, you can use the standard
editing keys (for example, Ctrl/Cmd+C to copy, and Ctrl/Cmd+V to paste). You can copy nodes to files
or memory. Copied nodes inherit the values of their parent, but these values, unlike those in cloned
nodes, are not actively linked. Copied nodes allow you to assign different values to the original and
the copy.
When you paste nodes, Nuke automatically connects them to the node that is selected before the
paste operation. If you don’t want to connect anything, click on a blank area of the workspace to
deselect any selected nodes before pasting.
Type of Edit How to Preform It
Copy nodes to the To copy nodes to the clipboard:

clipboard 1. Select the node or nodes you want to copy.
2. Choose Edit > Copy (or press Ctrl/Cmd+C).
Copy nodes to To copy nodes to files:

files 1. Select the node or nodes you want to copy.
2. Choose File > Export Comp Nodes...
3. Navigate to the directory where you want to store the node(s) as a script.
4. Type a name for the script at the end of the pathway, followed by the
extension .nk.
Cut Nodes To cut nodes:

1. Select the node or nodes you want to cut.
2. Choose Edit > Cut (or press Ctrl/Cmd+X).
Nuke removes the node(s) from the script and writes the node(s) to the
clipboard.
Paste nodes from To paste nodes from the clipboard:

the clipboard 1. Select the node that you want the pasted node(s) to follow.
2. Choose Edit > Paste (or press Ctrl/Cmd+V).
Nuke adds the nodes to the script, connecting them to the selected node.
Load nodes from To load nodes from files:

files 1. Select the node that you want the loaded node to follow.
USER GUIDE
158
Working with Nodes | Cloning Nodes
Type of Edit How to Preform It
2. Choose File > Insert Comp Nodes.

3. Navigate to the directory that stores the script.
4. Select the script, and click Open.
Nuke adds the nodes described by the file to the selected node.
Cloning Nodes
You can clone nodes directly into the Node Graph or copy clone nodes in preparation for pasting
pasting them elsewhere in a script. Cloned nodes inherit the values of their parent, but unlike copied
nodes, they also maintain an active link with their parents’ values. If you alter the values of a node
that has been cloned, the clone automatically inherits these changes.
Clones are helpful for maintaining consistent setups across multiple elements. For example, you
might use clones to apply an identical film grain setup to a series of elements shot on the same stock.
If you need to make changes to the setup, these changes would automatically ripple throughout the
script.
Note: You cannot clone gizmos. This applies to both gizmos created in your facility and the
nodes in the Nuke default Toolbar that are, in fact, gizmos, such as the LightWrap node, the
Grain node, and the IBK nodes. For more information on gizmos, see the Configuring Nuke
chapter.
Cloning a Node
To clone nodes, do the following:

1. Select the node or nodes you want to clone.
2. Click Edit > Clone.
Nuke clones the node(s), while maintaining an active link to the parental node(s). The clone has
the same name as the original node and is indicated with an orange line connecting the clone to
its parent node.
Copying a Node as a Clone
You can copy a node as a clone and paste it in a different location, by doing the following:
1. Select the node or nodes you want to clone.
USER GUIDE
159
Working with Nodes | Disabling and Deleting Nodes
2. Click Edit > Copy as Clones (or press Ctrl+K).
Decloning a Node
To declone nodes, do the following:

1. Select the node or nodes you want to declone.
2. Click Edit > Declone (or press Alt+Shift+K).
Nuke removes the clone status of the selected nodes.
Disabling and Deleting Nodes

Nuke allows you you to disable, re-enabe, and delete nodes in your Node Graph.
Action Instructions
Disable node To disable nodes:

1. Select the node or nodes you want to disable.
2. Select Edit > Node > Disable/Enable (or press D).
Nuke disbales the node’s effect on the data stream.
Re-enable node To re-enable nodes:

1. Select the node or nodes you want to re-enable.
2. Select Edit > Node > Disable/Enable (or press D).
Delete nodes To delete nodes:

1. Select the node or nodes you want to delete.
2. Select Edit > Delete (or press Delete).
Nuke removes the node(s) from the script.
Connecting Nodes
When you add or paste nodes into a script, Nuke automatically generates pipes between the currently
selected node and the new nodes. As you build up a script, you’ll need to move these pipes, or run
new pipes between nodes. In Nuke, you make such modifications by dragging on the back end of the
pipe (the end without the arrowhead).
USER GUIDE
160
Working with Nodes | Connecting Nodes
Disconnecting Nodes
You can disconnect nodes by either dragging the head or tail of the connecting arrow to an empty
area of the workspace, or selecting the lower node in the tree and pressing Ctrl/Cmd+D.
Reconnecting Nodes
You can reconnect a node by dragging the head or tail of the connecting arrow and drop it over the
center of the node that want to connect.
Note: Nuke distinguishes the dual inputs that may run into a Merge node with the labels A
and B. A refers to the foreground element, and B to the background element. Nuke always
copies from the A input to the B. This means that if you later decide to disable the node
associated with an A input, the data stream keeps flowing because, by default, it uses the B
input.
USER GUIDE
161
Working with Nodes | Connecting Nodes
Duplicating a Connecting Arrow
To duplicate a connecting arrow, hold Shift and drag the connecting arrow on top of the node you
want to create a connection to. Nuke duplicates the connecting arrow, leaving the original connection
untouched.
Adding Nodes Between Nodes
To add a node between two connected nodes, drag the node into the space between the already
connected nodes. As you do so, you see the link between these two nodes become active. When that
happens, simply release the node you are dragging and it is automatically placed and connected
between the two nodes.
Note: Dragging multiple nodes into a node tree only connects the first node in the selection
to the existing node tree.
Bending Connecting Arrows

To bend connecting arrows, do the following:
1. Select the node before the connector you want to bend.
2. From the Toolbar, select Other > Dot.
A dot appears after the selected node, causing a bend in the connector.
3. Drag the dot as necessary to reposition the bend.
Tip: You can also add a dot to an existing connection by pressing Ctrl/Cmd and clicking on
the yellow dot that appears on the connecting arrow.
USER GUIDE
162
Working with Nodes | Node Indicators
Node Indicators
There are several indicators that can appear on the nodes in the Node Graph, depending on what you
are doing. The following table describes what each indicator means.
Indicator Where it appears What it means
The wide rectangles indicate the channels the node

processes.
The thin rectangles indicate the channels that are passed

through the node untouched.
The node’s effect is limited by a mask from either the node’s

primary input or output.
The node has been disabled by pressing D or by selecting

Edit > Node > Disable/Enable.
The node has been disabled using an expression.
The node has been cloned. The indicator appears on both

the parent and the child node.
USER GUIDE
163
Working with Nodes | Displaying Node Information
Indicator Where it appears What it means
One or more of the node parameters are animated over

time.
One or more of the node parameters are being driven by an

expression.
You are working with a multi-view project and have split off
one or more views in the node’s controls.
You are working with a multi-view project and have split off
one or more views in the node’s controls, dots also appear
on the node to indicate which views have been split off. For
example, if you are using red for the left view and split off
that view, a red dot appears on the node.
The full effect of the node is not in use, because you have
adjusted the mix slider in the node’s controls.
Displaying Node Information

You can obtain more detailed information from any node by selecting that node and then pressing
the I key. This displays an information window associated with that node, particularly useful when
troubleshooting.
USER GUIDE
164
Working with Nodes | Customizing the Node Display
Customizing the Node Display

You can modify the color, name, and notes that a particular node displays. Doing so can make it
easier for other artists to decipher the intent of your script. For example, you might color all nodes
green that relate to keying.
Modifying Node Display Characteristics
To modify a node's display characteristics, do the following:

1. Double-click on the node in the Node Graph to display its properties in the Properties pane.
2. Click the Node tab at the top of the dialog in the Properties pane. Its attributes are displayed:
USER GUIDE
165
Working with Nodes | Customizing the Node Display
In the node's properties, you can also do the following:

• Select the hide input checkbox to conceal the node’s incoming pipe. This can enhance the
readability of large scripts.
• Select the postage stamp checkbox to display a thumbnail render of the node’s output on its
surface. You can also press Alt+P on the Properties pane or on the node, to toggle postage
stamps.
If you want the postage stamp to display a fixed frame (rather than update to match the current
frame), enter the frame to display in the static frame field. For this to work, you also need to
press Shift+S to open the Preferences dialog, go to Panels > Node Graph, and ensure the
postage stamp mode dropdown is set to Static frame. Note that if the frame number you use
is outside the frame range for the node, it is clamped to the first or last frame in the range.
Tip: You can also have Nuke automatically color code nodes for you based on their function.
See Appendix A: Preferences.
USER GUIDE
166
Working with Nodes | Creating Node Tool Sets
Creating Node Tool Sets

If you find yourself creating the same set of nodes repeatedly, you can create a custom tool set in the
Nuke toolbar. This allows you to quickly and easily create the set of nodes from the ToolSets menu.
Tool sets can be shared between artists if they are using a centralized .nuke folder. This needs to be
accessed through a NUKE_PATH environment variable that you can set up (see Configuring Nuke for
more information).
To create a tool set do the following:

1. In the Node Graph, select any number of nodes (or just one node). The nodes don’t have to be
connected to each other.
2. Click (ToolSets) on the toolbar and select Create. You can also right-click on the Node
Graph and select ToolSets > Create. The CreateToolSet dialog appears. You can also access your
tool sets by pressing Tab and searching for them in the search field.
3. In the ToolSets menu dropdown, select the menu where you’d like to place your new tool set.
Then give it a name in the Menu item field and click Create.
By default, your new tool set goes under the ToolSets menu, but if you’d like to create a subfolder
for the tool set, you can do that by specifying the folder name before the tool set name, separated
by a forward slash, in the Menu item field. For example, entering Roto/BasicRoto would create a
subfolder called Roto in the ToolSets menu, and place a new tool set by the name of BasicRoto in
it.
There are some tool sets for creating particles, for instance, that are built in to Nuke. These can
also be found in the ToolSets menu. To prevent confusion between built-in tool sets and tool sets
that you’ve created, if you happen to create a tool set folder with the same name as a built-in one,
Nuke adds [user] in front of the folder name for clarity.
4. To delete a tool set you’ve created, you can click ToolSets > Delete in the toolbar and select the
tool set you want to remove.
USER GUIDE
167
Using the Tab Menu | Adding Nodes by Sub-string
Using the Tab Menu

The Node Graph's Tab menu, opened by pressing Tab in the Node Graph, enables you to search for
and add nodes easily using partial names. Commonly used nodes are weighted so that they appear
higher up the list of choices and you can favorite nodes, pinning them to the top of the list. Weights
and favorites can be enabled, disabled, and cleared in the Preferences under Behaviors > Nodes.
Adding Nodes by Sub-string

Typing characters into the Tab menu displays a list of matches containing those characters. The more
characters you enter, the more the list of applicable nodes is refined. To display the menu, press Tab
when the Node Graph has either click or mouse-over focus.
Tip: See Panel Focus and Keyboard Shortcuts for more information about focus.
Typing in the Tab menu narrows the selection. For example, entering the string dea displays all
entries beginning with dea. In this case, you can see that there are no nodes beginning dea, so Nuke
looks instead for node beginning de with an a in the name.
USER GUIDE
168
Using the Tab Menu | Adding Nodes by Sub-string
The Tab menu parses capitalization as well. For example, if you want a DeepMerge node, start your
string DM rather than dm to refine the results.
You can also search by toolbar menu name. For example, if you know that you're searching for a
color correction node you could enter [Color].
USER GUIDE
169
Using the Tab Menu | Adding and Removing Favorites
Nuke's Bookmark feature also allows you to use sub-strings to locate the bookmark you need. See
Bookmarking Nodes for more information.
Adding and Removing Favorites

Favorites allow you to mark nodes that you use most often so that they are listed first in the Tab
menu. You can add and remove favorites by clicking the star next to their entry in the node list.
You can enable, disable, and clear all favorites in the Preferences under Behaviors > Nodes.
USER GUIDE
170
Using the Tab Menu | Managing Weighting
Managing Weighting
Weighting describes how often you use a node from the Tab menu. Commonly added nodes carry
more weight and are listed higher up in the node list.
Note: Adding nodes using a keyboard shortcut, such as pressing B to add a Blur node, does
not affect node weighting.
The more often you add a node form the Tab menu, the larger the green dot next to the node name.
Weighting can produce unexpected results when you enter strings in the Tab menu. For example, if
you use Backdrop nodes often, you may find that weighting causes Backdrop to appear above nodes
that start with the first letter of the search string.
USER GUIDE
171
Using the Tab Menu | Managing Weighting
You can enable, disable, and clear weighting information in the Preferences under Behaviors >
Nodes.
USER GUIDE
172
Navigating Inside the Node Graph | Panning
Navigating Inside the Node Graph

As scripts grow in complexity, you need to be able to pan to a particular cluster of nodes quickly. The
Node Graph offers a couple of methods for doing so.
Panning
You can pan using the mouse or the navigator map that appears in the lower right corner of the Node
Graph.
Panning with the Mouse
To pan with the mouse, press the middle mouse button and drag the mouse pointer over the
workspace (you can also use Alt+drag). The script moves with your pointer.
Note: In many Linux windows managers, the Alt key is used by default as a mouse modifier
key. This can cause problems in 3D applications where Alt is used for camera navigation in
3D environments.
You can use key mapping to assign the mouse modifier to another key, such as the
(Super or Windows logo) key, but the method changes depending on which flavor of Linux
you're using. Please refer to the documentation on key mapping for your particular Linux
distribution for more information.
Panning with the Map
If your script is larger than the visible workspace, a navigator map automatically appears in the
bottom-right corner.
USER GUIDE
173
Navigating Inside the Node Graph | Zooming, Fitting in the Node Graph
The map shows you a miniature view of the entire script and the pink rectangle shows the portion of
the script that you see within the workspace borders.
To pan with the map, drag the pink rectangle to pan to a different view of the script.
When the whole script is contained within the window border, then the map automatically
disappears.
Tip: The navigation map is resizeable. Drag on its upper-left corner to make it as large or
small as you like.
Zooming, Fitting in the Node Graph

Action How to perform it
Zooming in To zoom in, you can either:

• Move the cursor over the area you want to zoom in on, and press the
addition key (+) repeatedly until the workspace displays the script at the
desired scale.
OR
• Press Alt and drag right while holding down the middle-mouse button.
Zooming out You can zoom out by doing one of the following:
• Move your mouse pointer over the area you want to zoom out from, and
press the subtraction key (-) repeatedly until the workspace displays the script
at the desired scale.
OR
• Press Alt and drag left while holding down the middle-mouse button.
Note: In many Linux windows managers, the Alt key is used by default as a mouse
modifier key. This can cause problems in 3D applications where Alt is used for camera
navigation in 3D environments.
USER GUIDE
174
Navigating Inside the Node Graph | Bookmarking the Pan and Zoom Level
Action How to perform it
Fitting the selected To fit selected nodes in the Node Graph, click the middle-mouse button or
nodes within the press F.
visible workspace
Fitting the node To fit the entire node tree in the Node Graph, click on the Node Graph to make
tree within the sure no nodes are selected and click the middle-mouse button or press F.
visible workspace
Bookmarking the Pan and Zoom Level

To bookmark the pan and zoom level:
1. Pan and/or zoom the script as necessary.
2. To save the current pan and zoom level, navigate to Edit > Bookmark > Save Location 1 (or press
Ctrl/Cmd+F7).
You can save three more locations in a similar manner, using Save Location 2, Save Location 3,
and Save Location 4 (or by pressing Ctrl/Cmd+F8/F9/F10).
The saves are temporary and not saved to the script.
3. To restore a saved location later, select Edit > Bookmark > Restore Location1, Restore Location
2, Restore Location 3, or Restore Location 4 (or press Shift+F7/F8/F9/F10).
Note: Restoring a saved location doesn’t work across different Node Graphs - you can’t save
a location in one Node Graph and then restore it in another.
Bookmarking Nodes
One extremely useful function of Nuke nodes, is their ability to act as jump-to points throughout a
project.
1. Double-click on the node you want to bookmark to open its properties.
2. Go to the Node tab and select the bookmark checkbox (or press Ctrl/Cmd+Shift+B).
This adds the node to the bookmark list.
USER GUIDE
175
Navigating Inside the Node Graph | Searching for Nodes
By default, Backdrop nodes have their bookmark checkbox enabled. All other nodes have it
disabled.
3. Select Edit > Bookmark > Jump to Bookmarked Node (or press J on the Node Graph) to bring up
the bookmarks jump to menu.
4. Start typing the name of the node you wish to navigate to.
Nuke shows all matchingnodes that have the bookmark flag enabled. They are listed according to
the node name or, if a label is present, the label followed by the node name in square brackets.
5. To select the node to navigate to, you can either click on it, or scroll to it with the Up and Down
arrow keys, and press Return.
Searching for Nodes

Nuke’s Search feature allows you to search for nodes in your script and select any matches found. As
a search string, you can enter all or part of a node name. For example, you can search for all Blur
nodes in your script by entering bl as the search string.
You can also do more complex searches using regular expressions, such as searching for all the Read
and Write nodes in a script.
To Search for Nodes
1. Select Edit > Search (or press /) to bring up the search dialog.
2. In the search field, enter the string you want to search for.
If you want to search for all nodes in the script, enter * (an asterisk).
If you want to search for all Read nodes in the script, enter Read.
If you want to search for all the Read and Write nodes, enter the following expression:
(*Read*|*Write*)
3. Click OK.
Nuke searches for the nodes in the script and selects all matches it finds and focuses the Node Graph
the first of them. If you perform the same search again, the Node Graph view focuses on the next
node in order. For instance, if you have a ColorBars node, a ColorWheel node, and a ColorCorrect
node in your Node Graph, searching for “Color” three times in a row selects all of these nodes and
focuses on each of them in alphabetical order.
Note: When you enter expressions in the search dialog, remember that the search field
only takes regular expressions. Any characters that have specific meanings in regular
USER GUIDE
176
Navigating Inside the Node Graph | Cleaning up the Node Graph
expressions, such as square brackets ([ and ]), need to be preceded by the backslash
character (\).
Cleaning up the Node Graph

Sometimes you may find that your node tree becomes a bit disorganized as you’re creating and
removing nodes. There’s a quick fix you can use to tidy up your tree:
1. Select all the nodes you want to rearrange in the Node Graph.
2. Click Edit > Node > Autoplace. Alternatively, press L.
The Autoplace function automatically arranges the selected nodes in a neat tree formation. It
determines a sensible position for the nodes by ensuring they don't overlap with other nodes and
by keeping the input connections horizontal.
USER GUIDE
177
Properties Panels | Managing the Properties Panel
Properties Panels
When you insert a node, its properties automatically appear in the Properties panel with options to
define the node’s output. You can also open the properties panel later by doing any of the following:
• Double-click on the node in the Node Graph.
• Ctrl/Cmd+click on the node in the Node Graph.
• Select the node in the Node Graph and press Return.
Tip: To open a properties panel in a floating window, Ctrl/Cmd+double-click or

Ctrl/Cmd+Alt+click on the node.
Managing the Properties Panel

You can limit the number of node panels that can be open in the Properties panel. To do so, enter
the maximum number of nodes in the field on the Properties panel.
To lock the Properties panel and have all new panels appear in floating windows, click the lock
button on the Properties panel.
To empty the Properties panel and close all the node panels in it, click the remove all panels button.
Tip: You can also close all the nodes in the Properties panel by Alt+clicking on the close (X)
button of one of the nodes.
USER GUIDE
178
Properties Panels | Properties Panel Controls
Properties Panel Controls

These are the standard controls of every properties panel:
Control Function
Hide or show the node’s tabbed pages.
Centers the node in the Node Graph.
Centers one of the node’s inputs in the Node Graph. Select the input from the
dropdown menu that appears.
You can save, load, and manage node presets here.
name field (for You can enter a new name for the node here.
example,
Blur1)
Changes the color of the node. You can drag and drop this button on top of
(left)
another color button to copy the color. To revert to the default color defined in
your Preferences, right-click on the button and select Set color to default.
An X on the button indicates the color is unset, and the color defined in the
Preferences is used.
Changes the color used for the node’s controls in the Viewer. You can drag and
(right)
drop this button on top of another color button to copy the color. To revert to the
default color defined in your Preferences, right-click on the button and select Set
color to default.
An X on the button indicates the color is unset, and the color defined in the
Preferences is used.
Undoes the last change made to the node.
Redoes the last change undone.
USER GUIDE
179
Control Function
Reverts any changes made after the current node panel was opened.
Displays Help related to the node and its controls.
Note: If the Online Help is unavailable, a tooltip displays information

about the node.
Floats the panel. Clicking this button again docks the panel back into the pane if
the originating pane is open.
Closes the node panel. Alt+click this to close all the panels in the pane.
Ctrl/Cmd+click to close all panels except the one clicked on.
Floating control panels also include the following buttons:
Control Function
Reverts any changes made after the panel was opened.
Reverts any changes made after the panel was opened and closes the panel.
Hitting this button right after a node was created also deletes the node from the
Node Graph.
Closes the panel.
Many node panels also contain several tabbed pages.
USER GUIDE
180
On the Node tab, you can usually adjust the following controls:
Control Function
label Lets you add comments to the node. The comments are displayed on the
node’s surface.
If you like, you can use HTML in the label field. For example, to have your
comments appear in bold, you can enter <b>My Comment</b>. To add an
icon called
MyIcon.png to the node, you can use <img src="MyIcon.png"/>. Save the
icon in your plug-in path directory. (For more information on plug-in path
directories, see Loading Gizmos, NDK Plug-ins, and Python and Tcl Scripts.)
Most common image formats work, but we recommend using .png.
Note that the HTML has been changed to a slightly non-standard form
where newlines are significant. If there is a newline character in your data, a
new line is displayed in the label.
font Lets you change the font for any text displayed on the node.
Bolds any text displayed on the node.
Emphasizes any text displayed on the node.
Lets you change the font size of any text displayed on the node.
Lets you change the color of any text displayed on the node.
hide input Check this to hide the node’s incoming pipe. This control does not appear
on all nodes.
cached Check this to keep the data upstream from the node in memory, so that it
can be read quickly. When this is checked, a yellow line displays under the
node in the Node Graph.
disable Check this to disable the node. Uncheck to re-enable. (You can also disable
or re-enable a node by selecting it in the Node Graph and pressing D.)
dope sheet Check this to force the node to always display in the Dope Sheet. You can
also press Alt+D to toggle this on and off.
USER GUIDE
181
Properties Panels | Displaying Controls
Control Function
By default, the Dope Sheet displays all your Read nodes that have their
control panels open, and any nodes you’ve used to create keyframes.
bookmark Check this to show the node in the bookmark list. This allows you to quickly
navigate to the node. For more information, see Navigating Inside the Node
Graph.
postage stamp Check this to display a thumbnail render of the node’s output on its
surface. You can also press Alt+P on the Properties Bin to toggle postage
stamps.
static frame If you want the postage stamp to display a fixed frame (rather than update
to match the current frame), enter the frame to display here. For this to
work, you also need to press Shift+S to open the Preferences dialog, go to
the Node Graph tab, and make sure postage stamp mode is set to Static
frame.
Note that if the frame number you use is outside the frame range for the
node, it is clamped to the first or last frame in the range.
lifeteme range Enter a start and end frame to control the frame range when this node is
active. You can quickly toggle the lifetime on and off with use lifetime.
Displaying Controls
To display a node’s controls, double-click the node. It’s properties panel appears.
The image below shows the controls available for editing parameters. Note that the presence of each
control varies according to the parameter’s function.
USER GUIDE
182
Properties Panels | Using Input Fields
Using Input Fields

You can key values directly into a field, press the arrow keys to increment and decrement values, or
use the middle-mouse button to activate virtual sliders.
Changing Field Value
To key in field values:

1. Double-click in the field to select the whole value.
2. Type the value you want to replace the selection.
Tip: You can also enter expressions (programmatic instructions for generating values) into
fields. You always start an expression by typing =. See Expressions for information about
how to format expressions.
Tip: Nuke also allows you to enter formulas into fields, making it easy to do quick
calculations. For example, if you wanted to halve a value of 378, you could simply type 378/2
into a field and press Enter to get 189.
You can increment or decrement field values by hundreds, tens, tenths, hundredths, and so on. The
magnitude of change depends on the initial position of your cursor. For example if you wanted to
increment the initial value of 20.51 by ones, you would insert your cursor before the 0.
To increment or decrement a field value:

1. Click to insert the cursor just prior to the digit you want to increment or decrement.
2. Press the up arrow to increment by one unit, or the down arrow to decrement by one unit.
Tip: You can also increment and decrement values using the mouse wheel (if available) or by
pressing Alt while dragging on the value. The latter method is particularly useful for tablet
users.
Using Virtual Sliders
To use a virtual slider, you can hover the cursor over a numeric control, click the middle-mouse
button, and drag the field to adjust the associated control.
USER GUIDE
183
Properties Panels | Using Input Fields
Tip: Hold Shift to increase slider sensitivity.
USER GUIDE
184
Using the Color Controls | Using the Color Swatch
Using the Color Controls

In the Properties panels, there are usually four extra controls to the right of a control including color
swatch, color picker, channel selector and the animation menu. The first three are color controls.
For more information about using the animation menu, see Animating Parameters.
You can use the color swatch button to activate the eye dropper tool. You can copy a color from one
swatch to another by dragging and dropping the color swatch with the color you want to use over the
color swatch you want to replace.
The color picker button can be used to display and make color selections. Depending on the settings
in the Preferences dialog, you may use an in-panel color picker, or a floating color-picker window.
(See Appendix A: Preferences for more information.)
You can use the channel selector to toggle between the slider and manually enter values for each of
the channels.
Using the Color Swatch

You can use the color swatch to sample a color from the Viewer, by doing the following:
1. Click the color swatch to activate the eye dropper tool.
2. Move the cursor over the Viewer, over the color you want to sample.
You can zoom in and pan as necessary until the area you want to sample is clear.
3. Ctrl/Cmd+click to sample a color value, or Ctrl/Cmd+Alt+click to sample a color from the node's
input while viewing its output.
You can sample a region instead of just a single pixel by also pressing Shift+drag when you're
sampling the color. When sampling a region, Nuke calculates the average color of the region.
If you are not happy with the pixel or region you've sample, you can simply repeat the sample
procedure until you are happy with the selected pixel or region.
USER GUIDE
185
Using the Color Controls | Using the In-Panel Color Picker
Tip: You can discard sampled pixels by Ctrl/Cmd+right-clicking in the Viewer.
4. When you are happy with the color sample, click the color swatch again. This closes the eye
dropper tool and now displays your selected sample.
Using the In-Panel Color Picker

If your Preferences > Panels > Control Panels > Color Panel > color picker button opens
dropdown is set to in-panel color picker, you can display the color sliders and wheel within the
Properties panel by clicking the color picker button . You can have multiple in-pane color pickers
open simultaneously.
Tip: Holding Ctrl/Cmd and clicking the color picker button opens the alternate color picker
to the one specified in the Preferences dialog.
Using the In-Panel Color Wheel
You can adjust the hue by dragging the marker on the edge of the color wheel (or the marker inside
the wheel) around to the required hue. You can adjust the saturation by dragging the marker inside
the wheel in or out to the required saturation.
USER GUIDE
186
Using the Color Controls | Using the Floating Color Picker Window
Tip: You can also Ctrl/Cmd+click on the color wheel to only affect the hue, and Shift+click to
only affect the saturation.
To adjust the value (brightness) of the color, hold down Ctrl/Cmd+Shift and drag right or left.
By default, the color wheel uses relative cursor positioning. This means the cursor position can be
offset from the marker position, which allows for very fine adjustments. You can disable this behavior
and use absolute cursor positioning by holding Alt while moving the markers.
Note: If any red, green, or blue values are higher than 1, the marker inside the wheel turns
red. If the red, green, and blue channels all have a value of less than 0.1, the marker turns
blue. If at least one channel is negative and at least one is positive, the hue indicator turns
red.
Using the In-Panel Color Sliders
To increment the value by 0.01, left-click on the arrow button . To decrement the value by 0.01,
right-click on the arrow button. Use Shift+click for 0.1, and Alt+click for 0.001.
You can also click and drag right or left on the button to scrub the value up or down. Use Shift+drag
to scrub quickly, or Alt+drag to scrub slowly.
Using the Floating Color Picker Window

If your Preferences > Panels > Control Panels > Color Panel > color picker button opens
dropdown is set to floating color picker, you can display color sliders and wheel in a floating window
by clicking the color picker button .
Tip: Holding Ctrl/Cmd and clicking the color picker button opens the alternate color picker
to the one specified in the Preferences dialog.
You can change the floating window to be horizontal by dragging on one of the corners of the window
to resize it. When it is wide enough, the sliders automatically become horizontal.
USER GUIDE
187
Using the Color Controls | Using the Floating Color Picker Window
• From the TMI, HSV, and RGB buttons, you can select which slider you want to display.
• There are three states of the color wheel: hide color wheel, show color wheel, and show color wheel
with square. You can cycle through these states by simply clicking on the color wheel button
multiple times.
• You can choose to show the color swatches by clicking the color swatch button.
• If you want the background of the sliders to show what the value of the color would be if the sliders
were set to the current position, click the Dyn button.
Using the Color Wheel in the Floating Window
Using the color wheel in the floating window is exactly the same as using the in-panel color wheel;
drag the marker on the edge of the wheel to adjust the hue, and drag the marker inside the wheel to
adjust the saturation.
However, you can also pan and zoom on the color wheel in the floating window:
• To pan, press Alt and drag the cursor over the color wheel.
• To zoom, press Alt and drag left or right with the middle-mouse button.
• You can reset the zoom and/or pan by simply middle-clicking on the color wheel.
Using the Color Sliders in the Floating Window
Using the sliders in the floating color window is exactly the same as using the in-panel sliders.
Using the Color Swatches in the Floating Window
When you're happy with a color, you can save it by right-clicking on the swatch where you want to
save it. This replaces the original swatch that was there. You can also drag and drop a color you want
to save on any of the swatches to replace it.
USER GUIDE
188
Using the Color Controls | Using the Channel Selector
The rectangle above the sliders shows the original color on the right, the currently selected color on
the left. When you hover the cursor over this, an arrow appears allowing you to copy one to the other
by clicking once on the rectangle.
Opening Another Floating Color Picker Window
You can open multiple color picker floating windows, by pressing Ctrl/Cmd+Alt+click on another
parameter's color picker button.
Using the Channel Selector

By default, many parameters in Nuke automatically group channels for you. For example, if you drag
on the gain slider in the ColorCorrect node, you simultaneously affect the R, G, and B channels
(assuming you're processing the RGB layer.) You can use the channel selector button to display
and edit the individual channel values. The number displayed on the channel selector button
represents the number of available channels for editing.
You can edit an individual channel's value by doing the following:

1. Click the channel selector button.
The available channels appear with value fields for each. For example, when you press the gain -
channel selector button, four channel value fields appear (R,G,B, and A) as the channel selector
shows that there are four available channels for editing.
2. Enter the required values in the channel value fields.

After you are happy with the values, simply click the channel selector button again to close and
save your settings.
USER GUIDE
189
Using the Color Controls | Color Slider Functions
Color Slider Functions
RGBA Sliders
Slider Function
The red slider (R) This allows you to control the red channel's value (or the first channel in a layer
if you are processing another layer besides RGBA).
The green slider This allows you to control the green channel's value (or the second channel in a
(G) layer if you are processing another layer besides RGBA).
The blue slider (B) This allows you to control the blue channel's value (or the third channel in a
layer if you are processing another layer besides RGBA).
The alpha slider This allows you to control the alpha channel's value (or the fourth channel in a
(A) layer if you are processing another layer besides RGBA).
TMI Sliders
Slider Function
The temperature This allows you to control apparent color temperature by inversely affecting
slider (T) red and blue values (assuming you are processing the RGBA layer).
The magenta/green This allows you to control the mix of green and magenta hues. To add more
slider (M) magenta (increase the red and blue channels' values, while decreasing the
green channel's), drag up. To add more green (increase the green channel's
value, while decreasing the red and blue channels'), drag down.
The intensity slider This allows you to simultaneously control the red, green, and blue channel
(I) values. To increase the value of all channels by the same amount, drag up.
To decrease the value of all channels by the same amount, drag down.
USER GUIDE
190
Using the Color Controls | Color Slider Functions
HSV Sliders
Slider Function
The hue slider (H) This allows you to control the color's location on the traditional color wheel
(for example, whether the color is red, yellow, or violet).
The saturation This allows you to control the intensity or purity of the color.
slider (S)
The value slider (V) This allows you to control the brightness of the color (the maximum of red,
green, and blue values).
Note: The HSV sliders are only available in the floating color picker window. In the in-panel
color picker, you can use the color wheel to adjust hue, saturation, and value.
USER GUIDE
191
Customizing a Node’s Defaults with Node Presets | To Create a Node Preset:
Customizing a Node’s Defaults

with Node Presets
Sometimes the default control values on a node aren’t the ones you use most often. If you find
yourself frequently adjusting many control defaults, you may want to save your customized default
values as a node preset.
To Create a Node Preset:

1. Set the controls in a node’s control panel to the values you want.
2. Click the load node presets button on the control panel title bar and select Save as Preset.
3. Give your preset a name and click Create.

Your new preset now appears as an option when you click the load node presets button and
select Apply Preset.
Loading and Deleting a Preset

1. To load a preset on a node, click the load node presets button and select your preset under
Apply Preset. User presets are your own personal presets, and shared presets can also be used
by other users.
2. To delete a preset, click the load node presets button and select Delete Preset. In the dialog
that displays, select the preset you want to delete from the dropdown menu, and click Delete.
USER GUIDE
192
Animating Parameters
Animating a parameter refers to changing its value over time. You do so by setting keyframes (frames
at which you explicitly define a value) and allowing Nuke to interpolate the values in between. You can
animate most of Nuke’s parameters in this manner.
Working with Animated Parameters

The Animation menu allows you to set key frames, delete keys, and perform other editing operations
on the curves for animated parameters.
Setting Keyframes
To set keyframes, do the following:

1. Use a Viewer to navigate to a frame where you want to place a key.
2. Click the animation button next to the parameter you want to animate.
3. Select Set key from the dropdown menu. The parameter’s input field turns blue, indicating that a
keyframe has been inserted. Nuke switches to autokey mode: when you change the parameters
value at another frame, it automatically inserts another keyframe for you.
You can also set a key for all the controls in a node. To do so, right-click on a node's properties
panel and select Set key on all knobs.
4. Navigate to the next frame where you want to place a key.
USER GUIDE
193
Animating Parameters |
5. Edit the parameter’s value using the input field, regular slider, or color slider. The moment you
change the value, Nuke creates a keyframe.
6. Continue adding key frames as necessary.
7. Use the Viewer’s scrubber to preview the result.
Deleting Keyframes
To delete a single keyframe:

1. Use the Viewer’s next edit/keyframe and previous edit/keyframe buttons to navigate to the
keyframe that you want to remove. Notice that the scrub bar indicates key frames with a blue
mark.
2. Click the animation button .

3. Select Delete key from the dropdown menu.
Nuke removes the keyframe.
To delete all key frames from a parameter:

1. Click the animation button.
2. Select No animation from the dropdown menu.
A confirmation dialog appears.
3. Click Yes.
Nuke removes all key frames from the parameter, and sets the static value to match that of the
current frame.
Linking Animated Parameters with Tracker
You can link any controls with others by using expressions (see Expressions), but you can also quickly
and easily link with the Tracker node using the Link to option in the Animation menu. For example,
to link the translate control of the RotoPaint node with a Tracker node, do the following:
1. Create the Tracker node you want to link to.
2. On the Transform tab of the RotoPaint node’s control panel, click on the translate animation
menu.
3. Select Link to > Tracker linking dialog...
4. Select the Tracker node you want to use in the tracker node dropdown and in the link to
dropdown, select whether you want to link to the position of the track or the translate values of it.
5. Select which tracks you want to use by checking the track boxes.
The expression fields update with the appropriate expression syntax.
6. Then click OK, and your linking is done.
USER GUIDE
194
Animating Parameters | Animated Parameters in the Dope Sheet and the Curve Editor
Your Bezier shape’s translate value now changes when the Tracker value is changed.
Animated Parameters in the Dope Sheet and the Curve

Editor
As you add keyframes to a parameter, Nuke automatically both adds markers to the Dope Sheet for
them and plots a curve on its Curve Editor panel.
The Dope Sheet provides an easy way to edit keyframes as well as reposition, trim, and slip clips.
The Curve Editor allows you to adjust keyframes and the interpolation between them.
You can add key frames, delete key frames, and even adjust the interpolation between key frames
without ever looking at the Dope Sheet or Curve Editor. However, as the animation grows more
complex, you may find it easier to edit the animation by manipulating its keyframes directly. For
more information on how to do so, see Editing Clips in the Dope Sheet and Using the Curve Editor.
USER GUIDE
195
Animating Parameters | Using the Dope Sheet
Using the Dope Sheet

The Dope Sheet gives you access to all the keyframes you’ve created and provides an easy way of
editing them. You can also use it to reposition, trim, and slip Read nodes and certain Time nodes.
Also see Using the Curve Editor.
Using the Dope Sheet Interface

To show the Dope Sheet, click the Dope Sheet tab. If you can’t see the tab, right-click the Node Graph
title bar and select Windows > Dope Sheet.
In the Dope Sheet, you can see the following:
Hierarchy view Click through the hierarchy view to see your nodes and a hierarchical
list of animated controls. If you can’t see a node in the list, try opening
its properties panel.
To make a node appear in the Dope Sheet without having its properties
panel open, select the dope sheet checkbox in the bottom of the Node
tab of the node's properties panel.
You can display the Read and TimeClip nodes in the Dope Sheet when
their properties are closed, by clicking the middle button in the
bottom-left of the Dope Sheet.
Read nodes Any Read nodes that have their properties panels open display as gray
bars in the Dope Sheet.
USER GUIDE
196
Time nodes Any AppendClip, FrameRange, Retime, TimeOffset, TimeWarp, and

TimeClip nodes that have their properties panels open display as green
bars.
Keyframes Your keyframes display as gray markers in the keyframe grid. If you go
down to the lowest level of hierarchy, you can also view the keyframes
set for different views in stereoscopic or multi-view projects.
Current frame indicator The current frame indicator displays as an orange line.
First and last frame in The first and last frame in the project's frame range (as defined in the
the project Project Settings) display as white/gray lines.
You can also use the following controls at the bottom of the Dope Sheet:
Control Description
Synchronize the frame range of your project between the Dope Sheet
and the Curve Editor.
Display all Read and TimeClip nodes in the Dope Sheet. By default,
only the nodes that have their properties panels open show up on the
Dope Sheet. If you click this button, all Read and TimeClip nodes in
your script are displayed in the Dope Sheet. If you click the button
again, the default view is restored.
Move the selected keyframes by the number of frames specified. You

can enter negative values to move the keyframes backwards.
Note: You cannot move keyframes past other keyframes.
If you don't enter a number in the Move field, it shows the amount of
your previous move.
Set the frame range to display in the Dope Sheet.
Viewing Keys in the Dope Sheet

To view keys in the Dope Sheet, you need to create some Read nodes, Time nodes, or keyframes first.
The Dope Sheet displays:
USER GUIDE
197
• All the Read nodes that have their properties panels open.
• Any AppendClip, FrameRange, Retime, TimeOffset, TimeWarp, and TimeClip nodes that have their
properties panels open.
• Any nodes that have keyframes set on them and have their properties panels open.
Tip: To force the above nodes to display in the Dope Sheet even when their properties aren't
open, check the dope sheet box on the Node tab of the node’s properties panel or select
the node in the Node Graph and press Alt+D.
Tip: If you want to view an AudioRead node under your keyframes, you can display it by
right-clicking and selecting View > Audio > Draw Style > Below. For more information on
the audio options, see Audio in Nuke.
The Dope Sheet and Time Nodes

If you have an AppendClip, FrameRange, Retime, TimeOffset, TimeWarp, or TimeClip node
downstream of the node being animated, the Dope Sheet shows your keyframes in the context of the
currently active Viewer. Whenever you change the Viewer input, the keys in the Dope Sheet move
accordingly to reflect where their effect is actually seen in the Viewer:
• If you view keys before a node that moves them in time, you see them at their original time.
• If you view keys after the time manipulation, you see them shifted by the time operation.
For example, in the below script, the Transform node has been animated on frames 0, 10, 20, and 30.
When the Viewer is connected to the Transform node, the Dope Sheet shows keys on those frames.
There is also a TimeOffset node in the script, set to move the input clip 5 frames forward in time.
When the Viewer is connected to the TimeOffset node, the Dope Sheet shows the keys shifted by 5
frames - on frames 5, 15, 25, and 35.
USER GUIDE
198
Multiple Branches of Nodes in the Dope Sheet

If there are multiple paths from an animated node to the currently active Viewer, the keyframes for
the animated node appear in the Dope Sheet multiple times. For example, in the below script, there
are two paths from the animated Blur node to the Viewer:
• Blur > Merge > Viewer
• Blur > TimeOffset > Merge > Viewer
As a result, the keyframes for the Blur node appear in the Dope Sheet twice. Adjusting the keys in one
location also updates them in the other.
Selecting Keyframes
There are several ways to select and edit different keyframes:
Select multiple frames Drag a marquee over your keyframes in the keyframe view.
Add or remove Hold Shift and drag another marquee around some of the selected
keyframes from a keyframes.
marquee selection
Select a single keyframe Click it in the keyframe view. Its parent keyframe and all other keyframes
depending on it are selected too. The frame number of the keyframe is
displayed next to it.
Select all keyframes in Click the node name in the hierarchy view. All the child keyframes under
one node that node are selected.
If there are multiple paths from an animated node to the currently active Viewer, the keyframes for
the animated node appear in the Dope Sheet multiple times. When you select them in one location,
they turn orange, while the corresponding keys in other locations turn white.
USER GUIDE
199
Animating Parameters | Editing Clips in the Dope Sheet
Note: When selecting keys that appear in the Dope Sheet multiple times (using multiple
paths), marquee selection is constrained to one path only.
Copying and Pasting

You can copy and paste keyframes in the Dope Sheet. After selecting the keyframe(s) you want to
copy, use the Ctrl/Cmd+C keyboard shortcut to copy the keyframe(s), and Ctrl/Cmd+V to paste them
in another location of the Dope Sheet.
When using the Ctrl/Cmd+V keyboard shortcut, pasting keyframes is relative to the location of the
current frame indicator. In other words, if you have keyframes at 1, 10, and 20 and you copy this
selection before pasting the selection at frame 21, new keyframes are set at frames 21, 30, and 40.
To paste keyframes on the same frame as the source keys, select the control you want to receive the
keys, right-click on the Dope Sheet, and select Edit > Paste > Paste Absolute.
Deleting and Adding Keyframes

You can remove keyframes in the Dope Sheet. To delete a keyframe, just select it and press
Delete/Backspace button or right-click on the keyframe and select Edit > Erase.
You can also add keyframes in the Dope Sheet to add animation where there isn’t any at the moment.
Add keyframes by Ctrl/Cmd+Alt+clicking on an empty spot in the Dope sheet. A new keyframe is
created and you can adjust it as necessary.
Editing Clips in the Dope Sheet

The Dope Sheet allows you to easily reposition, trim, and slip your clips as well as set a frame range
for them:
• Repositioning clips refers to changing the position, but not the content or the duration, of the clip.
In the Dope Sheet, you can reposition clips using the Read, TimeOffset, and TimeClip nodes.
USER GUIDE
200
• Trimming clips refers to removing unwanted frames from the head or tail of the clip. In the Dope
Sheet, you can trim clips using the Read and TimeClip nodes.
• Slipping clips refers to changing the content of the clip that is seen, but not the position or duration.
In the Dope Sheet, you can slip clips using the Read and TimeClip nodes.
• Setting a frame range for a clip controls which frames a downstream AppendClip node uses from
the input, which frames are displayed in the Viewer when the timeline range dropdown menu is set
to Input, and which frames are sent to the flipbook. In the Dope Sheet, you can set the frame range
for a clip using the FrameRange node.
Repositioning, Trimming, and Slipping Clips Using Read and TimeClip Nodes
• To reposition a clip, place the cursor over the Read or TimeClip bar in the Dope Sheet and drag left
or right.
The new start and end frame numbers are displayed on either side of the bar, and the frameoffset
control in the Read or TimeClip properties panel is adjusted automatically.
Tip: You can also reposition a clip by selecting the bar, double-clicking on the first and last
frame numbers next to it, and adjusting them in the x fields that pop up.
• To trim a clip, place the cursor in the beginning or end of the bar and drag left or right.
The new start or end frame is displayed next to the bar, and an orange line appears to represent the
original frame range. The framerange control in the Read or TimeClip properties panel is adjusted
automatically.
USER GUIDE
201
Tip: You can also trim a clip by selecting the bar, double-clicking on the first or last frame
number next to it, and adjusting the number in the x field that pops up.
Tip: The number of frames you can trim from either end of a clip is constrained by the clip's
original frame range. However, if necessary, you can trim beyond the original range by
holding Ctrl/Cmd while trimming.
If necessary, you can change how the clip displays inside the clip handles (the unused frames
beyond the start or end of the clip). Adjust the before and after dropdown menus next to the
frame range control in the Read or TimeClip properties to change what happens before and after
the frame range limits:
• hold - select to show a still picture of the first/last frame of the frame range.
• loop - select to start over and keep looping the span of the frame range outside the first/last
frame of the frame range.
• bounce - select to play the span of the frame range backwards and forwards between the frame
range limits.
• black - select to display a black frame outside of the first/last frame.
• To slip a clip, first trim the clip to create clip handles. Then, place the cursor over the bottom half of
the bar and drag left or right.
The clip handles move to indicate what portion of the clip is visible. The frame range and frame
offset controls in the Read or TimeClip properties panel are adjusted automatically.
Tip: The number of frames you can slip a clip by is constrained by the clip's original frame
range. However, if necessary, you can slip beyond the original range by holding Ctrl/Cmd
while slipping.
Repositioning Clips Using TimeOffset Nodes
To reposition a clip, place the cursor over the TimeOffset bar in the Dope Sheet and drag left or right.
USER GUIDE
202
The new start and end frame numbers are displayed on either side of the bar, and the time offset
(frames) control in the TimeOffset properties panel is adjusted automatically.
Tip: You can also reposition a clip by selecting the bar, double-clicking on the first and last
frame numbers next to it, and adjusting them in the x fields that pop up.
To Set the Frame Range Using FrameRange Nodes
• To set both the first and the last frame for a clip, place the cursor over the FrameRange bar in the
Dope Sheet and drag left or right.
The new start and end frame numbers are displayed on either side of the bar, and the frame range
control in the FrameRange properties panel is adjusted automatically.
• To set only the first or the last frame for a clip, place the cursor in the beginning or end of the bar
and drag left or right.
The new start or end frame is displayed next to the bar, and the frame range control in the
FrameRange properties panel is adjusted automatically.
USER GUIDE
203
Animating Parameters | Synchronizing the Frame Range
Tip: You can also set the frame range for a clip by selecting the bar, double-clicking on the
first and last frame numbers next to it, and adjusting them in the x fields that pop up.
Synchronizing the Frame Range

You can synchronize the frame range of your project between the Dope Sheet and the Curve Editor.
This is useful when you’re using both of them to edit your animation curves and keyframes at the
same time. To synchronize the frame range:
1. Right-click anywhere on the Dope Sheet.
2. Select View > Synchronize frame range. Any changes you now make to the frame range in the
Dope Sheet are applied in the Curve Editor too, and any changes you make in the Curve Editor are
also reflected in the Dope Sheet. Zooming in and out in either also zooms the other view.
Tip: You can also do this by clicking the synchronize button at the bottom of the Dope
Sheet.
Using the Curve Editor

The Curve Editor enables you to edit curves without physically entering information in the Properties
pane.
Displaying an Animation Curve
To reveal an animation curve, do the following:
1. Click the animation button next to the parameter whose curve you wish to view.
2. Select Curve Editor. The Curve Editor panel appears with a focus on the selected parameter’s
curve.
The vertical, or y axis, denotes the value of the parameter.
The horizontal, or x axis, denotes time (in frame units).
USER GUIDE
204
Animating Parameters | Using the Curve Editor
Displaying Curves in the Editor
To display curves in the editor, do the following:

1. In the parameter tree on the left, click the + and - signs to expand and collapse the hierarchy as
necessary.
2. Click a parameter’s name to make its curve the focus of the editor. To focus on multiple curves at
the same time, either:
• Hold Shift and click on curves to create a consecutive list,
OR
• Hold Ctrl/Cmd and click on individual curves to add them to the selection.
3. To display separate curves for each channel, separate the channels for the relevant control in the
node’s properties panel.
The parameter tree on the left allows you to focus on any curve in the script.
Tip: If you want to view an AudioRead node under your animation curves, you can display it
by right-clicking anywhere in the curve editor and selecting View > Audio > Draw Style >
Below. For more information on the audio options, see Audio in Nuke.
Enabling/Disable a Curve in the Editor
To enable or disable a curve in the Editor, in the parameter tree on the left, click the + and - signs to
expand and collapse the hierarchy as necessary.
Zooming In or Out in the Editor
To zoom in or out in the Editor, do the following:
USER GUIDE
205
Animating Parameters | Using the Curve Editor
1. Click on the area you want to zoom in on or out of.

2. Press the + button to zoom in, or the - button to zoom out,
OR
Scroll up with the mouse wheel to zoom in, or down to zoom out.
Tip: Depending on your mouse preferences, you can zoom to a custom area in the Curve
Editor, middle-click on the Editor and drag to select an area with a marquee. When you
release the mouse button, the Editor zooms to fit the selected area in the Editor.
Panning in the Editor
To pan in the Editor, hold the middle-mouse button and drag over the Editor. You can also use
Alt+drag.
Resetting Zoom and Panning
To reset the zoom or panning:

1. Right-click on the Curve Editor.
2. From the menu that opens, select View > Frame All.
OR
Press A on the Editor.
OR
Click the middle-mouse button (dependent on your mouse preferences).
Nuke centers the curve in the Editor, resetting the zoom.
Centering a Portion of the Curve
To center a point of the curve in the Editor:

1. Select the points you want to center in the editor.
2. Right-click on the Editor, and select View > Frame Selected (or press F on the Editor).
Nuke centers the selected portion of the curve in the editor. If no points are selected, Nuke
centers the selected curve, or all curves.
USER GUIDE
206
Animating Parameters | Editing Curves
Editing Curves
You edit curves by moving the points on the curve to new locations. If necessary, you can add more
points to the curve. You can also sketch curves freely, use the usual editing functions such as copy
and paste, smooth curves with filtering, interpolate curves, loop, reverse or negate curves, and use
expressions to modify curves.
Adding Points to a Curve
To add points to a curve, do the following:

1. Click on the curve you want to edit. The curve turns yellow to indicate it’s selected.
2. Ctrl/Cmd+Alt+click on the part of the Curve Editor you want to add a point to. You can add points
both on the curve and outside the curve,
OR
1. Right-click on the Editor and select Edit > Generate. The Generate keys dialog opens.
2. In the Start at field, enter the first frame you want to use as a keyframe.
3. In the End at field, enter the last frame you want to use as a keyframe.
4. In the Increment field, enter the frame increment you want to use between the first and the last
keyframe. For example, if you want every tenth frame to be a keyframe, enter 10.
5. In the last field, enter the value you want to use for y. If you do not enter a value here, the key
frames are added to the current curve without modifying the curve shape.
6. Click OK.
Selecting Points on a Curve
You can select points in the following ways:

• To select individual points, click on the point you want to select.
• To select multiple points, Shift+click on the points, or drag a marquee around them.
A box is drawn around the points to indicate they have been selected.
• To select all points, press Ctrl+A (Mac users press Cmd+A).
A box is drawn around the points to indicate they have been selected.
Moving Points on a Curve
• To move a point along either the x or y axis only, drag the point to a new location.
USER GUIDE
207
• To move a point in any direction, Ctrl+drag (Mac users Cmd+drag) the point to a new location. You
can also nudge points using the numeric keypad arrows.
• To adjust the values of a point numerically, select the point and double-click on the x or y value that
appears next to it.
• By default, when you move a point, its position on the x axis is rounded to the nearest integer. To
disable this, you can right-click on the Curve Editor and select Edit > Frame Snap. You can also
momentarily disable the snapping by pressing Shift while moving a point.
• To move several points at the same time, select them and drag the selection box to a new location.
To add or remove points to or from the selection box, Shift+click on the points.
To resize and scale the selection box, drag its edges. If the selection box is very narrow, you can
press Ctrl/Cmd when resizing it. This allows you to resize the box in one dimension only. For
example, if you have a box that’s wide on the x axis but flat on the y axis, you can resize it in this
way along the x axis.
To avoid accidentally moving a point inside the selection box, press Ctrl/Cmd+Shift when dragging
the box to hide the points inside the box.
Adjusting the Slope Around the Points
To adjust the slope around the points, do the following:

1. Select a point on the curve. Red tangent handles appear on both sides of the point.
USER GUIDE
208
2. Drag the tangent handles to a new location. The curve follows the handles.
Sketching a Curve
You can sketch a curve freely by doing the following:
Press Alt+Ctrl+Shift (Mac users press Alt+Cmd+Shift) while drawing a curve on the editor. Nuke
sketches a curve that follows your mouse movements.
Cutting, Copying, and Pasting
You can cut, copy, or paste any selected points, expressions, or curves, by doing the following:
1. Right-click on the Curve Editor.
2. From the menu that opens, select Edit and the editing function you want to use on the entire
curve, for example:
• Edit > Copy > Copy Selected Keys to only copy the points you have currently selected.
• Edit > Copy > Copy Curves to copy an entire curve.
• Edit > Copy > Copy Expressions to copy the expression that creates the curve.
USER GUIDE
209
• Edit > Copy > Copy Links to copy a curve and keep its values linked to the original curve, so that
if you change the original, your changes also affect the copied curve.
Move Selected Points on the Curve
To move selected points on the curve by a fixed value:

1. Select all the points you want to move.
2. Right-click on the editor and select Edit > Move. The Move Animation Keys dialog opens.
3. In the x and y fields, define how you want to move the points along the x and y axes. For example,
to shift the selected points to the right by a value of 10, enter x+10 in the x field.
4. In the slope and left slope fields, define how you want to move the points’ tangent handles.
Smoothing the Curve
To smooth the curve with filtering, do the following:

1. Select the portion of the curve that needs smoothing.
2. Right-click on the editor and select Edit > Filter. The Filter Multiple dialog opens.
3. In the No. of times to filter field, specify how many times you want to filter the curve. Filtering
sets new values on each point based on the average values of their neighboring points. The more
filtering, the smoother the curve.
Interpolating Parts of a Curve
You can interpolate parts of a curve, by doing the following:

1. Select the point(s) between or around which you want to interpolate the curve.
USER GUIDE
210
2. Right-click on the Editor. Select Interpolation and the type of interpolation you want to use.
Select:
• Constant to force a constant value after each selected point.
• Linear to use linear interpolation. This produces sharp changes at key frames and straight lines
between them.
• Smooth to set the tangents’ slopes equal to the slope between the keyframe to the left and the
keyframe to the right if the selected point is between these two key frames along the y axis. If
the selected point is not between these key frames and has a larger or smaller value than both
key frames, the tangents’ slopes are made horizontal. This ensures the resulting curve never
exceeds the keyframe value.
• Catmull-Rom to set the tangents’ slope equal to the slope between the keyframe to the left and
the keyframe to the right regardless of where the selected point is located. The resulting curve
can exceed the keyframe values.
USER GUIDE
211
• Cubic to set the slope so that the second derivative is continuous. This smooths the curve.
• Horizontal to make the tangents horizontal, setting the slope around the selected points to
zero.
• Break to adjust the two tangents of a selected point independent of each other.
• Before > Constant or Linear to interpolate the parts of the curve that are on the left side of the
first point. This option only works if you have selected the first point on the curve.
• After > Constant or Linear to only interpolate the parts of the curve that are on the right side
of the last point. This option only works if you have selected the last point on the curve.
USER GUIDE
212
Repeating a Part of the Curve
To repeat a portion of the curve, throughout the curve, do the following:

1. Right-click on the editor and select Predefined > Loop. The Loop dialogue opens.
2. In the First frame of loop field, enter first frame of the portion you want to repeat throughout the
curve.
3. In the Last frame of loop field, enter the last frame of the portion you want to repeat.
4. Click OK.
The shape of the curve between these frames is repeated throughout the rest of the curve. The
solid line represents the actual curve, and the dotted line the original curve with the key frames.
Reversing a Curve
You can reverse a curve by right-clicking on the editor and select Predefined > Reverse.
This makes the curve go backward in time. Both the new curve and the original curve are displayed.
The solid line represents the actual curve, and the dotted line contains the key frames that you can
modify.
USER GUIDE
213
Negating a Curve
You can negate a curve by right-clicking on the editor and select Predefined > Negate.
The curve becomes the negative of the key frames. For example, a value of 5 turns into -5. Both the
new curve and the original curve are displayed. The solid line represents the actual curve, and the
dotted line contains the key frames that you can modify.
Modifying a Curve with an Expression
You can modify a curve using an expression by doing the following:

1. Enter the expression in the expression field at the bottom of the Curve Editor,
2. Click the Revert button to set the field back to previous valid expression.
OR
1. Right-click on the Editor, and select Edit > Edit expression.
2. In the dialog that opens, type the expression you want to use for the curve, for example, sin(x)/x.
3. Click OK.
USER GUIDE
214
Compositing Viewers
A Viewer node displays the render output of any connected process nodes in the Viewer panel and
does not edit any data. Viewer nodes allow you to quickly assign the right values to parameters as
they allow you to edit in context - that is, edit a given node’s parameters upstream in a script while
viewing the effect of those changes downstream.
Nuke's Viewer output is labeled with the bounding box size and the format. The bounding box is the
area the Nuke treats as containing image data and the format is the resolution of the input image.
These two are the same by default so that the output is the same as the resolution.
See Reformatting Elements and Adjusting the Bounding Box for more information on format and
bounding boxes.
You can place as many Viewer nodes in a script as you wish, which allows you to simultaneously view
multiple outputs. You can also pipe the output from up to ten process nodes into a single Viewer
node, and then cycle through the various displays. This allows you to easily compare an image before
and after processing by a given effect.
Note: The maximum image size the Viewer can display is 64k x 4k (or the equivalent
number of total pixels at other resolutions). Make sure though, that you have sufficient RAM
memory available if you want to use the maximum image size.
USER GUIDE
215
Compositing Viewers | Adding Viewer Nodes
Adding Viewer Nodes

Viewers have corresponding nodes that appear in the Node Graph. These nodes do not produce
output for rendering; they generate display data only. You can connect Viewer nodes as described in
Working with Nodes. In practice, it is faster to use the Viewer keyboard shortcuts described below.
To add a Viewer node:

1. Select the node that you want to view the output of.
2. Do one of the following:
• Using the menu bar, choose Viewer > Create New Viewer.
• Using the Toolbar, choose Image > Viewer.
• Using a keyboard shortcut, press Ctrl/Cmd+I.
Nuke connects a Viewer node to the node you selected in step 1, and displays the output of the
node in the Viewer panel. You can also insert a Viewer node and set up its first connection by
simply pressing 1 over the Node Graph.
Connecting Viewer Nodes

After you add a Viewer node to the script, you can quickly pipe any process node’s output to it simply
by selecting the process node then pressing any number key. Doing so, pipes the output to one of the
ten input ports available on every Viewer node (the 0 key represents the tenth slot).
USER GUIDE
216
Compositing Viewers | Toggling Views
Toggling Views
If a Viewer node has multiple inputs, like the one depicted above, you can press the up or down
arrow keys to quickly cycle through the views (your cursor needs to be in the Viewer window). To view
a particular node press the number key (1, 2, 3... 0) corresponding to the pipe number whose
contents you wish to view.
Panning and Zooming the Viewer Window

Nuke offers several ways, including some useful keyboard shortcuts, to focus on the area of the
Viewer you need to see.
Panning the Frame

To pan the frame, hold the middle-mouse button and drag on the display (you can also use Alt+drag).
The frame follows the mouse pointer.
Note: In many Linux windows managers, the Alt key is used by default as a mouse modifier
key. This can cause problems in 3D applications where Alt is used for camera navigation in
3D environments.
Re-centering the Frame

You can re-center the frame by clicking the middle-mouse button or pressing F.
Zooming In and Out

To zoom in on the frame:
1. Move your pointer over the area of the display on which you want to zoom.
2. Drag the mouse while pressing the middle mouse button and the left mouse button.
OR
Press the plus button (+) repeatedly until the frame attains the desired scale.
USER GUIDE
217
Compositing Viewers | Showing / Hiding Floating Viewers
OR
Select zoom in from the zoom dropdown menu in the top right corner.
To zoom out from the frame:

1. Move your pointer over the area of the display from which you want to zoom.
2. Drag the mouse while pressing the middle mouse button and the left mouse button.
OR
Press the minus button (-) repeatedly until the frame displays at the desired scale.
OR
Select zoom out from the zoom dropdown menu in the top right corner.
Restoring the Zoom to 100%

You can restore the zoom to 100% by pressing Ctrl/Cmd+1.
Showing / Hiding Floating Viewers

You can hide and show floating Viewers using a handy keyboard shortcut.
To hide or show a floating Viewer, press ‘ (the backtick key).
Showing / Hiding Viewer Toolbars

To hide or show the top toolbar, press {.
To hide or show the bottom toolbar, press }.
Locking the Viewer Zoom Level

You can choose to lock the Viewer zoom level, so that it doesn’t change when you switch between
inputs of different sizes. Right-click on the Viewer, and select Prevent Auto Zoom to toggle between
maintaining the same zoom level for all inputs (on) and changing it according to on-screen image
dimensions (off). Alternatively, you can also press Alt+Z on Viewer, or toggle prevent auto zoom in
the Viewer settings.
USER GUIDE
218
Compositing Viewers | Viewing Overscan in the Viewer
Viewing Overscan in the Viewer

If you want the Viewer to show overscan, that is any pixels extending beyond the left/right or
top/bottom of the frame, you can check the showoverscan box in the Viewer settings or check the
option in the Viewer right-click menu. Viewing pixels beyond the edges of the frame can be useful if
some of your nodes need to have access outside the frame. You can also control the amount of
overscan shown by specifying the pixel amount in the overscan field.
USER GUIDE
219
Using the Viewer Controls | Compositing Viewer Controls
Using the Viewer Controls

A Viewer’s on-screen controls let you navigate the timeline, display channels, zoom, choose cameras
(3D mode), and create display wipes and composites.
Compositing Viewer Controls

There are many useful tools at the top of the Viewer, some of which allow you to select layers and
channels, adjust gain and gamma, and zoom and scale down the image in the Viewer.
The tools at the bottom of the Viewer allow you to adjust the playback settings, including setting the
frame range, selecting the playback mode, and locking the Viewer playback range.
Drag the orange marker along the timeline to quickly cue to a specific frame. The number of the
current frame appears below the center of the timeline. You can also cue to a frame by typing its
number directly into this field.
Tip: The current frame and in an out point fields also accept simple mathematical functions,
such as +/-20 to jump forward or backward 20 frames.
USER GUIDE
220
Using the Viewer Controls | Compositing Viewer Controls
By default, Nuke automatically adjusts the timeline of every Viewer window to show the frame range
defined in your Project Settings. If no frame range is defined, the frame range of the first image you
read in is used as the global frame range.
Viewer timeline controls also have a frame range source dropdown menu that you can use to define
where the timeline gets its frame range from. You can set this menu to Global, Input, or Custom.
Global is the default setting described above.
The playback rate field (frames-per-second) displays the project’s playback speed. Nuke attempts to
maintain this speed throughout playback, although this adjusts depending on the resolution of the
imagery and your hardware configuration.
Note: The asterisk (*) denotes the playback speed selected using the Project Settings >
Root > fps control.
To have the Viewer adjust the timeline to show the “in” and “out” frames of the current input clip,
select Input from the frame range source dropdown menu. The number of the first frame in the clip
is shown in the left end of the timeline and the number of the last frame in the right end. If you
change the input of the Viewer, the frame range on the timeline is adjusted accordingly.
To manually adjust the frame range for the current Viewer window, pan and zoom on the timeline
until you see the desired frame range and Custom becomes selected in the frame range source
dropdown menu. Alt+drag to pan, and MMB+drag to zoom in. You can also zoom in on or out of the
timeline using the mouse wheel. To reset the zoom, press the middle mouse button over the
timeline.
To adjust the playback range for the current Viewer window, click+drag the red playback range
markers on the timeline to a new “in” and “out” frames as shown below, or enter a new playback
range in the playback range field and press the frame range lock button.
To toggle between the new playback range and the visible timeline range, click the lock frame button
again.
The fps field (frames-per-second) initially displays the project’s playback speed. Nuke attempts to
maintain this speed throughout playback, although this adjusts depending on the resolution of the
imagery and your hardware configuration.
The following table lists the functions of the playback buttons:
USER GUIDE
221
Using the Viewer Controls | Jumping to a Specific Frame
Buttons Functions
The Play backward and Play forward buttons play the sequence backward or
forward at the script’s frame rate. When you press a play buttons, it toggles to a
stop a button.
The Back 1 Frame and Forward 1 Frame buttons cue the sequence to the
previous or next frame.
The Previous keyframe and Next keyframe buttons cue the sequence to the
script’s previous or next keyframe.
The First frame and Last frame buttons cue the sequence to the first and last
frame.
The Frame Increment field allow you to specify the number of frames by
which the Previous increment/Next increment buttons cue the sequence. This
is set to 10 frames by default.
The Playback Mode button lets you control how many times and in what direction the Viewer plays
back the sequence. Click the button to toggle between the following modes:
Button Function
Repeatedly plays the sequence in a loop.
Repeatedly plays the image back and forth from head to tail.
Plays once through the section between in and out point and stops at the out
point. If these are not marked, then it plays from the beginning to the end of
the sequence.
Plays once from the beginning to the end of the sequence, ignoring any in and
out points.
Jumping to a Specific Frame

You can move quickly to a specific frame on the timeline by choosing Viewer > Go to frame (or by
pressing Alt+G), entering a frame number in the dialog that appears, and clicking OK.
USER GUIDE
222
Using the Viewer Controls | Synchronizing Viewer Playback
Synchronizing Viewer Playback

The 'Lock All Viewer Playback Ranges' button allows you to toggle synchronized playback of Viewer
windows. By default, all Viewers are locked - that is, if you cue to a frame in one Viewer, all other
Viewers follow suit.
When the lock icon changes from a closed lock to an open lock, that Viewer’s playback becomes
independent of other Viewers, and not cued to the other Viewers.
Pausing the Display

The 'Pause Viewer Update' button stops the Viewer from updating and holds the last frame rendered.
To reactivate display rendering for all frames, press the button again.
You can click the 'Refresh Viewer' button (or press U) to manually update the display while keeping
Viewer paused.
Displaying a Single Channel

You can press the R, G, B, and A keys on your keyboard to display the red, green, blue, and alpha
channels respectively. Or, you can also select a channel from the RGB dropdown menu in the top-left
corner.
USER GUIDE
223
Using the Viewer Controls | Layer and Channel Dropdown Menus
Press one of the channel keys again to toggle back and display all channels.
Tip: If you press Shift while selecting the channel, your selection only affects the currently
active input of the Viewer node. This way, you can display different channels from the
Viewer’s different inputs. For example, when keying it can be useful to view the RBG
channels from one input and the alpha channel from another, and toggle between the two.
To achieve this, do the following:
1. Create a Viewer with several inputs.
2. Activate one of the inputs by pressing its number (for example 1) on the Viewer.
3. Press Shift and select RGB from the channel dropdown menu.
4. Activate another input (for example, input 2) by pressing its number on the Viewer.
5. Press Shift and select A from the channel dropdown menu.
6. Toggle between the inputs by pressing their numbers or the up and down arrow keys.
Layer and Channel Dropdown Menus

The layer dropdown menu lets you choose a set of color channels to display in the Viewer. By default,
this is set to display the rgba layer, but you can choose any layer in the data stream.
USER GUIDE
224
Using the Viewer Controls | Displaying Alpha Mattes
The channel dropdown controls which channel appears when you view the “alpha” channel. The
default setting displays the alpha channel when you press the A key, but you can change this by
selecting any channel in the data stream, as shown below.
Displaying Alpha Mattes

When you’ve read in an image that has an alpha channel, you can display the alpha channel as a red
overlay on top of the image’s red, green, and blue channels.
To Display an Image's Alpha Channel on its RGB Channels
1. Select Image > Read to read in an image.

2. Connect a Viewer node to the Read node.
By default, Nuke displays the red, green, and blue channels in the Viewer.
USER GUIDE
225
Using the Viewer Controls | Viewer Info Bar
3. Click on the Viewer to make sure it’s the currently active panel.
4. Press M.
Nuke displays the image’s alpha channel as a red overlay on top of the RGB channels.
5. To return to the RGB display, press M again.
Viewer Info Bar

The info bar displays information about the image format, bounding box, and pixel underlying the
pointer, a sampled pixel or region of pixels. From left to right, the indicator displays the following
about the current image and pixel or sample: image resolution and the size of the bounding box, the
pixel's x and y position and its Red, Green, Blue, and Alpha values, and other values depending on the
color type you have selected from the color type menu on the right.
Color Type Description Values for RGB 0.1, 0.1,

0.1, 1.0 (Dark Gray)
Spotmeter Colorspace values measured in f-stops at 1/48th of a f/5.9 @48th sec
USER GUIDE
226
Using the Viewer Controls | Adjust Display Gain and Gamma
Color Type Description Values for RGB 0.1, 0.1,

0.1, 1.0 (Dark Gray)
second shutter speed.
8-bit RGB colorspace values. 25, 25, 25, 255
8-bit Hex The same colorspace values as 8-bit, but displayed in 1A 1A 1A FF

hexadecimal.
log Logarithmic colorspace values. 397, 397, 397
HSVL The default setting, shows colorspace values as Hue, 0, 0.00, 0.10, 0.10000
Saturation, Value, and Level.
You can sample a single pixel from the Viewer by pressing Ctrl/Cmd while clicking, a region from the
Viewer by pressing Ctrl/Cmd+Shift while dragging, a single pixel from the node’s input by pressing
Ctrl/Cmd+Alt while clicking, and a region from the node’s input by pressing Ctrl/Cmd+Alt+Shift while
dragging.
To cancel a pixel selection, hold Ctrl/Cmd and right-click in the Viewer.
Adjust Display Gain and Gamma

The gain and gamma sliders let you adjust the displayed image, without affecting your final output.
These controls are useful for tasks like spotting holes in mattes. You can boost or reduce gain by
entering a multiplier (exposure value), dragging on the slider, or using the F-Stop arrows. Boost or
reduce gamma by entering a gamma level or dragging the gamma slider.
The gain and gamma toggle buttons let you switch between the default values of 1 (no change) and
the last gain and gamma adjustments you made in the Viewer.
Press the 'Clipping Warning' button to apply stripes to all pixels outside the range 0.0 to 1.0.
USER GUIDE
227
Using the Viewer Controls | 2D / 3D Toggle and Camera Controls
2D / 3D Toggle and Camera Controls

The 2D / 3D dropdown menu lets you toggle between 2D and 3D display modes in the current Viewer.
This menu also lets you choose between different orthographic (non-perspective) views when working
in the 3D mode.
The camera dropdown menu on the right lets you choose which camera to look through when
multiple cameras exist in your 3D scene. For more information on these controls, see 3D
Compositing.
Selection Modes
Nuke features three selection modes, Rectangle, Ellipse, and Lasso. Selection modes work the same
in all Viewer contexts, whether you're in 2D selecting Roto splines or 3D selecting vertices, faces, or
objects. The selection mode dropdown is located above the Viewer at the top-right.
USER GUIDE
228
Using the Viewer Controls | 2D / 3D Toggle and Camera Controls
Rectangle Ellipse Lasso
Each mode uses the same selection modifiers:

• Shift - additive selection. Holding Shift and selecting vertices, faces, or objects adds the selection to
any existing selection.
• Alt + Shift - subtractive selection. Holding Alt + Shift and selecting vertices, faces, or objects
removes the selection from any existing selection.
USER GUIDE
229
Using the Viewer Controls | Monitor Output Toggle
Monitor Output Toggle

The monitor output button – which automatically appears when you connect an external monitor –
allows you to preview the current Viewer image on an external broadcast video monitor.
Note: For more information, see Previewing on an External Broadcast Video Monitor.
USER GUIDE
230
Controlling Zoom and Resolution | Using the Zoom Menu
Controlling Zoom and Resolution

Nuke includes a dedicated set of controls to determine what is rendered in the Viewer and at what
resolution, allowing you to work on specific areas of the image at the required quality level while
keeping the interface responsive.
Using the Zoom Menu

The Zoom dropdown menu lets you select the magnification factor by which the current image is
displayed. This menu also shows the keyboard shortcuts to press to quickly switch between the
different zoom settings.
Proxy Mode
Nuke can generate low-res proxies for displayed frames as needed. Press Ctrl/Cmd+P or click the
proxy mode toggle button on the Viewer to activate the proxy display mode.
By default, the proxy scale is set to 0.5. You can change the proxy scale in the Projects Settings,
which you can open by selecting Edit > Project Settings (or pressing S).
USER GUIDE
231
Controlling Zoom and Resolution | Lowering the Display Resolution of Individual Viewers
Proxy display resolution defined on the Project Settings

properties panel.
You can also read in rendered proxies using the Read nodes’ controls. The proxy file does not need to
match the proxy resolution in use. Depending on your settings, either the full-res or proxy file is
scaled to the required proxy size. For more information, see Read Nodes and Proxy Files.
Lowering the Display Resolution of Individual Viewers

Viewers also have a dropdown menu that allows you to easily switch to lower display resolutions,
regardless of whether you have activated proxy mode or not. Using this multiplier setting, you can,
for example, change the display resolution of an individual Viewer to 50% of the current (be it full-size
or proxy) resolution. This is useful if you want to have Nuke display your images more quickly without
having to touch the project settings. It also comes in handy if you have just a few very large plates in
your script, as you can choose to use lower resolutions when viewing just these plates.
To Lower the Display Resolution of Individual Viewers:
From the Viewer’s down-rez dropdown menu, choose the factor by which you want to lower the
display resolution.
For example, if you have a 4K plate and are using a proxy scale of 1:2, your plate is still 2K even in the
proxy mode. Setting the down-rez factor to 1:2 in the Viewer scales the plate down further to 50% of
the proxy resolution, that is to 1K. This gives you much faster (but less accurate) feedback.
Pixel Aspect Ratio

The pixel aspect ratio determines whether your images are displayed using square or rectangular
pixels. By default, the Viewer uses the pixel aspect ratio defined in your project settings. To see the
current setting, select Edit > Project Settings (or press S).
USER GUIDE
232
Controlling Zoom and Resolution | Full-frame processing
For example, a pixel aspect ratio of 2 accurately displays anamorphic footage the way it is projected,
as shown in the image:
The Viewer uses the pixel aspect ratio

defined for the script.
If you want to ignore the pixel aspect ratio, you can toggle it by pressing Ctrl/Cmd+Shift+P over the
Viewer window.
Press Ctrl/Cmd+Shift+P over the Viewer

window to ignore the pixel aspect ratio.
Full-frame processing
By default, when you display or play through a sequence in the Viewer, Nuke only calculates and
caches the scanlines for the visible area. It's not caching the full frame, but a subset of the scanlines
you're viewing. For example, if you have set the zoom level to be ÷4, Nuke is only caching 1/4 of the
scanlines that make up the frame. In a lot of cases, this allows for faster playback and may be what
you want.
However, if you play through a sequence and then pan around or change the zoom level, Nuke has to
calculate the scanlines it didn’t calculate before. This may not be what you want if you are performing
tasks that require you to constantly pan around or zoom in and out of the plate (such as paint and
roto). If this is the case, you can click on the full frame processing button in the controls at the top
of the Viewer to force Nuke to process every scanline in the image. Compared to the default mode,
USER GUIDE
233
Controlling Zoom and Resolution | Region of Interest (ROI)
this may take slightly longer initially and requires more space in the Viewer cache, but once Nuke has
cached the frames you require, you can pan around and change the zoom level without breaking the
cache or affecting playback.
Region of Interest (ROI)

The ROI button lets you enable rendering only through a region of interest - a portion of the image
you explicitly select. This is useful for quickly viewing render results in a process-heavy script.
To Define a Region of Interest:
1. Click on the ROI button in the Viewer controls. The ROI overlay appears.
2. Drag to resize and move the ROI overlay as necessary.
To Free-Draw a Region of Interest
1. Over the Viewer, press Alt+W once (do not hold the keys down). The ROI button turns red, but the
ROI overlay does not appear. This allows you to freely draw your own ROI rather than adjust the
default overlay.
2. Drag a marquee to draw the region of interest where you need it.
USER GUIDE
234
Controlling Zoom and Resolution | Region of Interest (ROI)
To Clear a Region of Interest:
1. After you’ve set a region of interest, you can clear it by pressing Alt+W over the Viewer. You can
then drag a new marquee to define a new region of interest.
2. To turn off the feature and update the whole Viewer with the recent changes, click the ROI button
again (or press Shift+W).
USER GUIDE
235
Viewer Overlays and Input Processes | Guides and Masks
Viewer Overlays and Input

Processes
Nuke's Viewer overlays and render modes help you to position elements correctly and compare shots
in the Node Graph. You can also use Input Process and Viewer Process to modify the image from
the viewed node before it is displayed on your monitor.
Guides and Masks

The Viewer guides and masks assist with placing effects and text within the current format. For
example, text placed with the Title Safe guide is visible to the audience. Use the Guideline and Mask
dropdown menus to select the required overlays.
Note: Guides and masks are not applied at render, they are simply Viewer overlays.
Guidelines are intended to highlight areas of the current format that won’t appear in the final render.
By default, no guidelines are selected. You can select any of the following guideline options:
• title safe - any text intended for the audience should reside within this zone.
USER GUIDE
236
• action safe - any visual elements intended for the audience should reside within this zone.
• format center - overlays a crosshair in the center of the format area.
USER GUIDE
237
• format - any formatting changes must be applied to the area outlined in red.
Tip: The above guideline options also exist in the Viewer Settings. Press S on the Viewer to
display its settings and adjust the safe zone and format center controls.
Masks can be used to simulate a particular format, for example, 4:3 or 16:9. You can also choose the
mask overlay type.
• none - no masking is applied to the Viewer. This is the default state.
USER GUIDE
238
• lines - any mask applied is highlighted using a pair of lines in the Viewer.
• half - any mask applied is highlighted using semi-transparent shading.
• full - any mask applied is highlighted using black shading.
• blanking ratio - select the masking ratio applied to the Viewer, for example, 4:3 or16:9.
Tip: The above mask options also exist in the Viewer Settings. Press S on the Viewer to
display its settings and adjust the mask region outside ratio and mask mode controls.
USER GUIDE
239
Viewer Overlays and Input Processes | Adding Custom Guides and Masks
Adding Custom Guides and Masks

You can add custom guides and masks to the Viewer dropdowns by creating a file called custom_
guides.py and placing it in your .nuke folder.
Tip: For information on locating your .nuke folder by platform, see Loading Gizmos, NDK
Plug-ins, and Python and Tcl Scripts.
The guides and masks that ship with the application are kept in the Nuke bundle, here:
<installation_directory>/pythonextensions/site-packages/custom_guides.py
Copy the custom_guides.py file to your .nuke folder and then add the guides and masks you require
to the existing Python code inside the .py to append them to the Viewer dropdowns. For example:
guides.Guide("myGuide", 0.75, 1, 0.8, 0.3, guides.kGuideMasked)
OR
guides.MaskGuide("5:3",5.0/3.0)
Using the Viewer Composite Display Modes

The wipe control provides an option for displaying a split-screen of two images, which can help you
compare before and after versions for color correction, filtering, and other image manipulation. This
control also includes display compositing options to overlay different images.
To Display a Comparison Wipe:
1. Select a node in your script and press 1 to display its output in the Viewer.
2. Select the node you want to compare and press 2.
The 2 keystroke connects the image to the Viewer (assigning the next available connection,
number 2).
3. From the A and B dropdown menus on top of the Viewer, select the images you want to compare.
The menus display a list of nodes most recently connected to the Viewer.
4. From the Viewer composite dropdown menu in the middle, select wipe.
USER GUIDE
240
Viewer Overlays and Input Processes | Using the Viewer Composite Display Modes
The two images are displayed split-screen in the Viewer. You can view their details in the A and B
information bars at the bottom of the Viewer.
5. Drag the handles of the crosshair to adjust the wipe:

• Drag the crosshair center to change its position.
• Drag the long handle (on the right) to rotate the wipe.
• Drag the “arc” handle to cross-dissolve the second image.
6. When finished with the split-screen, select none (-) from the Viewer composite dropdown menu.
Tip: If you press Shift while selecting a channel, your selection only affects the currently
active input of the Viewer node. This way, you can display different channels from the
USER GUIDE
241
Viewer Overlays and Input Processes | Input Process and Viewer Process Controls
Viewer’s different inputs. For example, when keying it can be useful to view the RBG
channels from one input and the alpha channel from another, and toggle between the two.
The display composite options - over, under, and minus - can also be selected to overlay two
images. When the two images are 2D, this allows you to create a quick comp.
When one image is 2D and the other is a 3D node, you can use under to line up the wireframe
preview with the 2D reference, and see how the 3D matches prior to a full render.
One example of this is when you want to preview a wireframe 3D scene with a background plate that
you are trying to match, as shown below. For more information, see the 3D Compositing chapter.
Special thanks to for use of the above footage.
Input Process and Viewer Process Controls

Input Process and Viewer Process operations can be used to modify the image from the viewed
node before it is displayed on your monitor. Both only affect the Viewer in which they are activated
and do not affect your rendered output. Input Process is a legacy system which uses a node
instantiated in the Node Graph to process the image. This is handy for script-specific, temporary, or
experimental use, but can be error prone due to the node accidentally being deleted or changed and
is limited to a single node. The Viewer Process system allows a gizmo (or compiled node) to be
registered from the Python programming language at start-up. The registered item appears in a
USER GUIDE
242
dropdown menu in the Viewer and the node is instantiated internally within the Viewer when the item
is selected so there is no danger of accidental deletion or modification. This also enables multiple
Viewer Processes to be registered at different points of start-up (as Nuke works through the NUKE_
PATH menu.py files).
The Viewer settings contain an option for the Input Process to be applied before or after the Viewer
Process, so the two may be used in conjunction, for instance, with the Input Process applying a
projection mask after the Viewer Process applies a film look profile. While you could combine the two
into a single Viewer Process node, it can be advantageous to keep operations separated. Having both
the Viewer Process and Input Process available provides a great deal of flexibility.
You can create an Input Process by creating a node in the Node Graph and naming it as an Input
Process using Nuke’s Edit menu. Once an Input Process has been named, the IP button appears in the
Viewer controls. When the IP button is activated, any image you view is passed through the Input
Process.
Unlike Input Processes, Viewer Processes are registered using Python. They can be session
independent and always appear in the Viewer’s Viewer Process dropdown menu. There are two
predefined Viewer Processes, sRGB and rec709, but you can also build and add your own. When a
Viewer Process is selected from the Viewer Process dropdown menu, any image you view is passed
through that Viewer Process.
Whenever possible, the Input Process and Viewer Process are executed on the GPU. 1D LUT and 3D
LUT (Vectorfield) have GPU implementations, so the built-in Viewer Processes run on the GPU (unless
gl buffer depth has been set to byte in the Viewer settings, in which case all processing is done on
the CPU). To get the GPU'd versions of the nodes for use in a custom Viewer Process gizmo, press x
over the Node Graph, enter ViewerGain, ViewerGamma, or ViewerClipTest in the command entry
window, and press Return.
The following table lists the differences between an Input Process and a Viewer Process.
Input Process Viewer Process
Set by selecting the node in the Node Graph and Registered using Python.
choosing Edit > Node > Use as Input Process.
Activated using the IP button in the Viewer Activated using the Viewer Process dropdown
controls. menu in the Viewer controls.
Requires that the node exists in the Node Graph. Is defined in a text file called menu.py that is
Can quickly and easily be modified by artists. Can run at start-up. Accessible for artists, but not
also be accidentally deleted, disabling the effect. likely to be accidentally modified or deleted.
USER GUIDE
243
Input Process Viewer Process
Script dependent. Unless your Input Process node Session independent. The Viewer Processes
is saved in the template.nk file that is loaded at registered in menu.py are always available in
start-up, the Input Process is lost when you each new session of Nuke.
restart Nuke.
There can only be one Input Process at a time. There can be an unlimited number of Viewer
Setting a new Input Process overrides any Processes available in the Viewer Process
previously used Input Process. dropdown menu. For example, it is possible to
register Viewer Processes in any menu.py file
at start-up, so Viewer Processes can be added
at any directory in your NUKE_PATH.
Useful for temporary or non-critical viewing Useful for viewing options that you often need
options that you want in the current shot for or that should not be modified by artists on a
convenience, or for testing Viewer Processes shot-by-shot basis.
before registering them. Can also be used for
other things, such as field charts or masks that
may be switched on or off and changed around in
the shot.
Note: Note that Input Processes and Viewer Processes are part of a built-in, fixed pipeline
of nodes that are applied to images before they are displayed in the Viewer. This pipeline is
either:
gain > Input Process > Viewer Process > gamma > dither > channels > cliptest (if input
process order has been set to before viewer process in the Viewer settings)
OR
gain > Viewer Process > Input Process > gamma > dither > channels > cliptest (if input
process order has been set to after viewer process in the Viewer settings).
However, depending on what the Input Process and Viewer Process are doing, the order in
the built-in pipeline may not be the correct order. Therefore, if your Input Process or Viewer
Process have controls that also exist for the Viewer, such as float controls named gain,
gamma, or cliptest, then the Viewer drives them from the corresponding Viewer controls
and does not do that image processing itself. This allows you to implement these controls in
your Input Process or Viewer Process node/gizmo using whatever nodes and order you
want. If your Input Process and Viewer Process do not have these controls, then the Viewer
applies the effects in its normal way according to the built-in pipeline.
In the built-in pipeline, dither is applied to diffuse round-off errors in conversion of floating
USER GUIDE
244
point data to the actual display bit depth. Although the cliptest is drawn at the end, it is
computed on the image as input to the Viewer.
Note: By default, the predefined Viewer Processes, sRGB and rec709, affect all channels.
However, if you want them to only affect the red, green, and blue channels, you can activate
apply LUT to color channels only in the individual Viewer Settings or on the Viewers tab of
the Preferences.
Input Process Controls
To activate or deactivate the effect of an Input Process, click the IP button in the Viewer controls. Note
that the IP button only appears if the input process field in the Viewer settings is not empty. The
button is also only enabled when a node in the Node Graph is set as an Input Process.
To open the Viewer settings, press S on the Viewer, or select Viewer Settings from the Viewer’s right-
click menu. By default, input process is set to VIEWER_INPUT. If a node called VIEWER_INPUT exists
in the Node Graph, it is automatically used as the input process for the Viewer. This ensures
backwards compatibility with pre-5.2 scripts.
However, the Input Process node does not have to be named VIEWER_INPUT. You can use any node
as an Input Process. Do the following:
1. Select the node in the Node Graph and choose Edit > Node > Use as Input Process.
Alternatively, you can press S on the Viewer to open the Viewer settings and enter the name of the
node in the input process field.
2. In the Viewer settings, you can also define whether the Input Process is applied before or after the
Viewer Process currently in use. To do so, set input process order either to before viewer
process or after viewer process.
The Input Process node should not be connected to other nodes in the Node Graph. If you attempt to
connect it, an error is displayed in the Viewer. If you delete the Input Process node from the Node
Graph, the effect of the Input Process is disabled.
Viewer Process Controls
To activate a Viewer Process, select it from the Viewer Process dropdown menu in the top right
corner of the Viewer. Any images you now view using this Viewer are passed through the selected
USER GUIDE
245
Viewer Process.
Nuke includes the following predefined Viewer Process gizmos: sRGB, rec709, and rec1886. By
default, sRGB is used because it is a best-guess default for a typical computer monitor.
In addition to using the predefined Viewer Processes, you can also add your own by registering a
node or gizmo as a Viewer Process. You can register as many Viewer Processes with custom Viewer
LUTs as you like. For more information on creating and registering custom Viewer Processes, see
Creating Custom Viewer Processes.
All available Viewer Processes (both predefined and custom ones) appear in the Viewer Process
dropdown menu in the Viewer controls. To disable the use of a Viewer Process, select None from the
Viewer Process dropdown menu.
To open the properties panel of the currently active Viewer Process, select show panel from the
Viewer Process dropdown menu.
Note that if the control you want to adjust has the same name as any of the Viewer controls (for
example, gain or gamma), you should adjust the control on the Viewer. This drives the control in the
following:
• the Input Process’ controls if an Input Process is in use
• in the Viewer Process’ controls if no Input Process is in use.
Tip: If you want to render out a file with the Viewer Process effect baked in, you can select
Edit > Node > Copy Viewer Process to Node Graph to create an instance of the Viewer
Process node in the Node Graph.
USER GUIDE
246
Using the File Browser | Input Process and Viewer Process Controls
Using the File Browser

Whenever you load or save files in Nuke, you’ll see a browser similar to the one shown in the image
below. The directory navigation buttons let you create or access the directory from which you wish to
read or write data.
The navigation controls let you move through the directory structure, bookmark favorite directories,
and create new directory folders.
Windows only: You can show/hide the drives that Windows auto creates by right-clicking the required
drive, selecting Show Defaults, and checking or unchecking the drive.
USER GUIDE
247
Using the File Browser | To Use the Navigation Controls
To Use the Navigation Controls

• Click the Create New Directory button to create a new directory at your current position in the
file hierarchy.
• Click Up one directory to go up one directory closer to the root.
• Click Previous directory to go back one directory.
• Click Next directory to go forward one directory.

• Click Home to access the directory defined as your local working directory.
• Click Root to ascend to the very top of your local drive or server’s file hierarchy.
• Click Nuke to access the directory you (or your system administrator) defined as your network
working directory.
• Click the + button to add a directory bookmark.
• Click the edit button to edit the name or path name to a bookmark.
• Click the - button to remove a directory bookmark.
Path Name Field

The path name field displays the current directory path, lets you navigate to a new path, and also
enter a file name for scripts and rendered images.
To Use the Path Name Field:

1. To navigate to a directory, type the path name in the field.
2. To enter a script name, browse to a directory path and enter the file name after the displayed
path.
To limit the file list to specific file types, use the filter dropdown menu and Sequences checkbox.
To Use the Filter Dropdown Menu and Sequences

Checkbox
• Select *.nk to display only Nuke script files.
USER GUIDE
248
Using the File Browser | To Preview Files in the File Browser
• Select * to display all files (except hidden files), regardless of whether they’re associated with Nuke.
• Select .* * to display all files, including hidden files.
• Select */ to display directory names, but not their contents.
• Check Sequences to display image sequences as single titles, as in fgelement.####.cin 1-50 rather
than fgelement.0001.cin, fgelement.0002.cin, fgelement. 0003.cin, and so on.
Note: File sequences with no file extension (for example, fgelement.0001, fgelement.0002,
fgelement.0003, and so on) are not displayed as single titles the first time you view the
directory in the File Browser. However, they are displayed as single titles once you have
navigated to another directory and back again.
Note: By default, Nuke may not be able to display custom file extensions (for example,
.cext) as single titles. To fix this, you can register your custom file extension as a sequence
type using Python:
1. Create a file called init.py in your plug-in path directory if one doesn’t already exist. For
more information on plug-in path directories, see Loading Gizmos, NDK Plug-ins, and
Python and Tcl Scripts.
2. Open the init.py file in a text editor and add an entry in the following format (replacing
cext with your custom file extension):
nuke.addSequenceFileExtension("cext")
• You can also split incomplete sequences into separate Read nodes using the split seq checkbox.
To Preview Files in the File Browser

1. Click the black arrow in the top right corner of the file browser.
The file browser expands to include a small viewer.

2. Select the file you want to preview. Nuke displays the file in the file browser.
USER GUIDE
249
Using the File Browser | To Select Multiple Files with the File Browser
To Select Multiple Files with the File Browser

1. Browse to the folder where the files are located.
2. Ctrl+click on all the files you want to open to select them (Mac users Cmd+click).
3. You can open files from multiple directories by clicking Next and browsing to the next file
location.
As you browse around, files that you have previously selected appear highlighted in the browser.
4. Click Open.
Nuke opens all the files you selected.
To Open Incomplete Sequences in Separate Read Nodes

1. Locate an incomplete sequence.
2. Check split seq and select all the resultant files.
3. Click Open to open the sequence in separate Read nodes.
USER GUIDE
250
Undoing and Redoing | To Open Incomplete Sequences in Separate Read Nodes
Undoing and Redoing

Nuke generally gives you an undo history that extends back to the first action of the application’s
current session.
• To undo an action in the workspace, select Edit > Undo (or press Ctrl/Cmd+Z). Repeat as necessary.
• To redo an action in the workspace, select Edit > Redo (or press Ctrl/Cmd+Y). Repeat as necessary.
• To undo a change in a properties panel, click the Undo arrow button in the properties panel.
• To redo a change in a properties panel, click the Redo arrow button in the properties panel.
• To undo all changes made after the properties panel was opened, click the Revert button .
OR
Right-click on the properties panel and select Revert knobs.

• To set all controls back to their default values, right-click on the properties panel and select Set
knobs to default.
USER GUIDE
251
Progress Bars | To Open Incomplete Sequences in Separate Read Nodes
Progress Bars
Nuke displays a progress bar for each active task it performs. By default, progress bars appear in a
pop-up dialog, but you can also display them in the Progress panel. To do so, click on a content menu
button and select Windows > Progress. This opens a Progress panel. The next time you get a
progress bar, it appears in the panel. If you delete the panel, progress bars appear in a pop-up dialog
again.
If you want to have the Progress panel appear in the same position in the future, you can save it as
part of a workspace. For more information, see Customizing Workspaces.
USER GUIDE
252
Handling Errors | To Open Incomplete Sequences in Separate Read Nodes
Handling Errors
Sometimes things may not go as you planned and you might face an error message in Nuke. When
this happens, an error alert displays in the Viewer and on the node that has a problem in the Node
Graph.
You can choose to view the error message itself in the Error Console tab next to the Properties pane.
If you can’t see the Error Console, click the content menu button and select Windows > Error
Console to display it. If you have an error in a Read or a Write node, or you are missing a plug-in, the
error message also displays in a pop-up window.
If you see an error alert on your node or in the Viewer, you can click the Error Console tab to open it
and view the error message.
In the Error Console error list, you can double-click on a message and, if possible, Nuke takes you to
the control panel of the node that’s in error. This isn’t always possible because of the nature of the
error. You can also click the clear output button on the Error Console to clear all error
messages on the tab.
Note: If you have a node in your Node Graph that is producing an error, but it’s not
connected to the node tree, Nuke won’t show a pop-up error message for the node and you
can still view the resulting image in the Viewer, if you have one connected. This enables
working without having to stop to close error messages in the event that you have erroring
nodes in the script that aren’t connected to your node tree.
USER GUIDE
253
Using Nuke Studio's Timeline
Environment
This chapter is designed to help you learn how to use Nuke Studio's Timeline environment, including
ingesting media and using the Timeline Viewer.
Achieving Real-time Playback

The following is a list of recommended hardware configurations for Windows and Linux that Foundry
have certified for real-time playback in the Nuke Studio and Hiero timeline Viewer (see the note
below). Nuke's compositing Viewer is not designed for real-time playback. Please note that playback
may also work on other machine configurations, but those listed below have been tested.
Note: For realtime playback, clips and sequences should be 16-bit RGBA DPX files, localized
to fast storage. See Localizing Media for more information.
In the case of complex or multi-format timelines, files must be cached to the fast storage
using the Timeline Disk Caching. See Caching Frames in the Disk Cache for more
information.
Windows and Linux Certified Hardware

Nuke Studio is currently tested on the HP Z workstation series. The following are the tested
components for an HP Z840.
Processor 2 x Intel Xeon E5-2687Wv3 CPU
RAM 128 GB DDR4-2133 (8x16 GB) 2 CPU Registered RAM
Internal system drives 1 x HP Z Turbo Drive G2 256 GB PCIe
GPU NVIDIA Quadro M6000 12 GB
Video monitor output AJA Kona 4K or BMD DeckLink 4K Extreme 12G
Real-time playback is currently tested with the following internal storage accessories:
USER GUIDE
254
Using Nuke Studio's Timeline Environment |
Drives 8 x Samsung SM863 960 GB
Dock 2 x HP B8K60AA - 4 Bay SSD carrier
Raid card LSI Megaraid ROC 9270-8i
Raid cable 2 x Avago 05-26113-00 0.6 Meter internal cable SFF8643 To X4 SATA
HDD (MINI SAS HD)
Raid type RAID 5, Always write-through, with a 256k Stripe size
The table below shows guidelines for the data rates required for real-time playback of single track,
16-bit RGBA DPX sequences, with no soft effects in Nuke Studio's timeline Viewer. The Data rate is
the required disk speed needed to achieve real-time streaming playback, for a given format, at a
given playback rate.
Note: The sample rates are taken from the AJA System Test application, with the Settings >
Disk Test set to File per frame, and should be used as guidelines only.
The AJA System Test application is available for download for Mac and Windows from
https://www.aja.com/products/aja-system-test
Frames per second (fps) Format Data rate MB/s
25 fps 4K Film (4096x3112) 2431.25
4K UHD (3840x2160) 1582.03
2K Film (2048x1556) 607.81
HD 1080p (1920x1080) 395.51
30 fps 4K Film (4096x3112) 2917.50
4K UHD (3840x2160) 1898.44
2K Film (2048x1556) 729.38
HD 1080p (1920x1080) 474.61
USER GUIDE
255
Using Nuke Studio's Timeline Environment |
Frames per second (fps) Format Data rate MB/s
60 fps 4K Film (4096x3112) 5835.00
4K UHD (3840x2160) 3796.88
2K Film (2048x1556) 1458.75
HD 1080p (1920x1080) 949.22
USER GUIDE
256
About Clips and Shots | Source Clips
About Clips and Shots

The interface sorts your clips into three broad categories: Audio and Video, Audio Only, and Video
Only. Clips are displayed differently depending on their content, location, and in the case of the
Viewer, the current mode.
Note: The timeline Viewer currently treats all alpha channels as premultiplied, which can
result in the Viewer background being "added" to the image. If you’re working with un-
premultiplied images, set the Viewer background to Black.
Note: There is currently no QuickTime audio support on Linux. Support for audio is
scheduled for a later release.
Source Clips
Source clips are representations of files on disk. Shots on the timeline refer to source clips, so
making changes to a clip in a bin affects all shots that refer to that source clip.
Note: The colored bars under the thumbnail represent the layers available in the clip, in this
case color. Other layers include alpha, depth, and motion, similar to Nuke.
USER GUIDE
257
About Clips and Shots | Shots
Shots
Shots are representations of source clips, they are not saved on disk. Making changes to a shot only
affects that instance. See Managing Timelines for more information.
A Source Clip Opened in the Timeline View

Opening a source clip in the timeline view allows you to control its output and add soft effects to the
source clip. Clips opened as timelines always contain all available frames, but you can adjust their
output using the original range controls in the clip's properties or in and out points. See Using In
and Out Markers and Soft Effects for more information.
USER GUIDE
258
Clip and Shot Properties | A Source Clip Opened in the Timeline View
Clip and Shot Properties

The Properties contain standard controls similar to Nuke's Read node Properties and format-
specific controls, depending on the selected source clip or shot. You can also control localization for
individual clips and shots using the localization policy dropdown.
See Localizing Media for more information on localization.
If the clip Properties are not displayed in your workspace, double-click a source clip or shot or
navigate to Window > Properties to open them in a floating window.
Note: The Properties panel allows you to override the Preferences and Project Settings
on a per file basis. See Appendix A: Preferences and Timeline Environment Project Settings
As an example, .mov files allow you to control the decoder and ycbcr Matrix for clips, as well as the
standard controls available for all clips. R3D and ARRIRAW media use their own software
development kits (SDKs) to control the extensive settings usually seen on RED and ARRIRAW Cameras.
Note: The RED Decode Resolution and ARRIRAW Resolution and Proxy dropdowns
control the maximum allowed Viewer resolution, overriding the Viewer Image Quality
setting.
USER GUIDE
259
Clip and Shot Properties | Setting Source Clip Ranges
In addition, source clip properties are accessible through the same Python API as Nuke, improving
scripting capabilities and integration into existing pipelines.
You can also add OCIO context Viewer processes to the Viewer color transform dropdown in the
same way as regular Nuke Read and OCIODisplay nodes. See Changing the Viewer Color Space for
more information.
Setting Source Clip Ranges

The original range controls in the source clip/shot Properties panel allows you to set the output
range for source clips when you add them to a sequence as shots. The entire frame range of the clip
is read on import, but altering the original range changes the output of the shot on the timeline.
Note: You can reset source clip to its original frame range at any time by clicking rescan.
Rescanning also appends new frames as they become available.
Changing the original range affects source clips and shots differently. Opening a source clip in the
Viewer, or in the right-click Timeline View, and adjusting the original range limits or extends the clip
to that range of frames.
Extending the range past the original range adds handles, shown in red.
USER GUIDE
260
Clip and Shot Properties | Using Relative Paths to Media
Adjusting the original range changes the output of any existing shots, but still allows you to trim and
slip the shot range using the available handles. New shots pick up the new original range specified in
the Properties.
Using Relative Paths to Media

Nuke Studio's timeline Read nodes support relative paths in the file control, allowing you to import
media relative to a specific project directory. The Project Settings control this behavior for the
current project.
USER GUIDE
261
Clip and Shot Properties | Using Relative Paths to Media
Tip: If you want to set the project path for all new projects, navigate to the Preferences
under Project Defaults > General and set the project directory there.
To set a Read file path to a relative location:

1. Open the Project Settings by clicking Project > Edit Settings.
2. Enter the required file path in the Project Directory field.
Tip: Click Hrox Directory to automatically enter an expression that evaluates to the .hrox
location.
3. Double-click the required source clip or shot to display its Properties.

4. In the file field, adjust the path to the relative location of the .hrox project file. For example, if the
project directory and absolute path to the source clip are:
C:/projects/timeline_read/scripts/
C:/projects/timeline_read/footage/renders/shot110/A009C005_101029_
L1O3.%04d.dpx
Then the relative path to the clip is:
./../footage/renders/shot110/A009C005_101029_L1O3.%04d.dpx
USER GUIDE
262
Ingesting Media | Using Relative Paths to Media
Ingesting Media
Adding media is as simple as drag-and-drop from a file browser or selecting File > Import File(s) or
Import Folder(s). The application imports your media into the bin view providing you with a
thumbnail of all of your clips and preserving the original folder and file hierarchy.
The media is soft imported, creating symbolic links to locations on disk. See Using the Copy Exporter
for information on how to quickly consolidate your media and projects, or Localizing Media to help
stabilize playback.
Note: Projects containing large amounts of movie files (for example .r3d and .mov) may
exceed the number of available file handles per process, causing problems opening new
files or projects and exporting.
You can increase the default limit of 1024 by entering the following command from the
terminal, then running the application from the same session:
ulimit -Sn 2048
Clips with no inherent frame rate information are assigned a frame rate at ingest as specified in the
Preferences.
1. Open the Preferences dialog by pressing Shift+S.
2. Select Behaviors > Timecode from the sub-menu on the left.
3. Use the RED file timecode dropdown to determine R3D clip behavior:
• Default from File - use the default set by the R3D file in question.
• Absolute Timecode - force the use of the Absolute Timecode as specified in the clip metadata.
USER GUIDE
263
Ingesting Media | Using Relative Paths to Media
• Edge Timecode - force the use of the Edge Timecode as specified in the clip metadata.
4. Use the other media timecode dropdown to determine clip behavior for all other clips:
• File Header - the file metadata header is used to derive the timecode, if it exists. This option
defaults to Frame Number if the header is missing.
• Frame Number - ignores the metadata header, even when present, and derives the timecode
from the frames in the clip.
5. Set the max valid timebase allowed from the image header, above which the value is clamped.
Image files are often created with application specific timebase values in the header description.
This can lead to reading in spuriously high frame rates, and the clamp aims to prevent this from
happening.
If your clips do have extremely high frame rates, increase this value as necessary to avoid
clamping.
6. Enable or disable EDL style spreadsheet timecodes:
• When disabled, the srcOut and dstOut values use the film convention, representing the last
frame of the cut.
• When enabled, the srcOut and dstOut values use the video convention, representing the frame
directly after the cut.
7. Click OK to save your settings.
USER GUIDE
264
Using Drag-and-Drop | Using Relative Paths to Media
Using Drag-and-Drop
Locate your media in a file browser and drag the frame range, clip, folder, or folders into the Project
tab.
Ingest behavior depends on the target:

• Dragging a folder into the Project tab automatically ingests all the contents of the folder, including
other folders and their contents.
• Dragging a movie file, such as a .mov or .r3d, automatically ingests the entire clip.
• Dragging a single file or file range, that is part of an image sequence, is controlled by the
Preferences > Behaviors > File Handling > Scan for file sequence range checkbox:
• Enabled - the default setting, dragging a single file or file range, that is part of an image
sequence, creates a clip in the bin view containing all available frames.
For example, dragging frames 1-5 and 11-20 ingests the entire frame range.
• Disabled - only the dragged frame or range is imported into the bin.
For example, dragging frames 1-5 and 11-20 ingests two distinct clips, one containing 5 frames
and one containing 10 frames.
USER GUIDE
265
Using the File Browser | Using Relative Paths to Media
Using the File Browser

If you prefer to work with menus, you can also import clips using the file browser. You can import
individual clip files, ranges, or entire folders, depending on the amount of media you intend to use.
Whenever you load or save files, a browser similar to the one shown below is displayed. The directory
navigation buttons let you create or access the directory from which you wish to read or write data.
The navigation controls let you move through the directory structure, bookmark favorite directories,
and create new directory folders.
Note: If you import folders, use the Import Options dialog to filter your ingest using
inclusion and exclusion parameters, separated by spaces. The dialog's Include patterns
field defaults to {supportedfiles}, which resolves to a list of all known supported file
extensions. To add your own custom extensions to this, you can use {supportedfiles} *.ext
(replacing .ext with your custom file extension).
Windows only: You can show/hide the drives that Windows auto creates by right-clicking the required
drive, selecting ShowDefaults, and checking or unchecking the drive.
USER GUIDE
266
Using the File Browser | To Use the Navigation Controls
To Use the Navigation Controls

• Click the Create New Directory button to create a new directory at your current position in the
file hierarchy.
• Click Up one directory to go up one directory closer to the root.
• Click Previous directory to go back one directory.
• Click Next directory to go forward one directory.

• Click the + button to add a directory bookmark.
• Click the edit button to edit the name or path name to a bookmark.
• Click the - button to remove a directory bookmark.
Path Name Field

The path name field displays the current directory path, lets you navigate to a new path, and also
enter a file name for scripts and rendered images.
• To navigate to a directory, type the path name in the field.

• To enter a script name, browse to a directory path and enter the file name after the displayed path.
• To limit the file list to specific file types, use the filter dropdown menu and Sequences checkbox.
USER GUIDE
267
Using the File Browser | To Use the Filter Dropdown Menu and Sequences Checkbox
To Use the Filter Dropdown Menu and Sequences

Checkbox
• Select *.nk to display only Nuke script files.
• Select * to display all files (except hidden files), regardless of what they’re associated with.
• Select .* * to display all files, including hidden files.
• Select */ to display directory names, but not their contents.
• Check sequences to display image sequences as single titles, as in fgelement.####.cin 1-50 rather
than fgelement.0001.cin, fgelement.0002.cin, fgelement. 0003.cin, and so on.
Note: File sequences with no file extension (for example, fgelement.0001, fgelement.0002,
fgelement.0003, and so on) are not displayed as single titles the first time you view the
directory in the File Browser. However, they are displayed as single titles once you have
navigated to another directory and back again.
Note: By default, the application may not be able to display custom file extensions (for
example, .cext) as single titles. To fix this, you can register your custom file extension as a
sequence type using Python:
1. Create a file called init.py in your plug-in path directory if one doesn’t already exist. For
2. Open the init.py file in a text editor and add an entry in the following format (replacing
cext with your custom file extension):
nuke.addSequenceFileExtension("cext")
• You can also split incomplete sequences into separate Read nodes using the split seq checkbox.
To Preview Files in the File Browser

1. Click the black arrow in the top-right corner of the file browser.
USER GUIDE
268
Using the File Browser | To Preview Files in the File Browser
The file browser expands to include a small viewer.

2. Select the file you want to preview in the file browser to view it.
USER GUIDE
269
To Select Multiple Files with the File Browser

1. Browse to the folder where the files are located.
2. Ctrl+click on all the files you want to open to select them (Mac users Cmd+click).
3. You can open files from multiple directories by clicking Next and browsing to the next file
location.
4. Click Open.
USER GUIDE
270
All the selected files open.
USER GUIDE
271
Sorting and Searching Media | Sorting the Project Panel
Sorting and Searching Media

Nuke Studio's Project panel includes several sorting and searching methods so you can organize,
manage, and navigate through media more easily. You can sort items alphabetically, by type, or
manually and select the appearance of media within bins. The search functionality allows you to enter
strings and apply searches on all or partial matches. You can control the poster frame displayed by
items in the project, selecting either an absolute or relative frame.
You can also color-code items in the Project panel and timeline to quickly sort and locate media. See
Color-coding Source Clips and Shots for more information.
Sorting the Project Panel

The Project panel can be sorted in the left-hand directory side and right-hand file side independently,
though files have more sorting and display options.
In the left-hand directory pane, items are sorted alphabetically in ascending order by default. To sort
project items manually:
1. Click and hold, or right-click, the sorting dropdown and select Manual.
2. Drag-and-drop items into your preferred order.
USER GUIDE
272
Sorting and Searching Media | Sorting the Project Panel
The right-hand bin pane is also sorted alphabetically in ascending order by default, and there are
more options available in the sorting dropdown. Click and hold the dropdown to display the various
options, including Duration, Resolution, and Type.
You can also change the layout and size of items in the right-hand bin pane. Choose a layout from
thumbnails, list, and details and use the slider to determine the size of the items. The slider is
disabled in details mode.
USER GUIDE
273
Sorting and Searching Media | Searching for Media
Searching for Media

Nuke Studio's search functionality allows you to enter strings and apply searches on all or partial
matches with the option to include metadata searches. Nuke Studio searches for items that match
any of the input string and displays only those items by default.
Tip: Nuke Studio also examines file metadata for the search string by default, but you can
disable this behavior by disabling Include Metadata in the search dropdown.
• Match All Criteria - media that matches all the entries in the search string are matched. For
example, entering A01 DPX only matches media containing both A01 and DPX.
USER GUIDE
274
Sorting and Searching Media | Setting Poster Frames
• Match Any Criteria - media that matches any of the entries in the search string are matched. Using
the same example, A01 DPX matches any media containing A01 or DPX.
Tip: This functionality extends to the media spreadsheet if you prefer to search for items
there. See About the Media Spreadsheet for more information.
Filtering and flagging produce the same search results, but they are presented differently. Filtering
only displays files that match some or all of the search string, whereas flagging displays all content
and flags files that don't match some or all of the search string.
Filtering ... ... and Flagging
You can also filter and flag media by tag or metadata. See Filtering and Flagging Media Using Tags
and Filtering and Flagging Media Using Metadata for more information.
Setting Poster Frames

You can assign poster frames to image items in the Project panel so that they represent the frames
you intend to use if your sequences. This is particularly useful if your source clips contain handles of
blank frames at the beginning of the footage.
To set a poster frame for source clips:

1. Select the clip or clips in the Project panel.
2. Right-click the selection, and select Set Poster Frame.
USER GUIDE
275
Sorting and Searching Media | Setting Poster Frames
3. Select a preset frame or click Custom to select an absolute or relative frame as the poster frame
for the source clip:
• Absolute - the poster frame number is derived from the file(s) on disk. For example, a .dpx
sequence myClip.####.dpx 1001-1500 can have an absolute poster frame between 1001 and
1500.
• Relative - the poster frame number is derived from the number of frames in the clip starting at
0. For example, a .dpx sequence myClip.####.dpx 1001-1500 can have a relative poster frame
between 0 and 499.
Note: If you select more than one source clip, you can't set an Absolute frame as the poster
frame.
USER GUIDE
276
Color-coding Source Clips and Shots | Setting Default Colors
Color-coding Source Clips and

Shots
In large projects, the Project bin and timeline can quickly become busy and difficult to manage.
Adding colors to items or file types can help you find what you're looking for in bins and sequences.
You can quickly enable and disable colors in the Project panel and timeline in the Preferences
under Panels > Project Items. See the Appendix A: Preferences page under Panels for more
information.
You can set general defaults for items such as bins, sequences, and source clips as well as for specific
file types such as .exr or .mov files. You can also set custom colors for individual selections or groups
using the right-click menu or color picker button at the top left of the Project panel.
Setting Default Colors

Project items have a default color assigned to them by type, such as orange for sequences and gray
for source clips. You can change these assignments using the Preferences > Panels > Project Items
dialog.
Note: You can disable the color scheme in the Project panel, timeline, and spreadsheet at
any time using the display in controls in the Preferences.
USER GUIDE
277
Color-coding Source Clips and Shots | Setting Colors by Selection
To change a default color, click a button in the Project Items preferences and use the color wheel to
change the color.
You can also set the label color for the Project panel and timeline or change the color used to
indicate the state of a shot on the timeline, such as offline.
Setting Colors by Selection

You can keep track of particular clips and shots by manually assigning them a color of their own. This
custom color overrides any color you set by type. You can set colors on items individually or by
selecting multiple instances in the Project panel or timeline.
USER GUIDE
278
Color-coding Source Clips and Shots | Setting Colors by File Type
1. In the Project panel or timeline, select the items you want to color.
Tip: The selection tools can help you make multiple selections quickly in the timeline. See
Using the Selection Tools for more information.
2. Right-click the selection and choose Color > Color Picker or click a color under Recent.
Tip: You can also color selections by clicking the button and picking the color.
3. Click Clear Color to remove any custom colors applied to the selection. The selection reverts to
the default item color set in the Preferences.
Setting Colors by File Type

Another way to organize your project is to assign colors to items by file extension. This custom color
overrides any color you set by item type.
1. Open the Preferences by pressing Shift+S.
2. Navigate to the Panels > Project Items sub-menu.
USER GUIDE
279
Color-coding Source Clips and Shots | Setting Colors by File Type
3. Under file types, click the button.

A new entry is added to the file types table.
4. Click the extension dropdown and select the required file type.
5. Double-click the color swatch and select the new color using the color wheel.
6. Click OK to apply your changes.
USER GUIDE
280
Reconnecting and Refreshing Clips | Setting Colors by File Type
Reconnecting and Refreshing

Clips
During the post process, media inevitably changes location or form. You can reload or replace your
media using the reconnect, refresh, and rescan functions.
Though all three options deal with reloading source clips, each has a particular use dependent on
context:
• Reconnect Media - allows you to redirect the filepath when the source file location changes.
• Refresh Clips (F8) - allows you to reload the clip when the source file location has not changed,
such as when work has been done on the clip offline. Selecting refresh only refreshes the clip’s
current frame range.
• Rescan Clip Range (Alt+F5) - similar to Refresh Clips, above, but rescan also checks for additional
frames that may have been added to the source file and adds them to the source clip’s frame range.
USER GUIDE
281
Localizing Media | Setting Localization Preferences
Localizing Media
Nuke Studio has the facility to cache source clips locally, either individually or by setting an
automatically localized folder (NUKE_TEMP_DIR/localize, by default), to help guarantee playback
stability. Local caching is controlled initially in the Preferences dialog, then on a clip-by-clip basis.
Setting Localization Preferences

The Preferences control how Nuke Studio deals with new source clips as they are ingested, but does
not affect existing source clips in the project. If you plan to localize source clips in your project, it's a
good idea to set the Preferences before ingesting your source clips.
1. Press Shift+S to open the Preferences dialog and navigate to Performance > Localization.
2. Set the required localization mode:

• on - checks for updates to source clips for all localization policies, and localizes those files set to
On or From auto-localize path automatically. On demand source clips must always be
localized manually.
• manual - checks for updates to source clips and prompts you to update them manually. Only
those with the policy set to off are not updated.
USER GUIDE
282
• off - no source clips are localized, regardless of the their localization policy.
Note: The current localization mode is displayed in the status bar at the bottom-right of the
interface.
3. Set the default localization policy for new source clips using the dropdown:
Note: The localization policy for existing source clips in the project must be set
individually. See Managing Localization for more information.
• on - always localize source clips with this policy.

• from auto-localize path - localize these source clips automatically if they reside in the auto-
localize from directory.
• on demand - only localize these source clips when you manually update them. See Updating On
Demand Clips for more information.
• off - never localize these source clips.
4. Enter a file path for auto-localize from, if required.
Any files that reside in this directory are automatically cached when ingested into Nuke Studio,
providing that the Localization Policy is set to From auto-localize path.
Note: Localization in both the timeline and Node Graph is paused during compositing
Viewer playback so that performance is not affected.
5. Enter a file path for localize to. Leaving this field as the default creates a sub-directory in the
Temp Directory as the local cache.
Note: On Windows, files saved to the localize to directory replace \\ (double back slashes)
and : (colon drive signifiers) with underscores so that the file path works as expected
between operating systems. For example:
\\windowspath\to\my\network\file.dpx is saved as __
windowspath\to\my\network\file.dpx
t:\my\network\path\file.dpx is saved as t_\my\network\path\file.dpx
6. Enter a value for limit to (GB) to control how much disk space is available in the cache directory.
USER GUIDE
283
Note: Negative values in this field reserve the specified amount of space at all times. For
example, -2 stops 2 GB of memory being used for caching.
7. You can specify the time interval (in minutes) before localized files are checked for updates using
the check for updated files every ## mins control.
The default setting checks for new versions of files every 30 minutes, but you can set this control
to any value. Any files that have changed since the last update check are flagged red in the Project
bin.
Tip: You can change the default localization indicators using the Appearance controls.
If read source files when localized files are out of date is enabled, source clips referencing cached
files that have changed since they were localized revert to reading the source files. Source clips that
are reading from the source files are marked with a striped red bar on the clip.
USER GUIDE
284
Localizing Media | Managing Localization
Enabling hide out of date progress bar hides the localization state of out of date files so that the
source clip appears the same as regular clips.
Note: The out of date localized files are not discarded, disabling read source files when
localized files are out of date picks up the out of date files instead of reading the source
files.
Managing Localization
As well as the overall Preferences for when new source clips should be localized, you can set
localization on a file-by-file basis. The localization policy for any existing source clips in your project
must be managed individually.
Tip: If you find that localization is slow to copy files, you can increase the number of threads
that Nuke uses to process jobs. Set the NUKE_LOCALIZATION_NUMWATCHERS environment
variable to the number of threads you want to use. See Environment Variables for more
information.
To set the localization policy for source clips:

1. Select the clip(s) in the bin view.
2. Right-click and select Localization Policy to display the available options:
• On - the files are localized, regardless of location, as long as the limit to (GB) limit is not
breached.
USER GUIDE
285
• From auto-localize path - the files are localized if they reside in the auto-localize from
directory, as long as the limit to (GB) limit is not breached.
• On Demand - the files are only localized when you update them manually. See Updating On
• Off - the files are never localized, regardless of location.
As clips are localized, an amber progress bar displays in the thumbnail. Fully cached clips are
marked with a green bar at the top of the thumbnail and out-of-date clips are marked in red.
Note: Container formats, such as .mov and .r3d, do not display progress bars during
localization. The source clip only shows the green indicator once the entire file is localized
successfully.
Note: Nuke Studio also features a playback cache and timeline disk cache, allowing frames
to be cached in RAM or disk. See Caching Frames in the Playback Cache and Caching Frames
in the Disk Cache for more information.
3. If you need to pause localization temporarily, navigate to Cache > Localization and select Pause.
USER GUIDE
286
4. If you find that your cache is filling up regularly, you can:

• Increase the amount of available space for localization by raising the limit to (GB) preference,
• Navigate to Cache > Localization > Clear Unused Local Files (see Clearing Localized Files), or
• Manually clear files from the cache directory in NUKE_TEMP_DIR/localize, by default.
5. You can force Nuke Studio to check for updated source files, rather than waiting for the check for
updated files every interval to expire, by selecting Force Update for All files, just Selected
source clips, or On Demand only.
Note: Each file has its own update time, which is reset whenever the source files are
checked.
The following table is a quick reference guide to when and how source clips are localized, where:
• green - clips are localized automatically.
• amber - clips are localized when updated manually.
• red - clips are not localized.
Source Clip Preference
System
on auto-path on demand off
Preference
on
USER GUIDE
287
Localizing Media | Updating On Demand Clips
Source Clip Preference
System
Preference
manual
off
Updating On Demand Clips

Source clips with their localization policy set to on demand are polled to check for updates at the
interval set in the Preferences. If the file has changed, the source clip is marked with a red bar to
show it is out of date.
To update a single on demand source clip:

1. Double-click the source clip in the Project bin to display its Properties.
2. Click Update.
Note: If a Read node's localization policy is set to on demand and it hasn't been localized
previously, clicking the Update button localizes the file.
The local copy of the source clip is updated from the remote source clip.
To update all on demand source clips:

1. Navigate to Localization > Force Update.
2. Click On demand only.
USER GUIDE
288
Localizing Media | Clearing Localized Files
The local copy of all on demand source clips is updated from the remote source clip.
Clearing Localized Files

Localizing a large amount of files can fill up the localization cache quite quickly if you leave the limit
to (GB) preference at the default 10 GB. When the cache runs out of space, a Failed to Localize File
dialog displays and localization pauses.
You can delete localized files by clicking Delete Unused Local Files (or by navigating to Cache >
Localization > Clear Unused Local Files). Nuke Studio displays a dialog containing all the files that
are marked for delete.
Tip: You can also open the Preferences from this dialog to adjust the localization behavior,
such as increasing the limit to (GB) preference.
USER GUIDE
289
Localizing Media | Clearing Localized Files
Click Continue to delete the localized files or Cancel to keep the cached files.
USER GUIDE
290
Using the Timeline Viewer | Clearing Localized Files
Using the Timeline Viewer

Nuke Studio's Timeline environment supports two distinct Viewer types: clip and sequence. This
chapter describes the difference between the two and how to use them.
Clip Viewers, sometimes referred to as source Viewers, are marked with the icon and deal
exclusively with source clips. You can set In and Out points and apply tags to the Viewer, but the
source clips are unaffected.
Sequence Viewers, also known as record Viewers, are marked with the icon and deal with
sequences and shots on the timeline. You can set In and Out points and apply tags here too, but you
can also edit the shots on the timeline by trimming, retiming, and so on. See Timeline Editing Tools
The Editing workspace combines both clip and sequence Viewers by default, enabling you to add
source clips to the timeline using insert and overwrite functions. See Insert, Overwrite, and 3-Point
Editing for more information.
To view your media in a Viewer, simply drag-and-drop a clip or sequence from the Project tab on to a
Viewer input, or double-click the item to send it to the appropriate Viewer.
Note: The Viewer currently treats all alpha channels as premultiplied, which can result in
the Viewer background being “added” to the image. If you’re working with un-premultiplied
images, set the Viewer background to Black.
USER GUIDE
291
Using the Timeline Viewer | Deleting Media
See Appendix A: Preferences for more information.
Deleting Media
To remove media from the bin view, select the clip(s) or bin and press Backspace or Delete.
If any of the media is in use in a sequence, the following warning displays:
Click Yes to delete the media from the bin view, but bear in mind that all instances of the deleted
media are removed from your current sequences.
USER GUIDE
292
Timeline Playback Tools | Deleting Media
Timeline Playback Tools

There are many useful tools at the top of the Viewer, some of which allow you to select channels,
adjust gain and gamma, and zoom and scale down the image in the Viewer.
For more information about the tools above the Viewer, see Timeline Viewer Tools.
The tools at the bottom of the Viewer allow you to adjust the playback settings, including setting the
frame range, selecting the playback mode, and locking the Viewer playback range.
Drag the orange marker along the timeline to quickly cue to a specific frame or timecode. The
number of the current frame or timecode appears below the center of the timeline. You can also cue
to a frame or timecode by typing its number directly into this field.
Tip: The current frame and in an out point fields accept simple mathematical functions,
such as +/-20 to jump forward or backward 20 frames or +/-00002000 to jump forward or
backward 20 seconds.
By default, Nuke Studio automatically adjusts the timeline of every Viewer window to show the frame
range defined in your Project Settings. If no frame range is defined, the frame range of the first image
you read in is used as the global frame range.
USER GUIDE
293
Timeline Playback Tools | In and Out Points
Viewer timeline controls also have a frame range source dropdown menu that you can use to define
where the timeline gets its frame range from. You can set this menu to Global, Input, or Custom.
Global is the default setting described above.
The playback rate field (frames-per-second) initially displays the project’s playback speed. Nuke
Studio attempts to maintain this speed throughout playback, although this adjusts depending on the
resolution of the imagery and your hardware configuration.
Note: The asterisk (*) denotes the Sequence playback speed selected using the Frame
Rate dropdown or, for new projects, the Project Settings > Sequence > Frame Rate
dropdown.
In and Out Points

In and Out markers enable you to alter the duration of a clip to just the portions of the source that
you require.
When a clip containing In and Out points is added to a timeline, you can slip the clip around the
markers to adjust the clip’s output. See Timeline Editing Tools for more information.
You can also use In and Out points to export certain portions of a clip or sequence. See Transcoding
To set In and Out markers:

1. Right-click on the required clip or sequence and select Open In > Timeline View.
Clips opened in a timeline have a purple background in the timeline.
2. Move the playhead to the location of the In point and press I on your keyboard.
The In point is marked by the In tab and the time is recorded in the playback controls.
3. Move the playhead to the location of the Out point and press O on your keyboard.
The Out point is marked by the Out tab and the time is recorded in the playback controls.
Note: You can also set markers by navigating to View > Mark In or Mark Out, by using the
Timeline menu to Mark Selection or Mark Clip dependent on clip selections on the
timeline, or by right-clicking a shots and selecting Open In > Viewer.
USER GUIDE
294
Timeline Playback Tools | In and Out Points
Click and drag the markers to adjust their position, or hold Ctrl/Cmd to move both markers at once,
retaining their relative positions. A Viewer preview shows the current frame for the selected marker(s)
and a timecode/frame popup helps to set the new position.
Clear the markers from your clip by navigating to Viewer > ClearIn Point (Alt+I) and Clear OutPoint
(Alt+O). The markers are removed completely, but you can reapply them by repositioning the
playhead and pressing I or O.
Tip: You can also press Alt+U to remove both markers at once.
When the playhead is positioned near In and Out markers, the top half of the timecode scale controls
the playhead and bottom half controls the markers.
USER GUIDE
295
Timeline Playback Tools | Playback Controls
Playback Controls
The playback rate field (frames-per-second) initially displays the project’s playback speed. The Viewer
attempts to maintain this speed throughout playback, although this adjusts depending on the
resolution of the imagery and your hardware configuration.
The following table lists the functions of the playback buttons:
Buttons Functions
The Play backward and Play forward buttons play the sequence
backward or forward at the script’s frame rate. When you press a play
buttons, it toggles to a stop a button.
The Back 1 Frame and Forward 1 Frame buttons cue the sequence to
the previous or next frame.
The Previous keyframe and Next keyframe buttons cue the sequence
to the script’s previous or next keyframe.
The First frame and Last frame buttons cue the sequence to the first
and last frame.
The Frame Increment field allow you to specify the number of frames
by which the Previous increment/Next increment buttons cue the
sequence. This is set to 10 frames by default.
The J, K, and L keyboard shortcuts also control playback. The K keyboard shortcut is mapped to
Pause/Play. J and L are mapped to backward and forward. Combinations are also supported:
• K+J - frame backward.
• K+L - frame forward.
• K+drag in the top third of the Viewer - standard jog controls. Dragging the cursor left and right
moves the playhead backward and forward, frame-by-frame.
• The jog controls also detect rotary motion to jog through frames. Clockwise motion in the top third
of the Viewer, while holding K, advances the playhead and anti-clockwise reverses the playhead.
• K+drag in the middle third of the Viewer - standard shuttle controls. Dragging the cursor left and
right plays backward and forward, with increased frame rate toward the edges of the Viewer.
• K+drag in the bottom third of the Viewer - skips the playhead to absolute timeline position.
USER GUIDE
296
Timeline Playback Tools | Playback Controls
The Playback Mode button lets you control how many times and in what direction the Viewer plays
back the sequence. Click the button to toggle between the following modes:
Button Function
Repeatedly plays the sequence in a loop.
Repeatedly plays the image back and forth from head to tail.
Plays once through the section between in and out point and stops at the out
point. If these are not marked, then it plays from the beginning to the end of
the sequence.
Plays once from the beginning to the end of the sequence, ignoring any in and
out points.
USER GUIDE
297
Timeline Playback Tools | Timeline Viewer Tools
Timeline Viewer Tools

The Viewer has two sets of tools for manipulating your media: the Viewer tools and the playback
tools. The Viewer tools, located at the top of the Timeline Viewer, are used to affect the mouse
pointer as you move over the Viewer, and to select Viewer preferences:
Icon Function Description
Layers Select the layer to output to the Viewer, for example forward
motion vectors or disparityL. Only layers available in the clip
are displayed - check the clip’s thumbnail to see at a glance
which layers are present:
- red color layer.
- green color layer.
- blue color layer.
- alpha layer.
- depth layer.
- forward motion vector layers.
- backward motion vector layers.
- all other custom layers, such as disparity.
Note: You can scroll through available layers using

PgUp or PgDn.
Channels Select the channel(s) to output to the Viewer, for example

RGB, single channel, Alpha, or Luma.
USER GUIDE
298
Viewer color Set the colorspace used to display images in the Viewer, for
transform example sRGB and rec709.
Note: If you have specified an OpenColorIO

configuration file in the Preferences, you may have
more colorspace choices available.
A/B Viewer Click the A or B dropdown and select what you want to view.
Output This can be selected tracks or tracks with selected tags.
When both Viewer buffers contain an image, enable wipe to

compare the two images. You can also use the center drop
down to set the blend mode between images in the Viewer,
for example Onion Skin or Difference, and the A/B buffer
configuration.
Guides Enable or disable Viewer overlays:

• title safe - any text intended for the audience should reside
within this zone.
• action safe - any visual elements intended for the audience
should reside within this zone.
• format center - adds a crosshair in the center of the format
currently in the Viewer.
• Format - adds a red, format-dependent box for the clip or
sequence in the Viewer. Sequences support multi-format
clips, see Viewing Multi-Format Timelines for more
information.
Mask Enable or disable a range of Viewer masks, for example 16:9

or 1.85:1.
Clipping Enable or disable Viewer warnings:

• No Warnings - all clipping warnings are disabled.
• Exposure - alerts you when the image is under (blue) or over
(red) exposed.
USER GUIDE
299
Annotations Click to enable the Annotations tool bar. Annotations allow

you draw and add text to clips in the Viewer. See Annotations
Note: The Annotations button also controls existing

annotation visibility.
ROI Click and drag to define a Region of Interest (ROI) in the

Viewer. The scopes only display information within the ROI,
when active.
Pause Pause or release Viewer playback caching, indicated by the

playback caching green bar under the Viewer.
Scale Set the scale applied to the clip in the Viewer, for example
25%, 75%, or Fit.
Image Quality Set the Viewer image quality, for example 1:1, 1:4, or 1:16. The
default setting, Auto, resizes the image dependent on the
Viewer zoom level, which may re-cache the image at a higher
resolution.
Note: Image quality, or proxy, for RED clips is

dependent on the clip’s Decode Resolution in the
Media panel.
For example, if you're viewing a 4K file and the
Decode Resolution is set to Half Premium, a 1:1
proxy value is equal to 2K, 1:2 is equal to 1K, and so
on.
Non RT Playback Sets the Viewer playback mode:

• Play All Frames - the default setting, plays all frames in real-
time (dependent on hardware).
• Skip Frames - plays frames in real-time skipping where
necessary to maintain the frame rate.
• Play All Frames, Buffering - plays all frames by buffering
USER GUIDE
300
and playing frames back as they become available.
See through missing When disabled, any offline media on a timeline is treated as a
media blank clip so the Viewer cannot display the track underneath.
This setting also applies to missing frames within a clip.
Note: This control only affects the current Viewer.
View Select the Viewer display mode, for example Audio and Video
or Video Only.
Obey Alpha Allows you to control the alpha channel independent of the
Viewer Blend Mode.
• Enabled - any alpha channel present in the image is treated
as premultiplied transparency.
• Disabled - the alpha channel is ignored.
Audio latency Sets the audio latency, in milliseconds, for the current Viewer
only. Audio latency allows you to correct audio and video
synchronization by changing the point at which audio
playback starts.
Positive values cause the audio track to start earlier in relation

to the video track, and vice versa.
Gain Adjusts the gain applied to the linear input image before
viewing, but doesn’t affect your exported image.
Gamma Adjusts the gamma applied to the image after the viewing
transform, but doesn’t affect your exported image.
Mute / Audio Click to mute audio output for the current Viewer or use the
slider to control the audio output level.
Tip: You can also control volume on a per track and

per shot basis. See Audio and the Timeline for more
information.
USER GUIDE
301
Timeline Playback Tools | Using In and Out Markers
Color Enable or disable the RGBA color information bar in the

Sample Viewer.
Note: The Color Sample tool displays color

information from the source file, not the colorspace
selected in the Viewer color transform dropdown.
See Working with Colorspaces for more information.
Using In and Out Markers

In and Out markers enable you to alter the duration of a clip to just the portions of the source or
sequence that you require.
Tip: You can use the source clip/shot properties original range controls in similar way to In
and Out points. See Setting Source Clip Ranges for more information.
markers to adjust the clip’s output. See Timeline Editing Tools for more information. You can also use
In and Out points to export certain portions of a clip or sequence. SeeTranscoding for more
information.

Tip: Source clips opened in the timeline view have a purple background in the timeline.
USER GUIDE
302
Clear the markers from your clip by navigating to View > ClearIn Point (Alt+I) and Clear OutPoint
USER GUIDE
303
Using In and Out Markers

In and Out markers enable you to alter the duration of a clip to just the portions of the source or
sequence that you require.
Tip: You can use the source clip/shot properties original range controls in similar way to In
and Out points. See Setting Source Clip Ranges for more information.
markers to adjust the clip’s output. See Timeline Editing Tools for more information. You can also use
In and Out points to export certain portions of a clip or sequence. See Transcoding for more
information.

Tip: Source clips opened in the timeline view have a purple background in the timeline.
USER GUIDE
304
Clear the markers from your clip by navigating to View > Clear In Point (Alt+I) and Clear Out Point
USER GUIDE
305
USER GUIDE
306
Working with Colorspaces | Using In and Out Markers
Working with Colorspaces

Colorspace changes are applicable to clips in bins and shots, as well as in the Viewer using the Media
tab.
To apply colorspace changes to clips in bins:

1. Select the clip or clips in the bin view.
2. Right-click a selected clip and navigate to Set Media Color Transform.
The current colorspace is highlighted with a tick mark.

3. Select the colorspace to apply to the clip selection.
Note: Only colorspaces applicable to the selection are displayed. For example, REDspace
and LogC - CameraNative are only available for R3D and ARRIRAW clips, respectively.
4. Selecting multiple formats supporting different colorspaces, for example R3Ds and ARRIRAW,
breaks the available LUTs into sub-menus:
USER GUIDE
307
Working with Colorspaces | To apply colorspace changes to shots:
To apply colorspace changes to shots:

1. Select the item(s) on the timeline.
2. Right-click a selected item and navigate to Set Media Color Transform.
3. Select the colorspace to apply to the selection.
USER GUIDE
308
Previewing on a Broadcast Monitor | To apply colorspace changes to shots:
Previewing on a Broadcast
Monitor
The Monitor Out feature allows you to preview Viewer images and audio on an external broadcast
video monitor to check the final result, including the correct colorspace and aspect ratio. This option
requires additional hardware, such as a monitor output card or a FireWire port.
Note: Audio scrubbing is not currently available through monitor output cards. Audio
scrubbing is only supported through internal audio output devices.
Our monitor out architecture interfaces directly with the AJA and BlackMagic device drivers, which are
unified across their respective hardware lines, meaning all current supported cards for the versions
detailed in Third-Party Libraries and Fonts should work. We've tested the following AJA and
Blackmagic hardware:
AJA Card: KONA LHi KONA 3G KONA 4 KONA iOXT
Formats
SD
HD
2K
UHD
4K
BNC
HDMI
Stereoscopic Support
No Yes Yes No
Platforms
USER GUIDE
309
Win, Mac, Linux Win, Mac, Linux Win, Linux Mac 10.9 and
10.10
Drivers
• Windows - 12.5.1 • Windows - 12.5.1 • Windows - 12.5.1 Mac - 12.5.1

• Mac - 12.5.1 • Mac - 12.5.1 • Linux - 12.5.2.7
• Linux - 12.5.2.7 • Linux - 12.5.2.7
Note: The following should be taken into account when using the Monitor Out functionality
with AJA cards:
If you're running AJA cards on Linux, you can contact www.aja.com/support to obtain the
correct drivers.
12-bit monitor output is only supported with dual connection cards, that is cards with two
physical connections, not dual links combining two separate streams of data.
Hiero is unable to send out the right eye separately using the 2nd output cable of KONA 3G
cards. Instead, both views are sent through the 1st output and can be viewed using the side-
by-side, anaglyph, and interlacing options.
Blackmagic Card: DeckLink SDI DeckLink HD DeckLink Intensity Pro 4K

Extreme 2 Extreme 3D+
Formats
SD
HD
2K
UHD
4K
BNC
HDMI
USER GUIDE
310

No No Yes No
Platforms
Win, Mac, Linux Win, Mac, Linux Win, Mac, Linux Win, Mac, Linux
Drivers
Desktop Video Desktop Video Desktop Video Desktop Video

10.7 to 10.8.4 10.7 to 10.8.4 10.7 to 10.8.4 10.7 to 10.8.4
Blackmagic Card: DeckLink Studio 4K DeckLink 4K Extreme DeckLink 4K Extreme

12G
Formats
SD
HD
2K
UHD
4K
BNC
HDMI
Yes Yes Yes
(Both views through

one output, so the Full
Resolution option is
not available.)
USER GUIDE
311
Previewing on a Broadcast Monitor | To preview output on an external broadcast video monitor:

12G
Platforms
Win, Mac, Linux Win, Mac, Linux Win, Mac, Linux
Drivers
Desktop Video 10.7 to Desktop Video 10.7 to Desktop Video 10.7 to

10.8.4 10.8.4 10.8.4
Some monitor out cards allow you to extend or mirror your desktop so that Hiero’s user interface is
visible on the monitor. Please refer to your card’s documentation for more information.
To preview output on an external broadcast video monitor:

1. Navigate to Window > Monitor Output.
The Monitor Output toolbar displays in a floating pane.
Note: If you're working with multi-view footage, additional controls display to determine the
stereo mode and view to output.
See Stereoscopic and Multi-View Projects for more information.
2. Select the external device you want to use from the output device dropdown. All available devices
are automatically detected and listed in this menu, along with the following default options:
• None - disables the monitor out feed.
• Floating Window - opens a pseudo output monitor window, without the need for a monitor
device and card. This is intended for full-screen use without displaying any of the interface.
3. Select the Viewer to feed to the output monitor using the source viewer dropdown. Selecting
Active Viewer always displays the current Viewer.
4. Select the view mode using the A/B selection dropdown.
USER GUIDE
312
Previewing on a Broadcast Monitor | To preview output on an external broadcast video monitor:
Note: For multi-view/stereo footage, selecting A/B mode in this dropdown forces the
monitor to output the timeline Viewer settings and the view to output controls are disabled.
5. Click to apply the active Viewer’s filtering, gamma, and gain to the monitor output.
6. Click to flip the output vertically.
7. Click to switch between full-range 0-255 (default) and 16-236 (ITU-R BT.610-4). This button can
correct the image output for certain monitor out cards.
8. Select the colorspace to apply to the image. If you’ve specified an OCIO configuration file in the
preferences, these custom LUTs are also applicable.
Note: If you plan to use the OCIO config file specified during exports, ensure that the
Preferences > Project Defaults > Color Management > Export > use OCIO nodes when
exporting to a Comp checkbox is enabled.
USER GUIDE
313
Using Scopes | Histogram
Using Scopes
Nuke provides scopes to help you evaluate your media. There are a number of global controls
(Preferences > Panels > Scopes) that affect how the Scopes display information:
• black point - sets the black out of range warning level.
• white point - sets the white out of range warning level.
• luma/chroma encoding - sets the video standard to use when converting RGB to luma or chroma
values in the scope displays, either REC601 or REC709.
• Include viewer color transforms - when enabled, scope data includes the applied Viewer color
transforms (gain, gamma, and LUT). When disabled, scope data does not include the applied Viewer
color transforms. This may slow down rendering, as it may require image calculation.
• Force full frame - When enabled, scopes display data for the full frame, regardless of what portion
of that frame is displayed in the Viewer. When disabled, scopes only display data for the current
area requested by the Viewer rather than the full frame.
To open a scope, navigate to Window > New Scope and select the required scope from the list.
Histogram
The Histogram provides three color channel and luma channel information that describes the
distribution of red, green, blue, and luma pixels throughout the current frame.
The Histogram graphs the number of pixels at each brightness level, and from left to right, the areas
of the Histogram represent shadow, mid tones, and highlights.
USER GUIDE
314
Tip: You can pan the view area by holding Alt, or the middle mouse button, and dragging in
the panel.
There are also Viewer and Channel selection controls on the Histogram tab:
• Viewer selection - if you have multiple Viewers open, use the dropdown menu to associate
Histogram output to the required clip.
The default value, Active Viewer, automatically displays details on the last Viewer you selected.
• Channel selection - select the channels to output. The default setting displays RGB, but you can
also view channels separately.
• Mode selection - select the mode to output. The default setting displays ganged RGB, but you can
also view the channels separately.
• Current View - describes the view currently displayed in the scope, whether it's the A or B buffer
and the view. The view defaults to main, unless main has been replaced in multi-view scripts or
projects.
Depending on which Viewer tools and views you have active, you can have up to four scopes
displayed at once.
For example, with two stereo Read nodes, one in each input buffer, and wipe and Side by Side
active, the scopes display something like this:
The scopes feature global customizable guides to help you view your clips. Navigate to Preferences >
Panels > Scopes and enter values between 0 and 1 for the Black and White points. Note that this
also sets the values for the Waveform display.
The guides at the edges of the Histogram turn red to warn you when the distribution is out of range:
USER GUIDE
315
Using Scopes | Waveform
Waveform
The Waveform scope provides information on clip luminance, or brightness, which you can use to
decide whether the clip is over or under exposed. The white traces represent luminance values from 0
- 100% (black through the spectrum to white). The higher the waveform, the brighter the image in the
Viewer.
the panel.
The upper white marker is used to measure when over exposure could be a problem. If your
waveform has a lot of traces over the white marker, you should consider reducing the brightness of
the clip. The opposite is true of the lower black marker.
There are also Viewer and Mode selection controls on the Waveform tab:
Waveform output to the required clip.
USER GUIDE
316
also view the channels separately.
projects.
displayed at once.
The scopes feature global customizable guides to help you view your clips. Navigate to Preferences >
Panels > Scopes and enter values between 0 and 1 for the Black and White points. Note that this
also sets the values for the Histogram display.
The guides at the top and bottom of the Waveform turn red to warn you when the distribution is out
of range:
USER GUIDE
317
Using Scopes | Vector
Vector
The Vector scope displays color, saturation, and hue information for the current frame. Similar to
color wheels, Vector scopes display information radially, from the center outward. The farther from
the center the data spans, the more saturation is represented.
In the image on the left, you can see that the frame represented contains mostly yellows and reds,
but the values are not oversaturated. The image on the right represents a badly saturated frame.
Notice the spill of red traces distributed toward the edge of the scope pass the target (the highlighted
square).
Normal saturation. High Saturation.
the panel.
There is also a Viewer selection control and Current View label on the Vectorscope tab:
• Viewer selection - if you have multiple Viewers open, use the dropdown menu to associate Vector
scope output to the required clip.
projects.
displayed at once.
USER GUIDE
318
Using Scopes | Vector
USER GUIDE
319
About Anamorphic Media | Vector
About Anamorphic Media

The Viewer automatically recognizes anamorphic clips and displays them with the correct aspect
ratio.
If for any reason you want to display an anamorphic clip with a 1:1 aspect ratio, right-click in the
Viewer displaying the clip and enable the Ignore Pixel Aspect checkbox, or use the Ctrl/Cmd+Shift+P
keyboard shortcut.
USER GUIDE
320
About QuickTime Media | Vector
About QuickTime Media

Working with .mov files can be unpredictable when compared to other formats, so Nuke gives you a
few QuickTime options when reading and writing .mov files.
Nuke attempts to select the ’best fit’ combination by reading an extended list of metadata key/value
pairs from the QuickTime header, including nclc atom, gama atom, and ProRes codec headers.
If you place a clip in the Viewer, or open a shot in the Viewer, and open the Media tab, you'll see that
Nuke has a number of media-specific controls that you can manually override if the 'best fit' is not
what you're looking for:
• YCbCr Matrix - sets the way Y’CbCr is converted to RGB. You can choose to use the new Rec 601 and
Rec 709 or the Legacy encoding methods, which are the methods used previously in Nuke.
• Codec - sets the codec used to read (write already had a similar control) the QuickTime file.
The codec dropdown defaults to a codec appropriate for the QuickTime in question, where
available, and only lists those that declare themselves able to read the file.
• Pixel Format - sets the read and write pixel format, which includes bit depth, colorspace, pixel
packing, and ranges.
This setting defaults to the best format accepted by the codec, allowing Nuke to perform the
conversion to RGB without the use of an unknown QuickTime transform, where possible. RGB pixel
types rely on QuickTime to do the conversion from Y’CbCr when dealing with a non-RGB codec.
In addition to the nclc, gama, and ProRes data Nuke, and by extension Nuke, also write additional
metadata into the file headers during export, retaining your QuickTime preferences. This combined
metadata represents myriad potential QuickTimes preferences, so Nuke reads the available metadata
in the following order, reverting down each level as the level above is unavailable or set to a reserved
or unknown value:
USER GUIDE
321
About QuickTime Media | Vector
• Foundry-specific metadata
• ProRes header data
• nclc atom data
• gama atom data
• The defaults associated with the chosen codec
In this way, the worst case scenario is that you end up with the chosen codec class' default values.
USER GUIDE
322
About RED Media | To modify the RED Rocket options:
About RED Media

When working with RED clips, using a RED Rocket card can increase the rendering speed significantly,
especially at higher resolutions.
Note: The RED Rocket icon is only visible if you have a RED Rocket installed.
The RED Rocket icon has three states:
Inactive - the RED Rocket card is inactive.
Firmware error - there is a problem with the card firmware. Hover the mouse over the icon
Active - the RED Rocket card is present and active.
To modify the RED Rocket options:

1. Click the icon in the Viewer.
Note: You must have use RED Rocket enabled in the Preferences > Performance >
Hardware dialog to access these options. See Appendix A: Preferences for more
information.
The RED Rocket Settings dialog displays.
2. Temporarily disable the RED Rocket card by deselecting Use RED Rocket card. Unlike the option
in the Preferences dialog, changing this setting does not affect the application at startup.
3. Click OK to save your settings.
USER GUIDE
323
About RED Media | To modify the RED Rocket options:
ulimit -Sn 2048
USER GUIDE
324
Compositing with Nuke
Each chapter in this section explains in detail a key feature of Nuke. You can use the section to
familiarize yourself with the features you are particularly interested in, or to get answers to specific
problems that arise during compositing. For information on the features in NukeX and Nuke Studio,
see Advanced Compositing with NukeX and Nuke Studio or Timeline Editing in Nuke Studio.
Organization of the Section

These are the topics covered by this section:
• Managing Scripts describes the basics for creating, saving, and loading scripts or comps.
• Reformatting Elements describes how you can reformat images through scaling, cropping, and pixel
aspect adjustments. This chapter also covers working with bounding boxes.
• Channels shows you how to manage image data using Nuke’s 1023-channel workflow.
• Merging Images teaches you how to layer background and foreground elements together, create
contact sheets, and copy rectangles from one image to another.
• Removing Noise with Denoise teaches you to use Denoise node to remove noise from your footage.
It uses spatial filtering to remove noise without losing image quality.
• Keying with ChromaKeyer teaches you to use the blue/greenscreen keyer ChromaKeyer in Nuke.
• Keying with Primatte teaches you to use the blue/greenscreen keyer Primatte in Nuke.
• Keying with Keylight teaches you to use the keyer tool Keylight in Nuke.
• Keying with Ultimatte shows you to use the Ultimatte keyer in Nuke.
• Using RotoPaint shows how to use Nuke’s RotoPaint node.
• Tracking and Stabilizing shows how to generate and edit 2D tracking data for purposes of removing
unwanted motion or applying it to other elements.
• Transforming Elements covers the tools for changing the size, location, and orientation of an image,
including how to translate, scale, rotate, and skew elements in 2D and 3D space. This chapter also
describes adding motion blur.
• Warping Images teaches you to use the GridWarp and SplineWarp nodes to warp and morph
images.
• Temporal Operations explains how to apply time-based effects like clip retiming and motion blur.
This chapter also explains how to perform editorial tasks, such as trimming and slipping.
• Working with Color explains a broad sampling of Nuke’s many color correction tools.
• Filtering and Spatial Effects deals with applying filters, such as convolves and blurs.
• Creating Effects describes how you can create effects, such as star filter effects, on your images.
USER GUIDE
325
Compositing with Nuke |
• Analyzing and Matching Clips explains how to use the CurveTool node to analyze and match image
sequences.
• 3D Compositing teaches you how to create and manipulate 3D scenes composed of objects,
materials, lights, and cameras.
• Stereoscopic Scripts describes how to composite stereoscopic material in Nuke.
• Deep Compositing goes through using the deep compositing node set in Nuke.
• Working with File Metadata describes how to use Nuke's MetaData nodes to work with information
embedded in images.
• Audio in Nuke covers using audio clips in Nuke.
• Previews and Renderingteaches you how to write out image sequences from scripts in order to
preview results or create final elements.
• Organizing Scripts is designed to help you organize your Nuke scripts in a clear and meaningful way.
• Configuring Nuke explains how to set up Nuke for multiple artists working on the same project.
• Expressions explains how to apply expressions or scripting commands to Nuke parameters.
• The Script Editor and Python takes you through using Nuke’s Script Editor for executing Python
commands.
USER GUIDE
326
Managing Scripts
In this chapter, you learn about Nuke’s project files called scripts or comps. The topics covered
include setting up, saving, and loading scripts. You’ll also learn about managing your node tree in the
Node Graph, using Precomp nodes, and working with file metadata.
Setting Up Your Script

When you start working on a script, you should first define the settings for it. This involves assigning
the script a name, frame range, frame rate, and default full and proxy resolution format.
Name, Time Span, and Frame Rate

To set the script name, frame range, and frame rate:
1. Select Edit > Project Settings, or simply press S over a blank portion of the workspace.
The Project Settings panel appears.
USER GUIDE
327
Managing Scripts |
2. On the Root tab, type a name for the script (say, firstcomp.nk) in the name field. Nuke’s scripts
always have the extension .nk.
3. Type the numbers of the first and last frames in the frame range fields to define length of time
for your “shot”.
4. In the fps field, enter the rate in frames per second (fps) at which you want your script’s Viewers to
play back footage. For film-based elements, 24 fps is appropriate.
Setting the Default Project Directory

In the Project Settings (Edit > Project Settings) you can define your default project directory. You can
then refer to it with ’./’ in your file paths, for example ./test.jpg for an image called test in your
USER GUIDE
328
Managing Scripts | Full-Size Formats
default project directory.
You can also click the Script Directory button to create an expression that sets your project
directory to be the directory where your current script is saved.
Full-Size Formats
When you start a new script in Nuke, you need to set up a full-size format. The full-size format
determines the size of the image that you get from any disconnected node inputs. It also sets the
default size of any script-generated elements, such as Constants and ColorBars.
Note: The full-size format does not affect the format of the elements you read into your
script. Nuke is resolution-independent, which means it respects and keeps the resolution of
your images. It won’t automatically crop or pad elements to match the project settings.
If you want elements you read in to conform to the project settings, you can do this
manually using the Reformat node. For more information, see Reformatting Image
Sequences.
The full-size format is also used to calculate proxy scaling if a proxy format is used. For
more information on the proxy mode and proxy formats, see Proxy Mode.
To Set Up a Full-Size Format

1. If it’s not already open, select Edit > ProjectSettings (or press S) to display the Project Settings
panel.
2. From the full size format dropdown menu, select the resolution for the final output of rendered
images. If the format you want to use is not in the menu, select new. The New format dialog
displays.
USER GUIDE
329
Managing Scripts | Proxy Mode
In the name field, enter a name for the new format.

In the file size fields, define the width and height of the format.
If you like, you can also define additional information, such as offsets and pixel aspect ratio.
Click OK to save the format. It now appears in the dropdown menu where you can select it.
Proxy Mode
When compositing with Nuke, you can work in two different modes: the full-size mode or proxy
mode. In the full-size mode, images are read in exactly as they are on the disk, and all positions are
in actual pixels in these images. This is the mode you want to use for accurate feedback and when
rendering the final output.
In proxy mode, instead, a proxy scale factor is used. All images and all x/y positions are scaled by this
factor. This produces the same (or at least very similar) composite at a different scale. For example,
you can use a fraction of the full output resolution to speed up rendering and display calculation.
In addition to the above, a separate proxy file can also be read in place of a full-size image, provided
you have specified one in the relevant Read node. This can further speed up the preview, by using a
smaller image that reads faster and also saves time by not needing to be scaled. For more
information, see Read Nodes and Proxy Files.
The proxy settings you define in the project settings affect both the proxies Nuke generates using the
proxy scale factor and proxies read from files. Below, we discuss setting a proxy format and/or a
proxy scale and defining how Read nodes use proxy files.
Note: Note that proxy versions of images are only used if you have activated the proxy
mode. When the proxy mode is off, Nuke always uses the full-res files.
Proxy Format and Proxy Scale

In the Project Settings panel, you have the option of defining a proxy format and/or a proxy scale
that you use in the proxy mode.
For the proxy format, you can define the image resolution as well as additional information about
offsets and pixel aspect ratio. When using the proxy format, the scaling is proportionate to the full-
size/proxy format relationship (not scaled to the proxy format).
USER GUIDE
330
Managing Scripts | Proxy Mode
For the proxy scale, you only define a simple scale factor by which your images are scaled down
whenever the proxy mode is activated. For example, you can use the scale factor of 0.5 to scale your
images down to half the size.
If you like, you can define both a proxy format and a proxy scale, and then choose which one to use
in proxy mode. A proxy scale is easier to set up, but a proxy format gives you more control over the
low-res versions of your images. Below, we first describe how to set up proxy formats and then how
to define a proxy scale.
To Set Up Proxy Formats

1. If it’s not already open, select Edit > Project Settings (or press S) to display the Project Settings
panel.
2. If you want to use a proxy format (rather than a proxy scale) whenever the proxy mode is
activated, select proxy mode > format from the dropdown.
3. From the proxy format dropdown menu, select the resolution to use while working to speed
things up. Notice that your images are not scaled to this resolution, but the scaling is
proportionate to the full-size/proxy format relationship. Nuke divides the proxy format width by
the full-size width and uses the result as the scale factor.
If the proxy format you want to use is not in the dropdown menu, select new. The New format
dialog displays.
• In the name field, enter a name for the new format.

• In the file size fields, define the width and height of the format.
Tip: You can type formulas in numeric fields to do quick calculations. For example, if your
full-size format width is 4096 and you want your proxy format width to be 1/2 of that, you
USER GUIDE
331
Managing Scripts | Read Nodes and Proxy Files
can enter 4096/2 in the New format dialog’s file size w field and press Enter. Nuke then
calculates the new width for you.
Click OK to save the format. It now appears in the dropdown menu where you can select it.
4. To activate the proxy mode and use the low-res format for calculations and display, check proxy
mode.
Alternatively, you can use the proxy toggle in the Viewer controls, or press Ctrl+P (Cmd+P on a
Mac). For more information, see Using the Viewer Controls.
To Set Up a Proxy Scale

panel.
2. Select scale from the dropdown in the Project Settings.
3. Using the proxy scale input field or slider, specify the factor by which you want to scale the width
and height of your images. For example, if you want to scale them down by 50%, use the value of
0.5.
4. To activate the proxy mode and use the low-res format for calculations and display, check proxy
mode.
Alternatively, you can use the proxy toggle in the Viewer controls, or press Ctrl+P (Cmd+P on a
Mac).
Read Nodes and Proxy Files

As an alternative to letting Nuke generate proxies on the fly, proxy files can be specified using a
second file name in the Read nodes (for how to do this, see Loading Image Sequences). If you don’t
USER GUIDE
332
Managing Scripts | Write Nodes and Proxy Files
have a proxy file, you can create one by activating the proxy mode and rendering your full-size
images using a Write node (see Rendering Output in the Previews and Rendering chapter).
The proxy file does not need to match the proxy resolution in use. Depending on your project
settings, either the full-res or proxy file is scaled to the required proxy size (that is, the size calculated
by taking the full-size format and scaling it by the current proxy settings). However, if your proxy
images match your typical proxy settings, you can save this time.
To Define Which File (Full-Res or Proxy) is Used in the Proxy Mode:

panel.
2. From read proxy files, select when to use the proxy file (rather than the full-res file) in a Read
node:
• never - Never use the proxy file in the proxy mode. Instead, scale the full-size file as necessary.
• if larger - Use the smaller of the two images if it is larger or equal to the desired size, scaling
down as needed. Otherwise, use the larger one, scaling down or up as needed. This is the
default option.
• if nearest - Use the image that is closest to the desired size, scaling up or down as needed.
• always - Always use the proxy image in the proxy mode, scaling it up or down as necessary.
The option you choose affects all Read nodes in your script, provided that a proxy file is named
and the proxy mode is on.
Write Nodes and Proxy Files

It is worth mentioning here that when a script is rendered in proxy mode, processing is done at the
proxy scale and image output goes to the file name in the Write node’s proxy field. If you do not
specify a proxy file name, the render fails with an error. It never resizes the proxy image, and it does
not write the proxy image over the full-size one.
For more information, see Rendering Output in the Previews and Rendering chapter.
Using Proxy Mode to Enlarge a Script

If you find it necessary to render a comp at higher than the original resolution it was intended for,
you can also use the proxy resolution to scale up from the root full size format. For example, you
could use a 2K composite to produce 4K or higher images in the proxy mode.
The reason you'd probably want to do this is because the comp you did needs to be re-run at higher
resolution for a different output target and possibly with some new higher resolution elements. For
USER GUIDE
333
Managing Scripts | To Toggle Between Full Resolution and Proxy Mode
example, you may need to re-render with new CG elements at print resolution rather than for film
out, but you don’t want to go through your script and modify everything that has an x/y position.
When scaling up the output of a script in the proxy mode, image generator nodes render at the larger
size, larger images can be specified in the proxy field, and as with scaling down, all x/y positions are
scaled to the proxy output resolution.
Using Small Proxy Files

If you have previously set up your script to use small proxy files, you do not have to remove these.
Make sure read proxy files in Project Settings is set to anything other than always, and Nuke reads
the larger original files and scales them up.
Using Large Proxy Files

If you actually have larger proxy files, you should enter them into the Read nodes’ proxy field, and set
read proxy files to anything other than never. Nuke then uses these larger files in the proxy mode.
For maximum quality, these should be at exactly the desired proxy size so that no scaling is done to
them. For example, if the required proxy size (defined by the project settings) is 4K, then the proxy
image should be exactly 4K. Otherwise, Nuke scales it to match the project settings, which reduces
the quality.
Rendering the Scaled-up Output

To render the larger output of a scaled-up script, you need to activate the proxy mode and enter a file
name in the proxy field of the Write nodes. The larger images are then written to these files.
To Toggle Between Full Resolution and Proxy Mode

It’s usually smart to work in proxy mode because most operations work quickly and more efficiently
under the low-res display. You can switch between low- and high-resolution when you need greater
precision (for example, when pulling a key or tracking), or when you’re ready for final rendering.
1. Click on an empty area of the Nuke window.
2. Press Ctrl+P to toggle the display mode (Cmd+P on a Mac).
Nuke automatically scales script elements - Bezier shapes, B-spline shapes, paint curves, garbage
masks, tracking curves, and so on - to keep the original placement on the image.
USER GUIDE
334
Loading Image Sequences | Importing Image Sequences
Loading Image Sequences

When you are ready to start compositing, you may want to begin by importing a background or
foreground image sequence. Typically, you would read in both full- and proxy-resolution versions of
the sequence. You can read in several image sequences in one go.
Importing Image Sequences

To import an image sequence using Nuke's file browser:
1. Select Image > Read (or press R over the Nuke Node Graph).
Tip: Pressing R with an existing Read node selected, opens the file browser at the location
specified by that node.
2. Browse to the image sequence you want to import. For instructions on using the file browser, see
Using the File Browser. Select the file you want to open. If you want to open several files at the
same time, Ctrl/Cmd+click on the files, and then click Open.
A Read node is inserted in the Node Graph.
Nuke imports the image sequence and displays it as a thumbnail on the Read node. Generally, the
Read node does not reformat or resize the sequence in any way, and the node’s properties panel
is updated to display the native resolution and the frame range for the sequence. Note that the
format and proxy format fields in the controls indicate the format of the images, they do not
cause the images read from files to be resized to this format.
3. You can cycle through the available versions of a file using the Alt+Up/Down arrow keys. If you
want to skip ahead to the highest version available, press Alt+Shift+Up arrow key.
USER GUIDE
335
Loading Image Sequences | Importing Image Sequences
Versions must written in the following format in order for Nuke recognize them:
../myFiles/grades/A003_C007_071105_v01_001.r3d
4. If your sequence has a red, green and blue channel but no alpha channel, check the auto alpha
box in the Read node control panel to set the alpha channel to 1. This prevents possible problems
from occurring if Nuke tries to read the alpha channel and one doesn’t exist. The auto alpha box
is unchecked by default.
Note: Nuke reads images from their native format, but the Read node outputs the result
using a linear colorspace. If necessary, you can change the Colorspace option in the Read
node’s properties panel, or insert a Color > Colorspace node to select the color scheme you
want to output or calculate.
Note: The maximum image size the Nuke Viewer can display is 2^28 = 268,435,456 pixels.
This is the same as 16k x 16k, 32k x 8k, 64k x 4k, or 128k x 2k. If your image is larger than
this, it is resized and you get the following warning:
“Viewer image is clipped to <size> x <size>!”
For example, if your image resolution is 60,000 x 4473, Nuke is able to display the image
because the number of pixels is less than 2^28. However, if the resolution is 60,000 x 4474
(more than 2^28 pixels), the image is resized to 59998 x 4474.
In addition to the Viewer, this limit is also applied to the bounding box of the images being
passed between each node.
5. If you have a proxy version of the image sequence, click the proxy field’s folder icon and
navigate to the proxy version. Select Open. If you don’t have a proxy version, don’t worry: Nuke
creates one on the fly according to the proxy scale or proxy format settings you specified in the
project settings.
The proxy file does not need to match the proxy resolution in use. Depending on your settings,
either the full-res or proxy file is scaled to the required proxy size. For more information, see
Proxy Mode.
6. You can access the metadata contained within the read file by clicking the Metadata tab. Once
you know which keys exist in the file, you can reference them in expressions. See Expressions for
more information.
The metadata displayed depends on the file type. For example, a .jpg might only contain input/
keys, whereas QuickTimes contain input/ and quicktime/ keys. See Working with File Metadata
USER GUIDE
336
Loading Image Sequences | To Import an Image Sequence from an External File Browser
To Import an Image Sequence from an External File

Browser
To load an image, you can also drag and drop the image into the Node Graph from an external file
browser (such as Windows Explorer or Mac Finder). To load an entire image sequence this way, drag
and drop the directory that contains the images into the Node Graph.
Notes on Importing QuickTime Files

When reading QuickTime .mov files, Nuke attempts to select the "best fit" combination by reading an
extended list of metadata key/value pairs from the QuickTime header, including nclc atom, gama
atom, and ProRes codec headers.You can manually override the following mov Options in the Read
node properties panel:
• decoder - sets the decode library used to read the file:
• mov32 - uses the full range of QuickTime codecs, but can be slow to process due to extra
complexity during decode.
• mov64 - uses its own packing and unpacking and streams decode/encode for extra processing
speed, but only supports a sub-set of QuickTime codecs.
Note: Nuke defaults to the fastest decoder for the codec used in the file - if you're reading
in a type supported by the mov64 sub-set, Nuke defaults to that reader. Otherwise, the
fallback mov32 reader is used.
• ycbcr matrix - This is only enabled when working with a Y’CbCr-based pixel type. Rec 601 and Rec
709 follow the ITU.BC specifications, whilst Nuke Legacy, Nuke Legacy Mpeg, and Nuke Legacy
YUVS are retained for backwards compatibility.
Note: The default option, Format-based, selects Rec 601 or Rec 709 automatically based
on the format size.
When you set the decoder to mov32, two additional controls are displayed:
• codec - sets the codec used to read the QuickTime file. The codec dropdown defaults to a codec
appropriate for the QuickTime in question, where available, and only lists those that declare
themselves able to read the file.
USER GUIDE
337
Loading Image Sequences | Notes on Importing AVI Files
Note: If you're using the Avid DNxHD codec, Avid AVDn, avoid setting the pixel format
control to r408 as there is a known issue within the codec causing frames to darken with
each frame progression in the sequence.
• pixel format - sets the read pixel format, which includes colorspace, bit depth, layout, and range.
This setting defaults to the best format accepted by the codec, allowing Nuke to perform the
conversion to RGB without the use of an unknown QuickTime transform where possible. RGB pixel
types rely on QuickTime to do the conversion from Y’CbCr when dealing with a non-RGB codec.
Note: When reading QuickTime files, Nuke looks for metadata in the following order and
uses it to govern the settings on the Read node, falling down to each level when the level
above is unavailable or set to a reserved or unknown value:
1. Foundry -specific user data
2. prores header
3. nclc atom
4. gama atom
5. defaults based on codec type
QuickTime .mov files may appear different in Nuke relative to Apple’s Final Cut Pro, because Final Cut
Pro introduces a gamma compensation based on assumptions about the content of the files and the
viewing environment.
To limit the number of background processes that Nuke can run when reading QuickTime files, go to
Preferences and set the number for QuickTime decoders to use in the Performance >
Threads/Processes tab.
Notes on Importing AVI Files

.avi files can be supported by default or only via Nuke’s reader that is based on the FFmpeg open
source library. If you get an error when using .avi files in Read nodes, you may need to use the prefix
mov64: before the file path and file name, for example, mov64:\z:\job\FILM\IMG\final_comp_
v01.####.avi.
Notes on Importing OpenEXR Files

Nuke supports multi-part OpenEXR 2.2 images, which allow you to store your channels, layers, and
views in separate parts of the file in order to speed up processing. You can load multi-part OpenEXR
files in exactly the same way as single-part OpenEXR files.
USER GUIDE
338
Loading Image Sequences | Notes on Importing PSD Files
The OpenEXR file format allows the display window to have the lower left corner in any position.
Unfortunately, Nuke needs all formats to have a lower left corner coordinate of 0,0. In the Read node
control panel, under exrOptions, you can check the offset negative display window box to tell
Nuke to offset the image so that the display window left side x coordinate is 0. If you uncheck the
box, Nuke shrinks the format from both sides the amount that it’s negative in the x coordinate, in
other words treat the area as overscan.
By default, the exr prefix is attached to metadata keys to make them distinct from other metadata in
the tree. If you’d rather read metadata in “as is” without attaching a prefix, enable do not attach
prefix.
When reading in .exr files, you can determine how pixels at the edge of the data window, or
bounding box in Nuke terms, are treated using the edge pixels dropdown:
• plate detect - if the bounding box and format match exactly, then repeat all edges. Otherwise, add
black at the edges.
• edge detect - for each matching edge, repeat the edge pixels. Add black at mismatched edges.
• repeat - always repeat the edge pixels outside the bounding box.
• black - always add black pixels outside the bounding box.
On Linux, the Nuke OpenEXR reader uses a memory-mapping function to improve performance
reading PIZ-compressed .exr files. However, some customers have experienced hanging when
reading large (frame size and number of channels) PIZ-compressed .exr files across an NFS network.
If you experience this problem, you can tell Nuke not to use the mmap function by enabling this
option. You can set it on a case-by-case basis or use a knobDefault in your init.py to always have it
disabled. For more information on knobDefault, see the Nuke Python documentation (Help >
Documentation).
Notes on Importing PSD Files

When loading a layered .psd file, a button called Breakout Layers appears under psd Options in the
Read node control panel. Clicking this “breaks out” the .psd file into separate layers using some
Shuffle nodes, and recombines these layers with a number of PSDMerge nodes. Each layer is grouped
and labeled using a Backdrop node for clarity. In addition, a Crop node is inserted between the
Shuffle and PSDMerge nodes, allowing you to adjust the bounding box for each layer individually.
The PSDMerge node is a type of merge node exclusive to this feature. In the control panel, there is an
operation dropdown menu, a mask field (with an invert checkbox) and a mix slider. If the blend
mode for a layer was set in Photoshop®, Nuke automatically sets this in the operation dropdown.
USER GUIDE
339
Loading Image Sequences | Naming Conventions
Note: The blend modes in PSDmerge are approximated and do not match Photoshop®
exactly.
Naming Conventions
The file names of image sequences generally end in a number before the extension, for example
image0001.rgb, image0002.rgb, image0003.rgb, and so on. When browsing for files like this, you may
notice that the sequence appears as image####.rgb. Here, #### is Nuke’s way of indicating that the
number is in a 4-digit incremental format. For a 3-digit format, such as image001.rgb, the frame
number variable would be ###.
Nuke’s file browser also understands unpadded file names, such as image1.rgb, image2.rgb,
image3.rgb, and so on. They appear as image#.rgb.
Changing the Relation Between the Current Frame and

the
Frame Read In
By default, Nuke assumes an exact relation between the current frame processed, and the frame read
in. For example, at frame 15, Nuke reads in image.0015.rgb. However, you can change this behavior
using the frame parameter on the Read node. For instance, if you have a sequence that runs from
image.0500.rgb to image.1000.rgb, you may want to read in image.0500.rgb at frame 1. Nuke lets you
do this using expressions, specified start frames, and constant offsets. Each method is described
below.
Using Expressions
1. Select Image > Read to import an image sequence.
2. In the Read node controls, set the frame dropdown menu to expression. Enter an expression in
the field on the right. The expression changes the relation between the current frame and the
frame read in.
USER GUIDE
340
Loading Image Sequences | Changing the Relation Between the Current Frame and the Frame Read
For example, if your clip begins from image.0500.rgb and you want to place this first frame at
frame 1 rather than frame 500, you can use the expression frame+499. This way, 499 frames are
added to the current frame to get the number of the frame that’s read in. At frame 1,
image.0500.rgb is read in; at frame 2, image.0501.rgb is read in; and so on.
Another example of an expression is frame*2. This expression multiplies the current frame by
two to get the number of the frame that’s read in. This way, only every other frame in the clip is
used. At frame 1, image.0002.rgb is read in; at frame 2, image.0004.rgb is read in; at frame 3,
image.0006.rgb is read in; and so on.
Specifying a Start Frame for a Clip

2. In the Read node controls, set the frame dropdown menu to start at. Enter a start frame number
in the field on the right. This specifies the frame where the first frame in the sequence is read in.
In other words, all frames are offset so that the clip starts at the specified frame.
For example, if your sequence begins from image.0500.rgb and you enter 1 in the field,
image0500.rgb is read in at frame 1. Similarly, if you enter 100 in the field, image0500.rgb is read
in at frame 100.
Offsetting All Frames by a Constant Value

2. In the Read node controls, set the frame dropdown menu to offset. Enter a constant offset in the
field on the right. This constant value is added to the current frame to get the number of the
frame that’s read in.
frame 1 rather than frame 500, you can use 499 as the constant offset. This way, 499 is added to
USER GUIDE
341
Loading Image Sequences | Changing the Relation Between the Current Frame and the Frame Read
the current frame to get the frame that’s read in. At frame 1, image.0500.rgb is read in; at frame 2,
image.0501 is read in, and so on.
You can also use negative values as the constant offset. For example, if you use the value -10,
Nuke subtracts ten from the current frame to get the frame that’s read in. At frame 20,
USER GUIDE
342
Reformatting Image Sequences | To Insert a Reformat Node:
Reformatting Image Sequences

When you import image sequences, Nuke stores their format settings and makes them available to
the Reformat node. You can then use the Reformat node to resize and reposition your image
sequences to a different format. Reformat nodes also allow you to use plates of varying image
resolution on a single script without running into issues when combining them.
To Insert a Reformat Node:

1. Make sure the Read node you added is currently selected.
2. Select Transform > Reformat.
The Reformat node is inserted in the script, and its properties panel opens.
3. From the output format dropdown menu, select the format to which you want to output the
sequence. If the format does not yet exist, you can select new to create a new format from
scratch. The default setting, [root.format], resizes the image to the format indicated on the
Project Settings dialog box.
4. You can now use the same Reformat node for any other Read nodes in the script. Simply select
the Reformat node and Edit > Copy. Select another Read node in the script and Edit > Paste.
USER GUIDE
343
Image Caching | To Insert a Reformat Node:
Image Caching
To ensure fast playback, Nuke uses several ways of caching data. Some of these include the following:
• The Viewer cache (also referred to as disk cache, which shares the DiskCache node's location) saves
the scanlines of the displayed image to the disk cache directory. This location can be set in the
Preferences panel (see The Cache Directory and Defining the Settings for Caching). When the
Viewer displays an image, it reads from this cache of pre-rendered scanlines. This way, Nuke can
quickly display image data when returning to a previously cached set of scanlines. You can specify
the total amount of space for disk cache in Preferences > Performance > Caching > comp disk
cache size . If the disk cache needs more space, the cache that was least recently used is disposed
of to free more space. You can clear all of the disk cache by selecting Cache > Clear Disk Cache
from the top menu bar.
Note: Clearing the Viewer cache (disk cache) also clears the DiskCache node data.
• The Buffer cache stores scanline data as calculated at selected intermediate stages in the Node
Graph. This speeds up the end result of the Node Graph by caching parts of the Node Graph that
haven't changed. Nuke automatically determines these intermediate cache points based on several
things, including the Node Graph structure, which properties panels are open, and so on.
Tip: You can also force this behavior manually by selecting the cached checkbox in the
node properties panel.
The Buffer cache uses RAM and you can specify the total amount of space in Preferences >
Performance > Caching > comp cache size (%). When the buffer cache needs more space, it
disposes of the first cached data that was stored (first in, first out). To clear all of the buffer cache
you can either select Cache > Clear Buffers from the top menu bar, or press F12 on the keyboard.
• The Playback cache also helps improve smooth playback. Playback cache uses RAM to temporarily
store the frames in the Viewer. When you first play back a clip in the Viewer, an orange bar appears
automatically displaying the progress of the playback cache. After the playback cache is complete
(and the orange line is complete) the frames play back smoothly. You can temporarily disable
playback caching by either selecting the pause button above the Viewer or by pressing P. To clear
playback cache select Cache > Clear Playback Cache.
• You can use the DiskCache node to cache parts of the Node Graph. This caches scanlines to disk
from its input as they are requested by its output. DiskCache data is stored in the same place as the
Viewer cache data and can share the data provided the Viewer is operating in full float (32-bit). See
Using the DiskCache Node for more information.
USER GUIDE
344
Image Caching | To Insert a Reformat Node:
Note: You can clear all caches by selecting Cache > Clear All from the menu bar.
• If you find that loading files from a remote server slows down your processing, you can localize the
files that you’re working with for speedier processing. See Localizing Files for Better Performance.
USER GUIDE
345
The Cache Directory | Defining the Settings for Caching
The Cache Directory

Both the automatic caching and the DiskCache node use the same cache directory defined using the
Preferences dialog. In the Preferences, you can also set the maximum size you allow the disk cache
to consume. For more information on the caching preferences, see Defining the Settings for Caching
below.
Note that the cached images have unique names reflecting their point of output location in the script.
This means that you can cache images from multiple nodes in the script without overwriting
previously cached images.
Defining the Settings for Caching

You can define the settings for caching in the Preferences > Performance > Caching.
1. Select Edit > Preferences (or press Shift+S).
The Preferences dialog displays.
2. Under Caching > Disk Caching, specify where you want Nuke to cache out data to disk. Pick a
local disk (for example, c:/temp), preferably with the fastest access time available.
The environment variable NUKE_DISK_CACHE can be used to override this setting. For more
information, see Environment Variables.
3. Using the comp disk cache size control, pick the maximum size the disk cache can reach. Ensure
there is enough space on the disk for this to be reached. The default value is 10 GB, but a larger
value, such as 50 GB, can often be used.
The environment variable NUKE_DISK_CACHE_GB can be used to override this setting. For more
information, see Environment Variables.
4. Click OK to update preferences and then restart Nuke.
Once these settings are defined, the automatic caching of images is enabled. The Viewer caches each
frame it displays in the directory specified. If you add a DiskCache node into your script, it also uses
the same directory.
The disk cache is preserved when you quit Nuke so it can be used again later. However, when the
cache becomes full, old items are automatically deleted to make room for newer items.
Clearing the Disk Cache

When the disk cache becomes full, old items are automatically deleted. If necessary, you can also
empty the disk cache manually. You may want to do this if, for some reason, wrong images are
USER GUIDE
346
The Cache Directory | Clearing the Disk Cache
displayed in the Viewer.
To Empty the Disk Cache

From the menu bar, select Cache > Clear Disk Cache.
USER GUIDE
347
Using the DiskCache Node | To Cache Images Upstream
Using the DiskCache Node

The DiskCache node caches to disk scanlines from its input as they are requested by its output. This
can be useful, for example, if:
• you are working on a large, complex node tree. Using the DiskCache node, you can break the node
tree into smaller sections and cache any branches that you are no longer working on.
• you are reading in images from a network. If you insert a DiskCache node after a Read node, the
image is cached locally and displayed faster.
• you are painting or rotoscoping. If you insert a DiskCache node before a RotoPaint node, flipping
frames becomes faster.
The cached images are saved in the same directory as the images the Nuke Viewer caches
automatically. You can set the location and size of this directory in the Preferences. For more
information, see Defining the Settings for Caching.
Note: Even though the DiskCache node and the automatic Viewer caching use the same
cache directory, they do not share the same cache files. Therefore, using a DiskCache node
does not create cache files for the Viewer and does not necessarily speed up playback.
Instead, if placed at strategic, expensive parts of a node tree, it can speed up calculations, as
Nuke can reference the cached data rather than recalculate it.
Unlike the images in the Viewer cache, the images created by the DiskCache node affect your
rendered output and are always saved as full floating point images.
If you make a change in the nodes upstream, the affected cached images are discarded and
automatically recalculated.
Note: When executing a script on the command line, the DiskCache nodes are NOT
automatically executed.
To Cache Images Upstream

1. Set the zoom level in the Viewer. By default, only the lines displayed in the Viewer are cached.
2. Select Other > DiskCache to insert a DiskCache node after the last node in the section of the
node tree that you want to cache.
3. From the channels dropdown menu, select which channels to cache.
USER GUIDE
348
Using the DiskCache Node | To Cache Images Upstream
Nuke caches the selected channels of the current frame at the current zoom level. From this point
on, Nuke references the cached data instead of constantly recalculating the output of the
preceding nodes.
As you pan and zoom around, new parts of the image are cached.
4. If you want to cache more than the current frame and zoom level, click the Precache button in
the DiskCache properties and enter a frame range in the dialog that opens.
This forces Nuke to cache all frames specified. All lines are cached regardless of what is shown in
the Viewer. Where the required images are partly cached already, Nuke only calculates what is
missing.
USER GUIDE
349
Localizing Files for Better Performance | Setting Localization Preferences
Localizing Files for Better

Performance
You may find that sometimes loading files from a remote server slows down processing. Nuke has
the facility to cache files locally, either individually or by setting an automatically localized folder
(NUKE_TEMP_DIR/localize, by default), to help guarantee playback stability. Local caching is controlled
initially in the Preferences dialog, then on a file-by-file basis.
Setting Localization Preferences

1. Press Shift+S to open the Preferences dialog and navigate to Performance > Localization.
2. Set the required localization mode:

• on - checks for updates to source clips for all localization policies, and localizes those files set to
On or From auto-localize path automatically. On demand source clips must always be
localized manually.
• manual - checks for updates to localized source clips and prompts you to update them
manually. Only those with the policy set to off are not updated.
USER GUIDE
350
• off - no source clips are localized, regardless of the their localization policy.
Note: The current localization mode is displayed in the status bar at the bottom-right of the
interface.
3. Set the default localization policy for new Read nodes using the dropdown:
Note: The localization policy for existing Read nodes in the script must be set individually.
See Managing Localization for more information.
• on - always localize Read nodes with this policy.

• from auto-localize path - localize these Read nodes automatically if they reside in the auto-
localize from directory.
• on demand - only localize these Read nodes when you manually update them. See Updating On
• off - never localize these Read nodes.
4. Enter a file path for auto-localize from, if required.
Any files that reside in this directory are automatically cached when brought into Nuke, providing
that the Read node's localization policy is set to from auto-localize path.
Note: Localization in both the timeline and Node Graph is paused during compositing
Viewer playback so that performance is not affected.
5. Enter a file path for localize to. Leaving this field as the default creates a sub-directory in the
Temp Directory as the local cache.
Note: On Windows, files saved to the localize to directory replace \\ (double back slashes)
and : (colon drive signifiers) with underscores so that the file path works as expected
between operating systems. For example:
6. Enter a value for limit to (GB) to control how much disk space is available in the cache directory.
USER GUIDE
351
Note: Negative values in this field reserve the specified amount of space at all times. For
example, -2 stops 2 GB of memory being used for caching.
7. You can specify the time interval (in minutes) before localized files are checked for updates using
the check for updated files every ## mins control.
The default setting checks for new versions of files every 30 minutes, but you can set this control
to any value. Any Read nodes that have changed since the last update check are flagged red in the
Node Graph.
Tip: You can change the default localization indicators using the Appearance controls.
If read source files when localized files are out of date is enabled, Read nodes referencing cached
files that have changed since they were localized revert to reading the source files. Read nodes that
are reading from the source files are marked with a striped red bar on the Read node.
USER GUIDE
352
Localizing Files for Better Performance | Managing Localization
Enabling hide out of date progress bar hides the localization state of out of date files so that the
Read node appears the same as regular Read nodes.
Note: The out of date localized files are not discarded, disabling read source files when
localized files are out of date picks up the out of date files instead of reading the source
files.
Managing Localization
As well as the overall Preferences for when new Read nodes should be localized, you can set
localization on a file-by-file basis. The localization policy for any existing Read nodes in your project
must be managed individually.
Tip: If you find that localization is slow to copy files, you can increase the number of threads
that Nuke uses to process jobs. Set the NUKE_LOCALIZATION_NUMWATCHERS environment
variable to the number of threads you want to use. See Environment Variables for more
information.
1. Open the Read or ReadGeo node's Properties panel.

2. Set the localization policy dropdown to one of the following options:
USER GUIDE
353
Localizing Files for Better Performance | Managing Localization
• On - the files are localized, regardless of location, as long as the limit to (GB) limit is not
breached.
• From auto-localize path - the files are localized if they reside in the auto-localize from
directory, as long as the limit to (GB) limit is not breached.
• On Demand - the files are only localized when you update them manually. See <XREF> for more
information.
• Off - the files are never localized, regardless of location.
As Read nodes are localized, an amber progress bar displays in the thumbnail. Fully cached clips
are marked with a green bar at the top of the thumbnail and out-of-date clips are marked in red.
Note: Container formats, such as .mov and .r3d, do not display progress bars during
localization. The Read node only shows the green indicator once the entire file is localized
successfully.
3. If you need to pause localization temporarily, navigate to Cache > Localization and select Pause.
USER GUIDE
354
Localizing Files for Better Performance | Updating On Demand Clips
4. If you find that your cache is filling up regularly, you can:

• increase the amount of available space for localization by raising the limit to (GB) preference,
• navigate to Cache > Localization > Delete Unused Local Files (see Clearing Localized Files), or
• manually clear files from the cache directory in NUKE_TEMP_DIR/localize, by default.
5. You can force Nuke to check for updated source files, rather than waiting for the check for
updated files every interval to expire, by selecting Force Update for All files, just Selected Read
nodes, or On Demand only.
Note: Each file has its own update time, which is reset whenever the source files are
checked.
The following table is a quick reference guide to when and how Read nodes are localized, where:
• green - Read nodes are localized automatically.
• amber - Read nodes are localized when updated manually.
• red - Read nodes are not localized.
Read Node Preference
System
Preference
on
manual
off
Updating On Demand Clips

Read nodes with their localization policy set to on demand are polled to check for updates at the
interval set in the Preferences. If the file has changed, the Read node is marked with a red bar to
show it is out of date.
To update a single on demand Read node:

1. Double-click the Read node in the Node Graph to display its Properties.
2. Click Update.
USER GUIDE
355
Localizing Files for Better Performance | Clearing Localized Files
Note: If a Read node's localization policy is set to on demand and it hasn't been localized
previously, clicking the Update button localizes the file.
The local copy of the files is updated from the remote source files.
To update all on demand Read nodes:

1. Navigate to Localization > Force Update.
2. Click On demand only.
The local copy of all on demand Read nodes is updated from the remote source files.
Clearing Localized Files

Localizing a large amount of files can fill up the localization cache quite quickly if you leave the limit
to (GB) preference at the default 10 GB. When the cache runs out of space, a Failed to Localize File
USER GUIDE
356
Localizing Files for Better Performance | Clearing Localized Files
dialog displays and localization pauses.
You can delete localized files by clicking Delete Unused Local Files (or by navigating to Cache >
Localization > Clear Unused Local Files). Nuke displays a dialog containing all the files that are
marked for delete.
Tip: You can also open the Preferences from this dialog to adjust the localization behavior,
such as increasing the limit to (GB) preference.
Click Continue to delete the localized files or Cancel to keep the cached files.
USER GUIDE
357
Saving Scripts and Recovering Back-Ups | Saving Scripts
Saving Scripts and Recovering

Back-Ups
You know the mantra: save and save often.
Nuke provides three ways to save your scripts, or comps, making it easy to version them. There's also
an automatic timed backup, which you can turn off if you're feeling brave - but we sure don't
recommend it.
Saving Scripts
There are three ways of saving scripts:
• To save a new script, select File > Save Comp as (or press Shift+Ctrl/Cmd+S).
• To update changes to a script already saved, File > Save Comp (or press Ctrl/Cmd+S).
• To save and upgrade to the next version, File > Save New Comp Version (or press Alt+Shift+S).
To Save a Script
1. Select File > Save Comp as.
The Save script as dialog opens.
2. Browse to the directory where you want to store the script. For instructions on using the file
browser, see Using the File Browser.
3. In the field in the bottom of the dialog, enter a name for the script after the folder path, for
example firstscript_v01.nk.
4. Click Save.
Tip: The _v01 string in the end of a script name allows you to use the Save New Comp
Version feature. Selecting File > Save New Comp Version saves the current version of your
script and increments its name (that is, saves the different versions under different names
using _v01, _v02, _v03, and so on, in the end of file names). This only works when the file
name includes a number that can be incremented.
USER GUIDE
358
Saving Scripts and Recovering Back-Ups | Automatic Back-Up of Scripts
Automatic Back-Up of Scripts

You can define where and how often Nuke makes automatic back-ups your files, or turn off the
autosave function.
To Define Autosave Options for a Script

1. Select Edit > Preferences.
The Preferences > General panel opens.
2. Edit the following settings:
• idle comp autosave after - to define how long (in seconds) Nuke waits before performing an
automatic back-up after you have left the system idle.
• force comp autosave after - to define how long (in seconds) Nuke waits before performing an
automatic back-up regardless of whether the system is idle.
• autosave comp filename - to define where and under what name Nuke saves your automatic
back-up files. By default, the files are saved in the same folder as your project files with the
extension .autosave. To change this, enter a full directory path name in the autosave filename
field.
3. Click Save.
Note: For the automatic back-up to work, you must save your script first so that the
autosave can reference the file. We’d hate for you to lose your work, so please do this early
on in the process!
To Turn off Automatic Back-Up

1. Select Edit > Preferences.
The Preferences > General panel opens.
2. Set the idle comp autosave after and force comp autosave after fields to 0.
From now on, Nuke does not perform any automatic back-ups, and you are more likely to lose
your work in the case of a system or power failure.
Recovering Back-Ups
After experiencing a system or power failure, you are likely to want to recover the back-up files
created by Nuke’s autosave function.
USER GUIDE
359
Saving Scripts and Recovering Back-Ups | Recovering Back-Ups
1. Relaunch Nuke.
A dialog opens that asks you if you want to recover the autosave file.
2. Click OK.
Nuke opens the back-up file for your use.
There may be times when you don’t want to load the autosave file and rather need to load the last
saved version. For example, consider a situation where you modified a script, but decided not to
commit the changes and so exited Nuke without saving. In all likelihood Nuke autosaved some or all
of your changes, in which case if you open the autosaved file you are not working on the original
script, as intended. If you accidentally open an autosaved script, then simply close it and reload the
last saved version.
Note: Breakpad crash reporting allows you to submit crash dumps to Foundry in the
unlikely event of a crash.
By default, crash reporting is enabled in GUI mode and disabled in terminal mode. You can
toggle reporting on and off using the --crashhandling 1 or 0 option on the command line or
by setting the environment variable NUKE_CRASH_HANDLING to 1 or 0.
When crash handling is enabled in GUI mode, you can control whether reports are
automatically submitted or not using the --nocrashprompt command line option or by
setting the environment variable NO_CRASH_PROMPT to 0.
Crashes in terminal mode are automatically submitted when crash handling is enabled.
USER GUIDE
360
Closing Scripts | Recovering Back-Ups
Closing Scripts
To close a script:
1. Select File > Close Comp (or press Ctrl/Cmd+W).
2. If you have made any unsaved changes to the script, Nuke prompts you to select whether to save
them. Click Yes to save your changes or No to ignore them.
Nuke quits and relaunches, as though you ran it again. It does everything it does at start-up, apart
from displaying the splash screen. Therefore, you can use Ctrl/Cmd+W as a quick way to clear
memory, reread plug-in paths, and reload the init.py and menu.py files. (The init.py and
menu.py files are files that Nuke runs at start-up and can be used for configuring Nuke. You may
want to reload these files if you have made any changes to them.)
USER GUIDE
361
Loading Scripts | Recovering Back-Ups
Loading Scripts
When you have built a script, or comp, and saved it and want to come back to it later, you need to
load in an entire script file. You recognize Nuke’s script files from the extension .nk (for example
firstscript.nk).
1. Select File > Open Comp (or press Ctrl/Cmd+O).
The Script to open dialog appears.
2. Browse to the script you want to open. For instructions on using the file browser, see Using the
File Browser.
3. Click Open.
Note: Some NukeX plug-ins are not supported by Nuke, and likewise, some Nuke nodes are
not supported in Nuke Assist. Unsupported plug-ins are displayed in the Node Graph with
errors, and unsupported nodes are outlined in red.
The Viewer renders the output of the node tree whether the components are supported or
not, but you cannot modify the output of unsupported plug-ins and nodes.
USER GUIDE
362
Defining Frame Ranges | Recovering Back-Ups
Defining Frame Ranges

Several dialogs in Nuke, such as the Frames to render and Frames to flipbook dialogs, prompt you
for a frame range. To define one, you need to enter a starting frame and an ending frame, separated
by a dash. For example, to restrict an action to frames 1, 2, 3, 4, and 5, you would use 1-5 as the
frame range.
The following table gives you more examples of frame ranges you can define.
Frame Range Expands To
3 3
-3 -3
1348 1, 3, 4, 8
1-10 1, 2, 3, 4, 5, 6, 7, 8, 9, 10
-3-4 -3, -2, -1, 0, 1, 2, 3, 4
-8--5 -8, -7, -6, -5
1-10×2 1, 3, 5, 7, 9
(frame range from 1 to 10 in

steps of 2)
1-10×3 1, 4, 7, 10
(frame range from 1 to 10 in

steps of 3)
1-4×1 8-10×1 12-14×1 1, 2, 3, 4, 8, 9, 10, 12, 13, 14

(multiple ranges separated by
spaces)
You can use the above ways of defining a frame range everywhere in Nuke. In addition to dialogs,
they can be used on the command line (where any frame ranges should be preceded by the -F switch)
and in Python statements. For more information, see Command Line Operations and the Nuke
Python documentation (Help > Documentation).
USER GUIDE
363
Reformatting Elements
These pages teach you how to reformat images through scaling, cropping, and pixel aspect
adjustments. You will also learn to adjust bounding boxes to minimize processing and rendering
times.
Using the Reformat Node

You can use the Reformat node for three different purposes:
1. To generate image sequences that match a desired image format in terms of both resolution and
pixel aspect ratio (the width to height ratio of the format’s individual pixels).
2. To create thumbnails (low resolution frames which you might post to the web in order to
storyboard a sequence). The node scales the frame until it fits inside a rectangle whose
dimensions you specify. It also sets pixel aspect ratio to one (square).
3. To scale images. The scale factor is rounded slightly so that the output image has an integer
number of pixels in the direction you select in the Reformat node’s controls.
Converting Images to a Desired Image Format

When you read in elements, Nuke stores their format settings and makes them available to the
Reformat node. You can then apply one of the existing formats to your images, or create, edit, and
delete formats yourself.
When creating a new format from scratch, you define the overall resolution, the cropped resolution
(optional) and the pixel aspect ratio. As you define these parameters, the Reformat operator
graphically displays them for you in the manner shown below.
USER GUIDE
364
Reformatting Elements |
To Create a New Output Format:

1. Click Transform > Reformat to insert a Reformat node at an appropriate place in your script
(generally before a Write node).
2. Connect a Viewer to the output of the Reformat node so you can see the effect of your changes.
3. Select new from the output format dropdown menu. The New format dialog appears.
4. Type a name for the new format in the name field.
5. In the file size fields, type the full output resolution (in pixels).
6. If you want to crop the full output resolution (for example, to create a letter box):
• Check image area.
• Increment the x field to define the left boundary of the crop. (The display updates to show you
the left boundary of the crop relative to the full-size input.)
• Increment the y field to define the bottom boundary of the crop.
• Increment the r field to define the right boundary of the crop.
• Increment the t field to define the top boundary of the crop.
7. If the destination display device for the image sequence uses non-square pixels, type the
appropriate pixel aspect ratio in the pixel aspect field (for example, if your destination is a digital
video display, type 1.1).
Note: You can also add formats to Nuke via entries to the menu.py file:
1. Open the menu.py file (located in same directory as your Nuke executable).
2. Add an entry similar to the following example:
nuke.addFormat ("720 486 0 0 720 486 0.9 NTSC_video")
where the numbers specify, respectively, the format’s full horizontal resolution, full vertical
resolution, left crop position, bottom crop position, right crop position, top crop position,
and pixel aspect ratio; and where the final text string designates the format’s name.
3. Save and close the menu.py file. The next time you launch Nuke the format is available
for selection from the Project Settings dialog, Reformat node properties panel, and
elsewhere.
To Edit a Format:
1. Select the format you wish to edit from the output format dropdown menu.
2. From the same dropdown menu, select edit. The Edit format dialog appears.
3. Edit the name, file size, image area, and pixel aspect fields as necessary.
4. Click OK to save the changes to the format.
USER GUIDE
365
Reformatting Elements | Creating Thumbnails
To Delete a Format:
1. Select the format you wish to delete from the output format dropdown menu.
2. From the same dropdown menu, select delete. The format is removed from the menu.
To Apply a Format:
1. If necessary, click Transform > Reformat to insert a Reformat node at an appropriate place in
your script (generally before a Write node).
3. From the type dropdown menu, select to format.
4. Select the format you wish to apply from the output format dropdown menu.
5. From the resize type field, select the method by which you want to preserve or override the
original aspect ratio. Select:
• width to scale the original until its width matches the format’s width. Height is then scaled in
such a manner as to preserve the original aspect ratio.
• height to scale the original until its height matches the format’s height. Width is then scaled in
• fit to scale the original until its smallest side matches the format’s smallest side. The original’s
longer side is then scaled in such a manner as to preserve original aspect ratio.
• fill to scale the original until its longest side matches the format’s longest side. The input’s
shorter side is then scaled in such a manner as to preserve original aspect ratio.
• distort to scale the original until all its sides match the lengths specified by the format. This
option does not preserve the original aspect ratio, so distortions may occur.
6. When cropping the output, check center to position the crop area at the center of the frame.
7. Select the appropriate filtering algorithm from the filter dropdown menu (see Choosing a
Filtering Algorithm).
8. When scaling an image with Key, Simon, and Rifmen filters, you may see a haloing effect which is
caused by pixel sharpening these filters employ. If necessary, check clamp to correct this
problem.
Creating Thumbnails
3. From the type dropdown menu, select to box.
4. In the width and height fields, type the output dimensions. The units are pixels.
USER GUIDE
366
Reformatting Elements | Scaling Image Sequences
5. Use the resize type dropdown menu to select the method by which you preserve or override the
original pixel aspect ratio. Select:
• width to scale the original until its width matches the value in the width field. Height is then
scaled in such a manner as to preserve the original aspect ratio (this means that the output you
specified in height may not match the result).
• height to scale the original until its height matches the value in the height field. Width is then
scaled in such a manner as to preserve the original aspect ratio (this means that the output you
specified in width may not match the result).
• fit to scale the original until its smallest side matches the corresponding value in width/height.
The longer side is then scaled in such a manner as to preserve the original aspect ratio.
• fill to scale the original until its longest side matches the corresponding value in width/height.
The smallest side is then scaled in such a manner as to preserve the original aspect ratio.
• distort to scale the original until its sides match the values in the width/height fields. This
problem.
Scaling Image Sequences

3. From the type dropdown menu, select scale.
4. In the scale fields, enter scale factors for the width and the height. To scale each direction
separately using different scale factors, click the 2 button.
5. Use the resize type dropdown menu to select the method by which you preserve or override the
original pixel aspect ratio. Select:
• width to scale the original so that it fills the output width. Height is then scaled in such a
manner as to preserve the original aspect ratio.
• height to scale the original so that it fills the output height. Width is then scaled in such a
• fit to scale the original so that its smallest side fills the output width or height. The longest side
is then scaled in such a manner as to preserve the original aspect ratio.
• fill to scale the original so that its longest side fills the output width or height. The smallest side
USER GUIDE
367
Reformatting Elements | Cropping Elements
• distort to scale the original so that both sides fill the output dimensions. This option does not
preserve the original aspect ratio, so distortions may occur.
problem.
Cropping Elements
To crop a frame is to cut out the unwanted portions of the image area.
The original image. Cropping the image.

1. Click Transform > Crop to insert a Crop node at an appropriate place in your script.
2. Connect a Viewer to the output of the Crop node so you can see the effect of your changes.
3. Define the crop boundaries:
• In the Viewer, drag on any side of the frame to reposition it.
• Or, in the Crop properties panel, increment or decrement the box field (x stands for left side, y
for bottom side, r for right side, and t for top side).
4. To fill the cropped portion with black, check black outside. To fill the cropped portion by
expanding the edges of the image, uncheck black outside. To adjust the image output format to
match the cropped image, check reformat.
5. If you wish to vignette the edges of the cropped portion, increment the softness field.
USER GUIDE
368
Reformatting Elements | Cropping Elements
USER GUIDE
369
Adjusting the Bounding Box
The bounding box defines the area of the frame that Nuke sees as having valid image data. The larger
the bounding box is, the longer it takes Nuke to process and render the image. To minimize
processing and rendering times, you can crop the bounding box. Occasionally, the bounding box may
also be too small, in which case you need to expand it.
Note: Other Nuke functions, such as Transforms and Merges, can also affect the size of the
bounding box, see Transforming in 2D and Merging Images for more information.
Resizing the Bounding Box

To adjust the bounding box, you can use the AdjBBox and CopyBBox nodes. The AdjBBox node crops
and expands the bounding box edges, whereas with the CopyBBox node you can copy a bounding
box from one input to another. If needed, you can also add a black outside edge to the bounding box
using the BlackOutside node.
The AdjBBox node expands or crops the edges of the bounding box by a specified number of pixels.
An expanded bounding box. A cropped bounding box.
Tip: You can enable a warning to indicate when the bounding box is greater that the format
in Nuke's Preferences. See Bounding Box Warnings for more information.
For example, if you have an image with lots of black (0,0,0,0), you can adjust the bounding box to
contain just the useful area so that Nuke won’t waste time computing results where there is no
change.
USER GUIDE
370
Adjusting the Bounding Box |
1. Select Transform > AdjustBBox to insert an AdjBBox node after the image whose bounding box
you want to resize.
2. Connect a Viewer to the AdjBBox node, so you can see the effect of your changes.
3. In the AdjBBox controls, adjust the Add Pixels slider to increase or decrease the size of the
bounding box. By default, 25 pixels are added to the edges of the bounding box.
Nuke expands or crops the edges of the bounding box. If the bounding box is cropped, whatever
is outside the bounding box area is replicated towards the edges of the image.
Copying a Bounding Box from One Input to Another

Some Nuke operations, such as a merge, can cause an expansion of the bounding box area because
Nuke does not know that the extra area is going to be black or another constant color. Often, you can
fix this by copying the bounding box from one of the inputs to the resulting image, thus cutting off
this extra area. For this, you can use the CopyBBox node.
1. Select Merge > CopyBBox to insert a CopyBBox node after the node whose bounding box you
want to use.
2. Connect the image whose bounding box you want to copy to the CopyBBox node’s input A, and
the image onto which you want to copy the bounding box to input B.
Nuke copies the bounding box from input A to input B. Whatever is outside the copied bounding
box area in image B gets replicated towards the edges of the image.
USER GUIDE
371
Adjusting the Bounding Box | Bounding Box Warnings
Adding a Black Outside Edge to the Bounding Box

If you adjust a bounding box with the AdjBBox or CopyBBox node, you may notice that whatever is
outside the bounding box area gets replicated towards the edges of the image. If necessary, you can
remove these replicated edge pixels and fill everything outside the bounding box area with black. To
do this, use the BlackOutside node.
A cropped bbox with replicated edges. The effect of the BlackOutside node.
1. Select the image whose edges outside the bounding box you want to fill with black.
2. Select Transform > BlackOutside to add a BlackOutside node in an appropriate place in your
script.
Nuke fills everything outside the bounding box area with black.
Bounding Box Warnings

Zooming into the Viewer to work on a shot means that you can't always see the extent of the
bounding box in relation to the format, which can result in unnecessary processing.
To make it easier to see the state of your bounding box, Nuke can display visual warnings on the
nodes that affect the bounding box.
Warning: If you enable the Bounding Box Warning, Nuke performs extra processing steps
to identify problematic nodes, which may result in performance degradation.
To enable the warnings, in Nuke's Preferences under Panels > Node Graph, enable Bounding Box
Warning:
• red rectangle with dotted stroke - the indicated node creates a bounding box greater than the
format.
USER GUIDE
372
• dotted stroke without the red rectangle - the bounding box size is greater than the format at the
indicated node, but the bounding box size has been set by an upstream node.
The bbox warning threshold controls how far past the edge of the format the bounding box can
expand before the warning is displayed in the Node Graph. For example, if you're working with UHD_
4K footage and the default 10% threshold, you can expand the bounding box horizontally by 384
pixels before the warning is displayed.
USER GUIDE
373
Tip: You can set the color of the warning rectangle in the Preferences under Panels >
Node Graph > Bounding Box Warning.
USER GUIDE
374
Channels
Digital images generally consist of the four standard channels: red, green, blue, and alpha. Nuke
allows you to create or import additional channels as masks, lighting passes, and other types of
image data.
Introduction
A Nuke script can include up to 1023 uniquely named channels per compositing script. For example,
you can combine multiple render passes from a 3D scene - an image from the red, green, and blue
channels, a depth mask (z-depth channel), a shadow pass, a specular pass, lighting passes, and
multiple mattes all stored within one image sequence in your composite.
Note: When a script is saved, any channels that are not referenced in the script are
discarded automatically.
When creating channels and layers, bear in mind these good practice guidelines:
• Ensure that all layers use the same channel names in the same order. This avoids complications
with multilayer .exr files imported into Nuke.
• Always use proper names for channels, never just a single letter.
• Always create a custom layer for custom channels, don't add to the existing default layers.
• Never use more than four channels per layer. Nuke only has a four channel interface.
A current Channel Count is displayed in the bottom-right of the interface, which changes color as the
number of channels increases. The default threshold is 1023, but you can set the limit in the
Preferences under Project Defaults > Channel Management > Channel Warning Threshold.
The Channel Count turns yellow if you exceed the Channel Warning Threshold and red if the
Channel Count is equal to or greater than the maximum channel value 1023.
Note: Nuke does not remove unused channels until you close and reopen a script, so the
Channel Count does not decrease when you remove Read nodes from the Node Graph.
USER GUIDE
375
Channels | Quick Start
Quick Start
Here's a quick overview of the workflow:
1. Channels in Nuke are always a part of a layer. You can create new channels and layers using the
new option in the channel selection dropdown menus (such as output and mask) in a node’s
properties panel. For more information, see Object Material Properties.
2. Using the channel selection controls you can select which channels the node is processing and
outputting, or using as a mask when color correcting for instance. For more information, see
Calling Channels and .
3. The channels can also be linked to other channel controls through the Link menu. For more
information, see Linking Channels Using the Link Menu.
4. Using the Shuffle and ShuffleCopy nodes, you can rearrange your input channels and apply the
result in the output. For more information, see Swapping Channels.
USER GUIDE
376
Understanding Channels and Layers | Channels
Understanding Channels and

Layers
At a very basic level, channels in Nuke carry image data and layers are containers for these channels,
up to the maximum of 1023 channels per script. When elements are rendered out in the OpenEXR
format, for example, you can operate on multiple channels from a single image.
The layer name, in this case rgba, depth, or masks,

followed by the channel name.
Channels
Think of a channel as a container that contains image data. Once created or read into your
composite, the image data stored in a channel is available downstream in the network until the value
is replaced with something else or the channel is removed. The channel may even be “empty” -
depending on where you reference it in the compositing network.
Layers
All channels in a script must exist as part of a layer (also called a channel set). You’re probably familiar
with the default layer - rgba - which includes the channels with pixel values of red, green, and blue,
and also the alpha channel for transparency.
All channels in a composite must belong to at least one layer. Some channels, like alpha, may be
available in other layers, too. Channel names always include the layer name as a prefix, like this:
USER GUIDE
377
Understanding Channels and Layers | Layers
layer_name.channel_name.
By default, every script has a layer called rgba. When you first import an image element, Nuke
automatically assigns its channels to the rgba layer - that is, the image channels are named rgba.red,
rgba.blue, rgba.green, and rgba.alpha.
The rgba layer allows for the standard four-channel workflow of most node-based compositing
systems. However, you’re not limited to these four channels. You can create new channels and assign
them to new layers up to the limit of 1023 channels per script.
USER GUIDE
378
Creating Channels and Layers | To Create a New Layer and/or Channel
Creating Channels and Layers

It’s important to understand that many types of nodes allow you to direct their output to a specific
channel and parent layer. You have the option of processing these channels in each subsequent
node, or leaving them unchanged.
Many nodes feature an output or channels setting, which lets you direct the output of the current
node to a specific layer and channel. You can also use the output or channels dropdown menu to
create new layers and channels.
Some nodes do not include an output or channels setting in their parameters. For these, you can
connect other nodes, such as Channel Copy or Shuffle, to create and manage channel output in the
node tree.
To Create a New Layer and/or Channel

1. Open the properties panel for the node whose output creates the new channel.
2. From the output or channels dropdown menu, select new.
3. Under Name, enter the name of the layer, and under Channels the new channel name.
Note: You can either use a new layer name to create a new layer, or enter a layer you’ve
created previously. You can’t create new channels into layers that are built into Nuke (such as
mask).
4. Click OK.
USER GUIDE
379
Creating Channels and Layers | To Create a New Layer and/or Channel
Note: You can also create new channels with the Shuffle and ShuffleCopy nodes. These are
explained later, under Swapping Channels.
USER GUIDE
380
Calling Channels | To Create a New Layer and/or Channel
Calling Channels
By default, most nodes in Nuke attempt to process the current channels in the rgba set and place the
output in those same channels. However, many nodes also contain an input dropdown menu which
lets you select the channels you want to process, and an output dropdown menu to select the
channel(s) where the results should be stored.
Some nodes also contain mask controls and a mask input connector, which let you select a channel
for use as a matte to limit operations such as color corrections. Using these mechanisms, you can
point the output of almost any node in the script to any available channel.
The script below attempts to clarify these concepts. Note the script generates six channels (though it
could just as well generate 1023). The steps below describe how each channel was created.
USER GUIDE
381
Calling Channels | Viewing Channels in the Viewer
A six-channel script.
1. The script reads in the foreground, creating three channels (red, green, and blue), which are by
default assigned to the rgba set. Channel count: 3
2. A low contrast key (soft) is pulled and assigned to a new layer called mattes. Channel count: 4
3. A high contrast key (hard) is pulled and also assigned to the mattes layer. Channel count: 5
4. The mattes.hard and mattes.soft channels are mixed to form the final matte (alpha), which is
assigned to the rgba layer. Channel count: 6
Suppose now that you wanted to perform a color correction using the output of the Soft Matte node
as a mask for the correction. There’s no need to pipe the output from that Soft Matte node - it already
exists in the data stream along with the other five channels that were created.
You simply attach a color correction node, such as HueCorrect, then select the appropriate channel
from the mask controls, for example mattes.soft.
Note: The mattes portion of the name indicates the parent layer.
Viewing Channels in the Viewer

You can view the individual red, green, blue, and alpha channels in the Viewer using the R,G,B, and A
keys on the keyboard. For more information, see Compositing Viewers and Using the Timeline
Viewer.
USER GUIDE
382
Calling Channels | Selecting Input Channels
Selecting Input Channels

A node’s channels field lets you select one or several channels for processing.
To Select a Single Input Channel

1. Open the properties panel of the node into which you wish to feed a channel.
2. From the channels field, select none.
3. From the right most channel field - the one which typically calls the alpha channel - select the
single channel you wish to process.
To Select Multiple Input Channels

1. Open the properties panel of the node into which you wish to feed channels.
2. From the channels field, select the layer containing the channels you wish to process.
The layer’s channels appear with check boxes.
3. Deselect those channels which you don’t wish to process. The node processes all those you leave
selected.
Selecting Masks
The mask controls in a node’s properties panel let you select a single channel for use as a matte in a
given process (typically, a color correction). The given process thereafter is limited to the non-black
areas of the selected channel.
You can use one of the script’s existing channels as the matte, or attach a mask to the node with a
mask input connector.
You can find mask input connectors on color correction and filter nodes, such as HueCorrect and
Blur. At first, they appear as triangles on the right side of the nodes, but when you drag them, they
turn into arrows labeled mask. You connect them the same way as any other connectors. If you
USER GUIDE
383
Calling Channels | Selecting Masks
cannot see a mask input connector, open the node’s properties panel and make sure mask is set to
none.
Before dragging the mask When dragging the mask

input connector. input connector.
To Select a Channel for Use as a Matte from the Mask Input

1. Connect a mask to the node with its mask input connector.
If you cannot see the mask input connector, open the node’s controls and make sure mask is set
to none.
By default, when a mask input is connected, the node uses the alpha channel from it as a matte.
2. If you don’t want to use the alpha channel as the matte, select the channel you want to use from
the mask dropdown menu.
3. If you want the mask from the mask input copied into the predefined mask.a channel, check
inject. This way, you can use the last mask input again downstream. You can also set a stream of
nodes to use mask.a as the mask, and then change the masking of all of them by simply
connecting a new mask into the mask input connector of the first node.
4. If necessary, check the invert box to reverse the mask.
Tip: The invert control also affects any injected masks from upstream nodes.
USER GUIDE
384
Calling Channels | Selecting Masks
5. If the overall effect of the node is too harsh, you can blend back in some of the input image by
adjusting the mix slider.
To Select a Channel for Use as a Matte from the Main Input

1. Make sure nothing is connected to the node’s mask input connector. If you disconnect a mask
input, the mask input connector disappears, as it is no longer being used.
2. Select the channel you want to use from the mask dropdown menu.
adjusting the mix slider.
USER GUIDE
385
Linking Channels Using the Link Menu | Selecting Masks
Linking Channels Using the Link

Menu
You can create expression links to connect channel and layer controls with other controls in various
nodes. Since these controls aren’t meant to be animated, you can’t use the full range of Nuke
expressions, nor can you use Python or Tcl languages. You can link controls using the Link menu
next to the control on the properties panel:

1. Click the Link menu and select Set link. An Expression dialog opens.
2. Enter your expression in the Expression field and click OK. For more information on expressions,
see Expressions.
3. You can edit an existing link by clicking the Link menu and selecting Edit link.
4. You can also Ctrl/Cmd+drag the Link menu to another control to create a link between the two.
5. To remove a link, click the Link menu and select Remove link.
USER GUIDE
386
Tracing Channels | Selecting Masks
Tracing Channels
You may have noticed that nodes visually indicate the channels which they are processing (that is,
treating in some way) and passing (that is, conveying without any treatment). This is done via a system
of colored rectangles, which allows you to trace the flow of channels throughout a script.
Look closer at this example node. The wide rectangles indicate channels which Nuke processes (in
this case, the red, green, blue, and alpha channels). The narrow rectangles indicate channels that
Nuke passes onto the next node without processing (in this case, the mattes.soft and mattes.hard).
Visual confirmation of extra

channels.
USER GUIDE
387
Renaming Channels | Selecting Masks
Renaming Channels
In the course of building your script, you may find it necessary to replace certain channels in a layer.
1. Open the properties panel for a node which has the channel selected on the channels, input, or
output dropdown menu.
2. Click on the dropdown menu where the channel is displayed and select rename.
The Rename Channel dialog appears.
3. Enter a new name for the channel and click OK.
USER GUIDE
388
Removing Channels and Layers | To Remove a Layer or a Channel Within a Layer
Removing Channels and Layers

When you are done using a layer or a channel within a layer, you may wish, for the sake of clarity, to
remove it so that it is no longer passed to downstream nodes. Note that leaving channels in the
stream does not itself cause them to be computed, only channels required are computed.
Note: When a script is saved, any channels that are not referenced in the script are
discarded automatically.
To Remove a Layer or a Channel Within a Layer

1. Click Channel > Remove to insert a Remove node at the appropriate point in your script.
2. In the Remove properties panel, select the layer you wish to remove from the channels fields.
3. If you don’t wish to remove the entire layer, uncheck the boxes corresponding to the channels
which you still wish to be able to call downstream.
4. Click OK to close the properties panel.
The layer and/or the channels you removed are no longer displayed in node parameters
downstream from the Remove node.
Note: Removing layers and or channels does not free up space for the creation of new
channels and layers. Once you create a channel, it permanently consumes one of the script’s
1023 available channel name slots. You are free, however, to rename channels and/or assign
them new outputs.
USER GUIDE
389
Swapping Channels | To Remove a Layer or a Channel Within a Layer
Swapping Channels
Nuke features two main nodes for channel swapping: Shuffle and ShuffleCopy. Shuffle lets you
rearrange the channels from a single image (1 input) and then output the result to the next node in
your compositing tree. ShuffleCopy lets you rearrange channels from two images (2 inputs) and
output the result. Let’s take a look at the ShuffleCopy node first.
The ShuffleCopy matrix.

• Channels from Input 1 - the first group of channel boxes are the channels supplied by input 1 on
the node. As shown above, the foreground element’s default rgba set is selected.
• Channels from Input 2 - the second group of channel boxes are the channels supplied by input 2
on the node.
To swap channels, do the following:

1. Click Channel > Shuffle or ShuffleCopy to insert a Shuffle or ShuffleCopy node. Remember you
use Shuffle when you only want to swap channels in a single upstream node, and ShuffleCopy
when you want to swap channels in two separate nodes, like a foreground and background
branch.
2. Select the incoming channels from the In 1 and In 2 (optional) dropdown menus. You can select
up to eight channels in this manner.
3. Select the layers to which you wish to direct the incoming channels from the dropdown menus on
the right. You can select up to eight channels in this manner.
4. If the outgoing layer to which you wish to direct channels does not yet exist, create it using the
new option on the dropdown menus on the right.
5. Click as necessary on the resulting matrix to swap channels.
USER GUIDE
390
Swapping Channels | Channel Outputs
Tip: If you just need to copy a channel from one data stream into another, use Channel >
Copy, instead of ShuffleCopy. Then, specify the channel to copy and the destination channel
for output.
Channel Outputs
The combination of the boxes checked in the channel matrix create the list of channels that are
output to the layer selected in the top dropdown menu on the right.
This four channel stream acts as the second set of outputs from the node. It allows you to output
another four channels from the node, for a total of eight channels of output to match the possible
eight channels of input.
In this case, this second set of outputs has not been utilized.
Tip: While not required, it’s good practice to use the first set of outputs for swapping
channels in the current data stream and the second set of outputs for creating new
channels. This protects the default rgba set from unintentional overwriting, and makes it
easier for other artists to understand the workings of your script.
The basic process then for swapping channels is to first select your incoming layers from the 1 in and
2 in (or, in the case of the Shuffle node, the in 1 and in 2) dropdown menus. Then, select your
outgoing layers from the dropdown menus on the right. Then make the actual channel swaps by
clicking on the resulting matrix.
For example, to take the simplest case, suppose, you wanted to copy the red channel of an rgba set
into its alpha channel. You would click on the matrix to create the following configuration.
Configuring the channel matrix.
USER GUIDE
391
Swapping Channels | Assigning Constants
You can see that the matrix makes use of the r channel (standing for red) twice. It goes out once as
the red channel, and another time as the alpha channel.
Assigning Constants
The shuffle nodes also include parameters that let you assign white (1) or black (0) constants to any
incoming channel. So, for example, to reset the alpha channel to a full-frame image, you would
configure the matrix as follows:
Assigning constants to channels.
Creating Swap Layers

Finally, note that if the layer to which you wish to output channels does not yet exist, you can create it
using the new option on the dropdown menus on the right. Once you select the new option, you
follow the same process for creating layers as is described under Creating Channels and Layers.
USER GUIDE
392
Merging Images
With Nuke, you can merge images in a wide variety of ways. In these pages, we teach you how to use
the Merge, ContactSheet, and CopyRectangle nodes.
Layering Images Together with the Merge Node

The Merge node with its compositing algorithms allows you to control just how your images are
combined.
Note: When using most of the available merge algorithms, Nuke expects premultiplied
input images. However, with the matte operation you should use unpremultiplied images.
To Layer Images with the Merge Node

1. Select Merge > Merge (or press M on the Node Graph) to insert a Merge node after the images
you want to layer together.
2. Connect your images to the Merge node’s A and B inputs.
3. If necessary, you can connect multiple A images to the Merge node. Once you have got the A and
B inputs connected as instructed in step 2, drag more connectors from the left side of the Merge
node to the images you want to use as additional A inputs.
Each input is merged in the order connected, for example A3, A2, A1, B.
4. Connect a Viewer to the output of the Merge node so you can see the effect of your merge
operation.
5. In the Merge node’s controls, select how you want to layer the images together from the
operation dropdown menu. The default and the most common operation is over, which layers
input A over input B according to the alpha of input A. For descriptions of all the available
operations, see Merge Operations.
6. Set which input's bounding box you want to use for the Merge output:
• union - resize the output bbox to fit both input bboxes completely.
• intersection - use only those parts of the image where the input bboxes overlap.
• A or B - use the selected input's bbox for the output.
See Adjusting the Bounding Box for more information on bounding boxes.
USER GUIDE
393
Merging Images |
7. Select which input's metadata and frame range is passed down the tree using the metadata from
and range from dropdowns.
8. Using the A channels and B channels dropdown menus, select which channels to use from the A
and B inputs and which channels to use as the A and B alpha. If you want to merge more channels
than these and output them into the same channels, select them from the also merge dropdown
menus and checkboxes.
9. From the output dropdown menu, select the channels you want to write the merge of the A and B
channels to. Channels named in the also merge dropdown menu are written to the same output
channels.
10. If necessary, you can also adjust the following controls:
• To select which input’s metadata to pass down the tree, use the metadata from dropdown
menu.
Note: When metadata from is set to All and there are keys with the same name in both
inputs, keys in B override keys in A.
For more information on file metadata, see Working with File Metadata.
• To dissolve between the original input B image (at 0) and the full Merge effect (at 1), adjust the
mix slider. A small light gray square appears on the node in the node graph to indicate that the
full effect is not used.
• If you want to mask the effect of the Merge operation, select the mask channel from the mask
dropdown menus. To invert the mask, check invert. To blur the edges of the mask, check
fringe.
Note that you should not use the alpha of the inputs for the mask. It produces erroneous results
(though the error is often hard to see); you can achieve better results by turning on alpha
masking.
• From the Set BBox to dropdown menu, select how you want to output the bounding box. The
default is union, which combines the two bounding boxes. You can also select intersection to
set the bounding box to the area where the two bounding boxes overlap, A to use the bounding
box from input A, or B to use the bounding box from input B.
• By default, Nuke assumes that images are in linear color space. However, if you want to convert
colors to the default 8-bit color space defined in the LUT tab of your project settings (usually,
sRGB), check Video colorspace. The conversion is done before the images are composited
together, and the results are converted back to linear afterwards. Any other channels than the
red, green, and blue are merged without conversion.
USER GUIDE
394
Merging Images | Bounding Box Warnings
Checking this option can be useful if you want to duplicate the results you obtained from an
application that uses the standard compositing math but applies it to non-linear images (for
example, Adobe® Photoshop®). In this case, you typically also need to make sure premultiplied
is not checked in your Read node controls.
• By default, the same math is applied to the alpha channel as the other channels. However,
according to the PDF/SVG specification, many of the merge operations (for example, overlay and
hard-light) should set the alpha to (a+b - ab). This way, the input images remain unchanged in
the areas where the other image has zero alpha. If you want to enable this, check alpha
masking.

nodes that affect the bounding box. To enable the warnings, in Nuke's Preferences under Panels >
Node Graph, enable Bounding Box Warning:
format.
USER GUIDE
395
Merging Images | Bounding Box Warnings
USER GUIDE
396
Merge Operations | Bounding Box Warnings
Merge Operations
When layering images with the Merge node, you need to select a compositing algorithm that
determines how the pixel values from one input are calculated with the pixel values from the other to
create the new pixel values that are output as the merged image.
The operation dropdown menu in the Merge node’s properties panel houses a large number of
different compositing algorithms, giving you great flexibility when building your composite. The
available algorithms are listed in alphabetical order.
Tip: With many compositing algorithms available, it may sometimes be difficult to find what
you’re looking for in the operation dropdown menu. Luckily, there’s a quick way of finding a
particular operation. With the menu open, you can type a letter to jump to the first operator
that starts with that letter. To move to the second operation that starts with the same letter,
press the letter again. For example, to select the soft-light operation, open the menu and
press S twice.
The following table describes each operation and its associated compositing algorithm. There are
example images to illustrate the effects, one that combines the letters A and B to a merged image
and another that has an image of fire merged with the familiar checkerboard. You may want to spend
some time familiarizing yourself with each algorithm in order to be able to determine which
operation to use in each situation.
Operation Algorithm Description Illustration Example Uses
atop Ab+B(1-a) Shows the shape of

image B, with A
covering B where the
images overlap.
USER GUIDE
397
average (A+B)/2 The average of the two

images. The result is
darker than the original
images.
color-burn darken B Image B gets darker

towards A based on the luminance
of A.
color-dodge brighten B Image B gets brighter

towards A based on the luminance
of A.
conjoint-over A+B(1-a/b), Similar to the over

operation, except that if
A if a>b a pixel is partially
covered by both A and
B, conjoint-over
assumes A completely
USER GUIDE
398
hides B. For instance,

two polygons where A
and B share some
edges but A completely
overlaps B. Normal over
produces a slightly
transparent seam here.
copy A Only shows image A. This is useful if you

also set the mix or
mask controls so
that some of B can
still be seen.
difference abs(A-B) How much the pixels Useful for comparing

differ. Also available two very similar
from Merge > Merges images. This mode
> Absminus. can also be used as a
difference keyer.
disjoint-over A+B(1-a)/b, Similar to the over This can be useful if

operation, except that if you want to merge
A+B if a+b<1 a pixel is partially element a over
covered by both a and element b, and
b, disjoint-over element a has
assumes the two element b already
objects do not overlap. held out. For
For instance, two example, you may
USER GUIDE
399
polygons that touch and have a CG character

share an edge. Normal whose hair, skin, and
over produces a slightly clothing are
transparent seam here. rendered separately
so that each object
has the other objects
held out of the
render.
In this case, using the

over operation would
produce dark lines
around the comped
objects. This is
because over does a
hold-out of the
background image,
meaning the
background is held
out twice.
divide A/B, Divides the values but This does not match
stops two negative any photographic
0 if A<0 and values from becoming a operation, but can be
B<0 positive number. used to undo a
multiply.
exclusion A+B-2AB A more photographic

form of difference.
USER GUIDE
400
from B-A Image A is subtracted

from B.
geometric 2AB/(A+B) Another way of

averaging two images.
hard-light multiply if Image B is lit up by a

A<0.5, very bright and sharp
light in the shape of
screen if A>0.5 image A.
USER GUIDE
401
hypot sqrt Resembles the plus and This is useful for

(A*A+B*B) screen operations. The adding reflections, as
result is not as bright as an alternative to
plus, but brighter than screen.
screen.
Hypot works with values

above 1.
in Ab Only shows the areas of Useful for combining

image A that overlap mattes.
with the alpha of B. Also
available from Merge >
Merges > In.
mask Ba This is the reverse of

the in operation. Only
shows the areas of
image B that overlap
with the alpha of A.
matte Aa+B(1-a) Premultiplied over. Use

unpremultiplied images
with this operation. Also
available from Merge >
Merges > Matte.
USER GUIDE
402
max max (A,B) Takes the maximum This is a good way to

values of both images. combine mattes and
Also available from useful for bringing
Merge > Merges > aspects like bright
Max. hair detail through.
min min (A,B) Takes the minimum

values of both images.
Also available from
Merge > Merges > Min.
minus A-B Image B is subtracted

from A.
USER GUIDE
403
multiply AB, A if A<0 Multiplies the values Used to composite

and B<0 but stops two negative darker values from A
values from becoming a with the image of B -
positive number. Also dark gray smoke
available from Merge > shot against a white
Merges > Multiply. background, for
example.
This is also useful for

adding a grain plate
to an image
regrained with F_
Regrain.
out A(1-b) Only shows the areas of Useful for combining

image A that do not mattes.
overlap with the alpha
of B. Also available
from Merge > Merges
> Out.
over A+B(1-a) This is the default This is the most

operation. Layers image commonly used
A over B according to operation. Used
the alpha of image A. when layering a
foreground element
over a background
plate.
USER GUIDE
404
overlay multiply if Image A brightens

B<0.5, image B.
screen if
B>0.5
plus A+B The sum of image A and Useful for

B. Also available from compositing laser
Merge > Merges > beams, but you’re
Plus. Note that the plus better off not using
algorithm may result in this one for
pixel values higher than combining mattes.
1.0.
screen A or B ≤1? If A or B is less than or This is useful for

A+B-AB: max equal to 1 the screen combining mattes
(A,B) else use the maximum and also for adding
example, resembles laser beams.
Plus. Also available
from Merge > Merges
> Screen.
soft-light B(2A+(B(1- Image B gets lit up. Not

AB))) if AB<1, as extreme as the hard-
2AB otherwise light operation.
USER GUIDE
405
stencil B(1-a) This is the reverse of

the out operation. Only
shows the areas of
image B that do not
overlap with the alpha
of A.
under A(1-b)+B This is the reverse of

the over operation.
Layers image B over A
according to the matte
of image B.
xor A(1-b)+B(1-a) Shows both image A

and B where the images
do not overlap.
USER GUIDE
406
Tip: If you have used older versions of Nuke, you may have seen Merge operations called
diagonal, nl_over, and nl_under. Diagonal has been renamed and is now called hypot. To get
the results of nl_over and nl_under, you can check Video colorspace and use over and
under.
USER GUIDE
407
Generating Contact Sheets | To Generate a Contact Sheet
Generating Contact Sheets

In order to demonstrate, document or manage what you are doing for a project, it can be useful to
generate a contact sheet that shows your frame sequence(s) lined up next to each other in a matrix.
For this, you can use the ContactSheet node. It generates a contact sheet from all its inputs or from
the frames of one input.
A contact sheet generated from

the frames of one image sequence.
To Generate a Contact Sheet

1. Select Merge > ContactSheet to insert a ContactSheet node in your script.
2. Connect the image(s) you want to include in your contact sheet to the numbered input(s) of the
ContactSheet node. If you want to include several different image sequences in the contact sheet,
use multiple inputs. If you want the contact sheet to include the frames of just one image
sequence, use only one input.
3. Connect a Viewer to the ContactSheet node so you can see the effect of your changes.
USER GUIDE
408
4. In the ContactSheet properties, define the Resolution (width and height) of the entire contact
sheet in pixels.
5. If you want to create a contact sheet from the frames of one input, check Use frames instead of
inputs. In the Frame Range field, define the frame range you want to include in the contact
sheet.
6. In the rows/columns field, specify into how many rows and columns you want to arrange the
input images or frames.
7. To adjust the size of the gaps between the images in the contact sheet, increment or decrement
the gap value.
The gap value set to 0. The gap value set to 50.

8. From the Row Order and Column Order dropdown menus, select how you want to order the
images or frames in the contact sheet:
Row order: TopBottom Row order: TopBottom

Column order: LeftRight Column Order: RightLeft
Row order: TopBottom Row order: BottomTop

Column Order: Snake Column Order: LeftRight
Row order: BottomTop Row order: BottomTop

Column Order: RightLeft Column Order: Snake
USER GUIDE
409
Tip: If you want to add any text, such as the frame number, on top of the images in the
contact sheet, insert a Text node between the input image(s) and the ContactSheet node.
USER GUIDE
410
Copying a Rectangle from one Image to Another | To Generate a Contact Sheet
Copying a Rectangle from one

Image to Another
With the CopyRectangle node, you can copy a rectangle from one input on top of another.
Input A. Input B.
The output of CopyRectangle.
The CopyRectangle node can also be used to limit effects, such as color corrections, to a small region
of an image. To do so, you need to use the same image in both input A and B and only perform the
color correction on one input.
A rectangle from input A color corrected

and copied on top of input B.
USER GUIDE
411
Copying a Rectangle from one Image to Another | To Copy a Rectangle from One Image to Another
The original image before Defining a rectangle with

CopyRectangle. Copy Rectangle.
The color corrected rectangle

on top of the original image.
To Copy a Rectangle from One Image to Another

1. Select Merge > CopyRectangle to insert a CopyRectangle node after the image that has a region
you want to copy (input A) and the image you want to copy the region to (input B). Create the
following setup:
2. In the CopyRectangle controls, use the channels dropdown menu to select the channels you want
to copy from input A.
USER GUIDE
412
Copying a Rectangle from one Image to Another | To Copy a Rectangle from One Image to Another
3. To define the rectangle you want to copy, resize and reposition the CopyRectangle overlay in the
Viewer. Drag the center of the overlay to reposition, and the edges to resize. If you cannot see the
overlay in the Viewer, open the CopyRectangle properties panel and double-click on the Viewer
Repositioning the rectangle. Resizing the rectangle.

4. To control how soft the edges of the rectangle seem, adjust the softness slider. The higher the
value, the softer the edges.
A low softness value. A high softness value.

5. To dissolve between the full CopyRectangle effect and input B, adjust the mix slider.
USER GUIDE
413
Removing Noise with Denoise
The Denoise node is an efficient tool for removing noise or grain from your footage. It uses spatial or
temporal filtering to remove noise without losing image quality.
Quick Start
1. Connect Denoise to the footage from which you want to remove noise.
See Connecting Denoise.
2. Position the analysis box over a suitable analysis region. Denoise automatically analyzes the noise
structure inside this region and removes noise from the footage.
See Analyzing and Removing Noise.
3. Review the results.
See Reviewing the Results.
4. If you’re not happy with the results, you can fine tune them by adjusting the noise profile,
frequencies, and channels.
See Fine Tuning.
Before Denoise. After Denoise.
Tip: You can check Use GPU if available to have the node run on the graphics processing
unit (GPU) rather than the central processing unit (CPU).
For more information on the minimum requirements, please see Windows, Mac OS X and
USER GUIDE
414
Removing Noise with Denoise |
macOS, or Linux or refer to the Nuke Release Notes available in Help > Release Notes.
You can select the GPU to use in the Preferences. Press Shift+S to open the Preferences
dialog, make sure you’re viewing the Preferences > Performance > Hardware tab, and set
default blink device to the device you want to use. You must restart Nuke for the change to
take effect.
If you are using a render license of Nuke, you need to add --gpu on the command line.
USER GUIDE
415
Connecting Denoise |
Connecting Denoise
1. Create a Denoise node by clicking Filter > Denoise.
2. Connect the Denoise node’s Source input to the image you want to denoise.
3. If you’re working with complex footage that doesn’t have a convenient area for analyzing noise (a
flat area free from image detail, edges, and luminance variations), you can attach an optional
noise clip to the Noise input. When a Noise clip is supplied, the noise is analyzed in this clip,
rather than the Source clip. The Noise clip should have similar noise characteristics to the
Source clip.
4. You can attach pre-calculated motion vectors to the Motion input. You can generate motion
vectors using VectorGenerator, NukeX's SmartVector, or an external vector generator. See
VectorGenerator or SmartVector for more information.
Denoise can generate motion vectors internally, but connecting this input may produce
significantly faster results.
5. Attach a Viewer to either the Source or Noise clip, depending on which you want to use for the
analysis.
6. Proceed to Analyzing and Removing Noise.
USER GUIDE
416
Analyzing and Removing Noise |
Analyzing and Removing Noise

1. In the Denoise controls, set Source to Film or Digital depending on the type of footage you’re
using. Digital is the default setting and it works fine in most cases even if the footage is in a film
format.
2. In general, you can also leave Noise Model set to Modulated. However, you may want to try
Constant if you’re working on either:
• film footage with lots of detail but not too much noise in dark regions, or
• digital footage with lots of detail but not too much noise in light regions.
3. In the Viewer, scrub to a frame with a suitable analysis region. This should be a flat area free from
image detail, so no textures, edges, or shadows. If this is not the case, you may get poor results,
as the algorithm thinks the image detail is noise and removes it.
4. You can enable Temporal Processing to use more than one frame to denoise the analysis frame.
Denoise blends frames either side of the analysis frame to improve the result. The number of
frames used either side of the current frame is determined by the Frames To Blend control.
Note: Temporal Processing requires the motion, forward, and backward channels from
the Motion input. You can't enable Temporal Processing unless the Motion input is
connected.
You can increase the Frame Blending value to force Denoise to blend more regions of the frame,
but higher values can lose image detail.
5. Position and resize the analysis box to cover the analysis region. Note that the minimum size for
the analysis region is 80x80 pixels. If the analysis region is too small, Denoise cannot analyze the
footage or remove any noise.
A good analysis region. A poor analysis region.

The analysis area selection automatically updates not only the Analysis Region parameter in the
Noise Analysis group, but also the frame from which the sample is taken (Analysis Frame). By
USER GUIDE
417
Analyzing and Removing Noise |
default, whenever the analysis box is altered, the internal analysis of the noise in that region
reoccurs.
6. Connect a Viewer to the Denoise node.
The output should now show the denoised frame.
7. Proceed to Reviewing the Results.
Tip: By default, Denoise starts analyzing the noise in your input footage when you move the
analysis box. If you scrub to a new frame, don't move the analysis box, and want to
reanalyze the noise from the new frame, you can force analysis by clicking the Analyze
Noise button in the Noise Analysis parameter group.
If you’d like to disable analysis, you can check Lock Noise Analysis.
Tip: You can export the analysis profile in an external file. Note that if you’ve set Profile to
Constant, only the controls that affect the analysis are saved in this file. By contrast, if you’ve
set Profile to Automatic, both the analysis profile and the automatically calculated noise
profile are exported.
To export the profile, use the Analysis file control under Noise Analysis to set the name
and location of the file. The file extension doesn’t really matter here - for example .txt is
fine. Once you have named the profile, click the Export button to export it.
To import the same profile later, use Analysis File again to navigate to it and click Import.
This disables any controls that are read from the profile. To re-enable them, you can
uncheck Lock Noise Analysis.
USER GUIDE
418
Reviewing the Results |
Reviewing the Results

1. Zoom in to review the results.
2. To compare the denoised image with the original, press D on the Denoise node repeatedly to
disable and re-enable it.
The original image. The denoised image.

3. It can also be useful to look at the noise that was removed from the original image. To do so, set
Output to Noise.
Only noise should be visible in this image. If you can see a lot of picture detail, it means the
current settings are making Denoise work too hard and remove too much of the image, which
leads to a soft result.
If you can see picture detail in the

Noise output, too much of the image
is being removed.
4. If your footage contains a lot of areas of sub-black, areas of black with a value less than 0, enable
Lift Blacks to lift the blacks towards white. You can also enable Preserve Edges to attempt to
sharpen the image at edges preventing over-smoothing. Preserve Edges can emphasize noise in
some cases.
5. To blend the denoised luminance with the image’s original luminance, you can temporarily
increase Luminance Blend. This brings back some of the image detail in the result. You might
want to have this set to 1, for example, when you’re working on denoising the footage, but for the
final result, you’ll want to decrease it. The default value is 0.7.
6. If you’re not happy with the results, you can try:
USER GUIDE
419
Reviewing the Results |
• moving the analysis box to a different, flat area of the image,

• analyzing on a different frame (by moving to a different frame and clicking Analyze Noise),
• enabling Temporal Processing to blend more than one frame to calculate the denoise effect, or
• tweaking the Denoise controls (Denoise Amount, Roll Off, and Smoothness in particular).
Proceed to Fine Tuning.
USER GUIDE
420
Fine Tuning |
Fine Tuning
1. Set Denoise Amount to adjust the overall amount of noise to remove. Increase this to remove
more noise, or decrease it if you want to keep more detail. A value of 0 removes no noise.
2. If the denoised image looks too sharp, use Roll Off to adjust the smoothness of the denoise
thresholding. A value of 1 equals hard thresholding. Any other value produces soft thresholding
between:
• the Denoise Amount value and
• the Roll Off value multiplied by Denoise Amount.
3. If you’re not getting the correct smoothness level by adjusting the Denoise Amount, try setting
Smoothness to a new value. This controls the smoothness of the denoised image, affecting the
underlying image rather than the noise detected. In most cases, the default value of 1 works fine
though.
4. If the results are too smooth and adjusting the above controls didn’t help, try setting Profile to
Automatic. Unlike the Constant profile, which looks at the analysis region and removes the same
amount of noise across all intensities, automatic profiling looks at the entire Profile Frame to
estimate a noise profile and removes different amounts of noise from the shadow, midtone, and
highlight areas of the Source image.
Constant profile. Automatic profile.

When you first switch to automatic profiling, Denoise uses the current frame to calculate the
profile. If you’d like to use a different frame, you need to scrub to that frame and click the
Recalculate Profile button on the Profile tab. This updates the Profile Frame control, which is
read only.
Note: Denoise always bases the noise profile on your Source footage even if you’ve
attached another clip to the Noise input.
USER GUIDE
421
Fine Tuning |
5. You can also tweak the noise profile yourself using the controls on the Profile tab. This works in
both the Constant and Automatic profiling mode.
• Denoise displays the noise profile curve in the curve editor. The x axis represents image
intensity, from low frequencies on the left to higher frequencies on the right. The y axis
represents the relative amount of noise removed.
A noise profile curve.

You can adjust the curve manually by dragging the points on the curve to a new location. To add
more points to the curve, Ctrl/Cmd+Alt+click on the curve.
If you are not happy with your changes, click Reset Profile to reset the curve to its original
shape (clicking reset works as well).
• Make sure you check Tune Profile to enable your changes. Then, adjust Low Gain, Mid Gain,
and High Gain to scale the denoising threshold in the shadow, midtone, and highlight areas of
the image. For example, a value of 2 multiplies the threshold by 2. Everything below the
threshold is considered noise and removed, while everything above the threshold is kept.
The noise profile after The noise profile after

decreasing Low Gain. decreasing High Gain.
6. Back on the Denoise tab, the Tune Frequencies section lets you enable, disable, and adjust the
denoising in different noise frequencies. This allows you to select the frequencies that contain
USER GUIDE
422
Fine Tuning |
noise, process them by as much as you think is best, and leave details in the other frequencies
untouched.
Normally, most of the noise occurs in the high and medium frequencies, so often you can disable
the low and very low frequencies altogether. Use the Gain sliders to remove noise (by increasing
the value) or to add more detail and noise (by decreasing the value).
7. Under Tune Channels, you can adjust the denoising threshold for the luma and chroma
channels. Increase the Luminance Gain and Chrominance Gain values to remove more noise,
or decrease them to remove less.
USER GUIDE
423
Keying with ChromaKeyer
This section explains how to use the blue/green screen keyer, ChromaKeyer, in Nuke. ChromaKeyer
can take advantage of modern GPUs and multi-core CPUs to accelerate the keying process when used
for compositing in Nuke's Node Graph. ChromaKeyer is also available as a soft effect in Nuke Studio's
timeline environment.
Quick Key
The images below show a green screen foreground and the background to be composited over.
Foreground Background
1. Start Nuke and read in the foreground and background images. From the Keyer menu, select
ChromaKeyer and attach a Viewer.
2. In the ChromaKeyer Properties panel, click the color swatch next to screen color to activate the
eye dropper.
3. Ctrl/Cmd+Shift+click and drag a rectangular area over the green pixels in the Viewer. This
averages the pixels in the area selected to produce a better key.
USER GUIDE
424
Keying with ChromaKeyer |
In some cases, this is all you need to do to perform a key, since selecting the screen color creates
a screen matte and despills the foreground.
4. Merge the foreground over the background to produce your final comp.
Picking the screen color may be enough for a lot of keys, but there are many more tools within Nuke
that can be used to tackle more complicated shots. See Improving Mattes on page 429, Despilling and
Color Replacement on page 432, and Multi-Pass Keying on page 434 for more information.
USER GUIDE
425
Picking the Screen Color |
Picking the Screen Color

The screen color is probably the most important control in keying, and you should always pick the
screen color before doing anything else. It should be set to the color of the green or blue behind the
foreground object.
1. Start Nuke and read in the foreground and background images. From the Keyer menu, select
ChromaKeyer and attach a Viewer.
2. In the ChromaKeyer Properties panel, click the color swatch next to screen color to activate the
eye dropper.
3. Pick the screen color directly from the image by Ctrl/Cmd+Shift and dragging a rectangle over the
green pixels. The average value of the pixels selected is used.
It’s worth picking different shades of green or blue from different parts of the screen to get the
best result.
Note: Picking colors repeatedly does not add to previous selections, keying more of the
image with each click. To key more of the image, pick different shades of green or blue in
additional ChromaKeyer nodes downstream, and set the inside mask control to source
alpha. See Multi-Pass Keying for more information.
Picking the screen color does two things:

• It creates the initial matte used to composite the foreground downstream. The matte can be
improved using the matte controls in the Properties panel. See Improving Mattes for more
information.
USER GUIDE
426
Picking the Screen Color | Adjusting Screen Gain and Balance
• It despills the foreground, but you can use the despill controls for greater accuracy. For more
information on despill, see Despilling and Color Replacement for more information.
Adjusting Screen Gain and Balance

The screen gain controls how much of the screen color is removed to make the matte. Increasing
this value keys more of the foreground. In the example image, a lower screen gain value adds too
much of the background back into the output whereas a high screen gain value erodes the
foreground too much and tends to tint the edges the opposite of the screen color (for green screens,
edges become magenta).
A screen gain value lower than 1 adds A screen gain value higher than 1
background color back into the image. removes background color from the
image.
The screen balance controls the bias toward the two non-primary colors after the screen color has
been chosen. A screen balance closer to 0 balances the output toward the stronger of the two colors
and a value closer to 1 toward the weaker color. In the example image, the background color is green
with RGB values of 0.04, 0.7, and 0.07, so values closer to 0 affect the blue component.
A screen balance closer to 0 affecting A screen balance closer to 1 affecting
USER GUIDE
427
Picking the Screen Color | Adjusting Screen Gain and Balance
the blue component. the red component.
USER GUIDE
428
Improving Mattes | Adjusting the Gain, White Point, and Black Point
Improving Mattes
The matte controls can improve the basic matte created after picking the screen color. The best way
to view the matte in ChromaKeyer is to switch the Viewer to display the alpha channel in the R or
Luminance display styles.
The matte in the alpha channel R The matte in the alpha channel
display. Luminance display.
Adjusting the Gain, White Point, and Black Point

Adjusting the chroma gain, white point, and black point either dilates or erodes the matte to
include more or less of the foreground image based on the control used. The example image,
concentrating on the area near the character's chin, shows a good example of where a matte can be
improved.
USER GUIDE
429
Adjusting the chroma gain controls how much of the chroma difference between the source image
and the screen color is used to correct the matte. Increasing the gain generates a matte that has less
transparent areas in the foreground, but can produce harder edges.
The matte with a low chroma gain The matte with a high chroma gain
value. value.
Adjusting the white and black point controls the pixels that appear as foreground and background.
Any pixels with values higher than the white point value are treated as foreground (alpha = 1) and
any pixels with values lower than the black point value are treated as background (alpha = 0).
USER GUIDE
430
The matte with the default white The matte with a white point value of
point value. 0.5.
The matte with the default black point The matte with a black point value of
value. 0.5.
The chroma gain, white point, and black point controls represent a balancing act between
removing too much foreground and retaining too much background. Once you're happy with the
matte, proceed to Despilling and Color Replacement.
For particularly tricky keying problems, you can improve the matte using multi-pass keying. This
method uses multiple ChromaKeyer nodes with all but the first one in the node tree set to inside
mask > source alpha, so that the mattes in the alpha channels are passed down the node tree. See
Multi-Pass Keying for more information.
USER GUIDE
431
Despilling and Color Replacement | Custom Despilling
Despilling and Color Replacement

Although ChromaKeyer automatically despills the matte when you select a screen color, you may find
that you can improve the matte by manually despilling the key using a despill bias or by replacing
color at the edges of the matte.
Custom Despilling
The despill bias allows you to specify a color from the image, separate from the alpha bias, to
improve the overall despill for the matte. Typically, you should pick skin tones or hair colors for the
despill bias.
1. Enable custom despill bias to allow you to use the despill bias controls.
2. Click the color swatch next to despill bias to activate the eye dropper.
3. Ctrl/Cmd+Shift+click and drag a rectangular area over the color you want to replace the
highlighted pixels in the Viewer. This averages the pixels in the area selected to produce a better
result.
Blue screen pixels present in a default Selecting the hair color despills the
matte. blue pixels, producing a better matte.
Color Replacement
Improving a matte by adjusting the alpha channel can remove the wrong amount of screen color
from the pixels where transparency has changed. The replace controls instruct ChromaKeyer how to
deal with these pixels. The replace mode controls which pixels inherit the replace color.
1. In the ChromaKeyer Properties panel, select the required replace mode using the dropdown:
• ignore - the despilled image is left untouched if the alpha is modified. This is the default
operation.
USER GUIDE
432
Despilling and Color Replacement | Color Replacement
• edge hard color - the despilled image has a corresponding amount of the replace color added
for any increase in alpha.
• edge linear color - the despilled image has a graded amount of the replace color, controled by
a linear curve, added for any increase in alpha. Pixels closer to the background have more
background bias and pixels closer to the foreground have more foreground bias.
• edge soft color - the despilled image has a corresponding amount of the replace color added
for any increase in alpha, however, it attempts to modulate the luminance of the resulting pixel
so that it matches the original pixel. This gives a more subtle result than the edge hard color
option.
2. Click the color swatch next to replace color to activate the eye dropper.
3. Ctrl/Cmd+Shift+click and drag a rectangular area over the color you want to replace the
highlighted pixels in the Viewer. This averages the pixels in the selected area to produce a better
result.
Note: You can enhance the despill effect by enabling add-in matte fix to also apply the
replace color to areas eroded or dilated by the white point and black point controls.
Image with edge soft color Highlighted pixels replaced by the color
highlighted. picked from the Viewer.
4. Use the replace amount slider to control how much of the replace color is applied.
5. Disable the premultiply control if you don't want the output premultiplied by the alpha channel.
USER GUIDE
433
Multi-Pass Keying | Color Replacement
Multi-Pass Keying
In production situations, you may not always get a uniform color across your blue or green screen
background, which can make pulling a key difficult. ChromaKeyer allows you to key different regions
of the screen additively by picking shades of blue or green in multiple ChromaKeyer nodes. The
example image shows a poor background containing different shades of blue as a screen color.
To perform a multi-pass key, do the following:

1. Pick an initial screen color as described in Picking the Screen Color on page 426.
You can see from the image that ChromaKeyer has done a decent job on the area chosen, but the
lower-half of the image still contains some noise.
2. Add a second ChromaKeyer node and in the Properties panel, set the inside mask control to
source alpha.
This takes into account the screen matte from the initial ChromaKeyer node when you pull
another key.
3. Pick another screen color in the problem area using the second ChromaKeyer node.
The key is improved by adding the second screen matte to the results of the first.
4. You can add as many ChromaKeyer nodes as you like, as long as you remember to set the inside
mask control to source alpha in all but the first ChromaKeyer node.
USER GUIDE
434
Keying with Keylight
This section explains how to use the blue/green screen keyer, Keylight, in Nuke.
Quick Key
Consider this shot from The Saint, pictures courtesy of CFC and Paramount British Pictures Ltd.
Blue screen.
The figure above is the blue screen foreground that should be composited over the background
shown below.
Background.
1. Start Nuke and read in both images. From the Keyer menu, apply Keylight and attach a Viewer.
2. Click the color swatch next to ScreenColor to activate the eye dropper. In the Viewer,
Ctrl/Cmd+Shift+Alt+click and drag a rectangular area over the blue pixels as shown below.
USER GUIDE
435
Keying with Keylight |
Picking the screen color also sets the ScreenBalance.

That’s it. In many cases, this is all you need to do to perform a key, since selecting the screen color
creates a screen matte and despills the foreground.
3. Switch output from FinalResult to Composite to see the foreground keyed over the background.
The final composite is shown below.
Picking the screen color may be enough for a lot of keys, but there are many more tools within Nuke
that can be used to tackle more complicated shots. These are described later in this chapter.
USER GUIDE
436
Basic Keying | Picking the Screen Color
Basic Keying
The following section describes the parameters you need to do basic keying. This gives you enough to
tackle most simple keys. A discussion of advanced parameters to fine tune keys and tackle complex
keys can be found under Advanced Keying.
Picking the Screen Color

The screen color is probably the most important parameter and you should always pick the screen
color before doing anything else. It should be set to the color of the green or blue curtain behind the
foreground object. Pick the screen color directly from the image by Ctrl/Cmd+Shift+Alt dragging a
rectangle over the blue pixels. The average value of the pixels selected is used.
Tip: If you press Alt when sampling a color, Nuke always samples the source image
regardless of what you’re looking at. This means that you can pick the blue screen color even
if you are viewing the matte, status, or composite.
Tip: Picking different shades of blue from the screen color can give quite different results.
It’s worth picking from different parts of the screen to get the best result.
Picking the ScreenColor creates the screen matte used to composite the foreground over the
background. It also sets the ScreenBalance (if this has not already been set manually) and despills
the foreground.
Screen Matte
Setting the screen color pulls a key, or in other words, creates a matte - the ScreenMatte. Setting the
screen color also despills the foreground, although you can also use the DespillBias to remove more
spill. In some cases this is enough to get a decent key. For more information on ScreenColor, see
Screen Color.
The image below shows a well-lit blue screen behind an actor.
USER GUIDE
437
Basic Keying | Viewing the Key
You should note that repeatedly picking colors does not add to previous selections and key more of
the image with each click. To key more of the image, try picking different shades of blue then use the
screen strength parameter. See Keying More.
Viewing the Key

After picking the screen color, you have created a matte (the screen matte) and despilled the
foreground. The result can be displayed in a number of different ways using the View control. You
can output the final composite of the foreground over the background as an rgba, or you can output
the premultiplied or unpremultiplied foreground for compositing elsewhere in the tree. The screen
matte and the status view are the other two options which are useful in fine-tuning the key rather
than as an output image in their own right.
The Status is one of the options in the View dropdown menu and shows an exaggerated view of the
key so that you can make a more informed decision when tuning the key. The image on the right
shows the Status display after the screen color has been picked from the image on the left.
Green screen. Status.
Three colors are displayed:

• Black pixels show areas that are pure background in the final composite.
• White pixels show areas that are pure foreground.
USER GUIDE
438
Basic Keying | Keying More
• Gray pixels are a blend of foreground and background pixels in the final composite. You need gray
pixels around the edge of the foreground to get a good key at the foreground edge. Pixels that are a
blend between the foreground and background are shown in just one shade of gray. This is done to
highlight potential problems with the key. These gray pixels may represent a
foreground/background blend of 50/50 or 99/1. No distinction is made as to this ratio.
You may occasionally see other colors in the Status view and these are covered under View.
Keying More
To improve the key by firming up the foreground so the background doesn’t show through, you
should adjust the ClipWhite parameter. To key more of the foreground so that the background is
clearer, you should use the ClipBlack parameter. Look at the ScreenMatte and the Composite while
you’re doing this. Don’t overdo either of these or the edges between foreground and background
become hard.
USER GUIDE
439
Advanced Keying | Under the Hood
Advanced Keying
The following section describes how Keylight works under the hood as well as the parameters you
need to fine tune keys and get the most out of Keylight. Basic parameters covered previously may
also be covered here in more detail.
Under the Hood

Keylight is a ’color difference keyer’, which means that for it to figure out a key, it compares every
pixel in the image against a single color, known here as the ScreenColor.
View
The View parameter allows Keylight to render the final composite of the foreground over the
background, or the foreground RGBA for compositing further down the tree. Two options,
ScreenMatte and Status, are for viewing the key rather than an output. The options are:
• Source - shows the blue/green screen foreground.
• SourceAlpha - shows the alpha channel on the foreground input.
• ScreenMatte - this is the matte created from picking the ScreenColor. It does not include any
inside or outside masks.
• InsideMask - shows the inside input. This is used to firm up the foreground matte to stop print
through.
• OutsideMask - shows the outside input. The outside mask is used as a garbage mask to reveal the
background.
• CombinedMatte - the screen matte, inside mask, and outside masks added together.
• Status - this renders an exaggerated view of the key so that minor problems are shown clearly.
• IntermediateResult - use this option on shots that can only be keyed using several different keys
on different parts of the image (multi-pass keying). This renders the original source image with the
ScreenMatte generated in this Keylight node. In Keylight nodes down the tree, you should set the
SourceAlpha in the InsideMask folder to AddToInsideMask.
• FinalResult - this creates a premultiplied RGBA foreground that can be composited later. There’s an
UnpremultiplyResult toggle you can use if you wish.
• Composite - this renders the foreground composited over the background using all mattes, spill
and color corrections.
USER GUIDE
440
Advanced Keying | Status
Status
Status is one of the options in the View dropdown menu and shows an exaggerated view of the key
so that you can make a more informed decision when fine tuning the composite. The figure on the
right shows the Status after the screen color has been picked from the image shown in the figure on
the left.

• Black pixels represent pure background in the final composite.
• White pixels are pure foreground.
• Gray pixels are a blend of the foreground and background pixels. The gray is just one color to
highlight any areas that are not pure foreground or background. Gray pixels do not mean the key is
poor - the final composite may be fine.
You may occasionally see other colors in the Status view. The figure on the left shows black, white,
gray, and green pixels.
Status showing processing Composite showing

of the alpha channel. ScreenReplaceColor.
USER GUIDE
441
Advanced Keying | Status
• Green pixels are a warning. They show you the parts of the alpha that have changed through
processing the alpha channel (clipped, softened, or eroded). These areas have had the correct
amount of spill removed, but the alpha has subsequently changed and the composite may no
longer look right. This can be corrected using the ScreenReplaceColor to put back color in these
areas. Above, the figure on the right is an extreme example to illustrate the point. The
ScreenReplaceColor has been set to pure red and you can see that this mirrors the green pixels in
the Status view.
Similarly, you may see blue pixels in the Status.
Status showing how Composite showing

the inside matte affects the inside replace color.
the foreground.
• Blue pixels represent processed pixels in the InsideMask that affect the despill of the foreground.
The InsideReplaceColor is used to modify these pixels. Another extreme example is shown above
in the figure on the right. The InsideReplaceColor is set to pure yellow and the InsideReplace is
HardColor.
• You may also see dark red pixels in the Status. Red pixels indicate areas where an outside mask has
been used to reduce the transparency of the image.
USER GUIDE
442
View | Status
View
The View parameter allows Keylight to render the final composite of the foreground over the
background, or the foreground RGBA for compositing further down the tree. Two options, Screen
Matte and Status, are for viewing the key rather than an output. The options are:
• Source - shows the blue/green screen foreground.
• Source Alpha - shows the alpha channel on the foreground input.
• Screen Matte - this is the matte created from picking the Screen Color. It does not include any
inside or outside masks.
• Inside Mask - shows the inside input. This is used to firm up the foreground matte to stop print
through.
• Outside Mask - shows the outside input. The outside mask is used as a garbage mask to reveal the
background.
• Combined Matte - the screen matte, inside mask, and outside masks added together.
• Status - this renders an exaggerated view of the key so that minor problems are shown clearly.
• Intermediate Result - use this option on shots that can only be keyed using several different keys
on different parts of the image (multi-pass keying). This renders the original source image with the
Screen Matte generated in this Keylight node. In Keylight nodes down the tree, you should set the
Source Alpha in the Inside Mask folder to Add To Inside Mask.
• Final Result - this creates a premultiplied RGBA foreground that can be composited later. There’s
an Unpremultiply Result toggle you can use if you wish.
• Composite - this renders the foreground composited over the background using all mattes, spill
and color corrections.
Status
Status is one of the options in the View dropdown menu and shows an exaggerated view of the key
so that you can make a more informed decision when fine tuning the composite. The figure on the
right shows the Status after the screen color has been picked from the image shown in the figure on
the left.
USER GUIDE
443
View | Status

• Black pixels represent pure background in the final composite.
• White pixels are pure foreground.
• Gray pixels are a blend of the foreground and background pixels. The gray is just one color to
highlight any areas that are not pure foreground or background. Gray pixels do not mean the key is
poor - the final composite may be fine.
You may occasionally see other colors in the Status view. The figure on the left shows black, white,
gray, and green pixels.
Status showing processing Composite showing

of the alpha channel. Screen Replace Color.
• Green pixels are a warning. They show you the parts of the alpha that have changed through
processing the alpha channel (clipped, softened, or eroded). These areas have had the correct
amount of spill removed, but the alpha has subsequently changed and the composite may no
longer look right. This can be corrected using the Screen Replace Color to put back color in these
areas. Above, the figure on the right is an extreme example to illustrate the point. The Screen
Replace Color has been set to pure red and you can see that this mirrors the green pixels in the
Status view.
USER GUIDE
444
View | Status
Similarly, you may see blue pixels in the Status.
Status showing how Composite showing

the inside matte affects the inside replace color.
the foreground.
• Blue pixels represent processed pixels in the Inside Mask that affect the despill of the foreground.
The Inside Replace Color is used to modify these pixels. Another extreme example is shown above
in the figure on the right. The Inside Replace Color is set to pure yellow and the Inside Replace is
Hard Color.
• You may also see dark red pixels in the Status. Red pixels indicate areas where an outside mask has
been used to reduce the transparency of the image.
USER GUIDE
445
Screen Color | Background Pixel
Screen Color
The screen color represents the color of the pure blue (or green) screen. The first thing you should do
when pulling a key is pick the Screen Color.
Note: If you press Alt when sampling a color, Nuke always samples the source image
regardless of what you’re looking at. This means that you can pick the blue screen color even
if you are viewing the matte, status or composite.
Picking the Screen Color creates the screen matte used to composite the foreground over the
background. It also sets the Screen Balance and despills the foreground.
The Screen Color is a single color. It has a primary component, blue or green, and that has a
saturation. Once the screen color has been picked, Keylight analyzes all the pixels in the image and
compares the saturation of the primary component in each of these pixels with the corresponding
saturation of the screen color. Keylight uses this comparison to do two things.
1. It calculates the transparency of that pixel and puts it in the alpha channel.
2. It removes the screen color from the pixel, a process known as despilling.
Tip: It’s worth sampling a selection of screen (blue or green) colors and viewing the result.
Picking different colors gives different results.
Background Pixel
If the saturation of the pixel in the image is as strong or greater than the screen color, then it’ll be a
pixel from the blue screen background, and that pixel is set to completely transparent and black. See
the figure below.
USER GUIDE
446
Screen Color | Edge Pixel
Blue screen pixel set alpha to zero.
Edge Pixel
If the saturation of the pixel is less than the screen color, then it’ll be the edge of the foreground
object, and we subtract some of the screen color from the pixel (despilling) and set the image to
semi-opaque. See the figure below.
Edge pixel gives partial alpha.
Foreground pixel
If the primary component in the pixel is not the same as the primary component of the screen color,
we have a foreground pixel, and the alpha is set to completely opaque. The pixel color is not
modified. See the figure below.
Foreground pixel gives full alpha.
USER GUIDE
447
Screen Color | Biasing
Note: You should note that the ScreenColor is a single color. You are not picking lots of
colors that are keyed out.
Biasing
What’s biasing all about? Biasing in Keylight was originally developed for a shot in the motion picture
"Executive Decision". The foreground consisted of reddish browns, but a combination of factors led
to the 'green screen' being lit so that its primary component was actually slightly red.
So what happens when we pick the screen color? Well because the screen was 'red', as is the
foreground, our pilot ends up being keyed out as shown below.
Not a great result, I’m sure you’ll agree, and much pressure was applied to the lowly programmers to
get around the problem.
A work around to this is to manually color correct the image so that the background is properly
green, pull the key from this corrected image, then 'un-correct' the result of that so that the
foreground colors match the original. A corrected image would look something like the one shown
below. The green screen is now strongly green and distinct from the foreground colors. Notice also
the red cast on the pilots mask has been removed and turned into a neutral gray.
USER GUIDE
448
Screen Color | The Bias Colors in Everyday Use
This is effectively how the Keylight developers got around the problem. They introduced the concept
of a 'bias' color, which is a color cast that is removed from the source image and screen color, then a
key is pulled from this modified image, then the color cast is put back. In essence, this automates the
work around described above, however, it is done in a way that does not slow Keylight down at all.
For our Executive Decision shot, an appropriate color is the red cast on the pilot's mask in the source
footage. Setting our bias to this now gives us the far better result as shown below.
The Bias Colors in Everyday Use

It also turns out that the bias color is actually useful for situations without strong casts, typically
where there is some color spill around the edge of keys. By setting the biases to the main color that
occurs near the edge of the foreground (typically flesh tones or hair tones), you allow Keylight to
better discriminate between foreground and background.
Picking a bias color

To pick a bias color, click the color swatch next to Alpha Bias to activate an eye dropper and
Ctrl/Cmd+Shift+Alt+drag a box over the image foreground. The average color under the box is used
for the bias you have selected.
Note: If you press Alt when sampling a color, Nuke always samples the source image
regardless of what you’re looking at. For instance, you may be looking at the blue screen
keyed over the background but you are picking colors from the Source image.
USER GUIDE
449
Why are there two bias colors?

Remember that Keylight does two things, calculates a transparency and removes the screen color
from the foreground. By default, one bias color, the Alpha Bias, is used for both operations. This
works fine in most situations, for example, the Executive Decision shot above.
However, sometimes you can pick a bias that gives a great alpha, but performs a poor despill, and
another bias that gives a great despill, but a poor alpha. Consider the blue screen from the TV series
Merlin, courtesy of CFC Framestore shown below in the figure on the left.
We pick the strong blue of the background without selecting an alpha bias, and end up with the lovely
alpha shown on the right, but the despill resulting from this key is poor as shown below.
Merlin blue screen. Nice Alpha.
Tip: There are several nodes in Nuke you can use for spill removal. For example, if you are
using a greenscreen image, you can add an Expression node after your foreground image
and set the expression field for the green channel to:
g>(r+b)/2?(r+b)/2:g
Similarly, you can despill a bluescreen image by setting the expression field for the blue
channel to:
b>(r+g)/2?(r+g)/2:b
USER GUIDE
450
You can also use the HueCorrect node for despill. For more information, see Correcting Hue
Only.
We can pick an alpha bias to get a better despill, but this destroys our nice alpha. The way around this
is to turn off the Use Alpha Bias for Despill, which gives you a separate bias factor to use solely for
despill calculations. If you then pick the Despill Bias to be something from Miranda Richardson's hair
or skin tone, you keep the nice alpha, and get a good despill as well (see the figure on the right).
Poor despill. Final key, using separate

Despill and Alpha Biases.
USER GUIDE
451
Clip Black and White | The Bias Colors in Everyday Use
Clip Black and White

The clip levels are adjusted using two parameters - Clip Black and Clip White. Any alpha value at or
below Clip Black is set to zero and any alpha value at or above Clip White is set to 1. The figure on
the left shows the original alpha of an image, and the figure on the right shows the result of clipping
it.
Clip Black = 0. Clip Black = 0.5.
Notice how the gray areas in the black background have been reduced and that the gray edges have
hardened up considerably. When compositing, the Clip Black control can be used to improve the
background image if parts of the foreground are showing through. The Clip White control, on the
other hand, can be used to firm up the center of the matte, making it less transparent to the
background.
Note: If you choose to use ClipBlack and ClipWhite, you need to be really careful that you
don't destroy the edges on your foreground. It is possible to use ClipRollback to
compensate for this.
USER GUIDE
452
Screen Gain | The Bias Colors in Everyday Use
Screen Gain
The screen gain controls how much of the screen color is removed to make the screen matte.
Increasing this value keys more. The figure on the left shows the Status after picking the Screen
Color.
Status after picking Status showing the

the Screen Color. increase in Screen Gain.
You can clearly see that parts of the background are gray where they should be black. When
composited, you may see faint pixels from the foreground where you should be seeing pure
background. Increasing the screen gain fixes this, as shown in the figure on the right (above), but
increasing it too much destroys your good work. Like many keying parameters it’s a balance - not too
much, not too little. Increasing the screen gain too much leads to the background showing through
the foreground and edge detail can be destroyed. Below, the figure on the right shows this quite well.
Screen Gain = 1.05 giving Screen Gain = 1.50 giving

a good Screen Matte. background show-through
and over eroded edges.
Note the steering wheel is black when it should be white. If you look at the composite, you can see
the background showing through here. Also, some of the fine hair detail on the actor, visible in the
figure on the left, has been eroded in the figure on the right.
USER GUIDE
453
Screen Gain | Screen Balance
Screen Balance
The ScreenBalance is set automatically after picking the ScreenColor.
Saturation is measured by comparing the intensity of the primary component against a weighted
average of the two other components. This is where the Screen Balance control comes in. A balance
of 1 means that the saturation is measured against the smallest of the other two components in the
screen color.
A balance of 0 means that the saturation is measured against the larger of the other two components.
A balance of 0.5 measures the saturation from the average of the other two components.
The appropriate balance point for each image sequence you key is different depending on the colors
in that image. Generally speaking, blue screens tend to work best with a balance of around 0.95 and
green screens with a balance of around 0.5. These values are selected automatically the first time you
pick the screen color. If the key is not working too well with these settings, try setting the balance to
about 0.05, 0.5 and 0.95 and see what works best.
USER GUIDE
454
PreBlur and Tuning | Tuning
PreBlur and Tuning

Some shots can be improved by softening the foreground image that is used to generate the key. The
original image is then used in the composite and color corrections. The Screen PreBlur parameter is
used to do this. DV footage or grainy shots may benefit from subtle use of this control.
Tuning
Keylight creates the screen matte after the screen color has been picked. You can make fine
adjustments to this matte using the Gain controls. Increasing the gain controls makes the screen
matte more transparent by increasing the amount of screen color showing through the matte. This
tends to tint the edges the opposite of the screen color (for blue screens, edges become yellow).
Decreasing the gain makes the main matte more opaque by reducing the amount of screen color
showing through the matte.
The matte can be adjusted independently in the shadows, midtones, and highlights, giving more
control than the clipping levels.
The level of the midtones can be adjusted too. For example, if you are working on a dark shot you
may want to set the midtone level to a dark gray to make the gain controls differentiate between
tones that would otherwise all be considered shadows.
USER GUIDE
455
Screen Processing | Two-stage keying
Screen Processing
Once you have picked the screen color and got the screen matte, you may wish to process this matte
using the parameters in the Screen Matte group. The matte can be adjusted using clipping levels; it
can be eroded or grown, despotted, and softened.
Two-stage keying
Consider this example. Having applied Keylight and picked the screen color, you have good edges to
your matte but the background is showing through the foreground. You could fix this by tweaking the
Clip White, but in doing so it ruins your edges. One way round this is a two stage key (another way is
using Clip Rollback). In the first key, you process the screen matte using the clipping levels to give a
harsh black and white matte, then soften and erode it. Switch View to Intermediate Result to output
the original green screen with the eroded matte as an RGBA. Then, use this as the input to another
Keylight node. In this second node, pick the screen color to give good edges but with transparent
foreground. Don’t process this matte, instead use the input alpha channel to fix the transparent
foreground. Just set Source Alpha in the Inside Mask folder to Add To Inside Mask.
Clip Rollback
Pulling a screen matte (the figure on the left) typically produces lots of transparency (gray) in the
matte at the edges. This is good since this is what you need to key hair well. You may also get
transparency in the foreground as shown in the figure on the right. This is bad as your subject appear
slightly see-through, and this should be corrected.
Screen matte highlighting Close up screen matte

the close up view as showing unwanted (gray)
shown in the figure on the transparency in the (white)
right. foreground.
USER GUIDE
456
Screen Processing | Dilate
You can do this by connecting a matte into the third (InM) input, or you can use the Clip White
parameter to turn these gray pixels white. This cleans up the foreground (the figure on the left) but it
also destroys the edge detail you want to keep. This is where Clip Rollback comes in. This is used to
put back the edges to restore the detail that was lost. A rather exaggerated clip rollback is shown in
the figure on the right to illustrate the point.
Clip White has been used Clip Rollback has been

to remove the unwanted used
gray pixels in the white to restore the unwanted
matte. erosion of the edge.
Dilate
This control should not normally be used as eroding the edges can produce a very poor key.
However, the Screen Dilate parameter allows you to grow (if greater than zero) or shrink (if less than
zero) the alpha in the Screen Matte. These controls are sub-pixel accurate.
Screen Matte. Eroded Matte.
Softness
Occasionally, it is useful to be able to blur the matte. Use Screen Softness for this. The most
common example would be to pull a very harsh matte that you would use as an inside matte further
down the tree. For this, you’d soften and erode the screen matte.
USER GUIDE
457
Screen Processing | Despot
Despot
This controls how much to simplify the matte. It coagulates similar regions so that, for example, black
specks in the white matte can be absorbed by the surrounding white areas. Increasing the Screen
Despot Black removes isolated spots of black in the white matte. Increasing Screen Despot White
removes isolated spots of white in the background up to that size.
Eroded matte. Despot.
USER GUIDE
458
Mattes | Despot
Mattes
There are 4 mattes in Keylight.
1. Screen Matte
2. Inside Mask
3. Outside Mask
4. Alpha (Composite Alpha)
The Screen Matte is generated by the Keylight algorithm after the screen color has been picked. It
can be processed (clipped, eroded, etc) by the screen matte processing tools.
The Inside Mask is the hold out matte. It is used to confirm areas that are definitely foreground. If
your subject has blue eyes and is being shot in front of a blue screen, this mask can be used to put
back the eyes. This mask is taken from the InM input to Keylight. The embedded alpha channel of the
foreground input can be added to this mask using the Source Alpha parameter in the Inside Mask
folder.
The Outside Mask is the garbage mask and is used to remove unwanted objects (lighting rigs, etc)
from the foreground. The mask is taken from the OutM input to Keylight. The luminance or the alpha
of this input is set using the OutM Component parameter.
The matte used to blend the foreground and background in the final composite is the alpha displayed
in the alpha channel of the composite. This matte is the combination of the screen matte, inside
matte, and outside matte.
USER GUIDE
459
Inside and Outside Masks | Despot
Inside and Outside Masks

If you can’t adequately improve the screen matte using the clip levels, you can create a matte in Nuke
round the pixels you definitely want to be foreground or background and use this as a mask input.
The inside mask makes the foreground less transparent and the outside mask is used to clean up the
background that might have bits of the foreground showing through. It is sometimes referred to as
the hold out mask.
The outside mask (garbage mask) is often used to clean up screens that are not a constant color or
have lighting rigs in shot by forcing the alpha transparent.
The inside mask can be used to keep elements in the foreground that you don't want to lose (an
actor’s blue eyes in front of a blue screen). These masks should normally be softened externally to
blend into the screen matte.
The below image shows the Bezier spline drawn around the lighting rig on the left side of the screen.
USER GUIDE
460
Inside and Outside Masks | Despot
Connect the mask to the OutM input of Keylight and switch the parameter OutM Component to
Alpha. The outside mask forces that part of the image to be in the background, thus keying out the
rig.
USER GUIDE
461
Source Alpha | Despot
Source Alpha
This parameter determines what to do with any embedded alpha in the original source image. You
need this if you are doing multiple keys on different parts of the image with the View output set to
Intermediate Result.
• Ignore - this does not add any embedded alpha to the screen matte.
• Add To Inside Mask - the embedded alpha is added to the inside mask. You should select this when
multipass keying with Output View set to Intermediate Result.
• Normal - the embedded alpha is used to composite the image.
USER GUIDE
462
Color Replacement | Inside mask
Color Replacement
Remember that Keylight does two things - it removes the screen color to despill the image and
generates an alpha (Screen Matte) to composite the foreground over the background layer.
If you then process the Screen Matte, for example, by eroding the alpha or changing the clip levels,
Keylight would be removing the wrong amount of screen color from the pixels whose transparency
has now changed. The Screen Replace instructs Keylight how to deal with such pixels. The Status
displays which pixels use a replace method. Those pixels that use a replace method because the
alpha processing tools modified the transparency are green, whilst those pixels whose transparency
was modified by the inside matte are blue. See the Status View under View.
There are four options to the replace method. These are:

1. None - the despilled image is left untouched if the alpha is modified.
2. Source - the image has a corresponding amount of the original pixel (screen color and all)
reintroduced/removed if the alpha is changed.
3. Hard Color - the despilled image has a corresponding amount of the Screen Replace Color
added for any increase in alpha.
4. Soft Color - the despilled image has a corresponding amount of the Screen Replace Color added
for any increase in alpha, however, it attempts to modulate the luminance of the resulting pixel so
that it matches the original pixel. This gives a more subtle result than the Hard Color option.
Inside mask
If the changes to the screen matte are due to an inside mask, the Inside Replace and Inside Replace
Color parameters can be used to modify the color in these areas just like the Screen Replace
parameters described above.
Edges
Built-in crop tools are included to quickly remove parts of the foreground at the edges of the image.
It can also be useful in tidying up a matte at the edges where luminance changes in the blue screen
are proving difficult to key out.
With X Method and Y Method set to Color and Edge Color, set to pure blue (for a blue screen), set
the Left to crop out the left-hand side of the image revealing the background. The figures below show
the changes to the Combined Matte with cropping.
USER GUIDE
463
Color Replacement | InM component
Left = 0. Left = 0.35.
InM component
The component (luminance or alpha channel) of the inside mask input that is used in the calculations.
Typically, this is a garbage matte that covers the area you know to be 100% foreground.
A bluescreen image. An inside matte.
Tip: To avoid having to roto the inside matte throughout the clip, you can connect another
Keylight node to the InM input and use it to create a hard, dilated key (set Screen Dilate to
a low value).
OutM component
The component (luminance or alpha channel) of the outside mask. Typically, this is a garbage matte
that covers the area you know to be 100% background.
USER GUIDE
464
Color Replacement | OutM component
A bluescreen image. An outside matte.
USER GUIDE
465
Keying with Primatte
This section explains how to use the blue/green screen keyer, Primatte, in Nuke.
Quick Start
1. Connect the Primatte node to your background and foreground images. See Connecting the
Primatte Node.
2. Click on the Auto-Compute button. Primatte attempts to automatically sample the backing
screen color and perform the cleanup phases of keying with Primatte.
If you get good results, jump ahead to spill removal. See Spill Removal - Method #1.
If you don't get the results you wanted from Auto-Compute, continue to step 3.
3. Click the operation dropdown, select Smart Select BG Color, and manually sample the targeted
background color by holding Ctrl/Cmd and clicking in the Viewer.
See Smart Select BG Color for more information.

4. Click the operation dropdown, select Clean BG Noise, and clean up any remaining white regions
in the dark, bluescreen area by sampling in the Viewer.
See Clean BG Noise for more information.
5. Click the operation dropdown, select Clean FG Noise, and clean up any dark regions in the
foreground by sampling in the Viewer.
See Clean FG Noise for more information.
6. Remove any spill using the spill removal tools. See Spill Removal - Method #1.
USER GUIDE
466
Connecting the Primatte Node |
Connecting the Primatte Node

1. Start up Nuke and create a Primatte node (Keyer > Primatte).
2. Connect a foreground image to the Primatte node’s fg input and a background image to the bg
input.
3. Add a Viewer node so you can see the result.
4. When you select the Primatte node, the Primatte properties panel displays.
USER GUIDE
467
Connecting the Primatte Node |
USER GUIDE
468
Primatte Basic Operation Tutorial | Auto-Compute
Primatte Basic Operation Tutorial

This describes the operation of the Primatte node in Nuke. A more detailed explanation of how the
Primatte algorithm actually works can be found under The Primatte Algorithm .
Auto-Compute
Primatte has a feature that attempts to eliminate the first three steps of the more standard keying
procedure. The Auto-Compute button is a good starting point and it may make your keying
operation much easier.
1. Click on the Auto-Compute button. Primatte attempts to automatically sense the backing screen
color, eliminate it, and even get rid of some of the foreground and background noise that would
normally be cleaned up in the Clean BG Noise and Clean FG Noise phases of keying with
Primatte.
2. If you get good results then jump ahead to the spill removal tools. See Spill Removal - Method #1.
3. If you don't get the results you wanted from Auto-Compute, please continue from this point on
to get the basic Primatte operation procedures. See Smart Select BG Color
The basic functionality for the Primatte interface is centered around the operation dropdown menu
and the Viewer window.
USER GUIDE
469
Primatte Basic Operation Tutorial | Auto-Compute
There are four main steps to using the Primatte and Select BG Color is the first step.
USER GUIDE
470
Smart Select BG Color | Auto-Compute
Smart Select BG Color

Ensure that the Smart Select BG Color action is selected (it should be at this time as it is the default
operation mode).
Position the cursor in the bluescreen area (or whatever background color you are using), usually
somewhere near the foreground object. Hold the Ctrl/Cmd key down and sample the targeted
background color. Release the mouse button and Primatte starts the compositing process. If the
foreground shot was done under ideal shooting conditions, Primatte has done 90-95% of the keying
in this one step and your image might look like this.
However, if you have a very unevenly lit backing screen, you may not be getting the results you’re
after. If this is the case, enable adjust lighting on the Primatte properties. For more information, see
Actions Section.
Note: Primatte works equally well with any color backing screen. It does not have to be a
specific shade of green or blue.
Tip: If you dragged the cursor in the blue area, Primatte averages the multi-pixel sample to
get a single color to adjust to. Sometimes Primatte works best when only a single pixel is
sampled instead of a range of pixels. The color selected at this point in the Primatte
operation is critical to the operation of the node from this point forward. Should you have
difficulties further along in the tutorial after selecting a range of blue shades, try the Smart
Select BG Color operation again with a single dark blue pixel or single light blue pixel. You
USER GUIDE
471
Smart Select BG Color | Auto-Compute
can also switch to the alpha channel view and click around in the bluescreen area and see
the different results you get when the initial sample is made in different areas.
Tip: If you would rather make a rectangular selection and not use the default 'snail trail'
sampling method, you can do a Ctrl+Shift+drag sample.
Tip: If the foreground image has a shadow in it that you want to keep it in the composite, do
not select any of the dark blue pixels in the shadow and the shadow comes along with the
rest of the foreground image.
The second and third steps in using Primatte require viewing the matte or alpha view in the Viewer
window. Press the A key on the keyboard to change to the alpha view. The image displayed changes
to a black and white matte view of the image that looks like this.
USER GUIDE
472
Clean BG Noise | Auto-Compute
Clean BG Noise
Set operation to Clean BG Noise. If there are any white regions in the dark, bluescreen area, it is
noise (or shades of blue that did not get picked up on the first sample) and should be removed.
Sample through these white noise regions and release the pen or mouse button to process the data
and eliminate the noise. Repeat this procedure as often as necessary to clear all the noise from the
background areas. Sometimes increasing the brightness of your monitor or the screen gamma allows
you to see noise that would otherwise be invisible.
Note: You do not need to remove every single white pixel to get good results. Most pixels
displayed as a dark color close to black in a key image become transparent and virtually
allow the background to be the final output in that area. Consequently, there is no need to
eliminate all noise in the bluescreen portions of the image. In particular, if an attempt is
made to meticulously remove noise around the foreground object, a smooth composite
image is often difficult to generate.
Tip: When clearing noise from around loose, flying hair or any background/foreground
transitional area, be careful not to select any of areas near the edge of the hair. Leave a little
noise around the hair as this can be cleaned up later using the FineTuning sliders.
Before background noise After background noise

removal. removal.
USER GUIDE
473
Clean FG Noise | Auto-Compute
Clean FG Noise
If there are dark regions in the middle of the mostly white foreground object, that is, if the key is not
100% in some portion of the targeted foreground, select Clean FG Noise from the operation
dropdown menu. Use the same techniques as for Clean BG Noise, but this time sample the dark
pixels in the foreground area until that area is as white as possible.
Before foreground noise After foreground noise

removal. removal.
Tip: After sampling the backing screen color and finding that the edges of the foreground
object look very good, you sometimes find that an area of the foreground object is
transparent. This is due to the foreground containing a color that is close to the backing
screen color. When this transparency is removed using the Clean FG Noise operation, the
edge of the foreground object picks up a fringe that is close to the backing screen color and
it is very hard to remove without sacrificing quality somewhere else on the image. To tackle
this problem, you can enable hybrid render. For more information, see Actions Section.
These were the steps necessary to create a clean matte or key view of the image. With this key, the
foreground can be composited onto any background image. However, if there is spill on the
foreground object from light that was reflected off the background, a final operation is necessary to
remove that background spill get a more natural looking composite.
For the fourth step in the Primatte operation, return the RGB view to the Viewer window by clicking
again on the A keyboard key. This turns off the alpha channel viewing mode; the Viewer window again
displays the RGB view with the background image (if you connected one to the Primatte node).
The sample image below has gone through the first three steps and has examples of spill. Notice the
blue fringe to her hair and a blue tint on her right cheek, arm and chest.
USER GUIDE
474
Clean FG Noise | Auto-Compute
Blue spill visible.
USER GUIDE
475
Spill Removal - Method #1 | Auto-Compute
Spill Removal - Method #1

There are three ways in Primatte to remove the spill color. The quickest method is to select Spill
Sponge from the operation dropdown menu and then sample the spill areas away. By just
positioning the cursor over a bluish pixel and sampling it, the blue disappears from the selected color
region and is replaced by a more natural color. Additional spill removal should be done using the
Fine Tuning tools or by using the Spill(-) operation. Both are explained further on in this chapter.
Note: All spill removal/replacement operations in Primatte can be modified using the Spill
Process replace with tools. Spill can be replaced with either the complement of the
background color, a solid color you’ve selected, or by colors brought from a defocused
background. Depending on the spill conditions, one of these options should provide the
results you are looking for. See the information in Replacing Spill for more details.
Note: Primatte's spill removal tools work on color regions. In the image in Clean FG Noise,
samples should be made on the light flesh tones, the dark flesh tones, the light blonde hair,
the dark blonde hair, and the red blouse color regions. One sample in each color region
removes spill from all similar colors in the foreground image.
If the spilled color has not been totally removed using the Spill Sponge or the result of the Spill
Sponge resulted in artifacts or false coloring, the Spill(-) tool should be used instead for a more
subtle and sophisticated removal of the spilled background color. This is discussed in Actions Section.
g>(r+b)/2?(r+b)/2:g
channel to:
b>(r+g)/2?(r+g)/2:b
Only.
USER GUIDE
476

1. Select the Fine Tuning Sliders in the operation dropdown menu. This activates the Fine Tuning
sliders.
2. In the Viewer, zoom into an area that has some blue edges or spill.
3. Using the cursor, sample a color region that has some spill in it. When you let up on the pen or
mouse button, Primatte registers the color selected (or an average of multiple pixels) in the
current color swatch.
4. For most images, the L-poly (spill) slider is all that is required to remove any remaining blue spill.
Move the slider to the right to remove spill color from the sampled pixels. Move it to the left to
move the selected pixels toward the color in the original foreground image.
When using the L-poly (spill) slider, spill color replacement is replaced based on the setting of the
Spill Process replacement with control. For more information on these tools, see the section of
this chapter on Replacing Spill .
Tip: It is better to make several small adjustments to the blue spill areas than a single major
one.
5. You can use the other two sliders in the same way for different key adjustments. The S-poly
(detail) slider controls the matte softness for the color which is closest to the background color.
For example, you can recover lost smoke in the foreground by selecting the Fine Tuning Sliders
action, sampling the area of the image where the smoke just starts to disappear and moving the
S-poly (detail) slider to the left. The M-poly (trans) slider controls the matte softness for the color
which is closest to the foreground color. For example, if you have thick and opaque smoke in the
foreground, you can make it semi-transparent by moving the Transparency slider to the right
after selecting the pixels in the Fine Tuning Sliders mode.
Tip: If the foreground image changed color dramatically during the fine tuning process, you
can recover the original color by selecting an area of the off-color foreground image and
moving the L-poly (spill) slider slightly to the left. This may introduce spill back into that
color region. Again, use the Fine Tuning Sliders option to suppress the spill, but make
smaller adjustments this time.
USER GUIDE
477

If these final spill suppression operations have changed the final compositing results, you may have
to return to earlier operations to clean up the matte. If the composite view looks good, it is a good
idea to go back and take a final look at the alpha channel view. Sometimes in the Primatte operation,
a 100% foreground area (all white) becomes slightly transparent (gray). You can clean those
transparent areas up by using the Matte Sponge tool.
1. Select the Matte Sponge tool in the operation dropdown menu.
2. Click on the transparent pixels and they become 100% foreground. All of the spill-suppression
information remains intact.
Note: The Matte(+) operation also works to solve this problem. For more information, see
Sampling Tools.
USER GUIDE
478
Sampling Tools | The Spill Sampling Tools
Sampling Tools
Primatte’s spill, matte, and detail sampling tools allow you to make fine adjustments to balance
between these aspects of your composite. You can find these tools in the operation dropdown menu
in the Primatte properties panel.
The Spill Sampling Tools

Using the Spill(+) and Spill(-) modes, you can gradually remove or recover the spill intensity on the
foreground object by sampling the referenced color region repeatedly. The difference to the Spill
Sponge tool is that Spill Sponge removes the spill component in a single action at one level and does
not allow sampling the same pixel a second time. Even though just a small amount of spill needs to
be removed, the Spill Sponge removes a preset amount without allowing any finer adjustment. To
use the Spill(+) and
Spill(-) tools:
1. Select the Spill(-) sampling tool from the operation dropdown menu.
2. In the Viewer, zoom into an area that has some blue edges.
3. Click on a pixel with some spill on it.
4. Repeated clicking incrementally removes the spill. Continue this operation until you reach a
desired result.
5. To add spill, select the Spill(+) tool and repeat steps from 2 to 4.
The Matte Sampling Tools

The Matte(+) and Matte(-) modes are used to thicken or attenuate the matte information. If you
want a thinner shadow on a foreground object, you can use the Matte(-) mode as many times as you
like to make it more transparent. On the other hand, you can use the Matte(+) mode to make the
matte thicker in that color region.
USER GUIDE
479
Sampling Tools | The Detail Sampling Tools
The Detail Sampling Tools

The Detail(+) and Detail(-) modes are a refined version of Clean BG Noise and Restore Detail. For
example, when you see some dilute noise in the backing area but don't want to remove it completely
because it affects some fine detail in a different area, try using Detail(-). It attenuates the noise
gradually as multiple samples are made on the pixel. You should stop the sampling when important
fine details start to disappear.
USER GUIDE
480
Replacing Spill | The Detail Sampling Tools
Replacing Spill
The proper processing of spill on foreground objects is one of the many useful features of Primatte.
You can move between four modes to see how they affect the image clip you are working with. Under
Spill Process, in the replace with dropdown menu, you can select the following options:
• no suppression - In this mode, no suppression is applied.
• complement - This is the default spill replacement mode. This mode maintains fine foreground
detail and delivers the best quality results. If foreground spill is not a major problem, this mode is
the one that should be used. The complement mode is sensitive to foreground spill. If the spill
intensity on the foreground image is rather significant, this mode may often introduce serious noise
in the resulting composite.
The complement mode maintains fine detail.
Serious noise in the composite.

• solid color - In the solid color mode, the spill component is replaced by a palette color that you can
pick. While the complement mode uses only the backing color complement to remove small
amounts of spill in the original foreground, the solid color mode tries to assuage the noise using
the user-defined palette color. Changing the palette color for the solid replacement, you can apply
good spill replacement that matches the composite background. Its strength is that it works fine
with even serious blue spill conditions.
On the negative side, when using the solid color mode, fine detail on the foreground edge tends to
be lost. The single palette color sometimes cannot make a good color tone if the background image
has some high contrast color areas.
Smooth spill processing with solid color replacement.
USER GUIDE
481
Replacing Spill | The Detail Sampling Tools
• defocused background - The defocused background mode uses a defocused copy of the
background image to determine the spill replacement colors instead of a solid palette color or just
the complement color. This mode can result in good color tone on the foreground object even with
a high contrast background. As in the example below, spill can even be removed from frosted glass
using this feature and still retain the translucency.
On the negative side, the defocused background mode sometimes results in the fine edge detail of
the foreground objects getting lost. Another problem could occur if you wanted to later change the
size of the foreground image against the background. Since the background/foreground alignment
would change, the applied color tone from the defocused image might not match the new
alignment.
Blue suppression of a frosted glass object.
USER GUIDE
482
Primatte Controls | Initialize Section
Primatte Controls
On the Primatte properties panel, you can further adjust several controls.
Initialize Section
In the Initialize section, you can select which algorithm Primatte uses to calculate your keying result:
Primatte algorithm dropdown menu.

• Primatte - The Primatte algorithm delivers the best results and supports both the solid color and
the complement color spill suppression methods. It is the algorithm that uses three multi-faceted
polyhedrons (as described further down in the this chapter) to separate the 3D RGB colorspace. It is
also the default algorithm mode and, because it is computationally intensive, it may take longer to
render.
• Primatte RT - is the simplest algorithm and therefore, the fastest. It uses only a single planar
surface to separate the 3D RGB colorspace (as described further down in this chapter) and, as a
result, does not have the ability to separate out the foreground from the backing screen as carefully
as the above Primatte algorithm. Other disadvantages of the Primatte RT algorithm is that it does
not work well with less saturated backing screen colors and it does not support the complement
color spill suppression method.
• Primatte RT+ - this is in between the above two options. It uses a six planar surface color
separation algorithm (as described further down in this document) and delivers results in between
the other two in both quality and performance. Other disadvantages of the Primatte RT+ algorithm
is that it does not work well with less saturated backing screen colors and it does not support the
complement color spill suppression method.
• Reset - Clicking this resets all of the Primatte properties to their initial values.
• Auto-Compute - This can be used as the first step in the Primatte operation. Its purpose is to try
and do the first three steps of the Primatte operation for you. It tries to automatically detect the
backing screen color, remove it, and do some clean-up on the foreground and background noise. If
the clip was shot with an evenly lit, well-saturated backing screen, the Auto-Compute button leaves
you with an image that may only need some spill removal to complete your keying operation. See
Auto-Compute.
• viewer - This opens a Primatte Viewer that displays a graphical representation of the Primatte
algorithms and allows you to see what is happening as the various Primatte tools are used. It is a
USER GUIDE
483
Primatte Controls | Initialize Section
passive feature that has no adjustment capabilities, but it may prove useful in evaluating an image
as you operate on it. See Primatte Viewer Tools.
When you select viewer, you are presented with a window that may look similar to one of these
images (depending on which Primatte algorithm you have selected).
Primatte algorithm. Primatte RT+ algorithm.
Primatte RT algorithm.
The different algorithms are described in more detail in a later section of this chapter, see The
Primatte Algorithm . For a description of the Primatte viewer tools, see Primatte Viewer Tools.
USER GUIDE
484
Primatte Viewer Tools | Initialize Section
Primatte Viewer Tools

To navigate within the Primatte Viewer, you can:
• drag to rotate the polyhedrons,
• Shift+drag to zoom in and out of the polyhedrons, and
• Ctrl/Cmd+drag to pan the polyhedrons.
At the top of the Primatte Viewer window, there are also three areas that you can click on:
• Clicking and dragging on the blue center area allows you to move the window around on the screen.
• Clicking and dragging on the triangular white region in the upper right corner allows you to scale the
Primatte Viewer window.
• Clicking on the square white region in the upper left of the window displays a dropdown menu that
looks like this:
Note: A selected feature has a solid yellow square next to it. An unselected feature has a
hollow yellow square next to it.
• Minimize - This feature, when selected, makes the Primatte Viewer window disappear. Only the blue
title bar at the top of the window remains.
• Large Surface - This feature, when selected, displays the large Primatte polyhedron in the Viewer
window.
• Middle Surface - This feature, when selected, displays the medium Primatte polyhedron in the
Viewer window.
USER GUIDE
485
Primatte Viewer Tools | Initialize Section
• Small Surface - This feature, when selected, displays the small Primatte polyhedron in the Viewer
window.
• Opaque - This feature, when selected, makes the selected polyhedrons opaque. De-selecting it
makes them semi-transparent.
• Samples - This feature, when selected, allows you to sample color regions on the image window
using the 3D Sample operation mode and see where those regions are in relation to the
polyhedron and the algorithm. The colors are displayed as a spray of pixels in the color selected.
This button only allows you to see or hide the sampled colors.
Note: The 3DSample mode must be selected in the operation dropdown menu for this
feature to work.
• Clear BG - This feature changes the background color of the Primatte Viewer window from black
(when unselected) to transparent (when selected).
• Sliced - This feature, when selected, slices open the large and medium polyhedrons so that the
inner polygons can be seen. When unselected, the largest polyhedron selected becomes a
completely closed polyhedron and you might not be able see the inner polyhedrons (unless the
Opaque feature is deselected).
• Wireframe - This feature, when selected, changes the polyhedrons from shaded-surface objects to
wireframe objects.
• Quad view - This feature, when selected, displays the different polyhedrons in different parts of the
Primatte Viewer:
• The upper left corner of the Viewer shows the medium polyhedron.
• The upper right corner of the Viewer shows the small polyhedron.
• The lower left corner of the Viewer shows the large polyhedron as a wireframe object.
• The lower right corner of the Viewer shows the large polyhedron.
USER GUIDE
486
Degrain Section | Degrain type
Degrain Section
The Degrain tools are used when a foreground image is highly compromised by film grain. As a
result of the grain, when backing screen noise is completely removed, the edges of the foreground
object often become harsh and jagged leading to a poor key. These tools were created to, hopefully,
help when a compositing artist is faced with a grainy image.
Degrain type
The Degrain type dropdown menu gives you a range of grain removal from none to large. If the
foreground image has a large amount of film grain induced pixel noise, you may lose a good edge to
the foreground object when trying to clean all the grain noise with the Clean BG Noise operation
mode. These tools allow you to clean up the grain noise without affecting the quality of the key. A
short tutorial explaining when and how to use these tools is at the end of this section. Select:
• none - When none is selected, you get the color of the exact pixel sampled. This is the default
mode.
• small - When small is selected, you get the average color of a small region of the area around the
sampled pixel. This should be used when the grain is very dense.
• medium - When medium is selected, you get the average color of a medium-sized region of the
area around the sampled pixel. This should be used when the grain is less dense.
• large - When large is selected, you get the average color of a larger region of the area around the
sampled pixel. This should be used when the grain is very loose.
• tolerance slider - Adjusting the tolerance slider should increase the effect of the Clean BG Noise
tool without changing the edge of the foreground object.
Degrain Tools Tutorial

If you have a noisy image as in the example below...
USER GUIDE
487
Degrain Section | Degrain Tools Tutorial
...you will find that the matte is also noisy:
You can use the Clean BG Noise operation to remove the noisy pixels, but this can also modify the
edge of the foreground object in a negative manner.
Using the Degrain tools in the following way may help you clean up the image and still get a good
edge on the matte:
1. Use the Clean BG Noise operation just a small amount to remove some of the white noise in the
alpha channel view but do use it so much that you affect the edge of the foreground object.
2. Then set the Degrain type dropdown menu to small as a first step to reduce the grain:
USER GUIDE
488
With the degrain tolerance slider set at 0, move it around some. This should increase the affect
of the Clean BG Noise tool without changing the edge of the foreground object.
Sometimes, this may not be enough to totally remove the grain, so by adjusting the degrain
tolerance slider, you can tell the Primatte algorithm what brightness of pixels you think
represents grain. You should try not to use too high of a value; otherwise, it affects the overall
matte. For an example of an over-adjusted image, see below.
The Primatte degrain algorithm uses a defocused foreground image to compute the noise.
Note: The small, medium and large settings for the degrain tools all produce defocused
foregrounds that have larger or smaller blurs respectively.
Note: It is important to make sure that the crop settings are correctly applied; otherwise,
when the defocus image is generated, if there is garbage on the edges of the images, then
that garbage is blurred into the defocused foreground.
As a review:
1. Select the Select BG Color operation mode and click on a backing screen color.
USER GUIDE
489
2. Select the Clean BG Noise operation mode and use it sparingly so that it has minimum affect to
the edge of the foreground object.
3. If there is still grain in the backing screen area, then use the Degrain type functionality starting at
the small setting to reduce the grain.
4. If the grain is still present, then try increasing the tolerance slider a little - not too much.
5. If grain is still a problem, then try changing the type to medium or large and also changing the
grain tolerance until the desired effect is achieved.
Note: The grain functionality does not always remove grain perfectly, but is sometimes
useful to minimize its effects.
USER GUIDE
490
Actions Section | Smart Select BG Color
Actions Section
In the operations dropdown menu, you can select from the following options:
Smart Select BG Color

This operational mode gets the sampled backing screen color, analyzes the original foreground
image, and determines the foreground areas using the Primatte foreground detection routine. Then,
using the newly determined foreground areas, performs a Clean FG Noise operation internally and
determines a more desirable shape for the middle and outer polyhedrons, before finally rendering
the composite using the generated polyhedrons.
For keying operations, this is the first step and should be followed by the steps described
immediately below.
Clean BG Noise
When you select this operational mode, you sample pixels in the Viewer known to be 100%
background. White noisy areas in the 100% background region become black. This is usually the
second step in using Primatte.
Clean FG Noise
When you select this operational mode, you sample pixels on the image window known to be 100%
foreground. The color of the sampled pixels is registered by Primatte to be the same color as in the
original foreground image. This makes dark gray areas in the 100% foreground region white. This is
usually the third step in using Primatte.
Matte Sponge
When this operational mode is selected, the color sampled in the Viewer becomes 100% foreground.
However, if the sampled color is already keyed out and removed, it leaves the current suppressed
color. It only affects the key or matte information. This tool is usually used to quickly remove stray
transparent pixels that have appeared during the chromakeying procedure. It is a quick and easy way
to make final adjustments to a composite.
USER GUIDE
491
Actions Section | Make FG Trans
Make FG Trans
When this mode is selected, the opaque foreground color region sampled in the Viewer becomes
slightly translucent. This operation is useful for the subtle tuning of foreground objects which are
otherwise 100% covered with smoke or clouds. It can only be used one time on a particular color. For
a more flexible way to thin out a color region and be able to make multiple samples, you should use
the Matte (-) tool. It expands the medium polyhedron slightly.
Restore Detail
With this mode selected, the completely transparent background region sampled in the Viewer
becomes translucent. This operation is useful for restoring lost hair details, thin wisps of smoke and
the like. It shrinks the small polyhedron slightly.
Spill Sponge
When you select this operational mode, the background color component in the sampled pixels (or
spill) within the image window is keyed out and removed for the color region selected. This operation
can only be used once on a particular color and the amount of spill suppression applied is not
adjustable. It is the fastest way to remove spill from a composite image. For more accurate spill
suppression, a Fine Tuning or Spill (+) operation should follow or be used instead. This can usually
be the fourth (and final) step in using Primatte unless additional adjustments are necessary.
Spill (-)
When this operational mode is selected, color spill is removed from the sampled pixel color (and all
colors like it) in the amount of one Primatte increment. If spill color remains, another click using this
tool removes more of the color spill. Continue using this tool until all color spill has been removed
from the sampled color region. This tool expands the Primatte large polyhedron in the color region
sampled.
Spill (+)
When this operational mode is selected, color spill is returned to the sampled pixel color (and all
colors like it) in the amount of one Primatte increment. This tool can be used to move the sampled
color more in the direction of the color in the original foreground image. It can be used to nullify a
Spill (-) step. This tool dents the Primatte large polyhedron in the color region sampled.
USER GUIDE
492
Actions Section | Matte (-)
Matte (-)
When this operational mode is selected, the matte is made more translucent for the sampled pixel
color (and all colors like it) in the amount of one Primatte increment. If the matte is still too opaque,
another click using this operational mode tool makes the sampled color region even more
translucent. This can be used to thin out smoke or make a shadow thinner to match shadows in the
background imagery. This tool expands the Primatte medium polyhedron in the color region
sampled.
Matte (+)
When this operational mode is selected, the matte is made more opaque for the sampled pixel color
(and all colors like it) in the amount of one Primatte increment. If the matte is still too translucent or
thin, another click using this operational mode tool makes the sampled color region even more
opaque. This can be used to thicken smoke or make a shadow darker to match shadows in the
background imagery. It can only make these adjustments to the density of the color region on the
original foreground image. It can be used to nullify a Matte (-) step. This tool dents the Primatte
medium polyhedron in the color region sampled.
Detail (-)
When this operational mode is selected, foreground detail becomes more visible for the sampled
pixel color (and all colors like it) in the amount of one Primatte increment. If detail is still missing,
another click using this operational mode tool makes detail more visible. This can be used to restore
lost smoke or wisps of hair. Sample where the smoke or hair just disappears and it returns to
visibility. This is for restoring color regions that were moved into the 100% background region. It may
start to bring in background noise if shooting conditions were not ideal on the foreground image.
This tool dents the Primatte small polyhedron in the color region sampled.
Detail (+)
When this operational mode is selected, foreground detail becomes less visible for the sampled pixel
color (and all colors like it) in the amount of one Primatte increment. If there is still too much detail,
another click using this operational mode tool makes more of it disappear. This can be used to
remove smoke or wisps of hair from the composite. Sample where detail is visible and it disappears.
This is for moving color regions into the 100% background region. It can be used to nullify a Detail (-)
step. This tool expands the Primatte small polyhedron in the color region sampled.
USER GUIDE
493
Actions Section | Fine Tuning Sliders
Fine Tuning Sliders

If the spilled color has not been totally removed at this point, use the Fine Tuning Sliders for more
subtle removal of spilled background color. For more information, refer to the section on Fine
Tuning.
3D Sample
When this operational mode is selected and viewer is enabled in the Primatte properties, the
sampled colors are displayed as a spray of white pixels in the Primatte Viewer. This allows you to see
where the selected backing screen colors reside within the 3D RGB colorspace.
Simple Select BG Color

Although using the Smart Select BG Color operation is preferred, the Simple Select BG Color
operation may still be useful in some cases. This operation uses the older method of taking the
sampled backing screen color, projecting a line in the opposite direction on the hue wheel and
generating artificial pixels that may represent the FG object. Then, using the artificially generated
foreground pixels, it performs a Clean FG Noise operation internally and creates the shape of the
middle and outer polyhedrons, before finally rendering the composite using the generated
polyhedron.
Current Color Chip

This shows the current color selected (or registered) by the Fine Tuning operational mode.
Adjust Lighting
This enables Primatte’s light adjusting feature, which, based on the currently selected BG color,
generates a clean, evenly lit backing screen to use in the keying operation. This can improve the
results if you have a very unevenly lit backing screen. For more information, see Adjust Lighting.
USER GUIDE
494
Actions Section | Hybrid Render
An uneven backing screen.
The resulting matte when The resulting key. Areas

adjust lighting is disabled. near
the sampled color on the
right look good, but on the
left the hair is chunky.
The resulting matte when The resulting key is clean

adjust lighting is enabled. on both sides.
Hybrid Render
To tackle this problem, you can enable hybrid render to have Primatte create two keys internally and
combine them for the best results. For more information, see Hybrid Matte.
USER GUIDE
495
Adjust Lighting | Hybrid Render
Adjust Lighting
When adjust lighting is enabled, Primatte first creates a grid on the foreground image and samples
the grid to determine which squares contain foreground information and which contain some shade
of the background color. Using the remaining shades of blue, it uses extrapolations of the existing
backing screen colors, and then uses this technique to fill in the area previously covered by the
foreground object. This results in a clean backing screen that has no foreground object (to see this,
you can set output mode to adjust lighting BG). Primatte uses this clean backing screen as the
reference data to process the original foreground image to generate light-adjusted foreground that
has a much more even shade of blue with the foreground object (this is displayed if you set output
mode to adjust lighting FG). This allows Primatte to perform the keying operation using the light-
adjusted foreground, resulting in a clean key around all areas on the foreground object.
The default settings under Adjust Lighting should detect all the areas of the grid that contain
foreground pixels and deliver a smooth, artificially created, optimized backing screen for the keying.
Should it fail to do this, you can adjust the settings of the algorithm by moving the threshold slider.
This determines if a grid pixel should be treated as a pure background sample, a simulated
background sample, or a foreground sample. Increasing the threshold value brings more of the
foreground into the adjusted lighting.
If necessary, you can also set the grid size used in the adjust lighting algorithm. Increasing this value
increases the grid resolution used in the adjusted lighting calculation.
USER GUIDE
496
Hybrid Matte | Hybrid Render
Hybrid Matte
After sampling the backing screen color and finding that the edges of the foreground object look very
good, you sometimes find that an area of the foreground object is transparent. This is due to the
foreground containing a color that is close to the backing screen color. When this transparency is
removed using the Clean FG Noise operation, the edge of the foreground object picks up a fringe
that is close to the backing screen color. This is very hard to remove without sacrificing quality
somewhere else on the image.
When hybrid render mode is enabled, Primatte internally creates two keys from the same image:
• core - This matte has the transparency removed, but suffers from the bad edges on the foreground
object.
• edge - This matte has a clean edge on the foreground, but suffers from transparency within the
foreground object.
The core matte with the bad edges is then blurred and eroded before it is composited over the edge
matte that has the transparency, resulting in a composite with the best of both options.
The controls under Hybrid Matte allow you to adjust the operations performed on the core matte:
• erode - Adjust the amount of erosion performed on the core matte. To view the results, set output
mode to hybrid core and view the alpha channel.
• blur radius - Adjust the blur radius used when blurring the core matte. To view the results, set
output mode to hybrid edge and view the alpha channel.
USER GUIDE
497
Fine Tuning | Spill or L-poly slider (Spill Removal)
Fine Tuning
When the Fine Tuning Sliders operational mode is selected, the color of the sampled pixel within the
Viewer window is registered as a reference color for fine tuning. It is displayed in the current color
chip in the Actions section. To perform the tuning operation, sample a color region on the image,
select a Fine Tuning slider and move the slider to achieve the desired effect. See the Fine Tuning
Sliders tool descriptions further on in this section for more details on slider selection.
Spill or L-poly slider (Spill Removal)

When in the Fine Tuning operation mode, this Spill slider can be used to remove spill from the
registered color region. After selecting the Fine Tuning Actions mode and registering a color region,
this slider can be moved to remove spill from the registered color region. The more to the right the
slider moves, the more spill is removed. The more to the left the slider moves, the closer the color
component of the selected region is to the color in the original foreground image. If moving the slider
all the way to the right does not remove all the spill, re-sample the color region and again move the
slider to the right. These slider operations are additive. This result achieved by moving the slider to
the right can also be achieved by clicking on the color region using the Spill(-) operational mode. This
slider bulges the Primatte large polyhedron near the registered color region.
Transparency or M-poly slider (Adjust Transparency)

When in the Fine Tuning operation mode, this Transparency slider can be used to make the matte
more translucent in the registered color region. After selecting the Fine Tuning operation mode and
selected a color region, moving this slider to the right makes the registered color region more
transparent. Moving the slider to the left makes the matte more opaque. If moving the slider all the
way to the right does not make the color region translucent enough, re-sample the color region and
again move the slider to the right. These slider operations are additive. This result achieved by
moving the slider to the right can also be achieved by clicking on the color region using the Matte(-)
operational mode. This slider bulges the Primatte medium polyhedron near the registered color
region.
Detail or S-poly slider (Add/Restore Lost Detail)

When in the Fine Tuning operation mode, this Detail slider can be used to restore lost detail. After
selecting the Fine Tuning operation mode and selected a color region, moving this slider to the right
makes the registered color region more visible. Moving the slider to the left makes the color region
USER GUIDE
498
Fine Tuning | Detail or S-poly slider (Add/Restore Lost Detail)
less visible. If moving the slider all the way to the right does not make the color region visible enough,
re-sample the color region and again move the slider to the left. These slider operations are additive.
This result achieved by moving the slider to the left can also be achieved by clicking on the color
region using the Detail(-) operational mode. This shrinks the small polyhedron (which contains all the
blue or green background colors) and releases pixels that were close to the background color. The S-
poly slider in the Fine Tuning mode is useful for restoring pixels that were lost because they were so
similar to the background color. This slider dents the Primatte small polyhedron near the registered
color region.
USER GUIDE
499
Spill Process Section | Replace with
Spill Process Section

The controls in the Spill Process section of the Primatte properties panel let you specify how to
replace the spill color.
Replace with
This allows you to select between the three methods of spill replacement:
• no suppression - No spill suppression is applied.
• complement - Replaces the spill color with the complement of the backing screen color.
• solid color - Replaces the spill color with a solid color of your choice.
• defocused background - Replaces the spill color with colors from a defocused version of the
background image.
Replace color slider

When solid color is selected, this sets the color to use to replace the spill.
Defocus slider
When defocused background is selected, this area allows you to adjust the amount of defocus
applied to the background buffer image.
USER GUIDE
500
Output Section | Defocus slider
Output Section
These are the formats for the output of the node (output mode):
• composite - outputs the composite result of the Primatte node, along with the calculated matte.
• premultiplied - outputs the premultiplied result of the Primatte node, along with the calculated
matte.
This can be useful if you want to do your compositing using a Merge node (with operation set to
over) rather than Primatte. This allows you to color correct, transform, and otherwise process your
image before compositing it over the background. Note, however, that Primatte works within the
sRGB colorspace, whereas Nuke works within a linear colorspace. This means you need to add a
Colorspace node after both Primatte and your original background image to convert their
colorspaces to sRGB, then do your color corrections or transforms, merge the images together, and,
finally, use another Colorspace node to convert the result back to linear.
USER GUIDE
501
• unpremultiplied - outputs the unpremultiplied result of the Primatte node, along with the
calculated matte.
This can be useful if you want to do your compositing using a Merge node (with operation set to
matte) rather than Primatte. This allows you to color correct, transform, and otherwise process
your image before compositing it over the background. Note, however, that Primatte works within
the sRGB colorspace, whereas Nuke works within a linear colorspace. This means you need to add a
Colorspace node after both Primatte and your original background image to convert their
colorspaces to sRGB, then do your color corrections or transforms, merge the images together, and,
finally, use another Colorspace node to convert the result back to linear.
• not premultiplied - outputs the unpremultiplied original foreground (rather than the result of the
Primatte node), along with the calculated matte.
• adjust lighting FG - outputs the light-adjusted foreground that the adjust lighting mode creates
(this has a more even shade of the backing color with the foreground object). If adjust lighting is
disabled, this option simply displays the un-optimized, original foreground image. For more
information, see Actions Section.
USER GUIDE
502
• adjust lighting BG - outputs the optimized artificial backing screen that the adjust lighting mode
creates (a clean backing screen that has no foreground object). For more information, see Actions
Section.
• hybrid core - outputs the internally generated core matte, used when hybrid render is enabled.
For more information, see Actions Section.
• hybrid edge - outputs the internally generated edge matte, used when hybrid render is enabled.
For more information, see Actions Section.
USER GUIDE
503
The Primatte Algorithm | Defocus slider
The Primatte Algorithm

There are three Primatte algorithms. Here is a chart that shows the main differences between them.
For a description of the Primatte algorithm, see Explanation of How Primatte Works.
For a description of the Primatte RT+ algorithm, see Explanation of How Primatte RT+ works.
For a description of the Primatte RT algorithm see Explanation of How Primatte RT works.
USER GUIDE
504
Explanation of How Primatte Works | Defocus slider
Explanation of How Primatte

Works
The Primatte chromakey algorithm is a sophisticated method of color space segmentation that can
be easily explained to help a user achieve maximum effectiveness with the tool. Basically, Primatte
segments all the colors in the foreground image into one of four separate categories. The result is a
spill suppressed foreground image and a matte which is used to apply the modified foreground to a
suitable background.
Primatte works in 3D RGB color space. Here is a visual representation of the Primatte algorithm after
an image has been processed.
By operating the Primatte interface, you essentially create three concentric, multi-faceted
polyhedrons. These can be pictured as three globes (or polyhedrons or polys), one within the other,
which share a common center point. The creation of these polyhedrons separates all possible
foreground colors into one of four regions; inside the small polyhedron (1), between the small and
medium polyhedrons (2), between the medium and the large polyhedrons (3) and outside the large
polyhedron (4).
USER GUIDE
505
The four regions created are described as follows:
Region 1 (inside the small polyhedron) - This region contains all of the foreground image colors that
are considered 100% background. These are the green or blue or whatever colors that were used as
the backing color of the foreground image.
Region 2 (between the small and medium polyhedrons) - This region contains all the foreground
colors that are at the edges of the foreground object(s), in glass, glass reflections, shadows, sheets of
water and other transparent and semi-transparent color regions. These color regions also have spill
suppression applied to them to remove color spill from the backing screen.
Region 3 (between the medium and large polyhedrons) - This region contains all the foreground
image colors that are 100% foreground but have spill suppression applied to them to remove color
spill from the backing screen. Otherwise they are 100% solid foreground colors.
USER GUIDE
506
Region 4 (outside the large polyhedron) - This region contains all the 100% foreground image colors
that are not modified from the original foreground image. There is no spill suppression applied to
these colors.
In the first step in using Primatte (Smart Select BG Color), you are asked to indicate the backing color
on the original foreground image. The sample should usually be taken from a medium shaded area
near the foreground object. By medium shaded area, it is meant that if green is the backing color and
the green area of the foreground image has many shades of green ranging from very pale green to
almost black, a shade of green in-between these extreme ranges should be chosen. If you’re not
getting good results using this sample, you should reset Primatte and take another sample using a
slightly darker or lighter shade of green. The first sample of Primatte often determines the final result
as the center point of all three polyhedrons is created based on this first sample.
A single pixel may be selected or a range of pixels (snail trail or rectangular sample). If a range of
pixels is taken, the sample is averaged to get a single color sample. This single pixel or averaged color
sample then becomes the center of the small polyhedron. A few other shades around that color are
included in the original small polyhedron.
USER GUIDE
507
Note: It is recommended that a single pixel be selected as the first sample as you then have
some idea where the center point of the polyhedrons is located. If a box sample or a long
snail trail sample is made, you can only guess at the average color that ends up being the
center point. You can get an idea how this sample affects the algorithm by resetting the
Primatte node, viewing the alpha channel, and clicking around on the green or blue screen
area while in the Smart Select BG Color operation mode. You can immediately see the
results of the initial settings of the polyhedrons in this way.
The second step in using Primatte is to clean up the backing color area by adding additional shades
of green or blue to the small poly. This second step (Clean BG Noise) is usually executed while
viewing the black and white alpha channel.
Before background After background

noise removal. noise removal.
While in the Clean Bg Noise sampling mode, you sample the white milky regions as shown in the
left-hand image above. As you sample these regions, they turn to black as shown in the right-hand
image above.
What is happening in the Primatte algorithm is that these new shades of green (the white milky areas)
are added to the small poly where all the shades of green or blue are moved. The advantage of this
technique is that the polyhedron distorts to enclose only the shades of green that are in the backing
screen. Other shades of green around these colors are left undisturbed in the foreground.
Now that you have created a small polyhedron, you need to shape the medium and large polys. A
default medium and large poly are both automatically created and are then modified based on the
next couple of Primatte operations. The third Primatte step (Clean FG Noise) is to sample and
eliminate gray areas in the 100% foreground area of the image.
USER GUIDE
508
Before foreground noise After foreground noise

removal. removal.
Again, you make several samples on the dark, grayish areas on the foreground object until it is solid
white in color. Primatte is shaping the large polyhedron with each color region that is sampled. Care
should be taken in both this and the previous steps to not sample too close to the edges of the
foreground object. Getting too close to the foreground object's edges results in hard edges around
the foreground object. Primatte uses these samples to modify and shape the medium and large polys
to the desired shape. At this point, the matte or key has been created and would allow the foreground
objects to be composited into a new background image.
If you now view the RGB channels, there is usually color spill on the edges (and sometimes the center)
of the foreground objects. When on the edges of the foreground object, this spill comes from where
the edges of the foreground object blended into the backing color. If it is on the center of the
foreground object, it usually results from reflected color from the backing screen. The next Primatte
step, either Spill Sponge, Fine Tuning or Spill(-), can now be used to eliminate this spill color.
Let's take a look at what is happening in the Primatte algorithm while this next step is performed.
Here is what the various tools in Primatte do to the Polyhedrons when they are used:
USER GUIDE
509
As you can see above, the Spill Sponge bulges the large polyhedron in the color region specified. A
color region is specified by clicking on the image in a particular area with spill present. For example,
if you click on some spill on the cheek of a foreground person, Primatte goes to the section of the
large polyhedron closest to that particular flesh tone and bulges the polyhedron there. As a result, the
flesh tones move from outside the large poly to in-between the medium and large polys. This is
Region 3 and, if you remember, is 100% foreground with spill suppression. As a result of the
suppression, the spill is removed from that cheek color and all other shades of that color on the
foreground. You would then continue to sample areas of the image where spill exists and each
sample would remove spill from another color region.
When all spill has been removed, you should have a final composite. As a last step, you should view
the alpha channel and make sure that gray, transparent areas have not appeared in the foreground
area. If there are any, you should select the Matte Sponge operation mode and sample those gray
pixels until they have all turned white again.
The Matte Sponge and Spill Sponge tools bulge or dent the polyhedrons a pre-selected amount. If
the desired results are not achieved or the results are too extreme for the image, a manual method
can be applied. You should select the Fine Tuning sliders, select a color region of interest and then
move the appropriate slider to get the desired results.
For example, to remove spill, select a region of the composite image with spill on it. Move the spill or
large poly slider to the right a little bit, the large poly bulges and the spill should disappear. Move it a
little more, if necessary. Moving this slider to the right removes spill (moves the colors from outside
USER GUIDE
510
the large poly to between the medium and large polyhedrons) and moving it to the left, dents the
large poly and moves that color region to outside the large poly.
If you sample a foreground object shadow and then move the M-poly (trans) slider to the right, the
shadow becomes more transparent. This is useful for matching composited shadows to shadows on
the plate photography. It can also be used to make clouds or smoke more transparent.
If some foreground detail disappears during the composite, you can select where the detail should be
and then move the S-poly (detail) slider to the left. This dents the small poly in that color region and
releases the detail pixels from the small poly into the visible region between the small and medium
polyhedrons.
The Spill Sponge and Matte Sponge tools are shortcut tools that automatically move the sliders a
pre-selected amount as a timesaving step for you. Other shortcut tools include the Make FG Trans.
tool and the Restore Detail tool.
These shortcut tools are one-step operations where you click on a color region of interest and
Primatte performs a pre-calculated operation. Hopefully, most operations using Primatte would only
require these tools, but the manual operation of the sliders is always an option.
The Spill(-) tool bulges the large poly a small amount incrementally in the color region that is clicked
on and the Spill(+) tool dents it a small amount with each click. The Matte(-) and Matte(+) tools do
the same to the medium poly and the Detail(-) and Detail(+) do it to the small poly.
USER GUIDE
511
Explanation of How Primatte Works | Explanation of How Primatte RT+ works
Explanation of How Primatte RT+ works

The Primatte RT+ algorithm differs from the Primatte algorithm in that it has a six surface color
separator instead of the 128-faceted polyhedrons. This makes the Primatte RT+ algorithm much
simpler and, therefore, faster to calculate. The results and performance of Primatte RT+ falls in
between the Primatte and Primatte RT options. Where the Primatte RT+ algorithm might not work
well is with less saturated backing screen colors and it also does not support the complement color
spill suppression method (which is the spill suppression method that delivers the best detail). For a
well-lit and photographed image or clip, this algorithm produces good results and render quickly.
Here is what a visual representation of the Primatte RT algorithm looks like after an image has been
processed:
Explanation of How Primatte RT works

Primatte RT is the simplest algorithm and, therefore, the fastest. It uses only a single planar surface
to separate the 3D RGB colorspace and, as a result, does not have the ability to separate out the
foreground from the backing screen as carefully as the above Primatte algorithm. Like the Primatte
RT+ algorithm, Primatte RT might not work well with less saturated backing screen colors and it too
does not support the complement color spill suppression method (which is the spill suppression
method that delivers the best detail). For a well-lit and photographed image or clip, this algorithm
produces good results and render very quickly.
Here is what a visual representation of the Primatte RT algorithm looks like after an image has been
processed:
USER GUIDE
512
Explanation of How Primatte Works | Explanation of How Primatte RT works
USER GUIDE
513
Contact Details | Main Office
Contact Details
Here are the contact details for the Photron main office and Primatte office.
Main Office
Photron USA, Inc., 9520 Padgett Street, Suite 110, San Diego, CA 92126, USA
Primatte Office
Photron USA, Inc., 3113 Woodleigh Lane, Cameron Park, CA 95682. Phone: 1-530-677-9980, FAX: 1-
530-677-9981, Cell: 1-530-613-3212, E-mail: sgross@photron.com, Website: http://primatte.com
Proprietary Notices
Primatte is distributed and licensed by Photron USA, Inc., San Diego, CA, USA. Primatte was
developed by IMAGICA Corp., Tokyo, Japan. Primatte is a trademark of IMAGICA Co IMAGICA Digix
Inc., Tokyo, Japan.
USER GUIDE
514
Keying with Ultimatte
This section explains how to use the blue/green screen keyer, Ultimatte, in Nuke.
Ultimatte Quick Start

1. Sample the screen (backing) color. For more information, see Sampling the Screen Color.
2. Refine the overlay using the overlay tools and, if needed, adjust the controls on the Ultimatte
tab. For more information, see Using Overlay Tools and Screen Correct.
3. Next refine the matte density using the matte tools and, if needed, adjust the controls on the
Density tab. For more information, see Adjusting the Density of the Matte .
4. If necessary, activate Shadow processing, and use Shadow tool and, if needed, adjust the controls
on the Shadow tab. For more information, see Retaining Shadows and Removing Noise.
5. If necessary, improve Spill Suppression using spill tools and, if needed, adjust the controls on the
Spill tab. For more information, see Adjusting Spill Controls.
6. If necessary, adjust the Cleanup controls, but in general you’ll get better results by using
ScreenCorrect and MatteDensity controls. For more information, see Retaining Shadows and
Removing Noise.
7. If necessary, you can adjust the controls on the Color tab to match your blacks, whites and
gammas between the foreground and the background. For more information, see Adjusting Color
Controls.
8. If necessary, activate film processing and adjust its settings on the Film tab. For more
information, see Adjusting Film Controls.
USER GUIDE
515
Connecting the Ultimatte Node | To use the Ultimatte inputs
Connecting the Ultimatte Node

Start up Nuke, create an Ultimatte node by clicking Keyer > Ultimatte and connect a foreground and
a background image (and any additional inputs you want) to it. Add a Nuke Viewer node so you can
see the result.
To use the Ultimatte inputs

In Ultimatte, you have a number of inputs you can use to connect images and mattes that you need
to key your footage. Connect the following inputs to your images and/or mattes as necessary:
• foreground (fg) - Connect your foreground image to this input.
• background (bg) - Connect your background image to this input.
• clean plate (cp) - Connect your clean plate to this input. For more information on clean plates, see
Using Overlay Tools and Screen Correct.
• garbage matte (gm) - Connect your garbage matte to this input. A garbage matte is often used to
clean up screens that are not a constant color or have lighting rigs in shot by forcing the alpha
transparent.
• holdout matte (hm) - Connect your holdout matte to this input. A holdout matte is used when
foreground objects are the same color or very close to the backing color. The holdout matte is used
to let Ultimatte know that the pixels in the holdout matte region should be considered 100%
opaque. Note that, unlike the other inputs, this input is hidden and appears as a small triangle on
the left hand side of the node.
USER GUIDE
516
Sampling the Screen Color | Adjusting the Controls on Ultimatte Tab
Sampling the Screen Color

The first step when keying with Ultimatte is to sample the screen color, or tell Ultimatte what color
your blue or green screen is. Do the following:
1. When you’ve connected the Ultimatte node, the foreground image is displayed in the Viewer.
2. Click Screen in the Ultimatte toolbar above the Viewer, and select the screen color by holding
down Alt+Ctrl/Cmd and clicking in the Viewer.
You should select an area on the green screen near important subject detail that is not obscured
in any way. The image is rendered and a composite is displayed. If further adjustments are
needed, use the controls described below.
Adjusting the Controls on Ultimatte Tab

After you’ve sampled your screen color, you can start adjusting your result. With the controls on the
Ultimatte tab you can select different sets of controls you want to enable on other properties panel
tabs. There’s also an enable checkbox on each tab that you can use to activate the controls on that
tab and the corresponding tools in the Viewer.
• Film - check this to reduce the effects of cyan undercutting on the Film tab.
• screen correct - check this to use the controls on Screen Correct tab.
• shadow - check this to use the controls on the Shadow tab.
• spill suppression - check this to use the controls on the Spill tab.
• cleanup - check this to use the controls on the Cleanup tab.
• color conformance - check this to use the controls on the Color tab.
USER GUIDE
517
Using Overlay Tools and Screen Correct | Adjusting the Controls on Ultimatte Tab
Using Overlay Tools and Screen

Correct
Screen Correct compensates for anomalies in the backing area such as uneven lighting, smudges,
seams, variation in the backing color, green set pieces, and unwanted shadows cast by set pieces,
boom arms, etc. This technique assumes that an exact copy of the problematic green screen element
with the subject matter omitted, usually called a Clean Plate, is supplied.
Although this technique gives great results by differentiating between foreground elements and
backing flaws, you often haven’t got the necessary reference materials. In that case, you can create a
synthetic Clean Plate with Ultimatte.
To achieve the best results, use a reference Clean Plate as an input to Ultimatte in addition to
allowing Ultimatte to generate a synthetic plate. In this way, the reference plate allows for the best
shadow separation, while the synthetic plate is used to reduce artifacts such as film grain or video
noise (which is rarely accurately reproduced when the clean plate is shot during the time of principal
photography). Switch the view to screen by clicking the overlay dropdown menu and selecting
screen.
With screen correct selected, use the add overlay tool to scrub on areas that represent the green
screen, including shadows. Usually, it’s best to press Alt+Ctrl/Cmd while scrubbing, so that you are
picking the color in the input image at that pixel, instead of the color of the processed image.
The overlay is used in the final process to fully eliminate these areas from the final composite.
Continue in this manner until the foreground subject and its shadows are surrounded with an
overlay. Make sure the overlay does not cover any foreground detail that is to be retained. If the
overlay does cover foreground subject matter, then use the remove overlay tool to scrub on those
areas that should not be covered by the overlay. Repeat these two processes until the overlay covers
the green screen areas and no subject matter. View the composite in the Viewer to see the screen-
corrected composite with anomalies in the backing area removed. To learn which controls were
automatically set by Ultimatte, click the Screen Correct tab and note the position of the controls.
Note: When scrubbing on the image using add overlay, the RGB value of the points
selected are accumulated in a keep list and removeoverlay points are accumulated in a
drop list. If both lists have equivalent values, a conflict may occur, resulting in no visible
change in the overlay. If a conflict occurs, try using Undo (Ctrl/Cmd+X) which flushes the last
USER GUIDE
518
Using Overlay Tools and Screen Correct | Adjusting overlay controls
set of points added to the appropriate list. If that doesn’t help, you can also press the Reset
button on the ScreenCorrect tab to clear the lists and start over.
Under some circumstances, it may be difficult to completely cover the screen area with the overlay.
This does not mean that the screen cannot be removed in that area, but that other controls may need
to be adjusted to help reduce anomalies in those areas that were not covered by the overlay. It may
be advantageous to resample the screen using the Screen tool above the Viewer in the area where
covering the screen with the overlay is unsuccessful, but be careful about selecting in shadow areas if
the final composite requires the retention of shadows.
Additionally, it may be difficult in some circumstances to remove the overlay from very dark areas of
foreground detail that may be close to values in dark or shadow areas of the screen. If the overlay
does include some dark foreground areas, these areas may be recovered by enabling the Shadows
controls.
Note: Since Ultimatte produces a synthetic clean plate using an interpolation process, it is
important to exclude non-video areas of the image, such as the black blanking lines that
may appear in video or the black areas in letterbox images. You can use other nodes, such
as Crop, to exclude these areas from the overlay and final composites. Otherwise the
synthetic clean plate may contain black (or other colors) which can result in an inaccurate
clean plate being generated.
Adjusting overlay controls

You can choose your overlay view, color and output mode in the properties panel, on the Ultimatte
tab.
• overlay - The overlay control is used to show the calculated overlay on the Viewer and it’s helpful for
debugging purposes. It helps tune the screen correct controls and also the add/remove overlay
tools. In this way you can see immediately if something is wrong and needs to be adjusted. In the
overlay dropdown menu, change
• off - to not view the overlay.
• screen - to view the subject in clear, and the preliminary matte area blended with the overlay
color.
• subject - show the subject blended with overlay color, and the preliminary matte area in clear.
• show image as monochrome - check this to make the input image appear in grayscale so that the
overlay areas can be more easily distinguished.
USER GUIDE
519
Using Overlay Tools and Screen Correct | Adjusting the screen correct controls
• overlay color - use this to change the color of the overlay. You can adjust the alpha channel to
modify the opacity of the overlay.
Adjusting the screen correct controls

• screen tolerance - This is used to adjust the color range or tolerance to be included or excluded
from the screen overlay.
• shrink - Use this control to shrink the screen overlay.
• darks(reds smaller) - Use this control to include or exclude dark areas from the screen overlay in
those areas where the blue value (when using green screen) is greater than the red value in the
original foreground image.
• darks(reds larger) - Use this control to include or exclude dark areas from the screen overlay in
those areas where the red value is greater than the blue value (when using green screen) in the
• brights(reds smaller) - Use this control to include or exclude bright areas from the screen overlay
in those areas where the blue value (when using green screen) is greater than the red value in the
• brights(reds larger) - Use this control to include or exclude bright areas from the screen overlay in
those areas where the red value is greater than the blue value (when using green screen) in the
• orphans - Use this to include or exclude neighboring "rogue" pixels from the screen overlay.
USER GUIDE
520
Adjusting the Density of the Matte | Adjusting Density controls
Adjusting the Density of the

Matte
The controls on the Density tab are used to adjust the density or opacity of the foreground objects.
The density of a foreground object is determined by its matte (alpha) value. A completely opaque
object's matte is white, a completely transparent object's matte is black, and a partially transparent
object's matte is gray. Use the add matte dropper to scrub on areas in the matte that appear gray,
but should be white (fully opaque) in the matte. These areas are described as "print-through",
meaning that the opacity of the subject is too low in this area and the background is visible through
the foreground in this area. Be careful not to select those objects that should have a gray matte value
such as fine hair, smoke or partially transparent objects, as they become opaque. To learn which
controls were automatically set by Ultimatte, click the Density tab in the properties panel and note
the position of the controls.
Note: If there are sections of the matte which should be opaque but are exhibiting gray
values that don’t look like typical transparency, then there is a chance that there is a
remainder of overlay in this area. If overlay exists on subject matter, return the Overlay
dropdown menu to Screen and use the removeoverlay tool to scrub on that area. Check
on show image as monochrome on the Ultimatte tab to aid in determining the extent of
the overlay.
Adjusting Density controls

Use these controls to adjust the density of your matte:
• brights - Use this control to adjust density in bright foreground objects. Advancing this control too
far can cause hard, dark edges around foreground subjects.
• darks - Use this control to adjust density in black glossy or dark foreground objects.
• edge kernel - Use this control to adjust number of pixels to use as a kernel to reduce dark edges
that may exist in transition areas due to an over-dense matte. Advancing this control too far, may
cause too much print-through at the edges.
• warm - Use this control to adjust density in warm colors (flesh tones). Note that reducing this
control too much can cause print-through in reddish foreground objects.
• cool - Use this control to adjust density in cool colors. Note that reducing this control too much can
cause print-through in bluish foreground objects.
USER GUIDE
521
Adjusting Spill Controls | Adjusting Density controls
Adjusting Spill Controls

Ultimatte automatically suppresses spill from the backing onto foreground subject matter. The spill
controls are used to suppress excessive spill or to retain color similar to spill that has been over-
suppressed. To adjust the spill controls, check the spill suppression box on the Ultimatte tab and
click Spill tab:
• cool - Use this control to adjust the amount of spill in cool colored foreground objects. Used to
reproduce blue, green or cyan colors that changed through the spill suppression algorithms.
• warm - Use this control to adjust the amount of spill in warm colored foreground objects. Used to
reproduce pink, purple and magenta colors for bluescreen, or yellow and orange colors for green
screen that changed through the spill suppression algorithms.
• midtones - Use this control to adjust the amount of spill in mid-range foreground objects.
• brights - Use this control to adjust the amount of spill on bright foreground objects.
• darks - Use this control to adjust the amount of spill on dark foreground objects.
• ambience - Use this control to select a color to subtly influence the foreground objects in areas that
may have contained spill.
• strength - Use this control to adjust the intensity of the selected ambience color.
• background veiling - This control is used to override the automatic suppression of the backing
color. Ultimatte uses the selected backing color to suppress the backing to black. An indication that
the automatic settings did not suppress enough backing is "veiling" or a colorized haze in some
background areas. An indication that the automatic settings suppressed too much backing is
darkened or mis-colored foreground edges and transparencies. In most cases this control should be
left at the default setting.
g>(r+b)/2?(r+b)/2:g
channel to:
b>(r+g)/2?(r+g)/2:b
USER GUIDE
522
Adjusting Spill Controls | Adjusting Density controls
Only.
USER GUIDE
523
Retaining Shadows and Removing Noise | Adjusting Shadows controls
Retaining Shadows and Removing

Noise
Use the hold shadow dropper (only available when screen correct and shadow are enabled) to
scrub on the shadows that you’d like to preserve. These shadows may best be seen in the foreground
image.
If unwanted shadows remain, then use the remove noise dropper to reduce or eliminate those
shadows. If the area that you scrub does not reside under the overlay, then erasing adjusts the
appropriate Cleanup controls to reduce anomalies in the screen area. This might result in losing
some fine detail. Repeat using these two tools until the shadows you want are retained and any
shadows or noise you don’t want to keep are removed.
The density, sharpness, and tint of the shadow may be altered by manually adjusting the controls in
Shadow controls. To learn which controls were automatically set by Ultimatte, click the Shadow
and/or Cleanup tab in the properties panel and note the position of the controls.
Adjusting Shadows controls

Check the shadow box on the Ultimatte tab and adjust the following controls:
• high - Decrease this to reduce or eliminate unwanted shadows (which must be lighter than those
shadows that are to be retained). All preserved shadows are lighter.
• low - Increase this to restore the density of the darkest part of the shadows that are to be retained.
• density - Use this to change the density of the shadows that are retained to better match shadows
in the background scene.
• blur - Use this to blur the shadows.
• tint - Use this to tint the shadows.
Adjusting Cleanup controls

The following controls are used to adjust the black and gray areas of the matte channel. This
dramatically affects the nature of foreground objects' edges, the opacity of transparent objects, and
the noise in the foreground image. Use these controls sparingly as they might result in the loss of
foreground detail. Using Screen Correction for dealing with imperfections in the screen is a good
alternative for using Cleanup.
USER GUIDE
524
Retaining Shadows and Removing Noise | Adjusting Cleanup controls
• cleanup - Use this control to reduce imperfections or small amounts of noise in the backing. Note
that adjusting this control too far results in a "cut and paste" look with a hard, unnatural edge.
Background noise (as well as some foreground detail) is reduced. An alternative method for dealing
with green screen imperfections is to use the Screen correct controls.
• shrink - Use this control to choke or reduce the size of the cleaned-up matte.
• blur - Use this control to soften the cleaned-up matte.
• recover - Use this control to set a threshold below which the Cleanup control does not have
influence.
USER GUIDE
525
Adjusting Color Controls | Adjusting Cleanup controls
Adjusting Color Controls

Check the color conformance box on the Ultimatte tab, and set these controls on the Color tab. A
good keying result requires matched blacks, whites, and gammas between the foreground and
background elements. With color conformance you can select blacks, whites, and gammas and can
automatically match the foreground to the background (or vice versa).
• darks - You can use this control to influence the darkest parts of the image. This is a global control
that affects the entire image, but the greatest effect is seen in the darkest areas.
• midtones - You can use this control to influence the mid-range parts of the image. This is a global
control that affects the entire image, but the greatest effect is seen in the mid-range areas.
• brights - You can use this control to influence the brightest parts of the image. This is a global
control that affects the entire image, but you can see the greatest effect in the brightest areas.
• hue - This control changes the color contents of the image without changing its brightness or color
intensity (purity) values. At default setting (0), the image hue is not altered. The range of the control
extends from -300 to +300.
• saturation - This control changes the color intensity or purity values of the image without altering
its color contents or brightness values. At default setting (0), the image saturation is not altered. At
minimum setting (-200), the color intensity is reduced to zero and the image is monochrome, or
shades of gray.
• brightness - This control changes the overall intensity of the image. At the default setting, the image
brightness is not altered.
USER GUIDE
526
Adjusting Film Controls | Adjusting Cleanup controls
Adjusting Film Controls

Use the controls on the Film tab to reduce the effects of cyan undercutting. These controls are only
available when Film box is checked on Ultimatte tab. You can view the results better when you select
subject in the overlay dropdown menu on the Ultimatte tab.
Due to the nature of film’s emulsion layers, a phenomenon know as cyan undercutting exists that
reveals itself as a smeared red edge in areas of sharp transitions, which can best be seen by viewing
the individual RGB channels of a film image. Normally, this phenomenon is not a problem until
bluescreen compositing techniques are applied. Since the red emulsion layer tracks at a slower rate
than the blue and green emulsion layers, the artificial red edge it produces are retained as
foreground detail resulting in an unacceptable composite.
• transparency - Use this control to adjust the amount of Film Correction in partially transparent
foreground objects (such as hair detail).
• correction - Use this control to adjust the amount of correction in the individual RGB channels of
the foreground image.
• strength - Use this control to adjust the overall amount of Film Correction that is applied to the
foreground image.
• shrink - Use this control to shrink the subject overlay.
• brights - Use this control to include or eliminate bright areas from the subject overlay.
• darks - Use this control to include or eliminate dark areas from the subject overlay.
USER GUIDE
527
Choosing an Output Mode | Adjusting Cleanup controls
Choosing an Output Mode

You can output the merged foreground and background as a final composite, or you can output the
premultiplied or unpremultiplied foreground for compositing elsewhere in your node tree. In the
output mode dropdown menu, select:
• composite - to output a merged layer of the background input with the extracted subject.
• premultiplied - to output the extracted subject premultiplied by the final matte.
• unpremultiplied - to output the extracted subject and final matte unpremultiplied.
USER GUIDE
528
Using RotoPaint
Nuke features a vector-based RotoPaint node for help with tasks like rotoscoping, rig removal,
garbage matting, and dustbusting. You can draw Bezier and B-Spline shapes with individual and layer
group attributes, including per-point and global feather, motion blur, blending modes and individual
or hierarchical 2D transformations. This chapter gives full instructions on its usage.
Roto or RotoPaint?
There are two similar nodes in Nuke for rotoscoping, Roto and RotoPaint. The main difference
between these two is that you can only create and edit Bezier and B-spline shapes with Roto, while
RotoPaint allows you to draw paint strokes too with various brushes. So the Roto node is an optimal
choice if you’re doing rotoscoping only, whereas RotoPaint gives you a broader scale of tools to use.
All tools and controls in the Roto node work the same way as they do in RotoPaint, so you can learn
about using them in the RotoPaint instructions in this chapter. For instance, see:
• Working with the Toolbars for information about the toolbars in Roto.
• Working with the Stroke/Shape List for information about the shape list in the Roto properties
panel.
• Drawing Shapes for information about using the Bezier and B-Spline Tools.
• Selecting Existing Strokes/Shapes for Editing for information about selecting shapes.
• Editing Shape-Specific Attributes for information about editing Bezier and B-Spline attributes.
• Editing Existing Stroke/Shape Timing for information about changing the timing of your shape.
• Animating Strokes/Shapes for information about editing your shapes.
RotoPaint Quick Start

1. Connect the RotoPaint node to a Viewer and possible backgrounds. For more information, see
Connecting the RotoPaint Node.
2. Select a stroke or a shape tool from the RotoPaint toolbar on the left side of the Viewer. For more
information, see Drawing Paint Strokes or Drawing Shapes.
3. Use the RotoPaint tool settings on the top of the Viewer to adjust the stroke/shape you’re about to
draw. For more information, see Working with the Toolbars.
USER GUIDE
529
Using RotoPaint |
4. Draw one or more strokes/shapes in the Viewer window. For more information, see for example
Using the Brush tool or Using the Bezier and Cusped Bezier Tools.
5. Select an existing stroke/shape using the Select tools or the stroke/shape list. For more
information, see Selecting Existing Strokes/Shapes for Editing.
6. Use the control panel controls to adjust your existing stroke/shape. For more information, see
Editing Existing Stroke/Shape Attributes.
In addition you can:

• Adjust the splines of your stroke/shape. For more information, see Editing Existing Stroke/Shape
Splines.
• Animate your strokes/shapes. For more information, see Animating Strokes/Shapes.
• Use RotoPaint in stereoscopic projects. For more information, see RotoPaint and Stereoscopic
Projects.
• Set your favorite RotoPaint tool as your default tool. For more information, see Setting Default
RotoPaint Tools and Settings.
USER GUIDE
530
Connecting the RotoPaint Node | To Connect the RotoPaint Node
Connecting the RotoPaint Node

The RotoPaint node accepts one primary input. Even if you have no inputs, you can still use RotoPaint
and in that case, you can use the format control to select your output format (by default, the format
control is hidden, but you can display it by clicking on the black triangle above color).
To Connect the RotoPaint Node

1. Click Draw > RotoPaint to add a new RotoPaint node. You can also press P on the Node Graph.
To create a Roto node, you can press O on the Node Graph.
2. Drag the bg input to the node that you want to apply RotoPaint to.
3. Connect any additional background elements you wish to use. If you plan to reveal pixels from a
background element, drag the bg1 input from the left side of the node to the node whose output
you wish to use.
USER GUIDE
531
Working with the Toolbars | To Connect the RotoPaint Node
Working with the Toolbars

In the RotoPaint node, you can use two toolbars to define the type of stroke/shape you want to start
drawing. These toolbars are placed in the Viewer. The vertical RotoPaint toolbar is for selecting the
tool you want to use and the horizontal one, RotoPaint tool settings, is for adjusting the currently
selected tool’s settings before drawing new strokes/shapes.
Note: You can’t use RotoPaint tool settings to adjust an already existing stroke/shape. For
any changes you want to make to a stroke/shape after you’ve created one, you can use the
controls in the RotoPaint properties panel.
RotoPaint toolbar and tool settings.
In the RotoPaint toolbar, you can select your stroke/shape tool. The tools are grouped under the
toolbar icons. You can click any tool to make it active and view a tool group by right-clicking (or left-
clicking and holding) the icon. The tool that is currently selected is highlighted.
Tip: When the Viewer has mouse-over focus, you can also use the S keyboard shortcut to
cycle through the modes of the currently selected tool.
In the RotoPaint tool settings on the top of the Viewer, you can define the settings for the tool that
you’ve selected. The controls in this toolbar change depending on which tool you have selected at any
given time.
USER GUIDE
532
Working with the Toolbars | To Connect the RotoPaint Node
Tip: You can hide the toolbars by clicking the hide button next to them. You can also press [
(square bracket) to hide the RotoPaint toolbar and { (curly bracket) to hide the tool settings
(and the entire top toolbar of the Viewer).
USER GUIDE
533
Working with the Stroke/Shape List | To Use the Stroke/Shape List
Working with the Stroke/Shape

List
After you’ve drawn strokes/shapes, you can edit their order and group them with the Stroke/Shape list
in the RotoPaint properties panel. By default, the newest stroke/shape/group appears on top of the
list, and your strokes/shapes are named according to their type (for example, "Bezier1" or "Smear2").
Using the list, you can also select strokes/shapes/groups and edit them with the various controls in
the properties panel. Some controls can be applied to groups, some can’t. Some controls also can
only be applied to strokes or shapes (these are grouped under the Stroke and Shape control tabs
respectively). If a control can’t be applied to a group, it is grayed out if you have a group selected in
the stroke/shape list.
The stroke/shape list provides an overview of existing parameters, showing, for example, whether the
item is locked, set invisible, or whether it has motion blur applied to it. Some controls can also be
edited directly in the overview by clicking on their icon.
To Use the Stroke/Shape List

You can edit the stroke/shape list in many ways, and use it to adjust strokes/shapes and how they’re
• You can reorder the columns in the stroke/shape list by dragging and dropping them.
• You can create groups for sets of strokes/shapes in the stroke/shape list by clicking the Add button
below the list. This creates a subfolder, named “Layer1” by default, and you can drag and drop
strokes/shapes to this folder to group them. After strokes/shapes have been grouped, you can edit
them as a group and they move together if you change their place in the Viewer. Every group also
has its own transformation overlay, which you can use to move the group.
• You can remove a stroke/shape or a group by clicking the Remove button under the
stroke/shape list.
• If you want to rename any of the strokes, shapes, or groups, double-click on the name while the
item is selected, and give your item a new name. The name has to be unique, so you can’t give two
items the same name.
• You can also cut, copy, and paste strokes and shapes by right-clicking on them in the control panel
and using the copy > curve, cut > curve, and paste > spline options in the menu that appears.
USER GUIDE
534
For information on copying, cutting and pasting shapes and points, see Copying, Pasting, and
Cutting Stroke Positions; or for animated shapes and points, see Animating Strokes/Shapes.
Alternatively, to copy, cut, or paste shape and point attributes only, see Editing Existing
Stroke/Shape Splines.
Note: You cannot use Ctrl/Cmd+C or Ctrl/Cmd+V in the control panel to copy and paste
shapes. You must use the right-click menu.
• You can duplicate strokes/shapes by right-clicking on them and selecting duplicate. A new
stroke/shape is created with the same spline, animation, and attributes as the one you selected.
• You can hide a stroke, shape, or group by clicking the Visible icon in the stroke/shape list. You
can still edit an invisible stroke/shape and view its position in the Viewer.
• You can lock strokes/shapes to prevent them from being edited. To lock an item in the stroke/shape
list, click the Lock column in the list. A lock icon appears next to the Visible icon.
• You can select the color in which you want the outline of your stroke/shape to appear in the Viewer.
Click the Overlay column and select your overlay color. To be able to see the splines for all
paint strokes in the Viewer, you need to activate one of the Select tools in the RotoPaint toolbar and
click in the tool settings.
• You can change the color of your stroke/shape in the stroke/shape list by clicking the Color column
and using the color picker to select the color.
• You can invert a shape using the Invert column . With your shape selected, click in the Invert
column to toggle between inverted and uninverted modes.
• You can select a blending mode for your stroke/shape using the Blending column . With your
shape selected, click the Blending column and select the mode you want to apply.
• You can apply shape motion blur using the Motion blur column . With your shape selected,
click the Motion blur column to toggle the motion blur effect.
See Adding Motion Blur for more information on shape and global motion blur.
• You can specify the frame range of certain strokes and shapes by navigating to the stroke/shape list
of the RotoPaint node, and right-clicking all under the Life column. Select frame range from the
pop-up menu, and specify the frame range desired in the Set frame range dialog.
By default, new shapes and strokes are set to all on creation.
USER GUIDE
535
Tip: To undo and redo any changes you’ve made with the RotoPaint node, use the Undo and
Redo buttons on top of the properties panel. Undo uses a cache memory of your changes,
so at any time you can undo and redo all the changes you’ve made since you last opened
your project.
USER GUIDE
536
Drawing Paint Strokes | To Use the Stroke/Shape List
Drawing Paint Strokes

Any given RotoPaint node can hold many paint strokes and shapes. You can apply paint strokes using
any of the following tools (see also Drawing Shapes).
Icon Tool Keyboard Shortcut Function
Brush N (toggles between Applies colors atop the current plate, or blends
Brush and Eraser) colors with the current plate. You can also clone
from surrounding frames.
Eraser N Removes pixels from existing strokes and brings

background back.
Clone C (toggles between Applies pixels from one region of the current
clone and Reveal) plate to another region of the current plate.
Reveal C Applies pixels from a source plate to a

destination plate in the corresponding place.
Blur X (toggles between Blurs the image in the area of the brush stroke.
Blur, Sharpen and
Smear)
Sharpen X Sharpens the image in the area of the brush

stroke.
Smear X Smears the area of the smear brush stroke,

stretching the selected pixels over their
surrounding area.
Dodge D (toggles between Brightens the background color on the area of

Dodge and Burn) the brush stroke to reflect the brush stroke.
Using this tool on black produces no change. No
part of the stroke area is darkened.
Burn D Darkens the background color on the area of

the brush stroke to reflect the brush stroke. No
part of the stroke area is lightened.
USER GUIDE
537
Drawing Paint Strokes | To Use the Stroke/Shape List
Tip: You can choose to have your RotoPaint always open with a particular tool selected. If
you want to open it with the Brush tool selected, for example, do the following:
1. Create a file called menu.py in your plug-in path directory if one doesn’t already exist. For
2. To select the Brush tool as your default tool, save the following in your menu.py:
nuke.knobDefault('RotoPaint.toolbox','brush')
3. Restart Nuke.
For more information on default tools, see Setting Default RotoPaint Tools and Settings.
Tip: If you are using a tablet, you can tie a new stroke’s opacity, size, or hardness to pen
pressure by clicking the corresponding pressure alters checkbox in the RotoPaint tool
settings. You can also later change which attribute the pressure alters for an existing stroke
on the Stroke tab of the properties panel.
Separate Select tools in the RotoPaint toolbar let you select strokes/shapes once they’ve been drawn
and after that you can make changes to them using the properties panel controls (see Selecting
Existing Strokes/Shapes for Editing, and Editing Existing Stroke/Shape Attributes).
These pages discuss the general steps for using each of these tools, and give an overview on editing
their attributes and timing.
USER GUIDE
538
Using the Brush tool | To Use the Brush Tool
Using the Brush tool

The Brush tool lets you apply colored or blended strokes to the current plate.
Painting with the Brush tool.
To Use the Brush Tool

1. Click the Brush tool in the RotoPaint toolbar.
2. Set color, opacity, brush type, brush size, and brush hardness in the RotoPaint tool settings at the
top of the Viewer. (For information on the available options, see Editing Existing Stroke/Shape
Attributes.) You can also change the size of the brush using Shift+drag directly in the Viewer as
shortcut.
3. Optionally, set the lifetime of the stroke in the RotoPaint tool settings. (For information on the
available options, see Editing Existing Stroke/Shape Timing.)
4. Apply strokes as necessary.
Tip: If you need to draw a completely straight line with the brush tool, try this: Draw a
freehand stroke, activate the Select All tool, enable show paint stroke splines mode in
the tool settings, and marquee select and delete all but the first and last point on the stroke.
USER GUIDE
539
Using the Eraser Tool | To Use the Eraser Tool
Using the Eraser Tool

The Eraser tool lets you remove pixels from existing paint strokes.
To Use the Eraser Tool

1. Set opacity, brush type, brush size, and brush hardness in the RotoPaint tool settings at the top of
the Viewer. (For information on the available options, see Selecting Existing Strokes/Shapes for
Editing.)
available options, see Selecting Existing Strokes/Shapes for Editing.)
3. Right-click the Brush tool in the RotoPaint toolbar and select the Eraser tool .
4. Apply strokes as necessary. You can also erase multiple strokes, if you have drawn more than one.
Unwanted paint strokes. Using the Eraser tool to

remove the unwanted paint
strokes.
Tip: If you’re using a graphics tablet, Nuke automatically switches to Eraser mode when you
use the erase end of the pen.
USER GUIDE
540
Using the Clone Tool | To Use the Clone Tool
Using the Clone Tool

The Clone tool lets you remove unwanted features from the plate or from a different input by
painting over them with pixels offset from the pointer or a transformation of the pointer.
Painting with the Clone tool.
To Use the Clone Tool

1. Click the Clone tool in the RotoPaint toolbar.
2. To view all the settings for the Clone tool, enable show clone settings in the RotoPaint tool
settings at the top of the Viewer.
3. In the RotoPaint tool settings, set the paint source dropdown menu to the input you want to clone
pixels from. (For information on the available options, see Editing Stroke-Specific Attributes.)
You can also use the transform controls in the clone settings to transform the clone source
and reset it back to original with the reset button.
4. Set opacity, lifetime, brush type, size, and hardness for your stroke in the tool settings. (For more
information, see Editing Existing Stroke/Shape Attributes and Editing Existing Stroke/Shape
Timing.) If you’ve got more than one view set up, you can check the single view box to only clone
in one view, or uncheck it to clone in all of them.
5. To set the clone offset, hold down Ctrl/Cmd and left-click at the source location, drag to where
you want to paint, and release. Alternatively, you can enter the offset numerically using the
translate controls in the RotoPaint tool settings. If you’d like the offset amount to be rounded to
an integer (whole number of pixels), check round. Rounding to a pixel can be useful if you don’t
want to soften the image by partially blending pixels together.
6. Start painting. The pointer overlay depicts the source of the offset as a crosshair within a circle
and the destination as a circle (the diameter of which represents the breadth of the stroke).
USER GUIDE
541
Using the Clone Tool | To Use the Clone Tool
You can use / (forward slash) and * (asterisk) on the numeric keypad to zoom your clone source in
and out, and 0 (zero) and . (decimal point) to rotate it right and left. You can also use the number
keys on the numeric keypad to nudge the clone source.
7. To reset the clone offset, you can use Ctrl/Cmd+drag to adjust the offset you set before, or
Ctrl/Cmd+Shift+drag to start a new offset from the brush pointer’s location.
Tip: If you’re cloning from the current plate (foreground), you’re also cloning all the
strokes/shapes you’ve previously drawn. If you want to clone from the original background
or a different picture, you need to set the paint source dropdown menu to pull from that
input.
Tip: To clone pixels from another frame of the input clip, you can use the time offset slider
to define which frame you want to clone from. See Editing Clone or Reveal Attributes.
USER GUIDE
542
Using the Reveal Tool | To Use the Reveal Tool
Using the Reveal Tool

The Reveal tool lets you pull pixels from background elements onto the current plate. The Reveal
tool requires at least two inputs (see Connecting the RotoPaint Node); otherwise, your strokes draw in
white. You can also view which input you are using as the source for your strokes in the Source
column in the stroke/shape list.
The current plate. The source input.
Painting with the Reveal tool.
To Use the Reveal Tool

1. Right-click the Clone tool in the RotoPaint toolbar and select the Reveal tool .
2. In the RotoPaint tool settings at the top of the Viewer, set the paint source dropdown menu to
the input you want to pull pixels from. (For information on the available options, see Editing
Stroke-Specific Attributes.)
3. Set opacity, brush size, and brush hardness in the RotoPaint tool settings. (For information on the
available options, see Editing Existing Stroke/Shape Attributes.)
5. You can also reveal pixels from another frame of the input clip by using the time offset slider to
define which frame you want to reveal from. See Editing Clone or Reveal Attributes.
6. If you want, you can view your revealing source image in a Viewer overlay. To do this, check the
onion box in the RotoPaint tool settings and enter an onion skin value to adjust the opacity of the
USER GUIDE
543
Using the Reveal Tool | To Use the Reveal Tool
overlay. This can help you better see what you are revealing from the source image. You can also
toggle the keyboard shortcut T to enable or disable onion skin.
7. Start painting.
Onion skin disabled. Onion skin enabled.
USER GUIDE
544
Using the Blur Tool | To Use the Blur Tool
Using the Blur Tool

The Blur tool lets you blur parts of the plate.
Painting with the Blur tool.
To Use the Blur Tool

1. Click the Blur tool in the RotoPaint toolbar.
2. Set opacity, brush type, brush size, and brush hardness in the RotoPaint tool settings. (For
information on the available options, see Editing Existing Stroke/Shape Attributes.)
4. Apply strokes by clicking on the part of image you want to blur.
Tip: You can use the effect control in the RotoPaint tool settings to adjust the strength of
the tool you’re using. With the Blur tool, it controls the amount of blur.
USER GUIDE
545
Using the Sharpen Tool | To Use the Sharpen Tool
Using the Sharpen Tool

With the Sharpen tool, you can sharpen the image within the area of the brush stroke.
Painting with the Sharpen tool.
To Use the Sharpen Tool

1. Right-click the Blur tool and select Sharpen tool.
the Viewer. (For information on the available options, see Editing Existing Stroke/Shape
Attributes.)
4. Apply strokes by clicking on the part of image you want to sharpen.
Tip: You can use the effect control in the RotoPaint tool settings to adjust the strength of
the tool you’re using. With the Sharpen tool, it controls how much the image is sharpened.
USER GUIDE
546
Using the Smear Tool | To Use the Smear Tool
Using the Smear Tool

With the Smear tool, you can smear or stretch pixels over the surrounding pixels.
Painting with the Smear tool.
To Use the Smear Tool

1. Right-click the Blur tool and select Smear tool .
Attributes.)
4. Apply strokes by clicking and dragging on the part of image you want to smear.
USER GUIDE
547
Using the Dodge Tool | To Use the Dodge Tool
Using the Dodge Tool

With the Dodge tool, you can lighten the pixels in the area of the brush stroke. This makes the
background color brighter on the area of the brush stroke to reflect the brush stroke.
Painting with the Dodge tool.
To Use the Dodge Tool

1. Click the Dodge tool .
Attributes.)
USER GUIDE
548
Using the Burn Tool | To Use the Burn Tool
Using the Burn Tool

With the Burn tool, you can darken the pixels in the area of the brush stroke. This makes the
background color darker on the area of the brush stroke.
Painting with the Burn tool.
To Use the Burn Tool

1. Right-click the Dodge tool and select Burn tool .
2. Set opacity, brush type, brush size, and brush hardness in the RotoPaint tool settings. (For more
information on the available options, see Editing Existing Stroke/Shape Attributes.)
USER GUIDE
549
Drawing Shapes | To Use the Burn Tool
Drawing Shapes
Any given RotoPaint node can hold several shapes, and you can draw them using any of the following
tools.
Icon Tool Keyboard Shortcut Function
Bezier V (toggles between Applies a Bezier shape. Bezier shapes are

Bezier, defined using control points and tangents.
B-spline, Ellipse, and
Rectangle)
Cusped V Applies a Bezier shape with sharp corners and

Bezier no tangents.
B-Spline V Applies a B-spline shape. Unlike Bezier

shapes, B-splines are created by only using
control points. The position of the points in
relation to each other determines what kind
of splines the shape consists of.
Ellipse V Applies an ellipse-shaped Bezier shape.
Rectangle V Applies a rectangle-shaped Bezier shape.
Cusped V Applies a rectangle-shaped Bezier shape with

Rectangle sharp corners and no tangents. You can see
the difference between rectangles and cusped
rectangles if you move at least one of the
control points.
Open Spline V Lets you draw curves in a similar way to other

shapes, except that they don't need form a
closed shape.
Separate Select tools in the RotoPaint toolbar let you make changes to a stroke/shape once it’s been
drawn (see Selecting Existing Strokes/Shapes for Editing).
USER GUIDE
550
Drawing Shapes | To Use the Burn Tool
Tip: You can choose to have your RotoPaint always open with a particular tool selected. If
you want to open it with the Bezier tool selected, for example, do the following:
2. To select Bezier as your default tool, save the following in your menu.py:
nuke.knobDefault('RotoPaint.toolbox','bezier')
3. Restart Nuke.
For more information on default tools, see Setting Default RotoPaint Tools and Settings.
Tip: If you begin drawing shapes and then decide you want to planar track these with the
PlanarTracker node, you can do that easily by right-clicking on the shape in the stroke/shape
list and selecting planar-track. This creates a new planar tracking layer for the shape and
attaches the shape to it. For more information about PlanarTracker, see Tracking with
PlanarTracker.
USER GUIDE
551
Using the Bezier and Cusped Bezier Tools | To Use the Bezier or Cusped Bezier Tools
Using the Bezier and Cusped

Bezier Tools
The Bezier and Cusped Bezier tools let you draw Bezier shapes. Cusped Beziers are Bezier shapes
with sharp corners and no tangents.
Drawing a Bezier shape.
To Use the Bezier or Cusped Bezier Tools

• Click the Bezier tool in the RotoPaint toolbar.
• Right-click the Bezier tool in the RotoPaint toolbar and select the Cusped Bezier tool .
2. Select color, blending mode, opacity, and other settings for the shape in the RotoPaint tool
settings at the top of the Viewer. (For information on the available options, see Editing Existing
Stroke/Shape Attributes.)
3. Optionally, set the lifetime of the shape in the RotoPaint tool settings. (For information on the
available options, see Editing Existing Stroke/Shape Timing on page 581.)
4. Draw a shape in the Viewer by clicking to create the outlines that make up your shape.
5. If you’re using the Bezier tool, you can click+drag while drawing to create a point and adjust its
tangent handles. With the tangent handles, you can adjust the spline of your shape.
• You can move individual handles to adjust their length, keeping the angle consistent.
• Press Shift while moving the tangent handles to move both handles at the same time, keeping
the angle consistent.
• Press Ctrl/Cmd to temporarily break the angle.
6. You can also use other shortcuts to adjust your shape while drawing:
USER GUIDE
552
Using the Bezier and Cusped Bezier Tools | To Use the Bezier or Cusped Bezier Tools
• Shift+click to create a sharp exit point on the previous point (this has no effect when using the
Cusped Bezier tool, as all the points are sharp anyway).
Curved exit point. Sharp exit point.

• Ctrl/Cmd+click to sketch your shape freely.
7. To close your shape, press Return or click the first point of your shape. Changing to a different
tool also closes a shape. By default, closing a shape activates the Select tool.
If you’re using the Bezier tool and close your shape by clicking its first point, you can also drag the
point to create tangent handles for adjusting it.
8. With the Select tool active, you can Shift+click on multiple shape points to bring up the transform
box, which you can use to further transform your shape or particular points in your shape.
9. You can also cusp and smooth a point in your Bezier shape to create a sharp corner on your
shape or smooth it again. To do this, right-click on a point in your Bezier shape and select
cusp/de-smooth or smooth.
Tip: You can also apply animation from a Tracker node to a point in your Bezier shape.
From the Tracker node, Ctrl/Cmd+drag the transformation information into the Viewer, on
the point you want to animate. See also Linking Expressions for creating links between the
Tracker node and RotoPaint controls.
USER GUIDE
553
Using the B-Spline Tool | To Use the B-Spline Tool
Using the B-Spline Tool

The B-spline tool lets you draw B-spline shapes.
Drawing a B-spline shape.
To Use the B-Spline Tool

1. Right-click the Bezier tool in the RotoPaint toolbar and select B-Spline tool .
Stroke/Shape Attributes)
4. Click in the Viewer to create the splines that make up your shape.
• Ctrl/Cmd+drag to sketch the spline freely.
• Click+drag the cursor left or right to adjust the tension (that is, the distance between the spline
line and a point) on the previous point.
• To adjust the tension of any point in the B-spline shape you’re drawing, select a point, press
Ctrl/Cmd+Shift, and drag left or right: right to increase the tension, left to decrease it.
5. To close your shape, press Return. Changing to a different tool also closes a shape. By default,
closing a shape activates the Select All tool.
6. With the Select All tool active, you can Shift+click on multiple shape points to bring up the
transform box, which you can use to further transform your shape or particular points in your
shape.
7. You can also cusp and smooth a point in your B-spline shape to create a sharp corner on your
shape or smooth it again. To do this, right-click on a point in your B-spline shape and select
cusp/de-smooth or smooth.
USER GUIDE
554
Using the B-Spline Tool | To Use the B-Spline Tool
Tip: You can convert your B-spline shape into a Bezier shape after creating it. Select the B-
spline shape using the Select All tool, right-click on it and select convert bspline to bezier.
You can now edit your shape in the same way as other Bezier shapes.
USER GUIDE
555
Using the Ellipse, Rectangle, and Cusped Rectangle Tools | To Use the Ellipse, Rectangle, and Cusped
Using the Ellipse, Rectangle, and

Cusped Rectangle Tools
The Ellipse, Rectangle, and Cusped Rectangle tools let you draw an ellipse- or rectangle-shaped
Bezier shape. Rectangles and cusped rectangles look exactly the same initially, but if you move one or
more of the control points, you can see that cusped rectangles always have sharp corners and no
tangents.
After creation, ellipses, rectangles, and cusped rectangles can be edited like normal Bezier shapes.
Here, an ellipse and a rectangle

have been used to create a quick
garbage matte.
To Use the Ellipse, Rectangle, and Cusped Rectangle

Tools
1. Right-click the Bezier tool in the RotoPaint toolbar and select the Ellipse tool , Rectangle tool
, or Cusped Rectangle tool .

Stroke/Shape Attributes.)
4. Click+drag across the Viewer to draw an ellipse or a rectangle shape.
If you want to draw a shape from the center out, press Ctrl/Cmd+Shift as you’re dragging.
USER GUIDE
556
Using the Ellipse, Rectangle, and Cusped Rectangle Tools | To Use the Ellipse, Rectangle, and Cusped
If you want to create a perfect circle with the Ellipse tool or a square with the Rectangle tool, hold
down Shift while drawing your shape.
Holding down Shift while drawing

an ellipse creates a perfect circle.
USER GUIDE
557
Using the Open Spline Tool | To Use the Open Spline Tool
Using the Open Spline Tool

The Open Spline tool lets you draw curves in a similar way to other shapes, except that they don't
need form a closed shape.
After creation, you can edit points on open splines using the standard smooth and transform
handles, but they also have individual thickness and feather handles. The image shows an open
spline and each point's smooth (orange), thickness (green), and feather (magenta) handles.
Here, an open spline is used to roto a creature's tail.
To Use the Open Spline Tool

1. Right-click the Bezier tool in the RotoPaint toolbar and select the Open Spline tool .
Tip: You can also press V, when the Viewer has focus, to cycle through the available tools.
2. Select color, blending mode, opacity, and other settings applicable to all RotoPaint tool at the top
of the Viewer. (For information on the available options, see Editing Existing Stroke/Shape
Attributes.)
RotoPaint has a few additional open spline-specific controls:
• width - controls the overall thickness of the spline.
• start type - sets the state of the first point in the spline to either rounded or square.
• end type - sets the state of the last point in the spline to either rounded or square.
3. Click in the Viewer to place the starting point of the spline and then add the required number of
points on the feature you're working on.
USER GUIDE
558
4. Press Return or select another tool to finish the open spline,

OR
Click on the first point in the open spline to close the shape. Closed splines are not filled in the
same way as other shape tools, such as Bezier and Ellipse. You can open a closed spline by right-
clicking a point on the spline and selecting open/close shape.
5. You can adjust the thickness and feather of individual points on the spline by selecting the
required points and using the handles displayed in the Viewer.
Drag the handle and release to apply the change to the selected points.
Decreasing and increasing spline thickness (green handle).
USER GUIDE
559
Decreasing and increasing feathering (magenta handle).

USER GUIDE
560
Setting Default RotoPaint Tools and Settings | To Set a Tool as Your Default Tool:
Setting Default RotoPaint Tools

and Settings
You can choose to have your RotoPaint always open with a particular tool selected, or open a
particular tool with the settings you want.
To Set a Tool as Your Default Tool:

1. Create a file called menu.py in your plug-in path directory if one doesn’t already exist. For more
information on plug-in path directories, see Loading Gizmos, NDK Plug-ins, and Python and Tcl
Scripts.
2. Select your default tool from the following list and add the Python line in your menu.py:
• Select All
nuke.knobDefault('RotoPaint.toolbox','selectAll')
• Select Splines
nuke.knobDefault('RotoPaint.toolbox','selectCurves')
• Select Points
nuke.knobDefault('RotoPaint.toolbox','selectPoints')
• Select Feather Points
nuke.knobDefault('RotoPaint.toolbox','selectFeatherPoints')
• Add Points
nuke.knobDefault('RotoPaint.toolbox','addPoints')
• Remove Points
nuke.knobDefault('RotoPaint.toolbox','removePoints')
• Cusp Points
nuke.knobDefault('RotoPaint.toolbox','cuspPoints')
• Smooth Points
nuke.knobDefault('RotoPaint.toolbox','curvePoints')
• Remove Feather
nuke.knobDefault('RotoPaint.toolbox','removeFeather')
• Open/Close Curve
nuke.knobDefault('RotoPaint.toolbox','closeCurve')
• Bezier
nuke.knobDefault('RotoPaint.toolbox','createBezier')
• Cusped Bezier
USER GUIDE
561
Setting Default RotoPaint Tools and Settings | To Set Your Default Tool Properties:
nuke.knobDefault('RotoPaint.toolbox','createBezierCusped')
• B-Spline
nuke.knobDefault('RotoPaint.toolbox','createBSpline')
• Ellipse
nuke.knobDefault('RotoPaint.toolbox','createEllipse')
• Rectangle
nuke.knobDefault('RotoPaint.toolbox','createRectangle')
• Cusped Rectangle
nuke.knobDefault('RotoPaint.toolbox','createRectangleCusped')
• Brush
nuke.knobDefault('RotoPaint.toolbox','brush')
• Eraser
nuke.knobDefault('RotoPaint.toolbox','eraser')
• Clone
nuke.knobDefault('RotoPaint.toolbox','clone')
• Reveal
nuke.knobDefault('RotoPaint.toolbox','reveal')
• Dodge
nuke.knobDefault('RotoPaint.toolbox','dodge')
• Burn
nuke.knobDefault('RotoPaint.toolbox','burn')
• Blur
nuke.knobDefault('RotoPaint.toolbox','blur')
• Sharpen
nuke.knobDefault('RotoPaint.toolbox','sharpen')
• Smear
nuke.knobDefault('RotoPaint.toolbox','smear')
3. Restart Nuke.
To Set Your Default Tool Properties:

1. Create a file called init.py in your plug-in path directory if one doesn’t already exist. For more
Scripts.
2. Define your RotoPaint tool and the default setting you want to set. To get the specific tool and
setting names, you can copy your RotoPaint node in Nuke and paste it into a text editor. The lines
with curved brackets after "toolbox" indicate your current tool settings.
For example:
USER GUIDE
562
Setting Default RotoPaint Tools and Settings | To Set Your Default Tool Properties:
• to set the brush size for paint to 30, and for clone to 280, add this Python line in your init.py:
nuke.knobDefault("RotoPaint.toolbox", '''clone {
{ brush bs 30 }
{ clone bs 280 }
}''')
OR
• to set the brush hardness for paint to 1.0 by default, add this Python line:
nuke.knobDefault("RotoPaint.toolbox", '''brush {
{ brush h 1 }
}''')
• to set the source transform round on clone to on by default, add this Python line:
{ clone str 1 }
}''')
3. Restart Nuke.
Note: As you can see from the above examples, the different RotoPaint tools have their own
default tool settings. If you attempt to set a default but don’t specify for which tool, the
default gets ignored as soon as you switch to a different tool. For example,
nuke.knobDefault('RotoPaint.toolbar_brush_hardness','1.0') sets brush hardness to 1.0
initially, but as soon as you activate a different tool, you get the default for that tool.
Note: You cannot set up multiple defaults through RotoPaint.toolbox. For example, if you
run the following three commands, the default is only set for the last one (sharpen):
nuke.knobDefault("RotoPaint.toolbox", '''clone {{ clone ltt 0}}''')
nuke.knobDefault("RotoPaint.toolbox", '''blur {{ blur ltt 0}}''')
nuke.knobDefault("RotoPaint.toolbox", '''sharpen {{ sharpen ltt 0}}''')
To set the defaults for clone and blur as well, you should instead use this:
{ brush ltt 0 }
{ clone ltt 0}
}''')
USER GUIDE
563
Selecting the Output Format and Channels | To Set Your Default Tool Properties:
Selecting the Output Format and

Channels
In the RotoPaint properties panel, you can select one output channel or many to indicate the
channels where the results of your changes should be stored.
1. From the output field, select the layer containing the channels you want to use. By default, rgba
is selected, and the red, green, blue, and alpha channels are checked on the right.
2. Uncheck any channels that you don’t want to process. The node processes all those you leave
checked. For more information on selecting channels, see Calling Channels.
3. If you want to, you can use the output mask dropdown menu to select a channel where
RotoPaint outputs a mask for what it rendered. By default, the channel is none, but if you select a
channel in the menu, the output mask box is automatically checked.
The mask can be useful, for example, if you need to apply grain to the areas you've painted, but
you don’t want to double up the grain in other areas.
By default, the output mask control is hidden, but you can display it by clicking on the black
triangle above color.
4. If necessary, select your premultiply value.
Premultiply multiplies the selected input channels with a mask representing the paint strokes
and shapes. For example, where there are no paint strokes or shapes (the paint matte is black or
empty) the input channels are set to black, and where the paint strokes or shapes are opaque (the
paint matte is white or full) the input channels keep their full value.
Note that selecting rgba premultiplies the alpha against itself (a*a). If you don’t want this to
happen, set premultiply to rgb instead.
Tip: You can use the Link Menu to link the channel controls with other controls. For
more information, see Linking Channels Using the Link Menu.
5. From the clip to dropdown menu, select how you want to restrict the output image:
• no clip - Do not restrict the output image.
• bbox - Restrict the output image to the incoming bounding box.
• format - Restrict the output image to the incoming format area (the default).
• union bbox+format - Restrict the output image to a combination of the bounding box and the
incoming format area.
USER GUIDE
564
Selecting the Output Format and Channels | To Set Your Default Tool Properties:
• intersect bbox+format - Restrict the output image to an intersection of the bounding box and
incoming format area.
You can also check the Replace box if you want to clear the channels to black before drawing into
them. You might find Replace useful, for instance, if you’re creating a mask in the alpha channel,
but the incoming image already has an alpha channel which you want to throw away.
6. If you haven’t connected an input to RotoPaint, select your format value. This is the format which
the node should output in the absence of any available input format. If an input is connected, this
control has no effect.
By default, the format control is hidden, but you can display it by clicking on the black triangle
above color.
USER GUIDE
565
Selecting Existing Strokes/Shapes for Editing | To Adjust Display Properties for Selected
Selecting Existing Strokes/Shapes

for Editing
If you’ve already drawn a stroke/shape but wish to make changes to it, you can select it or certain
parts of it with the RotoPaint selection tools. You can also toggle between all of them in the RotoPaint
toolbar using the shortcut Q.
When you click on a point in a stroke/shape, the point changes color to indicate whether it’s in focus
(green by default) and whether it has an expression set (red by default). You can change the default
colors on the Viewers tab in the Preferences.
You can also use the controls in the RotoPaint tool settings to display and hide information such as
points, point numbers, splines, and transform handles. See To Adjust Display Properties for Selected
Shapes/Strokes below.
To Adjust Display Properties for Selected Shapes/Strokes

Whenever you have a Select tool active, the RotoPaint tool settings allow you to control what
information is displayed for the visible shapes/strokes and points in the Viewer:
• To view or hide the numbers for visible shape/stroke points, toggle the label points button .
Feather points are marked with a bracketed number corresponding their shape point, so for
example, if a shape point is marked with the number 2, its feather point is marked with [2].
• To view or hide the splines in visible shapes, toggle the hide splines button .
• To view or hide the splines in visible paint strokes, enable show paint stroke splines and
toggle the hide splines button .
• To view or hide the points (and tangent handles) in visible shapes/strokes, toggle the hide points
button .
• To view or hide the transform handle jack or transform box for visible shapes/strokes, toggle the
hide transform handles button (or press T on the Viewer).
• To hide the transform handle jack or transform box when moving a selection, click the hide
transform handles on move button . This may make it easier to correctly position your selection.
USER GUIDE
566
Selecting Existing Strokes/Shapes for Editing | To Select an Entire Stroke/Shape
To Select an Entire Stroke/Shape

1. Click the Select All tool , or press the keyboard shortcut Q.
2. Select the stroke/shape you wish to edit either by clicking on it in the Viewer or by clicking on its
name in the stroke/shape list. To select several strokes/shapes, Ctrl/Cmd+click or Shift+click (to
select a range) their names in the stroke/shape list. You can have both splines and points selected
simultaneously.
When selecting strokes/shapes in the Viewer, you can invert your selection by right-clicking and
selecting invert selection. All strokes/shapes you didn’t have selected before are now selected.
If you want to move the selected stroke/shape, drag a selection box around the entire shape until
a box forms around the shape. Pulling on the cross at the center of the shape allows you to move
it around the Viewer. A keyframe is automatically added when you move a stroke/shape, if
autokey is enabled. For information on autokey and creating keyframes, see Animating
Strokes/Shapes.
Tip: To select points in a paint stroke and view your selection in the Viewer, you have to
enable show paint stroke splines in the tool settings.
Tip: By default, if you have selected a stroke/shape in the Viewer or the stroke/shape list,
clicking on an empty area in the Viewer does not deselect it. If you’d like to change this
behavior, you can disable constant selection mode in the RotoPaint tool settings.
To Select a Spline
1. Right-click the Select All tool and select the Select Splines tool .
2. Make sure hide splines is disabled in the RotoPaint tool settings.

3. Select the spline you wish to edit either by clicking on it in the Viewer or by clicking on its name in
the stroke/shape list. You can also click-and-drag a selection box around the shape with Select
Splines active. Selecting a spline only selects the spline, not points within it.
USER GUIDE
567
Selecting Existing Strokes/Shapes for Editing | To Select Only Points on a Stroke/Shape
Tip: Using the Select Splines tool, you can also duplicate the stroke/shape you’ve selected.
Just right-click on one of the points, and select duplicate. A new stroke/shape is created with
the same spline and attributes as the one you selected.
To Select Only Points on a Stroke/Shape

1. Right-click the Select All tool and select the Select Points tool .
2. Make sure hide points is disabled in the RotoPaint tool settings.

3. Select the stroke/shape you wish to edit by clicking on its name in the stroke/shape list and then
select a point in the Viewer. To select several points, Ctrl/Cmd+click on them in the Viewer or use
marquee selection to create a transform box.
Using the Select Points tool restricts selection to one stroke/shape only.
To Select Feather Points

1. Right-click the Select All tool and select the Select Feather Points tool .
2. Make sure hide points is disabled in the RotoPaint tool settings.

3. Select the stroke/shape you want to edit by clicking on its name in the stroke/shape list and then
select a feather point in the Viewer. To select several feather points, Ctrl/Cmd+click on them in the
Viewer or use marquee selection to create a transform box.
Using the Select Feather Points tool restricts selection to one stroke/shape only.
To Select Points Using a Transform Handle

1. Right-click the Select All tool and select the Select Points tool .
2. Make sure hide transform handles is disabled in the RotoPaint tool settings.
3. Select points in a stroke/shape with Shift+click or by clicking and dragging across the points you
want to select. A transform box appears.
4. You can also use shortcut T to toggle viewing the transform handle.
USER GUIDE
568
Editing Existing Stroke/Shape Attributes | Editing Color
Editing Existing Stroke/Shape

Attributes
After selecting a stroke/shape using the stroke/shape list or one of the Select tools, you can edit and
animate its attributes in the properties panel. For more information on selecting strokes/shapes, see
Selecting Existing Strokes/Shapes for Editing and Working with the Stroke/Shape List.
If you want to edit the attributes of a stroke/shape prior to drawing one, you should do that in the
RotoPaint tool settings in the Viewer. (See Working with the Toolbars.)
Editing Color
When drawing a stroke/shape, you can set the RGBA color values of the stroke/shape using the color
controls on the RotoPaint tab of the RotoPaint properties panel (for more information on the color
controls, see the Using the Compositing Environment chapter). You can also adjust color directly in
the stroke/shape list using the control in the color column.
Color set to white (the default). Color set to green.
Editing Opacity
You can set the opacity of the stroke/shape using the opacity slider. If you set the opacity of a shape
to zero, the outline for it won’t be drawn unless the shape is selected. You can also temporarily make
the stroke/shape invisible (that is, completely transparent) by toggling the visible box in the
stroke/shape list.
USER GUIDE
569
Editing Existing Stroke/Shape Attributes | Selecting a Source for Your Stroke/Shape
A low opacity value. A high opacity value.
Tip: When drawing brush strokes, you can tie their transparency to pen pressure. Just check
the opacity box next to pressure alters on the Stroke tab of the properties panel.
Selecting a Source for Your Stroke/Shape

You can select a source for your stroke/shape and define whether it’s a color, a background, or a
foreground image. With your stroke/shape selected, choose a source for your stroke/shape from the
source dropdown on the RotoPaint, Shape, or Stroke tabs.
Editing Blending Mode

By selecting different blending modes from the blending mode dropdown in the properties panel,
you can select how the colors in your strokes/shapes blend with the underlying image. You can also
apply blending modes to your strokes, shapes, or groups directly in the stroke/shape list using the
blending mode column .
Each of the blending modes blends the primary color, that is the color of the current
stroke/shape/group you’re editing with the secondary color, which is the combined color of your
background image and any previously rendered strokes/shapes/groups.
The different modes are as follows:

• Color burn - Darkens the primary color to reflect the secondary color by increasing the contrast. No
part of the image becomes lighter.
• Color dodge - Brightens the primary color to reflect the secondary color by decreasing the contrast.
No part of the image is darkened.
USER GUIDE
570
Editing Existing Stroke/Shape Attributes | Editing Blending Mode
• Difference - Subtracts either the secondary color from the primary color or vice versa, depending
on which is brighter. Blending with white inverts the primary color, while blending with black
produces no change. Similar colors return black pixels. Difference is a useful mode when working
with mattes.
• Exclusion - Creates a result similar to the Difference mode but lower in contrast. Like with
Difference, blending with white inverts the primary color. Blending with black produces no change.
• From - Subtracts the primary color from the secondary color.
• Hard Light - Lightens highlights and darkens shadows. If the secondary color is lighter than 50%
gray, the result lightens as if it were screened. If the secondary color is darker than 50% gray, the
result is darkened as if it were multiplied.
• Max - Selects the lighter of the two colors as the resulting color. Only areas darker than the
secondary color are replaced, while areas lighter than the secondary color do not change.
• Min - Selects the darker of the two colors as the resulting color. Any parts that are lighter than the
secondary color are substituted. Any parts of the image that are darker than the secondary color
don’t change.
• Minus - Subtracts the secondary color from the primary color.
• Multiply - Multiplies the primary color by the secondary color. The result is always darker. Blending
with black gives black, and with white returns the color unchanged.
• Over - This mode is the default. The colors of the two images do not interact in any way, and Nuke
displays the full value of the colors in the primary image.
• Overlay - Depending on the primary color, multiplies or screens the colors. The secondary color
brightens the primary color while preserving highlights and shadows.
• Plus - The sum of the two colors. Increases brightness to lighten the primary color and reflect the
secondary color. Plus is similar to the Screen blending mode, but produces a more extreme result.
• Screen - This is a soft Plus making everything brighter but ramping off the whites. Light colors have
more of an effect than dark colors. The result is always a lighter color. Blending with black leaves
the pixel unchanged, blending with white always returns white. The result is similar to projecting
multiple slides on top of each other.
• Soft Light - Depending on the primary color, darkens or lightens the colors. Less extreme than the
Hard Light mode.
Tip: Note that changing the stack order of your primary and secondary colors might have an
impact on your result. For example, if you have two Bezier shapes overlapping each other
with a blending mode active, the result depends on which shape is on top of the other. You
can change the stack order of strokes/shapes in the stroke/shape list.
USER GUIDE
571
Transforming Strokes/Shapes/Groups | Editing Blending Mode
Transforming
Strokes/Shapes/Groups
To apply spatial transformations to your strokes, shapes, or groups, you can use the controls under
the Transform tab. Select a stroke/shape/group from the stroke/shape list and adjust:
• translate - to move the stroke/shape on x and y axis.
• rotate - to spin a stroke/shape around the pivot point. Use center (or Ctrl/Cmd+drag the transform
jack) to position the pivot point.
• scale - to resize a spline. Use center (or Ctrl/Cmd+drag the transform jack) to position the pivot
point.
• skew X - to skew the spline of your stroke/shape along the X axis from the pivot point. Use center
(or Ctrl/Cmd+drag the transform jack) to position the pivot point.
• skew Y - to skew the spline of your stroke/shape along the Y axis from the pivot point. Use center
(or Ctrl/Cmd+drag the transform jack) to position the pivot point.
• skew order - to set the order in which skew X and skew Y are applied:
• XY - Skew X is applied before skew Y.
• YX - Skew Y is applied before skew X.
• center x, y - to set the center for rotation and scaling, adjust the values in center x, y.
• extra matrix - enter values you want to get concatenated with the transformation controls above.
For more information on concatenating, see How Nodes Concatenate.
Note: While you can drag the matrix to another node’s matrix to easily use the values
elsewhere, you shouldn’t try, for example, to drag a 4 by 4 matrix on a 3 by 3 matrix, as
doing that might have unexpected results.
Alternatively, you can also use the transform handle (shortcut T) in the Viewer to transform elements.
To transform an entire stroke/shape, you need to use the transform handle jack, and to transform
points in a stroke/shape, you should use the transform box.
The transform handle appears as a transform jack only when the Transform tab is active, when any
of the other tabs in the RotoPaint properties panel are active, the transform handle appears as a box.
USER GUIDE
572
Transforming Strokes/Shapes/Groups | To Transform a Stroke/Shape Using a Transform Handle Jack:
To Transform a Stroke/Shape Using a Transform Handle

Jack:
1. Click one of the Select tools in the RotoPaint toolbar, with the Transform tab active.
3. Select a stroke/shape by clicking it in the Viewer or by selecting it in the stroke/shape list. A
transform handle jack appears.
4. Use the jack to rotate, scale or skew your stroke/shape.
To Transform Points Using a Transform Box:

1. Click the Select All , Select Points , or Select Feather Points tool in the RotoPaint
toolbar, with the RotoPaint tab active.
3. Select points in a stroke/shape with Shift+click or by clicking and dragging across the points you
want to select. A transform box appears.
4. Use the box to rotate, scale or skew your stroke/shape, or points.
5. To corner pin using the transform box, press Ctrl/Cmd+Shift and drag the transform box points
to move them.
Note: Transforming points changes the actual point position, while transforming an entire
stroke/shape/group changes transformation applied to the point positions.
Tip: If you’d like to hide the transform box when moving a selection, enable hide transform
handles on move in the RotoPaint tool settings. This may make it easier to correctly
position your selection.
To Transform Onion Skin Source

When cloning or revealing, you can use the onion skin control on the RotoPaint tool settings to view
and transform your source input on top of your foreground. You can also use onion skinning if you’re
drawing a stroke/shape with a separate input as the source. To adjust onion skin:
USER GUIDE
573
Transforming Strokes/Shapes/Groups | To Transform Onion Skin Source
1. With your stroke/shape tool selected, check the onion box in the RotoPaint tool settings.
2. Adjust the opacity of the onion skin and transform your source by using the onion skin transform
overlay in the Viewer.
Onion skin transform handle.
USER GUIDE
574
Adjusting Mask Controls | To Transform Onion Skin Source
Adjusting Mask Controls

By default, mask is set to none, but if you want to use a mask, do the following:
1. Check the mask checkbox on the RotoPaint tab. By default, the mask control is hidden, but you
can display it by clicking on the black triangle above color.
2. Select a channel to use as a mask from the dropdown menu.
3. If you are using the mask input and want the mask copied into the predefined mask.a channel,
check inject. This way, you can use the mask input again downstream.
4. If necessary, check invert to reverse the mask and/or fringe to blur the edges of the mask.
5. If you find that the overall effect of the RotoPaint node is too harsh, you can also blend some of
the original input image back in by adjusting the mix slider.
USER GUIDE
575
Editing Shape-Specific Attributes | Adding and Removing Feather
Editing Shape-Specific Attributes

The RotoPaint properties panel includes a set of controls that you mainly need when you're editing
the attributes of a shape. You can find these on the Shape tab in the properties panel.
Adding and Removing Feather

To soften the edges of a shape, do the following:
1. With your shape selected, check the on box next to the feather slider on the Shape tab to apply
feather to your shape. If you don’t want any feather on your shape, uncheck the on box.
2. Use the feather slider to add outward or inward feather around the whole shape. With positive
feather values, your feather effect is outward and, respectively, if your feather values are negative,
the feather effect is inward.
3. Use the feather falloff slider to adjust the falloff profile. Falloff is measured in pixels. You can
also change the type of the falloff using the falloff type dropdown menu. Select either linear,
smooth0, smooth1 and smooth. Each of these produces a different rate of falloff that may be
helpful for example in matching the soft edge to motion blurred image content.
4. To add feather to a single point, right-click on the point in the Viewer and select increase
feather. The shortcut for this is E. If you press it several times, every press adds more feather.
5. Use the feather handle on the point to add feather into the point. By default, the point angle is
locked, and moving the point unlocks the angle.
You can also select several feather points and move them together.
6. To remove feather from a point, right-click on the point and select reset feather or use the
shortcut Shift+E.
7. If you disable feather link mode (enabled by default) in the RotoPaint tool settings, the
feather and shape points move independently.
• centered - to center the shutter around the current frame. For example, if you set the shutter
value to 1 and your current frame is 30, the shutter stays open from frame 29.5 to 30.5.
• start - to open the shutter at the current frame. For example, if you set the shutter value to 1
and your current frame is 30, the shutter stays open from frame 30 to 31.
• end - to close the shutter at the current frame. For example, if you set the shutter value to 1 and
your current frame is 30, the shutter stays open from frame 29 to 30.
• custom - to open the shutter at the time you specify. In the field next to the dropdown menu,
enter a value (in frames) you want to add to the current frame. To open the shutter before the
current frame, enter a negative value. For example, a value of - 0.5 would open the shutter half a
frame before the current frame.
USER GUIDE
576
Editing Stroke-Specific Attributes | Selecting a Source Image
Editing Stroke-Specific Attributes

The RotoPaint properties panel includes a set of controls that you mainly need when you're editing
the attributes of a paint stroke. You can find most of these under the Stroke tab in the properties
panel.
Selecting a Source Image

Both on the Stroke and RotoPaint tabs, you can set the source control to a specific color or the
input you want to pull pixels from for Clone and Reveal brushes. Select:
• color - to use a specific color in your stroke/shape.
• foreground - to pull pixels from the RotoPaint’s bg input, including any strokes/shapes drawn on it.
This input is mainly used with cloning.
• background - to pull pixels from the bg input, not including any strokes/shapes drawn on it.
• background 1, background 2 or background 3 - to pull pixels from the bg1, bg2, or bg3 input.
Editing Brush Type

On the Stroke tab, select the type of brush you want to use for the stroke. From the brush type
dropdown, select:
• paint - to use a normal paint brush.
• smear - to use a smear brush on the plate.
• blur - to blur your plate with the brush stroke. You can adjust the strength of the blur effect using
the effect slider.
• sharpen - to sharpen your plate with the brush stroke. You can adjust the strength of the
sharpening effect using the effect slider.
Editing Brush Size

On the Stroke tab, you can set the size of the stroke using the brush size slider. You can also tie a
stroke’s size to pen pressure by checking the size box next to pressure alters.
USER GUIDE
577
Editing Stroke-Specific Attributes | Editing Brush Spacing
A low brush size value. A higher brush size value.
Editing Brush Spacing

The brush spacing slider adjusts the distance between paint brush dabs. A higher setting increases
the space between dabs, creating a dotted line effect when painting. A lower setting decreases the
distance and creates a solid brush stroke.
A low brush spacing value. A higher brush spacing value.
Editing Brush Hardness

On the Stroke tab, you can set the hardness of the stroke using the brush hardness slider.
USER GUIDE
578
Editing Stroke-Specific Attributes | Adjusting Write On
A low brush hardness value. A higher brush hardness value.
You can also tie a stroke’s hardness to pen pressure by checking the hardness box next to pressure
alters.
Adjusting Write On
When you are animating a stroke or a part of it over a range of frames, you can use the write on
sliders under the Stroke tab in the properties panel to adjust the order in which the dabs on the
stroke appear over these frames. For more information on animating parameters, see the Using the
Compositing Environment chapter.
• write on start - slide to choose where along the stroke length the paint begins. 0 is the start of the
stroke, 1 is the end.
• write on end - slide to choose where along the stroke length the paint ends.
USER GUIDE
579
Editing Clone or Reveal Attributes | Adjusting Write On
Editing Clone or Reveal Attributes

When you are using the Clone or Reveal tool, you can adjust the controls under the Clone tab to
transform the input that's being cloned or revealed. Adjust:
• translate - to move the source image on x and y axis.
• rotate - to spin the source image around a pivot point.
• scale - to resize the source image by adding or removing pixels. Use center (or Ctrl/Cmd+drag the
transform jack) to position the pivot point.
• skew X - to skew the source image along the X axis from the pivot point. Use center (or
Ctrl/Cmd+drag the transform jack) to position the pivot point.
• skew Y - to skew the source image along the Y axis from the pivot point. Use center (or
Ctrl/Cmd+drag the transform jack) to position the pivot point.
• skew order - to set the order in which skew X and skew Y are applied:
• XY - Skew X is applied before skew Y.
• YX - Skew Y is applied before skew X.
• round to pixel - check this to round the translate amount to the nearest whole integer pixel. This
can help you avoid softening when using filtering. If source is set to color on the Stroke tab, this
control is disabled.
• filter - to select the appropriate filtering algorithm. For more information, see Choosing a Filtering
Algorithm.
• black outside - When rotating or translating the clone source, a part of the image area may get
cropped. To fill the cropped portion with black, check black outside. To fill the cropped portion by
expanding the edges of the image, uncheck black outside.
• time offset - to clone or reveal pixels from a different frame. Time offset is either relative to the
current frame (-1 is the frame previous to the current one) or absolute (1 is the first frame in the
clip).
• view - to select which view you want to clone from in a stereoscopic project.
USER GUIDE
580
Editing Existing Stroke/Shape Timing | Adjusting Write On

Timing
When editing an existing stroke/shape, you can edit the range of frames during which a stroke/shape
is visible in the properties Lifetime tab. The lifetime of a stroke/shape/group is also visible in the Life
column in the stroke/shape list. By default, a shape is visible on all frames, whereas a stroke is only
visible on a single frame, the frame it was painted on.
Icon Lifetime Type Description
all frames Click to make a stroke/shape visible for all frames (the default).
frame to end Click to make a stroke/shape visible from the current frame to the last
frame.
single frame Click to make a stroke/shape visible only on the current frame.
start to frame Click to make a stroke/shape visible from the first frame to the current
frame.
frame range Click to make a stroke/shape visible during a specified range of frames
using the frames dialog or from and to controls.
USER GUIDE
581
Editing Existing Stroke/Shape Stack Order | Adjusting Write On

Stack Order
When editing strokes/shapes after you’ve drawn them, you can edit their foreground to background
drawing order.
In the stroke/shape list, you can drag and drop strokes/shapes to change their drawing order, and to
group them under folders. For more information on using the stroke/shape list, see Working with the
Stroke/Shape List.
USER GUIDE
582
Editing Existing Stroke/Shape Splines | To Add a Point to a Stroke/Shape

Splines
To edit a stroke/shape’s position, you first need to select the stroke/shape in the Viewer or the
stroke/shape list. You can then modify the points that make up the stroke/shape’s position.
To Add a Point to a Stroke/Shape

1. Select the Add Points tool from the RotoPaint toolbar.
2. Make sure hide points is disabled in the tool settings.

3. In the Viewer, click on the selected stroke/shape to add a new point.
To Move a Point
1. With the Select All tool or Select Points tool active, select the stroke/shape in the Viewer
or the stroke/shape list.

3. In the Viewer, drag the points you want to move to a new location.
4. You can also nudge a point using the arrows on your numeric keypad. Nudging a point moves it
by one pixel to the direction you choose.
Tip: If you find it difficult to select a single point, you might want to increase the 2D/3D
handle size or the handle pick size in the Nuke Preferences dialog (Preferences > Panels
> Viewer Handles).
To Move Several Points Together

1. With the Select All tool or Select Points tool active, select the stroke/shape in the
Viewer or the stroke/shape list.
USER GUIDE
583
Editing Existing Stroke/Shape Splines | To Delete a Point
3. In the Viewer, drag a marquee around the points (or an entire stroke/shape) that you want to
move. If hide transform handles is disabled, a transform box appears.
4. Adjust the transform box as necessary.
Note: Moving a shape with Select All active moves the selected points or feather points to
the new location. Scrubbing the timeline shows the shape move across the Viewer between
keyframes. However, moving a shape with Select Feather Points active moves only the
feather points to the new location, displaying the feather falloff of the selected points.
To Delete a Point
1. Right-click on the Add Points tool and select Remove Points tool .
2. Select the stroke/shape in the Viewer or the stroke/shape list.

4. In the Viewer, click the point that you want to delete.
OR
In the Viewer, right-click on the point that you want to delete and select delete.
To Delete an Entire Stroke/Shape

1. In the stroke/shape list, or in the Viewer with the Select All tool , click on the stroke/shape
you want to delete.
2. Below the stroke/shape list, click on the minus button (-),
OR
1. Activate the Select All tool . In the Viewer, click on the stroke/shape.
2. Right-click on the shape and select delete or press the Delete/Backspace key.
USER GUIDE
584
Copying, Cutting, and Pasting Stroke Attributes | To Copy Stroke/Shape Attributes
Copying, Cutting, and Pasting

Stroke Attributes
Copying, cutting, and pasting stroke/shape attributes is achieved in a similar way to affecting position:
To Copy Stroke/Shape Attributes

1. With the Select All , Select Points , or the Select Feather Points tool active, select
the stroke/shape in the Viewer.

3. In the Viewer, drag a marquee around the points (or an entire stroke/shape) that you want to
copy. If hide transform handles is disabled, a transform box appears.
4. Right-click inside the transform box and select copy > curve (attribute values).
The attribute values of the stroke/shape, including position and shape, are copied to the
clipboard. Values only apply to a specific keyframe.
To Cut Stroke/Shape Attributes


3. In the Viewer, drag a marquee around the points or click on the stroke/shape for the points that
you want to cut. If hide transform handles is disabled, a transform box appears.
4. Right-click inside the transform box and select cut > curve (attribute values).
The attribute values of the stroke/shape, including position and shape, are cut from the Viewer
and placed in the clipboard. Values only apply to a specific keyframe. Note, that the shape of the
cut stroke/shape does not disappear from the Viewer.
USER GUIDE
585
Copying, Cutting, and Pasting Stroke Attributes | To Paste Stroke/Shape Attributes
To Paste Stroke/Shape Attributes


3. In the Viewer, drag a marquee around the points or click on the stroke/shape for the points that
you want to paste. If hide transform handles is disabled, a transform box appears.
4. Right-click inside the transform box and select paste > spline (attribute values).
This pastes the copied stroke/shape attribute values that were on the clipboard. If you have
copied another type of curve, for example, (values) or (animations), then all options except for
that one are disabled.
USER GUIDE
586
Point Cusping, Smoothing, Expressions, and Links | To Cusp or Smooth Points
Point Cusping, Smoothing,

Expressions, and Links
Roto and RotoPaint allow you manipulate curves, by cusping or smoothing the points on the curve,
add expressions, and link controls to points.
To Cusp or Smooth Points

You can cusp points on a shape to create sharp corners, and smooth points to replace sharp corners
with curved lines.
1. Select the shape you want to edit in the Viewer or the stroke/shape list.
2. Select the Smooth Points tool or Cusp Points tool in the RotoPaint toolbar, depending
on whether you want to cusp or smooth your points.

4. In the Viewer, click on the point that you want to cusp or smooth.
5. With your point selected, you can also use the shortcut keys Z and Shift+Z to smooth and cusp it
respectively.
OR
Right-click on the point you want to smooth or cusp, and select smooth or cusp/de-smooth.
To Add Expressions to Points

1. With the Select All , Select Points , or Select Feather Points tool active, make sure
hide points is disabled in the tool settings.

2. Select a point in your stroke/shape. Right-click and select add expression.
3. Enter your expression values to the fields in the Expression dialog. Alternatively, you can
Ctrl/Cmd+drag expression values from another node to the point in your stroke/shape.
You can add expressions to edit your point’s location or the shape of the point’s feather effect. For
more information on Expressions see Expressions.
USER GUIDE
587
Point Cusping, Smoothing, Expressions, and Links | To Link Other Controls to a Point’s Position
To Link Other Controls to a Point’s Position

You can copy a point link (that is, a point’s x and y position) and use it in other nodes.
To copy a point link:
1. With the Select All , Select Points , or Select Feather Points tool active, make sure
hide points is disabled in the tool settings.
2. Right-click on a point and select copy > point link.
3. Go to the field you want to use the link in, for example another node, right-click and select Copy >
Paste.
For example, you can copy a point link, click on a Transform node’s animation menu next to
translate, and select Copy > Paste. An expression arrow appears between the RotoPaint node
and the Transform node, and when you move the point, the Transform node’s translate values
update with your changes.
Note: When copying, cutting, or pasting from the right-click menu, numbers are displayed
next to the curve, spline, or point. These numbers represent how many of each you have
selected and thus, how many are copied or pasted. For example, 2 point (values) denotes
that 2 points are selected. Pasting two points into a shape with many more points replaces
the lowest two points in the shape, or the highest two splines in the stroke/shape list.
USER GUIDE
588
Animating Strokes/Shapes | To animate a stroke/shape using autokey
Animating Strokes/Shapes
All strokes/shapes that appear on more than one frame can be animated. By default, the autokey
option in the RotoPaint tool settings is on, which means your changes to a stroke/shape
automatically creates keyframes and animates your stroke/shape. You can also access all the curves
and shapes in the Curve Editor and Dope Sheet.
To animate a stroke/shape using autokey

1. Draw a stroke/shape that appears on more than one frame. By default, the autokey option in
the RotoPaint tool settings is selected and a keyframe is automatically created in the first frame
your stroke/shape appears.
2. Move to a new frame.
3. With one of the Select tools, select the points or the stroke/shape you want to animate.
4. Adjust the points in your stroke/shape position or change the stroke/shape’s attributes as
necessary. A new keyframe is automatically set. The frame marker on the timeline turns blue to
indicate the selected stroke/shape is animated.
5. Repeat steps 2, 3 and 4 for all the frames you want to set as key frames.
Tip: Note that if you are translating an entire stroke/shape using the Transform tab,
RotoPaint also draws a track of the stroke/shape’s animated position. You can, for example,
use the track in another node (such as Tracker or CameraTracker) by Cmd/Ctrl+dragging the
values from the translate animation button to an appropriate field in another node.
To animate strokes/shapes manually

If you choose to switch the autokey function off, you can still create key frames manually. You
can set key frames to the entire stroke/shape, or the stroke/shape’s spline, transformation or
attributes.
1. Move to the frame where you want to create a keyframe and select your stroke/shape.
• If you want to create a key that is set to animate the entire stroke/shape, right-click on the
stroke/shape and select set key > all.
USER GUIDE
589
Animating Strokes/Shapes | To animate strokes/shapes manually
• If you want to create a key that is set to animate a position, right-click on the stroke/shape and
select set key > spline.
• If you want to create a key that is set to animate transformation, right-click on the stroke/shape
and select set key > transform.
• If you want to create a key that is set to animate attributes, right-click on the stroke/shape and
select set key > attributes.
If you have autokey turned off, you can only adjust a point in a stroke/shape at a keyframe.
In other words, in order to make changes to a point, you either have to move to an existing
keyframe on the timeline, or you need to create a new keyframe first.
USER GUIDE
590
Viewing Spline Keyframes | To View Keyframes on the Timeline
Viewing Spline Keyframes

You can use the spline key controls on the RotoPaint tab to view whether there are keyframes set for
the spline of your stroke/shape. Do the following:
1. Select a stroke/shape on the stroke/shape list.
2. If there are spline keys set on your stroke/shape, the key boxes are highlighted blue and display
where you are currently on the timeline with regard to the keyframes that are set already. You can
move backwards and forwards between the keyframes using the arrow keys .
3. If you want to add or remove spline keys for the stroke/shape in a selected frame, use the Add
and Delete buttons.
To View Keyframes on the Timeline

You can view different types of keyframes you’ve set, either automatically or manually, on the
timeline. If you’ve set keyframes on the RotoPaint tab, these are visible on the timeline when you
have the RotoPaint tab open in the properties panel. Similarly, if you’ve created transformation
keyframes on the Transform tab, you can see those keyframes on the timeline when you have the
Transform tab open.
• If you want to delete a key that is set to animate a position, right-click on the stroke/shape and
select delete key > spline.
• If you want to delete a key that is set to animate a transformation, right-click on the
stroke/shape and select delete key > transform.
• If you want to delete a key that is set to animate attributes, right-click on the stroke/shape and
select delete key > attributes.
If no other keys are set for the selected stroke/shape, the frame marker on the timeline turns from
blue to the default color.
USER GUIDE
591
Deleting or Rippling Keyframes | To Delete a Keyframe
Deleting or Rippling Keyframes

You can delete keyframes individually or by stroke/shape, or use the ripple function to change a
keyframe and apply that same relative adjustment to the point across a range of frames.
To Delete a Keyframe
1. Using the Viewer timeline, scrub to the frame where you want to delete a keyframe.
2. In the stroke/shape list, select the stroke/shape whose key you want to delete.
• If you want to delete a key that is set to animate the entire stroke/shape, right-click on the
stroke/shape and select delete key > all.
• If you want to delete a key that is set to animate a position, right-click on the stroke/shape and
select delete key > spline.
• If you want to delete a key that is set to animate a transformation, right-click on the
stroke/shape and select delete key > transform.
• If you want to delete a key that is set to animate attributes, right-click on the stroke/shape and
select delete key > attributes.
If no other keys are set for the selected stroke/shape, the frame marker on the timeline turns from
blue to the default color.
To Delete All Keyframes for a Stroke/Shape

You can also delete all key frames you’ve set for a stroke/shape at one go. Do the following:
1. Select the stroke/shape from which you want to delete key frames in the Viewer.
2. Right-click on the stroke/shape and select no animation > all, spline, transform or attribute
depending on whether you want to delete all key frames for that stroke/shape or only key frames
animating shape, transform or attributes. All key frames are removed from the stroke/shape you
selected.
To Ripple Keyframes
Rippling keyframes allows you to adjust the position of a stroke/shape point on one frame and have
that same relative adjustment applied to the point across all frames or a specified range of frames.
USER GUIDE
592
Deleting or Rippling Keyframes | To Ripple Keyframes
This way, you can make non-animated changes to a stroke/shape which is being animated over a set
of frames.
1. Activate Select All tool in the RotoPaint toolbar and click the ripple edit button in the
RotoPaint tool settings. A red border displays around the Viewer to indicate that the ripple mode
is active. In the dropdown menu next to ripple edit , select:

• all - to ripple all frames in your sequence.
• from start - to ripple frames from the first frame to the current frame.
• to end - to ripple frames from current frame to the last frame.
• range - to ripple a particular range of frames.
2. Select the stroke/shape you want to edit from the stroke/shape list.
3. Make the necessary changes to the stroke/shape in the Viewer.
USER GUIDE
593
Copying, Cutting, and Pasting Animations | To Copy Animations
Copying, Cutting, and Pasting

Animations
Copying, cutting, and pasting can save you time, but can be destructive. For example, pasting over
shapes destroys all points and keyframes of the target shape in order to replace it.
To Copy Animations
If you want to copy a shape, including all animation attributes, animated points, and associated
keyframes, copying animations is the best way to do this. To copy animations for a stroke/shape:
1. With the Select All tool active, select the stroke/shape in the Viewer.
2. Right-click on the stroke/shape and select copy > curve (spline-key animations).
This copies the current selected animation and key frame information to the clipboard.
To copy only curve attribute information for animated shapes:
2. Right-click on the stroke/shape and select copy > curve (attribute animations).
This copies only the selected attributes for the animated shape to the clipboard.
To copy only the point information for animated shapes:
1. With the Select All tool active, select the stroke/shape in the Viewer. Make sure hide points
is disabled in the tool settings.
2. Right-click on the point or points and select copy > points (animations).
This copies only the selected animated points and keyframes to the clipboard.
To Cut Animations
If you want to move a stroke/shape and replace another with all the same animated points and
keyframes, cutting the shape is a better option than simply copying it. To cut animations for a
stroke/shape:
USER GUIDE
594
Copying, Cutting, and Pasting Animations | To Paste Animations
2. Right-click on the stroke/shape and select cut > curve (spline-key animations).
The shape is cut from the Viewer and placed in the clipboard. The transform box still appears on
the Viewer and allows you to amend the shape, until you click on another shape or elsewhere in
the Viewer.
To cut only animation attributes for a stroke/shape:
2. Right-click on the stroke/shape and select cut > curve (attribute animations).
This cuts the animation attributes for the selected stroke/shape and placed it them in the
clipboard.
To cut only the animated points for strokes/shapes:
1. With the Select All tool active, select the animated point(s) of a stroke/shape in the Viewer
that you want to cut. Make sure hide points is disabled in the tool settings.
2. Right-click on the point(s) and select cut > point (animation).
This cuts only the animated points and keyframes and places them in the keyboard.
To Paste Animations
If you want to paste copied or cut animated shape information, consider whether the shape you are
pasting over has values that need to be saved. Pasting animated points, spline-keys, and attributes
over another shape destroys all points and keyframes of the second shape in order to replace it. To
paste animations for a stroke/shape:
1. With the Select All tool active, select the stroke/shape in the Viewer that you want to paste
the animated shape over. If there are not the same number of points on the shape as on the
copied shape, pasting the animated shape may not paste as expected.
2. Right-click on the stroke/shape and select paste > spline (animations).
This pastes the animation and keyframe information for the copied or cut stroke/shape over the
previously selected shape. This shape may need to be moved away from the copied shape, in
order to see both.
To paste only animation attributes for a stroke/shape:
1. With the Select All tool active, select the stroke/shape in the Viewer that you want to paste
the animated attributes over.
2. Right-click on the stroke/shape and select paste > spline (attribute animations).
USER GUIDE
595
Copying, Cutting, and Pasting Animations | To Paste Animations
This pastes the animation attributes for the copied or cut stroke/shape over the previously
selected shape’s attributes.
To paste only the animated points for strokes/shapes:
1. With the Select All tool active, select the point(s) of a stroke/shape in the Viewer that you
want to paste the animated points over. Make sure hide points is disabled in the tool
settings.
2. Right-click on the point(s) and select paste > point (animation).
This pastes only the animated points and keyframes from the clipboard.
USER GUIDE
596
Adding Motion Blur | Adding Blur to Shapes
Adding Motion Blur

After animating your shapes, you can apply motion blur to individual shapes in the shape list or to
the all shapes in the current node using the controls in the properties Motion Blur tab.
Adding Blur to Shapes

Shape Blur determines the exposure for each moving shape and blends the resulting blurred
shapes. This may be more efficient than the global motion blur since each shape will only be blended
once.
Note: Shape motion blur may result in artifacts when shapes blur over the same region.
To apply shape motion blur:

1. Select the target shapes in the shape list or in the Viewer.
2. In the properties panel, select the Motion Blur tab and make sure that shape mode is enabled.
3. Check the on box next to the motionblur field to apply motion blur to your shape. If you don’t
want any motion blur on your shape, leave the on box unchecked.
Tip: You can also toggle motion blur on and off in the shape list by clicking in the motion
blur column for each shape .
4. In the Shape Blur > motionblur field, enter the sampling rate. This affects the number of times
the input is sampled over the shutter time. The higher the rate, the smoother the result. In many
cases, a value of 1.0 is enough.
A motionblur value of 0.5 produces A higher value, in this case 2.0,
USER GUIDE
597
Adding Motion Blur | Adding Blur to Shapes
less samples. produces a smoother blur.

5. In the shutter field, enter the number of frames the shutter stays open when motion blurring.
For example, a value of 0.5 would correspond to half a frame. Increasing the value produces more
blur, and decreasing the value less.
A shutter value of 0.2 produces less A higher value, in this case 1.0,
blur. produces more blur.
6. Adjust the shutter offset using the shutter offset dropdown menu. The different options control
when the shutter opens and closes in relation to the current frame value. Select:
Centering the shutter Opening the shutter at the Closing the shutter at the
around the current frame. current frame. current frame.
USER GUIDE
598
Adding Motion Blur | Adding Global Motion Blur
Adding Global Motion Blur

Global Blur correctly accounts for interaction between motion blurred shapes. This may be more
expensive than the shape motion blur since it may blend each shape for every sample, in the same
way as TimeBlur. See Applying a TimeBlur Filter for more information.
Global motion blur requires that shutter and sampling parameters are the same for all shapes and
has been optimized for consecutive shapes with the same properties using the over blend mode.
Note: Global motion blur overrides per-shape motion blur, applying your settings to all
shapes in the current node's shape list.
To add global motion blur:

1. In the properties panel, select the Motion Blur tab and make sure that global mode is enabled.
2. In the Global Blur > motionblur field, enter the sampling rate. This affects the number of times
the input is sampled over the shutter time. The higher the rate, the smoother the result. In many
cases, a value of 1.0 is enough.
A motionblur value of 0.5 produces A higher value, in this case 2.0,

less samples. produces a smoother blur.
USER GUIDE
599
Adding Motion Blur | Adding Global Motion Blur
A shutter value of 0.2 produces less A higher value, in this case 1.0,
blur. produces more blur.
Centering the shutter Opening the shutter at the Closing the shutter at the
around the current frame. current frame. current frame.
USER GUIDE
600
Viewing Points in the Curve Editor and the Dope Sheet | Adding Global Motion Blur
Viewing Points in the Curve Editor

and the Dope Sheet
You can edit your stroke/shape points, their feather points and tangents in the Curve Editor and Dope
Sheet. To read more about using the Curve Editor and Dope Sheet, see Animating Parameters.
1. Select your stroke/shape, and right-click on a point.
2. Select curve editor... and select:
• points - to add your point to the Curve Editor
• points+feathers - to add your point and its feather point to the Curve Editor
• points+tangents - to add your point and its tangent to the Curve Editor
• all - to add your point, feather point, and tangent to the Curve Editor.
3. Viewing your points in the Dope Sheet is very similar, just right-click on your point and select one
of the same options under dope sheet...
USER GUIDE
601
Copying, Pasting, and Cutting Stroke Positions | To Copy Stroke/Shape Positions
Copying, Pasting, and Cutting

Stroke Positions
After creating a stroke/shape, you can copy, paste, and cut its position to use the same shape in other
strokes/shapes. Copying, cutting, or pasting values applies to only one keyframe.
To Copy Stroke/Shape Positions

You can copy the stroke/shape position values for an entire selected shape. This enables you to use
the same values for another shape in the Viewer. Unlike cutting stroke/shape values, described
below, copying position values does not delete any keys set. To copy a stroke/shape:
1. With the Select All tool active, select the stroke/shape in the Viewer that you want to copy.
2. Right-click on the stroke/shape and select copy > curve (spline-key values).
This copies the position and point values into the clipboard.
To Cut Stroke/Shape Positions

You can cut the stroke/shape position values for an entire selected shape. This enables you to use the
same values to replace another shape in the Viewer. To cut a stroke/shape:
2. Right-click on the stroke/shape and select cut > curve (spline-key values).
This cuts the position and point values and places it into the clipboard. Cutting the position
deletes the keyframe, or, if there is no keyframe, deletes the points or spline.
Note: To cut stroke/shape position values, you must be cutting the stroke/shape when it is
on a keyframe, unless it has no keyframe. You cannot cut an animated stroke/shape not
currently on a keyframe.
USER GUIDE
602
Copying, Pasting, and Cutting Stroke Positions | To Paste Stroke/Shape Positions
To Paste Stroke/Shape Positions

You can paste copied or cut stroke/shape position values from the clipboard onto another
stroke/shape. This enables you to create a copy, or replacement, shape in the Viewer. To paste a
copied or cut shape:
2. Right-click on a selected stroke/shape and select paste > spline (values).
This pastes the copied or cut position and point values from the clipboard onto another shape. In
the case of copied shapes, the new shape may be created over the original. Simply select and drag
to move it away from the original shape. Spline key values must have the same number of points
between the copied or cut stroke/shape and the stroke/shape you are replacing.
USER GUIDE
603
Copying, Pasting, and Cutting Point Positions | To Copy Point Values
Copying, Pasting, and Cutting

Point Positions
You can copy point position values in a stroke/shape you have selected. This enables you to use the
same values for another point in a stroke/shape. Unlike cutting point values, described below,
copying position values does not delete any keys set.
Note: When copying, cutting, or pasting from the right-click menu, numbers are displayed
next to the curve, spline, or point. These numbers represent how many of each you have
selected and thus, how many are copied or pasted. For example, 2 point (values) denotes
that 2 points are selected. Pasting two points into a shape with many more points replaces
the lowest two points in the shape, or the highest two splines in the stroke/shape list.
To Copy Point Values

1. In the Viewer, scrub to the frame that contains the stroke/shape whose point position you want to
copy.
2. Activate the Select Points tool .

4. Right-click on the point a stroke/shape and select copy > point (values).
Nuke copies the positions of the selected point to the clipboard. Any keys set to animate these
positions are not affected.
USER GUIDE
604
Pasting Point Positions | To Paste Point Values
Pasting Point Positions

You can paste any position you’ve copied/cut from another point to a selected point in a
stroke/shape. If you have autokey turned on, this also sets a keyframe at the current frame to
animate this point. The attributes of the strokes or any keys set to animate the attributes are not
affected.
To Paste Point Values

1. In the Viewer, scrub to the frame that contains the stroke/shape to which you want to paste the
positions on the clipboard.

4. Right-click on the point and select paste > point (values).
Nuke pastes the positions (but not any attributes) on the clipboard to the selected point and, if
you have autokey turned on, sets the current frame as a keyframe.
Note: If you have copied more than one point into the clipboard but have only selected one
point to paste into, paste > point (values) only displays one point and may not look as you
expect. To paste both points, ensure that you have two point selected (or however many is
needed for your sample).
USER GUIDE
605
Cutting Point Positions | To Cut Point Values
Cutting Point Positions

You can cut the position values of a point in a stroke/shape. Cut also copies the positions at the
current frame to the clipboard. The point attributes or any keys set to animate the attributes are not
affected.
To Cut Point Values

1. In the Viewer, scrub to the frame that contains the stroke/shape whose point position you want to
cut.

4. Right-click on the point and select cut > point (values).
Nuke deletes any keys set to animate the positions of the selected point, and copies the position
to the clipboard. Cutting the position deletes the keyframe, or, if there is no keyframe, deletes the
points or spline.
Note: To cut stroke/shape point values, you must be cutting the stroke/shape when it is on
a keyframe, unless it has no keyframe. You cannot cut an animated stroke/shape not
currently on a keyframe.
USER GUIDE
606
RotoPaint and Stereoscopic Projects | Selecting the View to Draw On
RotoPaint and Stereoscopic

Projects
If you need to use RotoPaint in a multi-view or stereoscopic project, you may want to draw your
strokes/shapes or apply your edits to one view only (for example, the left view but not the right), or
create a stroke/shape in one view and have it automatically generated for the other in the correct
position.
For more information on multi-view or stereoscopic projects in Nuke, see Stereoscopic Scripts.
Selecting the View to Draw On

When you are using the RotoPaint node to draw a new stroke/shape in a stereoscopic or multi-view
project, you can toggle the singleview checkbox in the RotoPaint tool settings to draw your
stroke/shape on one view only or multiple views. For existing strokes/shapes/groups, you can use the
view control to select the view the stroke/shape is visible. If you’re working on a stereoscopic project,
the view you’re using for a particular stroke, shape, or group is also visible on the View column in the
stroke/shape list.
Selecting the View to Clone From

When cloning, you can select the view to use as the clone source. Go to the Clone tab and set the
view dropdown menu to the view you want to clone from. To use the view currently displayed in the
Viewer, select current.
USER GUIDE
607
Reproducing Strokes/Shapes in Other Views | Selecting the View to Clone From
Reproducing Strokes/Shapes in
Other Views
To create a stroke/shape on one view and have it automatically generated for the other:
1. Make sure there is a disparity field upstream from the image sequence you want to paint on. If
the image sequence is an .exr file, the disparity field can be included in its channels. Otherwise,
you can use a ShuffleCopy node or Ocula’s O_DisparityGenerator plug-in to add it in the data
stream.
2. In the RotoPaint properties, check all the views in the view dropdown menu. Display the view you
want to paint on in the Viewer.
3. Draw a stroke/shape in the Viewer.
4. Select the stroke/shape in the stroke/shape list in the RotoPaint node properties.
5. Right-click the stroke/shape and select either:
• Correlate points - to use the disparity at each point of the stroke/shape and translate each
point to the corresponding position in the other view.
• Correlate average - to take the disparity at each point, calculate the average disparity of the
stroke/shape and then translate it to the corresponding position in the other view.
The Correlate dialog displays.
6. In the Correlate dialog, select how to correlate the views in the correlate from dropdown. For
example, if your stroke/shape was in the correct position in the left view but not the right, you’d
set correlate from to left to right.
This adds the disparity vectors in the map to the current values, creating the corresponding
stroke/shape for the other view.
7. In the Viewer, switch between the two views to compare the original and the correlated
strokes/shapes.
8. If you want to adjust the stroke/shape further, you need to adjust both views independently.
Adjustments you make to one view are not automatically generated for the other.
USER GUIDE
608
Editing Strokes/Shapes in One View Only | To Edit Stroke/Shape Points
Editing Strokes/Shapes in One

View Only
Splitting a view off allows you to edit the stroke/shape data in that view only, without affecting any
other views that exist in your project settings.
To Edit Stroke/Shape Points

1. Display the view you want to edit in the Viewer.
2. Right-click on a spline or point and select split off [view], for example [split off left].
Any changes you now make to the stroke/shape points are only applied to the view you chose to
split off and are displaying in the Viewer.
Views that have not been split off can still be edited together. For example, if you have views
called left, middle, and right, and choose to split off the left view, any changes to the middle view
are also applied to the right view and vice versa.
3. To unsplit a view, right-click on a spline or point again and select unsplit [view].
The view is unsplit, and all changes you made after splitting it off are lost.
Note: To see splines for paint strokes in the Viewer, you need to activate show paint
stroke splines in the RotoPaint tool settings.
Tip: You can also split shapes off by right-clicking on them in the stroke/shape list and
selecting Split off [view], for example Split off left.
To Edit Stroke/Shape Attributes

1. Display the view you want to edit in the Viewer.
2. In the RotoPaint controls, click the Views menu button next to the control you want to edit
and select Split off [view name]. For example, to edit opacity in a view called left, select Split off
left from the Views dropdown menu next to opacity.
Any changes you now make to opacity are only applied to the view you chose to split off and are
displaying in the Viewer.
USER GUIDE
609
Editing Strokes/Shapes in One View Only | To Edit Stroke/Shape Attributes
Views that have not been split off can still be edited together. For example, if you have views
called left, middle, and right, and choose to split off the left view, any changes to the middle view
are also applied to the right view and vice versa.
You can also click on the small arrow on the left side of the control to see the values for each
view.
3. To unsplit a view, click the Views menu button and select Unsplit [view]. For example, to
unsplit a view called left, you’d select Unsplit left.
USER GUIDE
610
Applying a Stereo Offset | To Edit Stroke/Shape Attributes
Applying a Stereo Offset

The stereo offset control in the RotoPaint properties allows you to move the selected stroke/shape
on the x and y axes. This is an extra transform that is applied after all other transforms. Typically, you
would position the stroke/shape correctly in the hero view, then split stereo offset, and drag the
stroke/shape to its correct location in any other views. Note that you can also press Shift while
dragging to constrain the movement to the x or y axis only.
Stereo offset can be useful, for example, if you have a stroke/shape that is correctly positioned in
one view and you want to move it to its correct location in another view, but can’t use the translate
control on the Transform tab because that’s being driven by Tracker data.
USER GUIDE
611
Where Are the Bezier and Paint Nodes? | To Edit Stroke/Shape Attributes
Where Are the Bezier and Paint

Nodes?
The pre-6.0 Bezier and Paint nodes have been deprecated in favor of the RotoPaint node. With
RotoPaint you have the ability to add more strokes and shapes, group them, and so on. However,
Bezier and Paint are still in the application for backwards-compatibility with old scripts. Should you
find the need (or just feel nostalgic), you can create the Paint and Bezier nodes in a couple of easy
ways:
• Press X on the Node Graph, make sure Tcl is checked in the dialog that opens, enter Paint or
Bezier in the Command field, and click OK. Your node appears in the Node Graph.
• You can add a Paint or a Bezier node to your Toolbar menu with a statement in your menu.py file
like the following:
• For Bezier:
tb = nuke.toolbar("Nodes")
tb.addCommand("Draw/Bezier", "nuke.createNode(\"Bezier\")",
icon="Bezier.png")
• For Paint:
tb = nuke.toolbar("Nodes")
tb.addCommand("Draw/Paint", "nuke.createNode(\"Paint\")",
icon="Paint.png")
USER GUIDE
612
Tracking and Stabilizing
Nuke’s 2D Tracker allows you to extract animation data from the position, rotation, and size of a
pattern. You can then apply the data directly to transform or match-move another element. Or you
can invert the data values and apply them to the original element to stabilize the image.
Before Tracking
Before you track, it’s important to play back the image several times. This will help you identify the
best features for the process, as well as any problems with motion blur, occlusions, or features
moving out of frame.
For some images, you may need to apply filters or color-correction to boost the visibility of features
before you attempt to track them. Because of the procedural nature of Nuke, you can disable these
extra nodes after you get a successful track, or simply reconnect the Tracker node at the appropriate
place to apply the transform.
Quick Start
1. Connect a Tracker node to the image you want to track. See Connecting the Tracker Node.
2. Place the required number of track anchors on features in the image. See Adding Track Anchors.
3. Use Automatic Tracking for simple tracks or, for greater accuracy or difficult tracks, use Keyframe
Tracking.
4. Apply your tracking data depending on requirements. See Applying Tracking Data.
USER GUIDE
613
Connecting the Tracker Node |
Connecting the Tracker Node

To connect the Tracker node:
1. Read in and select the clip you want to track.
2. Click Transform > Tracker.
3. Click Image > Viewer to insert a Viewer node and connect it to the Tracker node.
USER GUIDE
614
Adding Track Anchors | Positioning Track Anchors
Adding Track Anchors

You can add as many track anchors as required, depending on which transformational components
you wish to track. For example, when tracking in areas of distortion or noise, it's a good idea to add a
lot of tracking anchors and then average the results to get a better overall track.
1. Enable the fast-add button and click in the Viewer to add tracking anchors, or click add
track on the Tracker tab in the properties panel to create the required number of anchors.
Note: Holding Ctrl/Cmd+Alt and clicking in the Viewer also enables fast-add mode.
New anchors always appear in the center of the current Viewer. You’ll also notice an anchor zoom
window in the top left of the Viewer. This allows you to accurately position tracking anchors
without zooming the entire Viewer.
2. Temporarily disable tracks using the checkbox on the left of the panel or remove tracks by
selecting them in the Tracks list and clicking delete tracks.
Positioning Track Anchors

Each track anchor consists of a pattern and search area:
• Pattern - the target pixels tracked across multiple frames. This pattern should be as distinct as
possible from the surrounding frame and remain visible throughout the majority of the sequence.
• Search - the area of the current frame in which the pattern is likely to be found in the next frame.
Large search areas can affect performance.
USER GUIDE
615
1. Drag the crosshair in the center of the anchor over the pattern to be tracked or manually adjust
the track_x and track_y fields in the Tracks list.
In the example, the corner of a building is directly under the crosshair.
Note: You can enable Settings > snap to markers to assist you when placing anchors.
Note: Whether you select blobs or corners, a guide displays in the Viewer as you drag the
anchor, highlighting likely patterns.
Release the mouse to snap the anchor to the marker.
2. Adjust the pattern and search areas as required using the following methods:
• Click and drag the boundaries to adjust on a single axis. For example, dragging a vertical
boundary adjusts the area on the x axis.
• Click and drag the corners of the boundaries to adjust on both axes. Hold Shift, click and drag
to scale the area.
Note: Hold Cmd/Ctrl to adjust boundaries asymmetrically, that is, irrespective of the anchor
position.
USER GUIDE
616
3. Reset the pattern and search areas by clicking in the Viewer tools.
USER GUIDE
617
Tracking Preferences and Viewer Tools | Positioning Track Anchors
Tracking Preferences and Viewer

Tools
Before tracking, use the Tracker’s properties panel and Viewer tools to control Viewer output and
determine tracking behavior.
1. Click the Settings tab to display the General controls.
2. Select the channels in the image to track using the track channels dropdown.
If this is set to anything other than all or none, use the checkboxes to select the required
channels.
3. Select the pre-track filter type to apply before pattern comparison using the dropdown:
• none - no filter is applied.
• adjust contrast - the default filter, stretches the image contrast to better suit the tracking
algorithm. This is the recommended setting and shouldn’t need changing in most
circumstances.
• median - helps to remove noise from the pattern.
4. Check adjust for luminance changes to force Tracker to calculate some extra pre-filtering to
compensate for changes in brightness. This option slows the tracking process and can reduce the
accuracy of tracks, so only enable this control if there are known changes in brightness.
Note: Enabling adjust for luminance changes can occasionally produce better tracks on
shots with no differences in luminance, particularly on jittery shots where sub-pixel accuracy
is vitally important.
5. Clamp super-white, sub-zero footage is enabled by default, restricting the pixel values within
the pattern between 0-1.
If you want to track the sequence using its full dynamic range, disable this control and increase
max_error to a value large enough to cover the range. For example, a sequence with pixel values
of 40 might require a similar max_error value.
6. Enable show error on track paths to color code tracks showing their pattern matching error
values, green for a good match through to red for a poor match.
Note: The Viewer toolbar button also toggles this control on and off.
7. Enable hide progress bar to free up Viewer real-estate while tracking.
USER GUIDE
618
Tracking Preferences and Viewer Tools | Positioning Track Anchors
8. Enable snap to markers to assist you when placing tracking anchors. See Positioning Track
Anchors for more information.
9. Set when the zoom window is displayed using the show zoom window dropdown:
• always - the window is always displayed.
• on track move - the window is only displayed when a track anchor is moved.
• when tracking - the window is only displayed during tracking.
• when tracking or track change - the window is displayed during tracking and when a tracking
anchor is moved.
• never - the window is never displayed.
10. Set the zoom widget’s size and magnification using the zoom size/mag. dropdowns. You can
select custom and enter values manually for greater flexibility.
11. A filter is applied to the zoom window on playback by default, but you can enable the filter
permanently, or disable it, using the zoom window filter dropdown.
Note: The filter applied is the same as that selected on the Transform tab, and can produce
a more visually stable track. It can make track positioning more difficult, however.
12. Proceed to Automatic vs. Keyframe Tracking to determine which tracking method suits your
needs.
USER GUIDE
619
Automatic vs. Keyframe Tracking | Positioning Track Anchors
Automatic vs. Keyframe Tracking

After placing track anchors and setting your preferences, you’re ready to calculate your tracks. Nuke’s
Tracker provides two calculation methods:
• Automatic Tracking - ideal for simple tracks, there are no extra preparation steps once you’ve set
your tracking anchors and preferences.
• Keyframe Tracking - a more involved method, requiring you to set keyframes on the sequence in
order to calculate tracks. Keyframe tracking may be the better option for more complicated patterns
and movement.
USER GUIDE
620
Automatic Tracking | Calculating Auto-Tracks
Automatic Tracking
Calculating tracks automatically uses the tools above the Viewer to control the direction and range of
frames to track. Tracking backwards can produce a better track than going forwards if the pattern is
clearly visible later in the clip. By default, Auto-Tracking grabs the reference pattern on the first frame
of the sequence, from within the pattern anchor, using this pattern throughout the track for
comparison between frames.
To help avoid clutter in the Viewer, you can enable or disable the Tracker overlay by right-clicking in
the Viewer and selecting Overlay, or by pressing Q to toggle between the available states:
• overlay off
• overlay on
• overlay on, no animation path
Calculating Auto-Tracks
1. In the Tracker properties panel, select each track you wish to calculate or click select all to use all
tracks.
2. For each track, select the type of movement the track is intended to output: translation, rotation,
or scaling. For example, tracking a feature across the sequence and toward the camera may
involve translation and scale.
Note: These controls deal with output from tracking data using the Transform controls and
are not the same as the Settings > warp type control, which deals with pattern recognition.
3. Using the tool bar above the Viewer, click either the frame backward (X) or forward (C) buttons
to move to the previous or next frame. Move through a few frames in this manner to
ensure that all enabled track anchors are “sticking” to their patterns.
If a particular track anchor doesn’t stick, experiment with a different position.
4. Once all track anchors stick, click the track backward (Z) or track forward (V) buttons to
analyze the whole sequence.
5. To track only a certain frame range, use the range buttons to enter the required frames.
6. Click stop (Esc) , to cease tracking in either direction.
USER GUIDE
621
Automatic Tracking | Troubleshooting Auto-Tracks
Tip: When calculating multiple tracks simultaneously, you may find that some tracks stick
accurately to the pattern, while others require resetting and reanalysis. When you’re happy
with a given track, deselect it in the Tracks list. This protects it from recalculation, and lets
you experiment with better placement for the wayward tracks.
See Troubleshooting Auto-Tracks for help with troublesome tracks.
Troubleshooting Auto-Tracks
No matter how sophisticated tracking becomes, some sequences are inevitably going to cause
problems. There are a number of pre-tracking checks you can perform to assist Auto-Tracking:
• Play through the sequence before placing your tracking anchors
• Look for features that are consistent throughout the majority of the sequence
• Avoid occluded features where possible - see Dealing with Occlusions.
You can also adjust the way patterns are tracked and how often they are re-sampled, or grabbed,
using the Settings tab and Auto-Tracking controls:
1. Try adjusting the max iterations, epsilon, and max_error controls to improve the track accuracy:
• max iterations - the maximum number of iterations before the tracking algorithm stops
searching for features.
• epsilon/resolution - the error level at which Tracker is assumed to have found the feature - no
further search for a better match is performed. Higher values may result in a faster but less
accurate track.
• max_error - the error level at which Tracker stops searching for features.
2. In the Auto-Tracking sub-menu, enable predict track to use the animation path to predict where
the pattern may appear in the next frame.
Note: If the track fails when prediction is enabled, click the Clear Forward button, or
re-tracking follows the same erroneous path.
3. Set the type of movement Tracker should expect in the pattern using the warp type dropdown:
• Translate
• Translate/Rotate
• Translate/Scale
• Translate/Rotate/Scale
• Affine
USER GUIDE
622
Automatic Tracking | Dealing with Occlusions
Translation only is the easiest to calculate, but can lose the pattern if it rotates or scales. Affine
can be used to attempt to preserve straight lines and relative distances, compensating for
sheering.
4. Try adjusting the pattern grabbing behavior, when or how often a new pattern should be grabbed
from the sequence:
• on first frame - the comparison pattern is grabbed on the first frame from with the pattern
anchor. You might select this option if the feature translates but doesn’t rotate, scale, or sheer.
• every frame - the comparison pattern is grabbed at every frame in the sequence. This option
takes longer to process, but can produce a smoother track.
• every n frames - allows you to set the frame interval between pattern grabs using the every n
frames control.
• if error above - the comparison pattern is grabbed when the error value is greater than that
specified by the when error > control. Setting this control to a low value grabs the pattern more
often.
• if error below - the comparison pattern is grabbed when the error value is less than that
specified by the when error < control.
• custom - this option enables all the pattern grab behavior controls, allowing you to fine-tune
when the comparison pattern in grabbed through out the sequence.
5. Enable when tracking is stopped to cause Tracker to re-grab the pattern at the current frame
when tracking stops.
6. Enable when tracker is moved to cause Tracker to re-grab the pattern at the current frame when
a tracking anchor is moved.
Dealing with Occlusions

Tracker’s offset capability allows you track an obscured feature using the relative position of another
feature, providing that the distance between the two points remains constant.
1. Track the pattern normally until the occlusion causes Tracker to fail.
The zoom window helps to identify the failure point.
USER GUIDE
623
Automatic Tracking | Dealing with Occlusions
2. Play though the sequence to identify a likely offset point - a pattern that remains equidistant from
the original pattern grab.
3. Hold down Ctrl/Cmd and drag the tracking anchor to the offset position.
The offset amount is recorded in the Tracks list and highlighted in yellow in the Viewer.
4. Continue tracking as normal by clicking the track backward (Z) or forward (V) button.
Tracker combines the two tracks into a single continuous track.
5. Use the clear backward and forward buttons to clear poor keyframes. Click clear all to
remove all selected tracks and keyframes, excluding manually placed keyframes.
Note: You can reset tracking anchor pattern and search areas by clicking .
USER GUIDE
624
Keyframe Tracking | Calculating Keyframe Tracks
Keyframe Tracking
Calculating tracks using keyframes can be the better option for more complex patterns and
movement. It requires a little more work to set up, but can produce more reliable tracks.
Unlike auto-tracks, keyframe tracking compares the current pattern anchor to the patterns of the two
nearest keyframes.
To help avoid clutter in the Viewer, you can enable or disable the Tracker overlay by right-clicking in
the Viewer and selecting Overlay, or by pressing Q to toggle between the available states:
• overlay off
• overlay on
• overlay on, no animation path
Calculating Keyframe Tracks

To calculate keyframe tracks:
1. In the Tracker properties panel, select each track you wish to calculate in the Tracks list or click
select all.
2. For each track, select the type of movement the track is intended to output: translation, rotation,
or scaling. For example, tracking a feature across the sequence and toward the camera may
involve translation and scale.
3. Scrub through the sequence a few frames and adjust the position of the tracking anchor by
dragging the anchor to the location of the pattern. You can use the zoom window to fine-tune
your positioning. Continue on through the sequence as required.
USER GUIDE
625
Keyframe Tracking | Importing Tracking Data
Tip: You can change the magnification of zoom windows by holding Shift and dragging the
magnifying glass cursor away from the center of the window.
At each frame, a new keyframe window is added to the right of the zoom window. The keyframe
closest to the current playhead frame is highlighted in orange.
It’s a good idea to place more keyframes around areas of complexity or greater movement and
fewer on straight forward translation. Generally speaking, a greater number of keyframes
produces a better track, but at the expense of processing time.
4. When you’re satisfied with your keyframes, make sure all your tracks are selected in the Tracks
list and then click to track all keyframes.
Tip: When calculating multiple tracks simultaneously, you may find that some tracks stick
accurately to the pattern, while others require resetting and reanalysis. When you’re happy
with a given track, deselect it in the Tracks list. This protects it from recalculation, and lets
you experiment with better placement for the wayward tracks.
5. You can also force the selected tracks to recalculate between the two nearest keyframes by
clicking in the Viewer toolbar.
See Troubleshooting Keyframe Tracks for help with troublesome tracks.
Importing Tracking Data

You can import tracking data from third-party software using a plain text file containing three values
in each line, frame, x, and y coordinates. Tracker can read files with values separated by spaces,
commas, or colons and any blank lines or lines starting with #, ;, and / are ignored. For example,
importing a file containing:
# Tracking Data
1,1,1
10,240,240
USER GUIDE
626
Keyframe Tracking | Importing Tracking Data
50,1000,700
90,1800,1200
100,2048,1556
Produces five keyframes at frames 1, 10, 50, 90, an 100 with the relevant x,y coordinates.
Tip: If you use a particular file format on a regular basis, you might want to create you own
importer to parse the .txt file. You can use the <install_dir>/plugins/import_discreet.tcl file
as a guide on how to do this.
To import tracking data:

1. Add a tracking anchor as described in Adding Track Anchors.
This adds a track to the tracks table in the Properties panel.
2. Right-click the new track and select File > Import Time+value Ascii.
The Import discreet dialog displays.
3. Enter the file path and file name in the File field.
4. Set the required column containing the x and y coordinate data.
5. Click OK to import the track.
The points from the file are converted into keyframes and displayed in the Viewer. The track
between keyframes is interpolated as normal. The five keyframe example file described earlier
produces a track similar to the following image.
USER GUIDE
627
Keyframe Tracking | Troubleshooting Keyframe Tracks
Troubleshooting Keyframe Tracks

Again, even with preset keyframes, some sequences are inevitably going to cause problems. There are
a number of pre-tracking checks you can perform to assist Auto-Tracking:
• Play through the sequence before placing your tracking anchors,
• Look for features that are consistent throughout the majority of the sequence,
• Avoid occluded features where possible - see Dealing with Occlusions.
Keyframe tracking won’t generally stop when a problem is encountered. Tracker attempts to continue
using the next keyframe as a reference, which is why placing a lot of keyframes around problem
areas is recommended.
Tip: Tracking areas of distortion or noise can produce unreliable results due to the
movement of the pixels in the pattern matching box. One way to deal with this is to seed
multiple tracks in and around the problem area and then average the resulting tracks
together, producing a single more reliable track, by clicking average tracks in the
properties panel.
1. First, turn on the color-coded error indicator by clicking the traffic light Viewer tool.
Each keyframe is colored on a sliding scale from green (good match) to red (poor match).
USER GUIDE
628
Keyframe Tracking | Dealing with Occlusions
Bear in mind that a red keyframe doesn’t necessarily mean that the tracking result is poor, only
that Tracker couldn’t reliably match the pattern from one keyframe to the next.
2. Move the tracking anchor to the first of the poor frames, just about the center of the image in the
example.
3. Tracker defaults to adding and deleting keyframes automatically when certain conditions are met,
but you can toggle these features on and off in the Properties Tracker > Settings tab Keyframe
Tracking controls:
• re-track when keyframe is moved - disable this control if you plan to manually position
multiple keyframes before re-tracking.
• re-track on creation of a new keyframe - disable this control when placing multiple new
keyframes, such as when the track encounters problem areas.
• create new key when track is moved - you could disable this control if you wanted to use the
zoom window to examine the sequence more closely without triggering a re-track.
• auto-tracks delete keyframes - when this control is enabled, custom keyframes are deleted
during automatic re-tracking.
4. Using the zoom window, drag the anchor to the correct location of the grabbed pattern.
Tracker attempts to recalculate the track by including your correction.
5. Advance the playhead to the next poor keyframe and repeat until the track is complete.
Dealing with Occlusions

Tracker’s offset capability also applies to keyframe tracking, allowing you to track an obscured feature
using the relative position of another feature, providing that the distance between the two points
remains constant.
1. Place keyframes normally until you reach the occlusion.
USER GUIDE
629
Keyframe Tracking | Dealing with Occlusions
2. Play though the sequence to identify a likely offset point - a pattern that remains equidistant from
the last keyframe.
3. Hold down Ctrl/Cmd and drag the tracking anchor to the offset position.
The offset amount is recorded in the Tracks list and highlighted in yellow in the Viewer.
4. Continue tracking as normal by clicking the track backward (Z) or forward (V) button.
Tracker combines the two tracks into a single continuous track.
5. Use the clear backward and forward buttons to clear poor keyframes. Click clear all to
remove all selected tracks and keyframes, excluding manually placed keyframes.
Note: You can reset tracking anchor pattern and search areas by clicking .
USER GUIDE
630
Applying Tracking Data | Applying Tracking Data Using Tracker’s Controls
Applying Tracking Data

You can apply your tracking data to the input image using either the Tracker node’s controls, linking
expressions, or other Nuke nodes.
Applying Tracking Data Using Tracker’s Controls

The simplest way to apply tracking data to the input image or other nodes is to use the controls of the
Tracker node itself. Here, we look at using these controls to stabilize, match-move, and remove or
apply jitter.
Stabilizing Elements
The Tracker node’s controls let you remove motion, such as unwanted camera shake, from the node’s
input clip. You can use a single track to stabilize horizontal and vertical motion across the 2D plane,
or two or more tracks to eliminate rotation and scale.
1. Create the track(s) you want to use for stabilizing the footage:
• If you’re using a single track, ensure T is selected in the Tracks list, so that Tracker only calculates
translation.
On the Transform tab, select transform > stabilize 1-pt to lock the Filter to Keys. This filter
produces the best results when using a single track.
• If you’re using more than one track in the Tracks list, select the transformations that you want
Tracker to calculate when stabilizing the image, Translation, Rotation, and/or Scale.
On the Transform tab, select transform > stabilize.
2. Set the reference frame if you don’t want to use the first frame as the transform control frame.
3. Use the smooth controls to average frames together to smooth the transforms applied.
For example, if you’re stabilizing using more than one track, you might average frames together
for translation and rotation by entering the number of frames in the t and r fields.
4. Select the required filter. See Choosing a Filtering Algorithm for more information.
Nuke stabilizes the footage, locking its elements to the same position within the composite.
Note: You can export the transform information to a linked or baked Transform node by
selecting the required type from the Export dropdown and clicking create.
USER GUIDE
631
Match-moving Elements
You can use the Tracker node’s controls to apply the tracked motion to another image, that is, to
match-move an image.
1. Use a Tracker node to track the feature you intend to match.
2. Copy the Tracker node and paste it after the footage you want to match-move.
3. In the second Tracker node’s controls, go to the Transform tab.
4. From the transform dropdown menu, select match-move.
For example, if you’re using more than one track, you might average frames together for
translation and rotation by entering the number of frames in the t and r fields.
Nuke applies the tracked movement to the footage you want to match-move. A simple script
might appear as follows, where Tracker2 is a copy of Tracker1:
A simple match-move script.
Note: You can export the transform information to a linked or baked Transform node by
selecting the required type from the Export dropdown and clicking create.
USER GUIDE
632
Removing or Adding Jitter

Tracker can be used to remove high frequency jitter from a sequence, exaggerate existing jitter, or
add jitter to another sequence for a consistent look.
To remove jitter:
1. Create the tracks you want to use for jitter removal.
2. In the Tracker node’s controls, go to the Transform tab.
3. From the transform dropdown menu, select remove jitter.
5. Use the jitter period to average together frames, adjusting the jitter to achieve the required
stability.
For example, if you’re removing jitter using more than one track, you might average frames
together for translation and rotation by entering the number of frames in the t and r fields.
8. Nuke removes jitter from the footage, locking its elements to the same position within the
composite.
To exaggerate or add jitter:

1. Create the tracks you want to use for the required jitter operation.
2. In the Tracker node’s controls, go to the Transform tab.
3. From the transform dropdown menu, select add jitter.
5. Use the jitter period to average together frames, adjusting the jitter to achieve the required
instability.
For example, if you’re adding jitter using more than one track, you might average frames together
for translation and rotation by entering the number of frames in the t and r fields.
Nuke exaggerates the tracked jitter in the footage.
Note: If you want to transfer jitter to another sequence, copy the Tracker node and paste it
after the footage to which you want to add jitter, then follow steps 3-7 above.
USER GUIDE
633
Applying Tracking Data | Applying Tracking Data Using Linking Expressions
Applying Tracking Data Using Linking Expressions

Nuke’s CornerPin2D and Stabilize2D nodes are specifically designed to receive tracking data by linking
expressions, but you can apply tracking data in this manner to virtually any Nuke node. For example,
you might animate a Bezier or a B-spline shape with tracking data by entering linking expressions
into the RotoPaint node’s transformation parameters. You can also apply tracking data to individual
points.
This section explains the basic procedure for applying tracking data to any node via linking
expressions.
Creating Linking Expressions

The Tracker node’s Tracker panel displays data related to the position of each track anchor over time
(tracks’ x and y fields). This is the data that you most typically apply to other nodes.
To Drag and Drop Tracking Data:

1. Display both the Tracker parameters and the parameters to which you wish to apply the tracking
data (the destination control - for example, a RotoPaint node’s translate control).
2. Select the required track from the Tracks list. Only one track can be linked per control.
3. Ctrl+drag (Cmd+drag on a Mac) from the source control animation button to the destination
control animation button.
When you release, the destination control will turn blue, indicating an expression has been applied.
In this case, the drag and drop action has created a linking expression resembling the following
example:
Tracker1.tracks.1.track_x
Tip: You can also apply tracking (or other transform) data to individual RotoPaint,
SplineWarp, or GridWarp points (this is sometimes called per vertex tracking). To do so,
Ctrl/Cmd+drag and drop the track’s animation button on a RotoPaint, SplineWarp or
GridWarp point in the Viewer.
You can add other components to this linking expression as necessary. For example, you might add a
spatial offset to the linking expression by subtracting out the initial frame’s tracking values, in which
case the final expression would resemble the following:
Tracker1.tracks.1.track_x-Tracker1.tracks.1.track_x(1)
USER GUIDE
634
Applying Tracking Data | Transforming Masks with Tracking Data
See Expressions for more information. Once you enter the linking expression, the destination
parameter turns blue.
To Link Animated Parameters with a Tracker Node

You can also link controls with the Tracker node if you use the Link to option in the Animation menu.
For example, to link the translate control of the Roto node with a Tracker node, do the following:
1. Create the Tracker node you want to link to.
2. Go to Tracker’s Transform tab and enable live-link transform.
This control enables Tracker to update dynamically as the expression link changes.
3. On the Transform tab of the Roto node’s properties panel, click on the translate animation
menu.
4. Select Link to > Tracker linking dialog.
5. Select the Tracker node you want to use in the tracker node dropdown and in the link to
dropdown, select whether you want to link to the position of the track, the translate values of the
track, or treat the translate value as an offset.
6. Select which tracks you want to use by checking the track boxes. The Expression field updates
with the appropriate expression syntax. If you select more than one track, the tracks are averaged,
for example:
(Tracker1.tracks.1.track_x + Tracker1.tracks.2.track_x)/2
7. Then click OK, and your linking is done.
Your Bezier shape’s translate value now changes when the Tracker value is changed.
Transforming Masks with Tracking Data

Creating animated masks using Roto and keyframes can be a very time-consuming process, but
Nuke's Tracker node can do some of the initial work for you, especially with garbage mattes.
Once you have some solid tracking data, you can drive a roto shape without keyframing individual
points:
1. Track a feature in the area you intend mask. In the example, the figure's head serves as the
driving point for the matte.
USER GUIDE
635
Applying Tracking Data | Transforming Masks with Tracking Data
2. Add a Roto node to the script (keyboard shortcut O) and draw the shape you intend to drive with
the tracking data. In this case, we don't need to be too accurate as we're creating a garbage matte.
3. In the Roto properties panel, click the Transform tab and select the matte in the shapes list.
4. Right-click the translate control's animation icon and then select Link to > Tracker 1 > track
1.
The Tracker's track_z and track_y keyframes are copied into the Roto's translate control,
applying the same translate and offset to the matte shape.
5. To compensate for this, select the Root item in the shape list and reposition the Roto shape
correctly using the transform handle in the Viewer.
USER GUIDE
636
Applying Tracking Data | Using the CornerPin2D Node
The matte with offset. The same matte after repositioning.

6. Scrub the playhead to see the matte following the tracked path.
Using the CornerPin2D Node

The CornerPin2D node is designed to map the four corners of an image sequence to positions
derived from tracking data. In practice, this node lets you replace any four-cornered feature with
another image sequence. For example, suppose you needed to replace the monitor image in the fast-
panning shot shown below.
Fast-panning shot requires four-corner

tracking.
First, use the Tracker to calculate four separate tracks, one for each corner of the feature.
Generating the four tracks.
USER GUIDE
637
Next, attach a CornerPin2D node to the image sequence you want to use as the replacement for the
feature, and apply the tracking data. This remaps the image sequence’s corners to the correct
positions over time. You can create the node manually, or by using Tracker’s Export dropdown.
The final step is to layer the result over the original element.
The composited image.
The steps below summarize the use of Tracker’s Export CornerPin2D workflow.
To Use the CornerPin2D Node

1. Generate four tracks, one per corner, on the feature requiring replacement.
2. Use the current frame or reference frame field to specify the frame to use as the starting point.
You can also decide whether the CornerPin2D node is expression linked or baked using the
Export dropdown:
• CornerPin2D (use current frame)
• CornerPin2D (use transform ref frame)
• CornerPin2D (use current frame, baked)
• CornerPin2D (use transform ref frame, baked)
3. Click create to add the CornerPin2D node to the script.
4. Attach the image or sequence to replace the feature tracked to the input of the CornerPin2D
node.
5. If necessary, select a different filtering algorithm from the filter dropdown menu. See Choosing a
Filtering Algorithm for more information.
6. When filtering with Keys, Simon, or Rifmen filters, you may see a haloing effect caused by the
pixel sharpening these filters employ. If necessary, check clamp to correct this problem.
7. In most cases, you will keep black outside checked. This renders black pixels outside the image
boundary, making it easier to layer the element over another. (If you uncheck this parameter, the
outside area is filled with the outermost pixels of the image sequence.)
8. The final step is to layer the result over the original element.
A simple script might appear as follows:
USER GUIDE
638
USER GUIDE
639
Transforming Elements
These pages explain how to perform a range of 2D and 2.5D spatial transformations. You learn how
to apply geometric transformations (including translations, rotations, scales, and skews) to elements,
and how to add motion blur using the nodes in the Transform menu.
USER GUIDE
640
Transforming in 2D | Using the 2D Transformation Overlay
Transforming in 2D
This section describes how to apply 2D transformations including translations, rotations, scales, and
skews to elements using a number of Nuke nodes. All transforms, except translations, change the
size of the image bounding box.
Using the 2D Transformation Overlay

Several of the nodes discussed in this section display a Viewer overlay for executing spatial
transformations. This overlay is often a faster alternative to the properties panel. The figure below
shows you how to use Nuke 2D transformation overlay.
Transformation overlay.
• A) Drag to skew the frame (see Skewing Elements).
• B) Drag to scale the frame uniformly - simultaneously on x and y (see Scaling Elements).
• C) Drag to translate the frame (seeTranslating Elements).
Shift+drag to constrain the translation to x or y.
Ctrl/Cmd+drag to reposition the pivot point (the point that acts as the center to transformation
operations).
• D) Drag to scale the frame on x.
USER GUIDE
641
Transforming in 2D | Translating Elements
• E) Drag to rotate the frame around the pivot point (see Rotating Elements). The transform overlay
snaps to typical values. To prevent the snapping, press Shift while dragging.
• F) Drag to scale the frame on y.
Translating Elements
To translate an element is to slide it on x or y.
You can use the Transform, TransformMasked, or Position nodes to translate elements.
Using the Transform Node

The Transform and TransformMasked nodes let you not only translate elements, but also rotate,
scale, and skew them from a single properties panel.
TransformMasked is identical to Transform except that it offers controls for assigning a mask to
protect certain areas of the frame from translations. For the sake of brevity, this chapter only
discusses the use of Transform, but keep in mind you can use TransformMasked any time you need
to process a transformation through a mask. The mask controls work in the same fashion as those
described in Masking Color Corrections.
To translate an element using the Transform node:

1. Click Transform > Transform to insert a Transform node at the appropriate place in your script.
2. Connect a Viewer to the output of the Transform node so you can see the effect of your changes.
3. In the Transform properties panel, increment or decrement the translate x and y fields to slide
the element along either axis.
Or drag on the center of the transformation overlay.
USER GUIDE
642
Transforming in 2D | Rotating Elements
Using the Position Node

The Position node gives you just bare-bones parameters for translating an element.
To translate an element using the Position node:

1. Click Transform > Position to insert a Position node at the appropriate place in your script.
2. Connect a Viewer to the output of the Position node so you can see the effect of your changes.
3. In the Position properties panel, increment or decrement the translate x and y fields to slide the
element along either axis.
Rotating Elements
To rotate an element is to spin it around the pivot point.
To rotate an element using the Transform node

3. In the Transform properties panel, select the appropriate filtering algorithm from the filter
dropdown menu (see Choosing a Filtering Algorithm).
4. Position the pivot point as necessary:
• Increment or decrement the center x and y fields to move the axis in either direction.
• Or press Ctrl (Cmd on a Mac) while dragging on the center of the transformation overlay.
5. Increment or decrement the rotate field.
Or drag on the horizontal bar of the transformation overlay.
USER GUIDE
643
Transforming in 2D | Scaling Elements
Scaling Elements
To scale an element is to resize it by adding (upsampling) or removing (downsampling) pixels.
Nuke offers several nodes for scaling elements. Transform is designed primarily for scaling the
background plate up or down in a composite. The scaling functions for transform are described
below.
Reformat is designed for writing out elements with specific resolutions and pixel aspect ratios.
Adding Motion Blur describes the use of this node.
To scale an element using the Transform node

• Or press Ctrl (Cmd on a Mac) while dragging on the center of the transformation overlay.
5. To scale the frame uniformly (on both x and y):
• Increment or decrement the Transform node’s scale field.
• Or drag the circle-portion of the transformation overlay.
6. To scale the frame asymmetrically (on x or y):
• Click scale parameter’s channel chooser to reveal the x and y fields, then increment or
decrement each individually.
USER GUIDE
644
Transforming in 2D | Skewing Elements
• Or drag any of the four points on the circle-portion of the transformation overlay. The top and
bottom points scale on y; the left and right points, on x.
Skewing Elements
To skew an element is to rotate its pixel columns around the pivot point.
To skew an element using the Transform node

• Or Ctrl+drag (Cmd+drag on a Mac) on the center of the transformation overlay.
5. Increment or decrement the skew field to rotate the pixel columns around the pivot point.
Or drag the vertical bar of the transformation overlay.
To Invert a Transform Effect

You can invert the effect you’ve created with the Transform node by checking the invert box in the
Transform properties panel. This uses the inverse values of your translate, rotate, scale and skew
values. When the box is checked, a small transform handle appears next to the standard transform
handle in the Viewer.
USER GUIDE
645
Transforming in 2D | Bounding Box Warnings

nodes that affect the bounding box. To enable the warnings, in Nuke's Preferences under Panels >
Node Graph, enable Bounding Box Warning:
format.
USER GUIDE
646
Transforming in 2D | Bounding Box Warnings
USER GUIDE
647
Choosing a Filtering Algorithm | Bounding Box Warnings
Choosing a Filtering Algorithm

Spatial transformations involve remapping pixels from their original positions to new positions. The
question arises as to what values to assign remapped pixels. In the simplest case, they retain their
original values, but this can create problems with image quality, particularly in high contrast areas of
the frame. For example, the figure below shows a close up of a high-contrast feature that has been
rotated clockwise by 45 degrees. The remapped pixels have retained their original values, but the
result is a highly aliased, or jagged, edge:
The solution is to apply a more sophisticated filtering algorithm to determine the values of remapped
pixels - one that takes into account, in some fashion, the values of neighboring pixels.
For example, applying Nuke’s cubic algorithm to the above rotation, results in a softer, less jagged
edge:
When executing spatial transformations, Nuke lets you select from the filtering algorithms described
in the table below.
Note that the curves shown in the table plot the manner by which each algorithm samples from
neighboring pixels. The center of each curve represents the value of the remapped pixel itself, and
the rising and falling portions of each curve represent the amount of sampling that occurs across a
five pixel radius.
Tip: When using filters that employ sharpening, such as Rifman and Lanczos, you may see
a haloing effect. Some nodes, such as Transform and Tracker, include a clamp control to
correct this problem.
USER GUIDE
648
Filter Description Sampling Curve and Output
Impulse Remapped pixels carry original values.
Cubic Remapped pixels receive some smoothing.

(default)
Keys Remapped pixels receive some smoothing,

plus minor sharpening (as shown by the
negative -y portions of the curve).
USER GUIDE
649
Simon Remapped pixels receive some smoothing,

plus medium sharpening (as shown by the
Rifman Remapped pixels receive some smoothing,

plus significant sharpening (as shown by the
Mitchell Remapped pixels receive some smoothing,

plus blurring to hide pixelation.
USER GUIDE
650
Parzen Remapped pixels receive the greatest

smoothing of all filters.
Notch Remapped pixels receive flat smoothing

(which tends to hide moiré patterns).
Lanczos4 Remapped pixels receive minor sharpening

(as shown by the negative -y portions of the
curve), good for scaling down.
The filter number at the end of the filter

name denotes the width of the filter (4
pixels).
USER GUIDE
651
Lanczos6 Remapped pixels receive some sharpening


pixels).
Sinc4 Remapped pixels receive a lot of sharpening


pixels).
Note: Lanczos and Sinc filters exhibit haloing as the filter becomes sharper.
Note: The numbers appended to the filter name denote the width of the filter, similar to the
naming scheme used by PRMan. See http://www.renderman.org/RMR/st/PRMan_
Filtering/Filtering_In_PRMan.html for more information.
USER GUIDE
652
How Nodes Concatenate | Bounding Box Warnings
How Nodes Concatenate

Concatenation is behavior that some Nuke nodes perform when you have several nodes transforming
your image one after another. When nodes concatenate, they pass on these adjacent transformation
operations to the last transforming node in the row, and the last node then performs all the
transformations at once.
For example, if you have three Transform nodes in a row, instead of performing each transformation
separately and filtering the input image three times, Nuke combines the transformations into one
operation and only filters the image once. As filtering unavoidably destroys some image detail, it’s
important to concatenate as many nodes as possible in order to preserve image quality.
Three Transform nodes A Crop node between the

that concatenate. The image Transform nodes breaks
is only filtered once. concatenation. The image
is filtered twice, resulting
in poor image quality.
USER GUIDE
653
In order to concatenate, the concatenating nodes have to be adjacent. So, if you have a node that
doesn’t concatenate (a Crop node for example) between two concatenating nodes (for example
Transform nodes), no concatenation occurs.
If you’re using more than one filtering method in the nodes that concatenate, the last filtering method
in the series of concatenating nodes is applied on the result.
When nodes concatenate, Nuke only

uses the filtering method set on the
last concatenating node (in this case,
the filtering method used is Mitchell).
As a rule of thumb, the only nodes that concatenate are usually transform nodes. In addition to these,
Dot, NoOp, Switch nodes and disabled nodes do not break concatenation.
Color nodes do not concatenate because Nuke works in a 32-bit float, which is enough to avoid
banding and visible round-off errors in color.
Please be aware that some transform nodes do not concatenate and should not be interspersed with
those that do. There are also transform nodes that receive concatenation, but don’t pass anything on.
Nodes that concatenate Only concatenate upstream Do not concatenate
Card3D BasicMaterial AdjustBbox
CornerPin Diffuse BlackOutside
USER GUIDE
654
Nodes that concatenate Only concatenate upstream Do not concatenate
Reconcile3D DisplaceGeo Crop
Reformat Displacement Mirror
Stabilize Emission PlanarTracker
Tracker Environment PointsTo3D
Transform Flare Position
Dot GridWarp TVIScale
NoOp IDistort
Switch LensDistortion
Disabled nodes MotionBlur2D
Phong
Sparks
Specular
SphericalTransform
SplineWarp
STMap
Tile
TransformMasked
USER GUIDE
655
Applying Core Transformations in 2.5D | Using the 3D Transformation Handles
Applying Core Transformations in

2.5D
Nuke’s Card3D node lets you apply the same geometric transformations possible with the Transform
node, but gives you an additional axis of operation, z.
Just to be clear, the Card3D node’s transformations are not truly 3D, but rather what is sometimes
called “2.5D” - meaning that you can move an element back on the z axis, but doing so does not
convey the sense that it is behind or in front of another element. 2.5D transformations are useful for
tasks like “cheating” the perspective of an element or “faking” a camera zoom.
Remember, however, that Nuke doesn’t limit you to 2.5 dimensions. If you need true 3D capabilities,
you can construct a 3D scene. See 3D Compositing for more information.
Using the 3D Transformation Handles

You’ll note when viewing the output of a Card3D node that it displays an overlay for executing spatial
transformations. This overlay is often a faster alternative to the properties panel. The figure below
shows you how to use it.
• A) Drag to translate the frame on the y axis (see Translating Elements).
Press Ctrl/Cmd while dragging to rotate the frame on any axis (see Rotating Elements).
• B) Drag to translate the frame on the z axis.
Press Ctrl/Cmd while dragging to rotate the frame on any axis.

• C) Drag to translate the frame on the x axis.
USER GUIDE
656
Applying Core Transformations in 2.5D | Adding a Card3D Node
Press Shift while dragging to constrain the translation to x.
Press Ctrl/Cmd while dragging to rotate the frame on any axis.
Adding a Card3D Node

To add a Card3D node:
1. Click Transform > Card3D to insert a Card3D node at the appropriate place in your script.
2. Connect a Viewer to the output of the Card3D node so you can see the effect of your changes.
Translating Elements
When using the Card3D node, you can translate elements on z in addition to the other axes.
To Translate an Element Using the Card3D Node

In the Card3D properties panel, increment or decrement the translate x, y, and z fields to slide the
element along any axis,
OR
Drag on any axis or transformation overlay.
Rotating Elements
When using the Card3D node, you can rotate elements around the x and y axes, in addition to the z.
This is useful for cheating the perspective.
USER GUIDE
657
Applying Core Transformations in 2.5D | Scaling Elements
To rotate elements using the Card3D node

1. Position the pivot point as necessary by incrementing or decrementing the pivot x, y, and z fields
to move the axis in any direction.
Alternatively, you can position the pivot point by pressing Ctrl/Cmd+Alt while dragging.
2. Increment or decrement the rotatex, y, and z fields to spin the element around the pivot point.
Alternatively, you can press Ctrl/Cmd while dragging on any axis on the transformation overlay.
Scaling Elements
To scale an element using the Card3D node:
2. To scale the frame simultaneously on x, y, and z, increment or decrement the uniform scale
field.
3. To scale the frame asymmetrically, increment or decrement the scale x, y, and z fields.
Skewing Elements
Whereas the Transform node lets you rotate pixel columns only around the z axis, Card3D permits
you to so around all three axes.
To Skew an Element Using the Card3D Node

USER GUIDE
658
Applying Core Transformations in 2.5D | Specifying the Order of Operations
2. Increment or decrement the skew x, y, and z fields to rotate the pixel columns around the
corresponding axes.
Specifying the Order of Operations

The order by which Nuke executes operations can affect the outcome. The Card3D node lets you
select the order by which Nuke executes scales, rotations, and translations, as well as the order by
which it executes rotation on individual axes.
To Choose the Operation Order for Scales, Rotations, and Translations

In the Card3D properties panel, select an option from the transform order dropdown menu, which
displays all possible combinations (S signifies scale, R, rotation; and T, translation).
To Choose the Operation Order for Rotations

Select an option from the rotation order dropdown menu, which displays all possible axial
combinations.
Choosing a Filtering Algorithm

Filtering algorithms let you specify the degree of smoothing and sharpening that remapped pixels
receive during transformation. The Card3D node offers the same filter algorithms as the Transform
node. See Choosing a Filtering Algorithm for more information.
To choose a filter algorithm, select the desired algorithm from the filter dropdown menu.
USER GUIDE
659
Adding Motion Blur | To Add Motion Blur
Adding Motion Blur

The following nodes under the Transform menu have their own controls for adding motion blur to
transformations:
• Transform
• TransformMasked
• Card (3D)
• CornerPin2D
• Reconcile3D
• Tracker
• Stabilize2D
These controls allow you to create motion blur without adding a separate node for it. The output is
similar to a TimeBlur node (see Applying a TimeBlur Filter), but rather than averaging the results of
several whole images computed at steps over the shutter period, a number of samples are taken at
many random times over the shutter period. This effectively gives many more "steps" and thus a
smoother-looking result for a smaller total number of computations.
Before rotation and motion After rotation and motion

blur. blur.
When using several of these nodes in a row, the motion blur is concatenated, and the last transform
in the chain defines the motion blur applied.
To Add Motion Blur

1. Open the transform node’s controls.
USER GUIDE
660
Adding Motion Blur | To Add Motion Blur to an Image Rendered in a Third-party Application
2. Create a transform and animate it. For instructions on how to do this, see the Using the
Compositing Environment chapter.
3. In the motionblur field, enter the sampling rate. This affects the number of times the input is
sampled over the shutter time. The higher the rate, the smoother the result. In many cases, a
value of 1.0 is enough. Setting the value to 0 produces no motion blur.
5. From the shutteroffset dropdown menu, select when the shutter opens and closes in relation to
the current frame value:
• to center the shutter around the current frame, select centered. For example, if you set the
shutter value to 1 and your current frame is 30, the shutter stays open from frame 29.5 to 30.5.
• to open the shutter at the current frame, select start. For example, if you set the shutter value
to 1 and your current frame is 30, the shutter stays open from frame 30 to 31.
• to close the shutter at the current frame, select end. For example, if you set the shutter value to
1 and your current frame is 30, the shutter stays open from frame 29 to 30.
• to open the shutter at the time you specify, select custom. In the field next to the dropdown
menu, enter a value (in frames) you want to add to the current frame. To open the shutter
before the current frame, enter a negative value. For example, a value of -0.5 would open the
shutter half a frame before the current frame.
To Add Motion Blur to an Image Rendered in a Third-party

Application
Another way to add motion blur to your image is to use the VectorBlur node. VectorBlur takes each of
your image’s pixels and blurs them in a straight line, using the u and v channels to determine the blur
direction.
VectorBlur expects the values from your input plates to be pixel space screen units, in other words
one unit should be equal to one pixel. Nuke uses this information to calculate the distance that one
pixel travels between two frames. So, in order to get a working motion blur result, you should make
sure Nuke is getting correct values to work with. You might have files using varying values, particularly
if you’ve used a third party application to create your input files. The following is an example of
creating motion blur with the VectorBlur node using files written from a third party application.
USER GUIDE
661
Adding Motion Blur | To Create Motion Blur with the VectorBlur Node
To Create Motion Blur with the VectorBlur Node

1. Read in your footage and motion blur files, for example an .exr file with a spinning donut and a
.sgi file with motion blur vectors that are normalized to have values between 0 and 1.
2. Using the ShuffleCopy node, select which channels VectorBlur should read from your motion
vector file (node input 1) and color image file (node input 2). In this case, you would use the
motion vector file’s red and green channels as the motion u and v channels, and its alpha channel
as the alpha channel. Meanwhile, the image file would output the red, green and blue channels
for the main color image.
With this setup, your ShuffleCopy node controls would look like this:
3. Connect the VectorBlur node to the ShuffleCopy node. You also need to tell VectorBlur which
motion vector channels to use, so change the uv channels control to motion.
4. If your motion vectors have been normalized to be between 0 and 1, you can set the u and v
values in the add control to -0.5 to offset the motion blur center. This is usually necessary for any
motion vectors stored in an integer file format like 16 bit TIFF or TARGA. Vectors that go to
negative x or y directions use half the numbers in the range and vectors that go positive use the
other half.
5. With the multiply and offset controls, you can further adjust the amount of motion blur you want
to produce. The offset value allows you to correct for normalization of your vector values, and the
multiply value controls the magnitude of them.
6. If the vectors have been premultiplied with the alpha channel, their value is not accurate in places
where the alpha is not 1.0. You’ll want to check the alpha checkbox to use the input image’s alpha
channel to help VectorBlur to deal with motion vectors that have been premultiplied by this alpha
channel.
USER GUIDE
662
Adding Motion Blur | To Create Motion Blur with the VectorBlur Node
Alpha disabled. Alpha enabled.

7. Getting a good, even motion blur result largely depends on selecting the right calculation method.
In the method dropdown, select:
• backward - backward method is effective and fast but may not be accurate if you don't have
motion vector values at all pixels throughout the whole frame
• forward - the forward method is slower, but it gives you a more accurate result, especially in
cases where the vectors don't cover the whole frame. In this case, we know the motion vectors
are not continuous, so selecting forward is a good option.
USER GUIDE
663
Replicating the Input Image Across the Output | To Use the Tile Node
Replicating the Input Image

Across the Output
The Tile node produces an output image that contains scaled-down, tiled copies of the input image.
The output image is the same format as the input.
Before the Tile node. After the Tile node.
To Use the Tile Node

1. Select the image you want to replicate and select Transform > Tile.
A Tile node is inserted in the Node Graph.
2. Attach a Viewer to the Tile node.
3. In the Tile node properties, use the rows field to define how many times the image is replicated
vertically. Note that the value can be fractional.
4. In the columns field, enter the number of times you want to replicate the image horizontally.
Note that the value can be fractional.
The original input image. The output of the Tile node

with rows set to 4 and
columns set to 5.
If you want to flip adjacent tiles vertically to form mirror images, check mirror.
USER GUIDE
664
Replicating the Input Image Across the Output | To Use the Tile Node
The output of the Tile node The same image with

without mirroring. vertical mirroring.
If you want to flip adjacent tiles horizontally to form mirror images, check mirror.
The output of the Tile node The same image with

without mirroring. horizontal mirroring.
5. From the filter menu, select an appropriate filtering algorithm. For more information, see
Choosing a Filtering Algorithm.
USER GUIDE
665
Warping Images
Nuke’s GridWarp and SplineWarp nodes allow you to distort elements in an image and morph one
image into another. There are other types of warp available in Nuke, but in this chapter, we focus on
the GridWarp and SplineWarp nodes.
Quick Start
1. For both warping and morphing, you can use either GridWarp node (Transform > GridWarp) or
the SplineWarp (Transform > SplineWarp). Working with the GridWarp is sometimes slightly
faster, whereas the SplineWarp node allows for more fine adjustment. With both nodes, you first
connect your source image to the src or A input and, if morphing, the destination image to the
dst or B input. GridWarp also allows you to connect an additional background image to the bg
input.
2. Depending on which node you’re using, you can then go on to set your source and destination
points, and tweak them until you’re happy with your results. For more information, see Warping
and Warping an Image Using the SplineWarp Node.
3. If you’re looking to morph an image into another, you can do this with both warper nodes as well.
For more information, see Morphing and Morphing One Image into Another Using the
SplineWarp Node.
4. If you are warping or morphing moving images, rather than stills, have a look at Transforming
and Animating Warps and Morphs.
USER GUIDE
666
Warping | Warping Images Using the GridWarp Node
Warping
Warping refers to manipulating an image so that elements in the image are distorted. Unlike many of
the transformations described under Transforming Elements, warps are transformations that only
affect some of the pixels in an image rather than all of them. For example, you might make an
animal’s eyes bigger or a person’s smile wider without affecting the rest of their features.
This is not to say that the pixels around the area you are moving do not move with the area. They do,
because accommodating the change this way often produces more realistic results. However, the
distortion lessens the further you get from the moved pixels. You also have some control over which
pixels are moved and which are not, and can isolate the warp to a small area. Still, in an ideal
situation, the subject you are going to warp is a subject you can key out or rotoscope to isolate it
from its background before you create the warp. This way, you can be sure that the background stays
intact.
In addition to performing creative manipulations on the shapes of the subjects in your images, you
can also use warping to simulate different types of film or video lenses or to remove unwanted lens
distortions.
Below, we discuss how to warp images, first using the GridWarp node and then the SplineWarp node.
Finally, we also teach you to animate the warps. Again, we start with the GridWarp node and then
show you how to do the same with the SplineWarp node.
Warping Images Using the GridWarp Node

The GridWarp node allows you to warp images by transferring image information from one Bezier
grid onto another. When using this node, you first position the source grid, which defines where to
warp from. Next, you position the destination grid, which defines where to warp the image to. This
grid can be a duplicate of the source grid, or you can define it separately. When you manipulate the
destination grid, the corresponding warp is applied to the source image.
The GridWarp node also includes controls for animating the warp and selecting the level of filtering
used to remove any artifacts the warp may have caused.
To Warp an Image Using the GridWarp Node

1. Select Transform > GridWarp to insert a GridWarp node after the image you want to warp.
2. Connect the src input and a Viewer to the image.
USER GUIDE
667
3. When the GridWarp properties panel is open, by default the destination grid overlay appears in
the Viewer. You can show or hide the source and destination grids using the and buttons in
the Viewer toolbar or the visible checkbox in the GridWarp tab. Use the destination grid to define
the warp areas.
Note: GridWarp automatically attempts to resize the grids to the image size as long as you
have not modified any of the control points.
If the grids are not the same size as the input image, click the Resize to image buttons
under both Source Grid and Destination Grid.
USER GUIDE
668
4. Use the Viewer tools to control the following aspects of the grids:
Note: You can use the copy and paste buttons in the grid controls to copy control
point keyframes between the source and destination grids.
Control What it does
output Controls the output displayed in the Viewer:

• source - the source image and source grid.
• sourcewarped - the source image and destination grid.
• destination - the destination image and destination grid.
• destinationwarped - the destination image and source grid.
• morph - the morphed image, controlled by the warp and mix parameters,
and both grids.
When enabled, the grids shown in the Viewer depends on the output setting.
For example, if output is set to source warped, only the destination grid
appears in the Viewer.
When enabled, the source grid is displayed in the Viewer. This control can be
overridden by the button.
When enabled, the destination grid is displayed in the Viewer. This control can
be overridden by the button.
When enabled, changes to points in a grid are automatically keyed. You can
disable this and set your keyframes manually, which is particularly useful if you
intend to use the Curve Editor. See Animating Warps.
ripple Rippling keyframes allows you to adjust the position of a stroke/shape point on
one frame and have that same relative adjustment applied to the point across
all frames or a specified range of frames.
• off - ripple edit is disabled.
• all - ripple all frames in the sequence.
• from start - ripple frames from the first frame to the current frame.
• to end - ripple frames from current frame to the last frame.
• range - ripple all frames within the from and to fields.
label points When enabled, points selected on the grid are labeled x,y measured from the
USER GUIDE
669
origin.
When enabled, the transform handle overlays all selected points.
When enabled, a low resolution preview displays when points are moved on a
grid. Once the render is complete, the low-res image is updated.
divisions Enter the number of divisions required in the field to modify the grid. The
number of divisions must be between 3 and 20.
You can also click and hold the slider to overlay a preview of the subdivisions.
GridWarp attempts to modify the grid based on the current control point
distribution.
Note: When using the preview, accuracy may suffer at lower divisions
and additional smoothing may be required at higher divisions.
Click to enable Edit mode. Select individual points, multiple points using shift-
click, or marquee groups of points.
Edit mode also allows you to adjust the curve between points to produce
distortion.
Click to enable Insert mode. Click on a horizontal line to add a vertical line to
the grid and vice versa.
Click to enable Delete mode. Click on a grid line to remove it from the Viewer.
Click to enable Draw Boundary mode. The cursor changes to a crosshair and
you can drag a marquee in the Viewer to create a custom grid.
Click to subdivide the grid columns across the currently selected area.
Click to subdivide the grid rows across the currently selected area.
Click to subdivide the grid columns and rows across the currently selected
area.
USER GUIDE
670
5. Modify the grid around the area you want to warp. Usually, you want the grid to conform to the
subject of the source image. For example, if you are warping an animal’s eyes, you need to create
grid lines that follow the edges of the eyes.
Note: If you have both grids visible when you move a point, and the same point for both
grids are on top of each other, both points are moved and you won’t see any distortion.
You can use the grid lines to isolate the areas you do not want to warp. You do this by adding grid
lines between the area you intend to warp and the area you don’t want to change.
An unlimited warp. The same warp limited to

a small area with grid lines.
When you select a point, four tangent handles appear around it. You can use these handles to
modify the curves connecting the points.
To move several points together, draw a marquee around them and use the transformation
overlay that appears.
USER GUIDE
671
You can also use the Draw Boundary tool in the Viewer to quickly set a user defined grid. Click
Draw Boundary and draw a marquee over the required area of the image.
The Draw Boundary The resulting user-defined

marquee. grid.
Note: You can also use the Curve Editor to edit curves by right-clicking a control point and
selecting Curve Editor > points, tangents, or both.
The curves appear in the Dope Sheet as long as the GridWarp Properties panel is open.
6. In the areas where you want to warp the image, drag the points on the grid to a new position.
When you click on a point, the point changes color to indicate whether it’s in focus (green by
USER GUIDE
672
default) and whether it has an expression set (red by default). You can change the default colors
on the Viewers tab in the Preferences.
The pixels in these areas are moved in the direction you moved the points. Pixels in the nearby
areas are also moved to accommodate the change, but the distortion lessens the further you get
from the repositioned points. If you don’t want a nearby area distorted, add more grid lines
between the area and the points you want to move before you drag the points to a new location.
Tip: You can nudge selected control points by a single pixel using the numeric pad 4 and 6
to nudge left and right, or 8 and 2 to nudge up and down.
Holding down Shift and using the numeric pad moves the selected points by 10 pixels, for
example, Shift+6 moves the selected points 10 pixels to the right.
7. To better see what the warped image looks like, press Q on the Viewer to toggle the overlay off.
USER GUIDE
673
To compare the original and warped images, press D on the GridWarp node to disable and enable
it. If you see changes in the areas you don’t want to warp, go back to modifying the grid.
8. If necessary, animate the grid to match any movement in the source image. For more information
on how to do this, see Animating Warps.
9. You can adjust the controls described in the following table to enhance your results.
GridWarp Tab
channel Sets the channels affected by the distortion.
mask Connect a mask input and set the channel to use as a mask. By default, the
mask is limited to the non-black areas of this channel.
Use the checkboxes to modify the mask properties:

• inject - copies the mask input to the predefined mask.a channel. Injecting
the mask allows you to use the same mask further downstream.
• invert - inverts the use of the mask channel so that the mask is limited to the
non-white areas of the mask.
• fringe - blurs the edges of the mask.
background The warped image is rendered on top of an unwarped background. This control
sets what to use as that background:
• on black - render the warped image on top of a constant black image.
• on src - render the warped image on top of the image connected to the src
input of the GridWarp node.
USER GUIDE
674
Warping | Warping an Image Using the SplineWarp Node
• on dst - render the warped image on top of the image connected to the dst
input of the GridWarp node.
• on bg - render the warped image on top of a background image connected to
the bg input of the GridWarp node.
background mix Blends between the output of the GridWarp node (at 0) and whatever you have
selected from the background dropdown menu (at 1).
Set bbox to Sets the boundary box properties.
Render Tab
submesh Sets the number of subdivisions that are created between Bezier curves in the
resolution grid. The higher the value, the more accurate the distortion between the grid
lines, but rendering time increases.
filter Select the appropriate filtering algorithm (see Choosing a Filtering Algorithm).
Options Tab
source color Sets the source grid color.
destination color Sets the destination grid color.
Warping an Image Using the SplineWarp Node

The SplineWarp node deforms an image based on multiple Bezier or B-spline curves that you create.
Source curves define where to warp from, while destination curves define where to warp the source
image to. Unlike the GridWarp node, you can draw these curves anywhere on the image, rather than
only adding points on the existing grid lines, then join them to create the source and destination. The
controls for adding and modifying points are similar to the RotoPaint node controls.
To Warp an Image Using the SplineWarp Node

1. Select Transform > SplineWarp to insert a SplineWarp node after the image you want to warp.
2. Connect the A input to the image and attach a Viewer to the SplineWarp node.
USER GUIDE
675
3. Use the Viewer tools to control the following:
Control Name What it does
autokey When enabled, changes to points and shapes are

automatically keyed. You can disable this and set your
keyframes manually, which is particularly useful if you intend
to use the Curve Editor. See Animating Warps for more
information.
label points When enabled, points selected in the Viewer are numbered
sequentially.
hide curves When enabled, curve lines are hidden.
hide points When enabled, points are hidden.
hide transform handle When enabled, the transform handle is hidden when a
selection is made.
hide transform handle When enabled, the transform handle is hidden when a
when moved selection is moved.
constant selection When enabled, clicking in the Viewer does not deselect the
current shape(s).
ripple Rippling keyframes allows you to adjust the position of a

stroke/shape point on one frame and have that same relative
adjustment applied to the point across all frames or a
specified range of frames.
USER GUIDE
676
• all - ripple all frames in the sequence.

• from start - ripple frames from the first frame to the
current frame.
• to end - ripple frames from current frame to the last
frame.
• range - ripple all frames within the from and to fields.
add/remove Add and remove keyframes at the current playhead position

keyframe using these buttons.
output A When enabled, output the image attached to the A input and
any shapes in the curves list labeled with A.
output A warped When enabled, output the image attached to the A input
warped by any shapes in the curves list labeled with A.
output AB morph When enabled, output the morphed A and B inputs and any
shapes in the curves list labeled with AB.
output B When enabled, output the image attached to the B input and
any shapes in the curves list labeled with B.
output B warped When enabled, output the image attached to the B input
warped by any shapes in the curves list labeled with B.
show source curves When enabled, joined source curves are displayed in the
Viewer.
show destination When enabled, joined destination curves are displayed in

curves the Viewer.
show correspondence When enabled, all correspondence points are displayed in

points the Viewer when a Correspondence tool is selected.
show boundaries When enabled, all boundary curves are displayed in the
Viewer.
show hard When enabled, all hard boundary curves are displayed in
boundaries the Viewer.
USER GUIDE
677
Note: Hard boundary is only useful on closed

curves.
visibility Controls visibility for the selected shape.
lock Controls the lock state of the selected shape. Locked shapes
cannot be modified.
toggle boundary Toggles the Boundary state of the selected shape(s).
Clicking toggle boundary on a standard shape defines it as a

Boundary and vice versa. See Using Boundary Shapes and
Pins to Limit Warp for more information.
update mode Sets the Viewer update mode:

• off - no updates are applied when points are moved. Use
this option for complex warps where the Viewer update can
be slow.
• on - the Viewer updates in real-time to help you position
points. With complex scripts, this mode maybe
prohibitively slow.
• persistent - similar to on, but displays a preview until the
render is complete, rather than showing real-time
scanlines.
Select tool Click to enable Select mode allowing you to select points
and shapes, or marquee select multiple points and shapes.
Points tool Click to enable Add mode or toggle between Add, Remove,
Cusp, Smooth, Open/Close, and Remove Destination
modes.
Click on existing shapes to add or modify points.
Draw tool Click to enable Draw mode or toggle between Bezier,

Cusped Bezier, B-spline, Ellipse, Rectangle, and Cusped
Rectangle mode.
Click on the Viewer to draw the selected shape.
USER GUIDE
678
Correspondence tool Click to enable Correspondence mode or toggle between

Add, Modify, and Remove mode:
• Add - clicking on a joined curve creates a correspondence
point. These points are used to improve the warp accuracy
in isolated areas, but overuse can cause performance
issues.
• Modify - allows you to move an existing correspondence
point along its curve, in either direction, as far as the next
point on the shape.
• Remove - click to remove an existing correspondence
point.
Pin tool Click to enable Pin mode. You can use pins as single point
joined curves to create warp or to secure specific points in
place to isolate them from warping.
Join tool Click to enable Join mode or toggle between Join, Reverse,
and Unjoin mode.
4. Click to select source warped mode.
5. Select the Draw tool and draw a source, or warp from, curve as shown below.
Note: If you’re creating multiple open splines, draw your first spline and then press Esc or
select another tool to let SplineWarp know you’re done. Reselect Draw to begin drawing
another spline.
USER GUIDE
679
6. You can modify shapes by adding points using the Add tool and adjusting the tangent handles
that appear around it.
When you mouse over a point, it turns green. Clicking on a point changes its color to indicate it’s
selected (white by default) and whether it has an expression set (red by default). You can change
the default colors on the Viewers tab in the Preferences.
To move several points together, draw a marquee around them and use the transformation
overlay that appears.
To close an open curve or open a closed curve, right-click on a point and select open/close
curve.
To remove a point, right-click on the point and select delete.
Note: You can also use the Curve Editor to edit curves by right-clicking a control point and
selecting curve editor > points or points+tangents.
The curves appear in the Dope Sheet as long as the SplineWarp Properties panel is open.
7. Once you’re happy with the source curve, either:
• draw a destination curve and use the Join tool to connect the two curves.
OR
• right-click a point on the source curve and select duplicate in A and join to produce the
destination curve.
8. Select the destination curve and drag points or the entire shape to a new position. The pixels in
these areas are warped in the direction of movement.
USER GUIDE
680
The image A and warp The warped image.

curves.
9. To better see what the warped image looks like, press Q on the Viewer to toggle the overlay off.
To compare the original and warped images, press D on the SplineWarp node to disable and
enable it. If you see changes in the areas you don’t want to warp, go back to modifying your
shapes.
10. Using the curves list, you can quickly build up a complex warp hierarchy using the add/remove
buttons.
Using the root, layer, and pair warp controls, you can affect different elements as required:
• root warp - affects warp globally, irrespective of layer and pairing.
• layer warp - affects warp on the selected layers only, irrespective of pairing. This control is
disabled if anything other than layers is selected.
• pair warp - affects warp on the selected pairings only. This control is disabled if anything other
than pairs is selected.
Note: The pair warp value is multiplied by the layer warp value, which is in turn multiplied
by the root warp value.
11. If necessary, animate the source warped shapes to match any movement in the source image. For
more information on how to do this, see Animating Warps.
12. You can adjust the controls described in the following table to enhance your results.
USER GUIDE
681
SplineWarp tab
crop to format When disabled, the input image is not cropped to the Project Settings > full
format size. As a result, warping can introduce pixels from the image outside
the format size, rather than black.
Note: Disabling crop to format can affect performance as Nuke has

to calculate the warp for the entire image, not just the area inside the
format size.
bbox boundary When enabled, a boundary curve is added around the input format, limiting
curve warp at the edges of the image.
Note: In AB morph mode, the format of input A is taken as the

boundary.
Render Tab
curve resolution Adjusts the accuracy of the warp/spline match. Higher values increase accuracy,
but sacrifice speed and vice versa. For example, with open splines at low curve
resolution, image tearing may occur. You can raise the curve resolution value
to compensate for tearing, but the render times increase.
Note: Correspondence points may be used to improve the warping

accuracy in a specific part of the curve if turning this value up too high
causes performance problems.
boundary curve Adjusts the number of interpolated points on boundary and hard boundary
resolution curves. Higher values stop the warp from filtering through the boundary curves
but sacrifice speed, and vice versa.
preview Improves the accuracy of the preview at higher values and the rendering speed
resolution at lower values.
Classic warping Set the type of warp function to use:

• disabled - uses an updated quadratic warp function that copes well with
USER GUIDE
682
overlapping control points, but has a more local warp effect.

• enabled - employs a warping function from previous versions of Nuke that
has a more global warp effect, but doesn’t cope well with overlapping control
points.
filter Select the appropriate filtering algorithm (see Choosing a Filtering Algorithm).
Using Boundary Shapes and Pins to Limit Warp

You may notice that pixels in areas around warping are also moved to accommodate the change, but
the distortion lessens the further you get from the repositioned points. If you don’t want a nearby
area distorted, you can:
• Designate a curve or shape as a boundary to control or eliminate warp.
• Place pins in the areas you want to limit warp.
Note: Boundaries only affect the input they are drawn on, A or B, not both.
• Boundary - generally open splines helping to control warp rather than eliminate it.
To define a boundary for an existing curve, enable boundary in the curves list, as shown below.
Using the example curves from Mr. Lion above, you could control the warp by adding a curve and
designating it as a boundary:
USER GUIDE
683
The source and destination The same curves, but with

curves. a boundary above the
destination to limit warp.
• Hard Boundary - closed shapes that eliminate warp outside the area that they enclose.
To define a hard boundary for an existing shape, enable hard boundary in the curves list, as shown
below.
The effect of a hard boundary on a warp is quite distinct - the hard boundary, colored white by
default, limits the warp effect to entirely within the closed shape:
• Pins - placing a pin creates a single point source and destination curve, effectively limiting warp in a
very localized way.
The effect of a pin on a single joined curve is shown below. The warp direction is shown on the left
and the pin’s effect on the right.
USER GUIDE
684
USER GUIDE
685
Morphing | Warping an Image Using the SplineWarp Node
Morphing
Morphing refers to dissolving two images together so that the subject of one image seems to change
shape and turn into the subject of the other through a seamless transition. A morph can be easily
noticeable or very subtle. An example of a noticeable morph would be a man turning into a woman
or one animal turning into another, whereas a transition from an actor to his stunt man would result
in a much more subtle morph.
An image of a monkey turning

into an image of a lion.
Morphing can be a time-consuming task, but it can be made easier by good advance planning of the
shots. The more similar the characteristics, position, and movement of the subjects you want to
morph are, the easier it is to morph them together. If the position and the movement of the subjects
do not match, however, you can try to reposition and retime the clips before morphing them
together. To do so, use the nodes described in Transforming Elements and Temporal Operations. You
can also use the Tracker and PlanarTracker nodes to track the features you want to morph over time
or to stabilize your clips before morphing them together. See Transforming Warps for more
information.
Below, we first discuss morphing images using the GridWarp node and then using the SplineWarp
node.
USER GUIDE
686
Morphing | Morphing One Image into Another Using the GridWarp Node
Morphing One Image into Another Using the GridWarp

Node
To morph one image into another using the GridWarp node:
1. Select Image > Read to import the two images you want to morph together.
2. If the images do not have the same resolution, insert Reformat nodes after them. For more
information, see Reformatting Image Sequences or Reformatting Elements.
3. Select Transform > GridWarp to insert a GridWarp node into your script.
4. Connect the source image (the image you want to turn into another image) to the src input of the
GridWarp node, and the destination image (the image you want the source image turned into) to
the dst input. Connect a Viewer to the GridWarp node.
5. To make the grids the same size as the input images, click the Resize to Image buttons under
both Source Grid and Destination Grid.
6. In the GridWarp Settings controls, set output to source to display the source image and grid.
7. Adjust the grid to follow key features on the source image. For example, if you’re morphing an
animal, you might consider the eyes, nose and mouth.
USER GUIDE
687
Morphing | Morphing One Image into Another Using the GridWarp Node
8. In the GridWarp Settings controls, set output to destination to display the destination image
and grid.
Note: If the source and destination images are similar, you can copy the source grid onto
the destination grid using the copy and paste buttons.
9. Adjust the grid to follow key features on the destination image.
10. In the GridWarp Settings controls, set output to morph to display both grids and activate the
warp and mix sliders.
USER GUIDE
688
Morphing | Morphing One Image into Another Using the SplineWarp Node
11. Scrub to the frame where you want the morph to begin. Bring the warp slider down to 0 (the
source image). Click on the animation button and select Set key. Then, scrub to the frame
where you want the morph to end and set warp to 1 (the destination image).
12. Animate the mix control in the same way. Scrub to the frame where the morph starts, and set mix
to 0 (the source image). Click the animation button and select Set key. Scrub to the frame where
the morph ends, and enter 1 (the destination image) as the mix value.
Play through the sequence in the Viewer and you’ll see that the source image is morphed into the
destination image.
Morphing One Image into Another Using the SplineWarp

Node
To morph one image into another using the SplineWarp node:
1. Select Image > Read to import the two images you want to morph together.
2. If the images do not have the same resolution, insert Reformat nodes as required. For more
information, see Reformatting Image Sequences or Reformatting Elements.
3. Select Transform > SplineWarp to insert a SplineWarp node into your script.
4. Connect the source image (the image you want to turn into another image) to the A input of the
SplineWarp node, and the destination image (the image you want the source image turned into)
to the B input. Attach a Viewer to the SplineWarp node.
5. In the SplineWarp node’s controls, switch output between A and B or click and in the
Viewer tools. You should see both images in the Viewer.
6. Identify some features that are similar in the source and the destination images. For example, if
you are morphing together images of people or animals, these features might include their eyes,
noses and mouths as well as the outlines of their faces and heads.
7. Switch to input A to draw curves around the key features you identified in the previous step. For
more information on how to create the curves, see Warping an Image Using the SplineWarp
Node.
USER GUIDE
689
You may find it useful to name the entries in the curves list, left_eye, right_eye, etc. Meaningful
names can help you later on when you’re using the Join tool to link the source and destination
shapes.
8. Switch to input B and draw curves in a similar fashion. Draw the same amount of curves, on
similar features, to make the morph as accurate as possible.
9. Switch the Viewer to output ABmorph , which displays both sets of curves, and activate Join
in the Viewer tools.

10. Work through the A input shapes, linking them to their B input counterparts. Try to link similar
points, such as the corners of the eyes or the side of the mouth.
Note: You can invert the source and destination shapes using the Reverse tool.
USER GUIDE
690
As you join shapes, the curves list updates showing the linked pairs.
11. Scrub to the frame where you want the morph to begin. Bring the mix slider down to 0 (input A).
Click on the animation button and select Set key.

12. Set keyframes for root warp in the same way.
13. If you’re using stills, as in this example, scrub the playhead to the number of frames required for
the morph. Otherwise, scrub to the end of the sequence.
14. Set the mix and root warp sliders to 1 (input B).
15. Press Q in the Viewer to disable the spline overlay and press play.
Input A is morphed into input B.
Using the cookie-cutter to create traveling masks

The curves list cookie-cutter is primarily designed to overlay a morph between two images, which
may not include an alpha channel, on top of a sequence by creating a “traveling mask” automatically
between closed cookie-cutter shapes.
For example, the monkey-lion morph demonstrated previously could be merged with a sequence,
replacing a face with a monkey face that is morphed into a lion over time.
USER GUIDE
691
Note: The cookie-cutter only works with closed shapes from the curves list.
1. Designate a pair of closed shapes as cookie-cutters by clicking the cookie-cutter icon in the curves
list.
In this instance, enabling cookie-cut for the monkey’s head, m_head, also designates the lion’s
head, l_head, as a cookie-cutter shape.
2. Use the output mask dropdown to select the channel where the traveling mask is stored, or use
the mask_splinewarp.a default.
3. Animate the morph to follow the required background element as described in Transforming and
Animating Warps and Morphs.
4. Finally, merge the output of the SplineWarp node with your sequence. A simple script might
appear as follows:
USER GUIDE
692
You can confirm the presence of the traveling mask by selecting the output mask channel in the
Viewer dropdown and then pressing M to display the Mat channel. You should see an
automatically created mask that matches the morphed shape.
USER GUIDE
693
Transforming and Animating Warps and Morphs | Transforming Warps
Transforming and Animating

Warps and Morphs
Unless you are warping or morphing a still image, you’ll probably need to transform or animate warp
operations. The Tracker and PlanarTracker nodes can help automate transformations, and you can
use Expressions and the GridWarp and SplineWarp Properties panel controls as well. See
Transforming Warps and Animating Warps.
Transforming Warps
You can transform grids and splines in the same way as any other element in Nuke (see Temporal
Operations), but you can also link control points to the Tracker and PlanarTracker nodes to automate
any transforms you require. Once you have a suitable track, you can use the animation data to
control points on grids and curves.
Using the Tracker Node

1. To generate tracking data, see Tracking and Stabilizing .
2. Make sure both the Tracker and the GridWarp/SplineWarp properties panels are open, then
either:
• Ctrl/Cmd+drag the animation button from the Tracker node properties on top of the individual
grid/curve point in the Viewer. When you release the mouse, the point follows the animation
from the Tracker node,
OR
• Right-click on the required control point and select Link to > Tracker1 and select the required
transform type.
Using the PlanarTracker Node

1. To generate PlanarTracker data, see Tracking with PlanarTracker.
2. Use the Draw Boundary button to create a source grid roughly the same shape as the tracked
plane, or draw similar splines.
3. On the GridWarp/SplineWarp and RotoPaint Transform tabs, open up the Transform tab extra
matrix.
4. On the RotoPaint Transform tab, select PlanarTrackerLayer1 in the curves list.
USER GUIDE
694
Transforming and Animating Warps and Morphs | Transforming Warps
5. Copy and paste the extra matrix animation from RotoPaint to the GridWarp/SplineWarp extra
matrix.
6. Play through the sequence to check that the grid or splines follow the plane.
Using Expressions with GridWarp

As well as using other nodes to control points, you can do the inverse, using expressions to link
control points and their tangents to other operations.
The expressions take the following form:

• GridWarp3_1.destination_grid_col.1.2.pos.x
• GridWarp3_1.destination_grid_col.2.2.tangent.1.x
Note: Control points are indexed starting from the bottom left 1.1, in x.y order, and tangent
points are labelled 0, 1, 2, and 3 corresponding to north, south, east, and west respectively.
It’s worth mentioning that in order to get the true position of a tangent, not its offset value, you need
to concatenate the pos and tangent expressions. For example:
GridWarp3_1.destination_grid_col.2.2.pos.x + GridWarp3_1.destination_grid_col.2.2.tangent.4.x
USER GUIDE
695
Transforming and Animating Warps and Morphs | Animating Warps
Note: You can only modify tangent positions using the Curve Editor.
For more information, refer to Expressions.
Animating Warps
In the GridWarp and SplineWarp node’s properties panels, there are controls for animating both the
source and the destination grids or shapes. Here, we first look at the GridWarp node and then the
SplineWarp node. The instructions assume you know how to warp a still image using these nodes (if
not, refer to Warping and Warping an Image Using the SplineWarp Node).
To Animate a Warp Using the GridWarp Node

1. While viewing the source image, adjust the source grid as necessary (see Warping for information
on how to do this).
If autokey is enabled, key frames are added every time you adjust the grid. Otherwise, click the
add key button under Source Grid. This saves the current grid as a key shape.
2. Move to a new frame and adjust the source grid accordingly. A new key shape is set automatically.
3. Repeat the previous step as necessary. If you need to delete a key shape, scrub to the frame
where you set it and click delete key under Source Grid.
4. Hide the source grid and display the destination grid.
5. While viewing the output of the GridWarp node, adjust the destination grid until you are happy
with the warp.
If autokey is enabled, key frames are added every time you adjust the grid. Otherwise, click the
add key button under Destination Grid. This saves the current grid as a key shape. Click the
set button under Destination Grid. The current grid is saved as a key shape.
6. Move to a new frame and adjust the destination grid again. The modified grid is automatically set
as a key shape.
7. Repeat the previous step until you are happy with the animated warp.
To Animate a Warp Using the SplineWarp Node

1. Create the required curves as described under Warping an Image Using the SplineWarp Node.
If autokey is enabled, key frames are added every time you adjust a shape. Otherwise, click the
add key button with A warped selected. This saves the current shape as a key frame.
2. Move to a new frame and adjust the shapes accordingly. New key shapes are set automatically.
USER GUIDE
696
Transforming and Animating Warps and Morphs | Animating Warps
Repeat the previous step as necessary. If you need to delete a key shape, scrub to the frame
where you set it and click delete key with A warped selected.
Linking Transforms Using the SplineWarp Node

Transforms applied in the properties panel Transform tab can be expression linked with other
shapes in the curves list using the transform linked toggle.
1. Create the required transform on the source shape.
2. Click the transform linked toggle in the curves list.
A menu displays the available shapes and transform types.
You can link all transform types associated with the source shape or be selective, such as all but
extra matrix.
3. In the example, any scale keyframes added to the l_left_nose shape are applied equally to the l_
head shape.
4. If you want to remove a link, click the transform linked toggle for the required shape, select the
link, and click remove link.
USER GUIDE
697
Temporal Operations
This chapter explains the temporal or time-based operations in Nuke. Learn how to distort time (that
is, slow down, speed up, or reverse clips), apply motion blur, and perform editorial operations like
slips, cuts, splices, and freeze frames.
Quick Start
1. If you want to speed up or slow down your clip:
• You can use the Retime node (Time > Retime). If you want, you can also add a FrameBlend
(Time > FrameBlend) node before your retime to create smoother retiming results. For more
information, see Simple Retiming and Interpolation.
• For more advanced retiming operations and adding motion blur, you can use the OFlow node
(Time > OFlow), or if you have a license for NukeX, the Kronos node (Time > Kronos). For more
information, see OFlow Retiming and Kronos.
• Sometimes you’ll want to create speed-up and slow motion within a clip without altering its
actual length. This is called warping and you can use the Retime node (Time > Retime) to do it.
For more information, see Warping Clips.
After retiming your clip, you may need to adjust the script’s global frame range and playback
speed (frames-per-second). For more information, see Global Frame Range and Speed .
2. If necessary, you can use the TimeBlur node to apply motion blur to garbage masks that are
tracked to a fast moving feature. For more information, see Applying a TimeBlur Filter.
3. If you're looking to perform editorial operations:
• You can use the TimeOffset (Time > TimeOffset) and TimeClip (Time > TimeClip) nodes to slip
clips. For more information, see Offsetting and Slipping Clips.
• The FrameRange (Time > FrameRange) node allows you to cut clips. For more information, see
Cutting Clips.
• You can use the AppendClip (Time > AppendClip) node to splice your clips. For more
information, see Splicing Clips.
• Using the FrameHold (Time > FrameHold) node allows you to set the output of a clip or scene to
a particular frame or frame interval. For more information, see Freeze Framing Clips.
USER GUIDE
698
Temporal Operations |
Note: Most temporal operations in Nuke can only be applied to 2D compositions, but the
FrameRange, TimeOffset, and FrameHold nodes are applicable to 3D nodes, such as
geometry.
USER GUIDE
699
Distorting Time |
Distorting Time
Time distortion changes the length of time required to playback a clip in your composite. These
operations generally fall under one of two categories: retiming and warping.
Retiming is the process of slowing playback by adding frames, or accelerating playback by subtracting
frames.
Warping is the process of slowing down, speeding up, or even reversing playback on a clip without
necessarily altering the overall length.
Tip: When working with temporal operations, it helps to attach a Viewer to the retiming or
warping node so you can see the effect of your changes.
USER GUIDE
700
Simple Retiming | To Retime All Frames in a Clip
Simple Retiming
Nuke’s Retime node lets you change the playback time for all the frames in a clip or for range of
frames within the clip. You can also use it to reverse the clip playback. It does this by dropping or
duplicating frames. For higher quality retiming see OFlow Retiming.
To Retime All Frames in a Clip

1. Select Time > Retime to insert a Retime node into your script.
2. Enter a value in the speed parameter. Values higher than 1 increase playback speed; values less
than 1 decrease playback speed.
3. Check the reverse box if you want to play the clip backwards - making the last frame the first, the
first frame the last, and so on.
4. Increase the shutter parameter to enable frame-blending (For more information, see
Interpolation).
To Retime a Range of Frames in a Clip

1. Select Time > Retime to insert a Retime node into your script.
2. Check the boxes for input range and enter the “in”and “out” frames.
For example, if your original clip is 50 frames, but you want to only retime the last half, you would
input 25 for the in point and leave the out point at 50.
3. Check the box for output range and enter the “in” and “out” frames to retime to a specific clip
length.
or
Enter a factor in the speed parameter and Nuke calculates the output range for you. Values
higher than 1 increase playback speed; values less than 1 decrease playback speed.
4. Check the reverse box to invert the selected frame range.
5. Increase the shutter parameter to enable frame-blending.
USER GUIDE
701
Interpolation | Frame-Blending
Interpolation
Time distortions that slow down a clip require the creation of additional, or interpolated, frames. For
example, suppose you want to slow down the collision, shown in the clip below, by a factor of two.
This requires the creation of one interpolated frame for every existing frame in the clip.
Interpolating frames.
The simplest way for Nuke to interpolate is to duplicate existing frames and increase the length of the
clip - this is the default method of interpolation. However, this method can create jittery playback,
especially when the image depicts very fast motion and the clip is retimed to be considerably longer
than its original length. For such cases, Nuke provides different nodes for smoothly interpolating the
frames.
Frame-Blending
The FrameBlend node interpolates frames by generating an additive composite of the frames that
precede and follow it, rather than creating mere copies between the existing frames.
USER GUIDE
702
Interpolation | To Insert a FrameBlend Node
Here is an example of the difference between frame-copy and frame-blend interpolation. In the first
frame of the figure below, you see a copy of the preceding frame. In the second frame, you see a new
image generated by blending the previous and subsequent frames.
Frame-copied versus frame-blended interpolation.
The latter method creates “ghosting” around all fast moving features (the window frame and the
pages on the desk, for example). This may look odd when viewed as part of a still frame, but
contributes to smoother motion during actual playback.
You can enable frame-blending by manipulating the shutter value of a retiming node. Higher shutter
values generate more frame-blending. Or, you can insert a FrameBlend node before the temporal
effect you want to influence. The below figure shows an example of frame-blending with the Retime
node.
To Insert a FrameBlend Node

1. Select Time > FrameBlend from the toolbar.
Remember to place it upstream from the temporal effect you want to influence.
USER GUIDE
703
Interpolation | To Insert a FrameBlend Node
2. Enter the number of frames to blend in the Number of frames field.

or
Check the Custom box and enter the starting and ending frames that you want to blend. To use
the input range as your custom frame range, click Input Range.
3. If necessary, check Foreground matte and select the channel to limit the blending effect.
The output Image count to option saves a floating point alpha image to a channel you specify; the
result indicates the number of images that contributed to each pixel of the matte. To normalize the
alpha, divide the number 1 by the number of frames averaged, and then multiply the alpha channel
by this result. You can also use the inverse of this matte for additional degraining.
USER GUIDE
704
OFlow Retiming | To Retime with OFlow
OFlow Retiming
The OFlow node generates high-quality retiming operations analyzing the movement of all pixels in
the frames and then rendering new “in-between” images based on that analysis. This node can also
add motion blur or enhance the existing motion blur in the image.
Tip: We recommend using NukeX's Kronos node instead, as it includes an improved motion
estimation algorithm. For more information, see Kronos.
To Retime with OFlow

1. Select Time > OFlow to insert an OFlow node after the sequence you want to retime.
The Input Range is set automatically by the source clip when you first create the node. After that,
it is only updated if you click Reset.
2. If you intend to use Output Speed or Frame timing, attach a Viewer to the output of the OFlow
node,
OR
If you intend to use Input Speed timing, attach a Viewer before the OFlow node.
3. By default, OFlow is set to perform a linear half-speed slow down using Timing > Output Speed.
4. Enable Use GPU if available to render output on the Local GPU specified, if available, rather than
the CPU.
For more information on the minimum requirements, please see Windows, Mac OS X and macOS,
or Linux or refer to the Nuke Release Notes available in Help > Release Notes.
5. Select the channels you want to apply the retime to using the channels dropdown. OFlow retimes
all channels by default.
6. To adjust the slow down or to speed the sequence up, enter a new value for the Output or Input
Speed control. Values below 1 slow down the clip. Values above 1 speed up movement. The
default value, 0.5, created the half-speed slow down. Quarter-speed would be 0.25.
7. Alternatively, you can describe the retiming in terms of ‘at frame 100 in the output clip, I want to
see frame 50 of the source clip‘. To do so, set Timing to Source Frame. Go to a frame in the
timeline, and set Frame to the input frame you want to appear at that output position. You’ll need
to set at least two key frames for this to retime the clip. For example, to slow down a 50 frame clip
by half, you can set Frame to 1 at frame 1, and to 50 at frame 100. To do a four times slow down,
set Frame to 1 at frame 1, and to 25 at frame 100.
8. If you’d like to add motion blur to your retimed footage, proceed to Adding Motion Blur.
USER GUIDE
705
OFlow Retiming | Refining the Results
Tip: Some Nuke users may remember nodes called OpticalFlow and OptiFlow. OFlow
replaces these nodes. It can be used for retiming and adding motion blur, but it does not
have the capability to output motion vectors. To output motion vectors, you could use
VectorGenerator (included in NukeX and Nuke).
For retiming and motionblur, we recommend using the Kronos node instead of OFlow. For
more information, see Kronos.
Refining the Results

To refine the results:
1. To have the motion vectors displayed in the Viewer, expand the Advanced control and enable
Overlay Vectors. Forward motion vectors are drawn in red, and backward motion vectors in
blue.
Overlay Vectors enabled.
Note: Motion vectors displayed in the Viewer are added to your output if you don’t turn off
the overlay before rendering.
2. To set the spacing between motion vectors displayed on the Viewer, adjust Vector Spacing. The
default value of 20 means every 20th vector is drawn. Note that Vector Spacing only affects the
Viewer overlay rather than the retimed result.
USER GUIDE
706
A Vector Spacing value of 4. A Vector Spacing value of 40.

3. Adjust Vector Detail to vary the density of the vector field. The larger vector detail is, the greater
the processing time, but the more detailed the vectors should be. A value of 1 generates a vector
at each pixel. A value of 0.5 generates a vector at every other pixel. For some sequences, a high
vector detail near 1 generates too much unwanted local motion detail, and often a low value is
more appropriate.
Areas of unwanted local motion detail. Lower Vector Detail is more

appropriate
in this case.
4. Vector fields usually have two important qualities: they should accurately match similar pixels in
one image to another and they should be smooth rather than noisy. Often, it is necessary to trade
one of these qualities off against the other. A high Smoothness misses lots of local detail, but is
less likely to provide you with the odd spurious vector. A low Smoothness concentrates on detail
matching, even if the resulting field is jagged. The default value of 1.5 should work well for most
sequences.
USER GUIDE
707
Jagging as a result of a low Minimal jagging as a result of a high

Smoothness Smoothness value.
value.
5. Set the type of resampling applied when retiming:
• Bilinear - the default filter. Faster to process, but can produce poor results at higher zoom
levels. You can use Bilinear to preview a retime before using one of the other resampling types
to produce your output.
• Lanczos4 and Lanczos6 - these filters are good for scaling down, and provide some image
sharpening, but take longer to process.
6. If the overall brightness in your Source footage changes between frames, enable Flicker
Compensation. This allows OFlow to take into account variations in luminance and overall
flickering, which could otherwise cause problems with your output.
Examples of variable luminance include highlights on metal surfaces, like vehicle bodies, or
bodies of water that reflect light in unpredictable ways.
Note that using Flicker Compensation increases rendering time.
7. In order to reduce processing time, much of the motion estimation is done on luminance only -
that is, using monochrome images. In most cases this is perfectly acceptable, but the parameters
in the Tolerances group allow you to concentrate on a particular feature in an image by adding
bias to individual colours. You may, for example, wish to increase Weight Red to allow the
algorithm to concentrate on getting the motion of a primarily red object correct, at the cost of the
rest of the objects in a shot.
USER GUIDE
708
Warping Clips | Refining the Results
Warping Clips
Warping refers to slowing down, speeding up, or even reversing select frames in a clip without
necessarily altering its overall length. Otherwise stated, warps add, subtract, or reverse the temporal
detail in a range of frames within a clip. For example, the figure below depicts a snowmobile clip
(down-sampled to just ten frames for easy representation) that we might want to warp.
Planning the time warp.
One way - in fact, kind of the classic way - to warp this clip would be to play the frames just prior to
the collision at their original rate, the frames involving the collision in slow motion, and the frames
after the collision in fast motion.
You could achieve such a warp by sculpting the curve in Nuke’s TimeWarp curve, which is a part of the
Retime node’s parameters, to look something like the one depicted below.
Editing the warp curve.
USER GUIDE
709
Warping Clips | To Warp a Clip
The basic “rules” for editing the warp curve are as follows:
• To slow down motion, decrease the slope of the curve.
• To speed up motion, increase the slope of the curve.
• To reverse motion, create a downward sloping portion on the curve (a dip, in other words).
To Warp a Clip
1. Click Time > Retime to insert a Retime node into your script.
2. Click the TimeWarp tab to reveal the TimeWarp curve.
3. Attach a Viewer to this node, so you can see the effect of your changes.
4. Sculpt the TimeWarp curve according to the rules above. (Ctrl/Cmd+Alt click to insert keyframe
knots on the curve; Ctrl/Cmd+drag to reposition keyframe knots; Ctrl/Cmd+drag to rotate a
keyframe knot control handles.)
5. If you want to enable frame blending on the output, either input a value larger than one in the
Retime node’s shutter parameter, or insert a FrameBlend node prior to the Retime node.
USER GUIDE
710
Global Frame Range and Speed | To Warp a Clip
Global Frame Range and Speed

Nuke automatically adjusts the timeline of every Viewer window you open to show the "in" and "out"
frames for the clip you’re viewing.
After you retime a clip in your compositing script, you may need to adjust the script’s global frame
range and playback speed (frames-per-second), to account for the retiming operations.
Adjusting the global frame range.
Select Edit > Project Settings (or press S over the Nuke window) and then enter the new frame
range and fps in the Project settings properties panel.
USER GUIDE
711
Applying a TimeBlur Filter | Applying Motion Blur to a Clip
Applying a TimeBlur Filter

When a fast-moving subject is recorded on film or video, its edges appear to smear as a result of the
object's movement while the shutter is open. The longer the shutter remains open at each frame
interval, the more obvious this effect. TimeBlur simulates this phenomenon by sampling its input at
"divisions" times over "shutter" frames starting at the current frame plus "offset".
Time blur is commonly applied to garbage masks that are tracked to a fast moving feature. The time
blur averages the incoming mask image over the shutter period, to better match the motion blur in
the original image and creating a more convincing integration.
Tip: You can also use the Roto and RotoPaint Global Blur controls to achieve the same
results. See Adding Motion Blur for more information on Roto and RotoPaint motion blur.
Applying Motion Blur to a Clip

To apply motion blur to a clip:
1. Click Time > TimeBlur to insert a TimeBlur node into your script. Place it downstream from the
element to which you want to apply motion blur.
2. In the TimeBlur properties panel, set divisions to the number of times you want to sample the
input over the shutter time. For images with fast-moving content higher values are necessary to
eliminate "steppiness" or banding in the output.
3. Set shutter to equal the span of time (in frames) over which the input should be sampled. A
shutter time of .5 is typical and would correspond with a camera shutter of 180 degrees.
4. Set the shutteroffset to control when the sampling of the input starts relative to the frame being
rendered, analogous to when the camera shutter opened to capture corresponding film or video
footage you might have at the same frame. You may need to adjust this by eye to align, for
example, a garbage mask with an underlying feature.
Tip: You may find that using TimeBlur on all the upstream nodes in your composition can
be unnecessary and very time consuming. In these cases, you can use NoTimeBlur node to
limit the number of nodes to which you’re applying TimeBlur. Just insert the NoTimeBlur
node in your node tree above the TimeBlur and any nodes you want the TimeBlur node to
process.
USER GUIDE
712
Editing Clips | Offsetting and Slipping Clips
Editing Clips
As a node-based system, Nuke doesn’t have a timeline. Nevertheless, you can still perform editorial
operations that you might associate with a timeline. You can slip clips (move them forward or
backward in time), cut them, or splice them to other clips.
Offsetting and Slipping Clips

Offsetting a clip refers to moving the clip backwards or forwards in time. In Nuke, you can offset clips
using the TimeOffset and TimeClip nodes.
Unlike TimeClip, the TimeOffset node can also be used with 3D nodes, for example, if you want to
offset camera times.
To Offset a Clip Using the TimeOffset Node

1. Click Time > TimeOffset to insert a TimeOffset node into your script. (Place it downstream from
the element to which you want to offset.)
3. In the TimeOffset properties panel, check reverse input if you want to invert the clip (make the
last frame the first, and so on).
4. In the time offset (frames) field, type the number of frames by which you want to offset the clip.
Enter a negative value to subtract frames from the head of the clip. Enter a positive value to add
frames to the head of the clip.
5. Adjust the script length for the new output range. Select Edit > Project Settings, and enter frame
range values that match the output range you specified.
Note: It’s not mandatory that you adjust the script’s frame range after slipping the clip. If
you don’t, the Viewer fills the empty frames at the tail of the clip by holding on the last
frame.
Tip: If you are using 3D nodes, such as cards, spheres, or cubes, together with a TimeOffset
node, you can view the object in just a specified frame range by setting the frame range
control for the object, and attaching a Viewer node to the TimeOffset node. Be sure to set
the timeline range to Input to view the offset.
USER GUIDE
713
Tip: Using TimeOffset, you can also offset clips directly in the Dope Sheet. See Animating
Parameters for more information.
Slipping a clip refers to changing the content of the clip that is seen, but not the position or duration.
In Nuke, you can slip clips using the TimeClip node.
Just like TimeOffset, the TimeClip node also lets you move clips forwards or backwards in time and
reverse the order of frames in a clip. In addition to this basic functionality, you can set the frame
range for the clip, set what happens to frames outside of this frame range, fade the clip to or from
black, and set expressions to adjust the node’s behavior. Using TimeClip, you can also slip the clip
directly in the DopeSheet, which can be handy if you prefer editing clips on a timeline.
To Slip a Clip Using the TimeClip Node

1. Click Time > TimeClip to insert a TimeClip node into your script. Place it downstream from the
element that you want to adjust.
2. Attach a Viewer to the TimeClip node to see the effect of your changes.
3. In the Fade In and Fade Out fields of the TimeClip properties panel, type the number of frames,
if any, you want to fade to or from black.
4. Set the new frame range for your clip by entering the first and last frame numbers.
5. In the before dropdown menu, set how Nuke should treat the frames before the first frame in the
range (examples refer to a 20-frame sequence with a first frame value of 5):
• hold - the first frame in the sequence is held until the first frame is reached. Example: 1, 1, 1, 1,
1, 2, 3, 4, etc.
• loop - substitutes an equal number of frames, effectively creating a clip loop. Example: 17, 18,
19, 20, 1, 2, 3, 4, etc.
• bounce - substitutes a reversed equal number of frames, creating a clip bounce. Example: 5, 4,
3, 2, 1, 2, 3, 4, etc.
• black - frames are black until the first frame is reached.
6. In the after dropdown menu, set how Nuke should treat the frames after the last frame
(examples refer to a 20 frame sequence with a last frame value of 5):
• hold - the last frame in the sequence is held until the end of the sequence is reached. Example:
16, 17, 18, 19, 20, 20, 20, 20, etc.
• loop - substitutes an equal number of frames, effectively creating a clip loop. Example: 16, 17,
18, 19, 20, 1, 2, 3, 4, etc.
• bounce - substitutes a reversed equal number of frames, creating a clip bounce. Example: 16,
17, 18, 19, 20, 19, 18, 17, etc.
• black - frames are black from last until the end of the sequence is reached.
USER GUIDE
714
7. Use the frame dropdown to set the frame mode:

• expression - Lets you enter an expression in the field on the right. The expression changes the
relation between the current frame and the frame read in.
frame 1 rather than frame 500, you can use the expression frame+499. This way, 499 frames are
added to the current frame to get the number of the frame that’s read in. At frame 1,
two to get the number of the frame that’s read in. This way, only every other frame in the clip is
used. At frame 1, image.0002.rgb is read in; at frame 2, image.0004.rgb is read in; at frame 3,
image.0006.rgb is read in; and so on.
• start at - Lets you enter a start frame number in the field on the right. This specifies the frame
where the first frame in the sequence is read in. In other words, all frames are offset so that the
clip starts at the specified frame.
For example, if your sequence begins from image.0500.rgb and you enter 1 in the field,
image0500.rgb is read in at frame 1. Similarly, if you enter 100 in the field, image0500.rgb is read
in at frame 100.
• offset - Lets you enter a constant offset in the field on the right. This constant value is added to
the current frame to get the number of the frame that’s read in.
frame 1 rather than frame 500, you can use 499 as the constant offset. This way, 499 is added to
the current frame to get the frame that’s read in. At frame 1, image.0500.rgb is read in; at frame
2, image.0501 is read in, and so on.
You can also use negative values as the constant offset. For example, if you use the value -10,
Nuke subtracts ten from the current frame to get the frame that’s read in. At frame 20,
8. To reverse the clip within the specified frame range, check reverse.
9. While the original range fields don't actually affect the output of the node, they help make things
clear in the Dope Sheet. Open the Dope Sheet and enter some values here and observe how they
affect the TimeClip lines.
10. If necessary, adjust the script length for the new output range. Select Edit > Project Settings, and
enter frame range values that match the output range you specified.
Tip: Using TimeClip, you can also offset, trim, and slip clips directly in the Dope Sheet. See
Animating Parameters for more information.
USER GUIDE
715
Editing Clips | Cutting Clips
Cutting Clips
The FrameRange node allows you to set a frame range for a clip. This controls which frames a
downstream AppendClip node uses from the input, which frames are displayed in the Viewer when
the timeline range dropdown menu is set to Input, and which frames are sent to the flipbook.
To make an edit, first use this node to cut portions out of input sequences and then append the
results together using the AppendClip node.
To Cut a Clip
1. Click Time > FrameRange to insert a FrameRange node into your script.
3. In the frame range fields, then enter the appropriate in and out point values.
For example, if your original clip is 50 frames but you want to use only the last 25 frames in your
composite, you would enter 25 as the In point and leave the Out point at 50.
4. Adjust the script length for the new output range. Select Edit > Project Settings, and enter frame
range values that match the output range you specified.
Note: It’s not mandatory that you adjust the script’s frame range after cutting the clip.
Tip: Using FrameRange, you can also set the frame range for a clip directly in the Dope
Sheet. See Animating Parameters for more information.
Splicing Clips
Splicing refers to joining clips head-to-tail, thus allowing action to flow from one shot to the next.
When you splice clips, you have options for:
• Fading to or from black.
• Dissolving from the first to second clip.
• Slipping the combined clip in time.
To Splice Slips
1. Click Time > AppendClip to insert an AppendClip node into your script.
USER GUIDE
716
Editing Clips | Splicing Clips
2. Attach its 1 and 2 inputs to the clips you want to join. (The clip attached to input 1 precedes the
one attached to input 2.)
4. If necessary, expand the script length to accommodate the total length of the newly merged clip:
• Click Edit > Project Settings. The project settings properties panel appears.
• Enter frame range values that matches the total length.
5. In the Fade In and Fade Out fields of the AppendClip properties panel, type the number of
frames, if any, you want to fade to or from black.
For example, typing a 5 in the Fade In field would result in the following effect at the head of the
merged clip.
(The inverse of this effect would occur at the tail of the merged clip were you type 5 in the Fade
Out field.)
6. In the Cross Dissolve field, type the number of frames, if any, of overlap you want between the
first and second clip.
For example, leaving Cross Dissolve at the default of 0 creates a simple cut - the transition from
the first to second clip is instantaneous. Typing in a 5 creates a time span of five frames in which
the first clip’s gain ramps downward to zero, while the second’s ramps upward to 100%.
Dissolve.
Cut.
7. In the First Frame field, type the number of frames, if any, by which you want to slip the clip.
Enter a negative value to subtract frames from the head of the merged clip. Enter a positive value
to add frames to the head of the clip.
8. Slipping the merged clips may create empty black frames at its head or tail. As appropriate, select
First frame or Last frame if you want these empty frames to appear as copies of the first or last
frame.
USER GUIDE
717
Editing Clips | Freeze Framing Clips
Freeze Framing Clips

Using the FrameHold node, you can set the output of a clip to a particular frame or frame interval,
rather than the entire clip.
To Output a Freeze Frame

1. Click Time > FrameHold to insert a FrameHold node into your script.
2. Attach its input to the clip you want to affect.
3. Attach a Viewer to the FrameHold node, so you can see the effect of your changes.
4. Enter the first frame number to freeze as the output of the FrameHold node.
Note: The first frame control defaults to 0, setting the freeze frame as the first frame of the
clip, regardless of the actual frame number.
To Output a Frame Interval

1. Click Time > FrameHold to insert a FrameHold node into your script.
2. Attach its input to the clip you want to affect.
3. Attach a Viewer to the FrameHold node, so you can see the effect of your changes.
4. Enter the increment frames to skip at output. For example, with a first frame of 1 and an
increment value of 5, the frames passed down the node tree would be 1, 6, 11, 16, etc.
USER GUIDE
718
Working with Color
These pages explain how to use Nuke’s color correction nodes and tools to adjust the appearance of
the images in your composites. When starting out, this information provides a good overview of
Nuke’s scopes and color-correction nodes, however not all options are covered here.
USER GUIDE
719
Using Scopes
Nuke provides scopes to help you evaluate your footage. Scopes are accessed from the contents
menu, Windows > New Scope sub-menu.
There are a number of global controls (Preferences > Panels > Scopes) that affect how the Scopes
display information:
• black point - sets the black out of range warning level.
• white point - sets the white out of range warning level.
• luma/chroma encoding - sets the video standard to use when converting RGB to luma or chroma
values in the scope displays, either REC601 or REC709.
• Include viewer color transforms - when enabled, scope data includes the applied Viewer color
transforms (gain, gamma, and LUT). When disabled, scope data does not include the applied Viewer
color transforms. This may slow down rendering, as it may require image calculation.
• Force full frame - When enabled, scopes display data for the full frame, regardless of what portion
of that frame is displayed in the Viewer. When disabled, scopes only display data for the current
area requested by the Viewer rather than the full frame.
Histogram
The content menu Windows > New Scope > Histogram provides three color channel and luma
channel information that describes the distribution of red, green, blue, and luma pixels throughout
the current frame.
The Histogram graphs the number of pixels at each brightness level, and from left to right, the areas
of the Histogram represent shadow, mid tones, and highlights.
USER GUIDE
720
the panel.
There are a number of view controls on the Histogram tab:

• Viewerselection - if you have multiple Viewers open, use the dropdown menu to associate
Histogram output to the required Viewer.
The default value, ActiveViewer, automatically displays details on the last Viewer you selected.
also view the channels separately by checking Parade.
projects.
displayed at once.
For example, with two stereo Read nodes, one in each input buffer, and wipe and Anaglyph active,
the scopes display something like this:
The scopes feature global customizable guides to help you view data. Navigate to Preferences >
Panels > Scopes and enter values between 0 and 1 for Blackpoint and Whitepoint.
The guides at the edges of the Histogram turn red to warn you when the distribution is out of range:
USER GUIDE
721
Using Scopes | Using the Histogram Node
Using the Histogram Node

The properties panel for the Histogram node includes a window that graphs the number of pixels at
each brightness level. This is a useful gauge to see whether an image has a good distribution of
shadows, midtones, and highlights.
The histogram maps the distribution

of shadows, midtones, and highlights.
To Define Tonal Range with the Histogram Node

1. Click Color > Histogram to insert a Histogram node at the appropriate place in your script.
2. Connect a Viewer to the output of the Histogram node so you can see the effect of your changes.
3. Drag the leftmost input range slider till it roughly lines up with the initial boundary of the
histogram.
4. Drag the rightmost input range slider till it roughly lines up with the final boundary of the
histogram.
5. Drag the middle input range slider to define the midtone, or neutral, value.
USER GUIDE
722
Waveform
The content menu Windows > New Scope > Waveform scope provides information on luminance,
or brightness, which you can use to decide whether the footage is over or under exposed. The
information represents luminance values from 0 - 100% (black through the spectrum to white). The
higher the waveform, the brighter the image in the Viewer.
the panel.
The upper white marker is used to measure when over exposure could be a problem. If your
waveform has a lot of traces over the white marker, you could consider reducing the brightness of the
image. The opposite is true of the lower black marker.
There are also Viewer and Channel selection controls on the Waveform tab:
Waveform output to the required Viewer.
also view the channels separately by checking Parade.
projects.
USER GUIDE
723
Using Scopes | Vectorscope
displayed at once.
For example, with two stereo Read nodes, one in each input buffer, and wipe and Anaglyph active,
the scopes display something like this:
The scopes feature global customizable guides to help you view your data. Navigate to Preferences >
Panels > Scopes and enter values between 0 and 1 for Blackpoint and Whitepoint.
The guides at the top and bottom of the Waveform turn red to warn you when the distribution is out
of range:
Vectorscope
The content menu Windows > New Scope > Vector display color, saturation, and hue information
for the current frame. Similar to color wheels, vector scopes display information radially, from the
center outward. The farther from the center the data spans, the more saturation is represented.
In the image on the left, you can see that the frame represented contains mostly yellows and reds,
but the values are not oversaturated. The image on the right represents a badly saturated frame.
USER GUIDE
724
Notice the spill of red traces distributed toward the edge of the scope pass the target (the highlighted
square).
Normal saturation. High Saturation.
the panel.
There is also a Viewer selection control on the Vectorscope tab:

• Viewerselection - if you have multiple Viewers open, use the dropdown menu to associate vector
scope output to the required Viewer.
The default value, ActiveViewer, automatically displays details on the last Viewer you selected.
projects.
displayed at once.
USER GUIDE
725
USER GUIDE
726
Using the Pixel Analyzer | Vectorscope
Using the Pixel Analyzer

Nuke’s Pixel Analyzer enables you to analyze single and multiple pixels, or the entire image, and
compare color values between Viewers. The analyzer stores current, minimum and maximum,
average, and median values which can then be copied by value to controls on other nodes. For
example, you might use the minimum and maximum values from an analysis to set the black and
white points on another image. The analyzer is accessed from the contents menu Windows sub-
menu.
The Pixel Analyzer has two modes:

• pixel selection - the default mode, allows you to make single or multiple pixel selections in the
Viewer for analysis. You can also analyze an area of the Viewer using a Region of Interest tool.
• full frame - analyzes the contents of the current frame, regardless of any selections you’ve made in
the Viewer.
By default, full frame samples the visible region of the Viewer. As a result, actions that change the
visible area, such as zooming in and out, alter the available color values.
Note: The Preferences >Panels > Scopes > Include Viewer color transforms control
does not affect the Pixel Analyzer.
USER GUIDE
727
Using the Pixel Analyzer | Analyzing Pixel Selections
Analyzing Pixel Selections

The analyzer’s pixel selection mode allows you to select pixels in the Viewer, individually or as a
group, and display the color information in swatches. To make selections in the Viewer:
1. Connect a Viewer to the output you intend to analyze. You can connect multiple Viewers to a
single output, and vice versa, for comparison.
2. Click the contents menu and select Windows > Pixel Analyzer.
The Pixel Analyzer panel displays.
3. Use the sample dropdowns to control which Viewer and layer provides the swatch information:
• viewer - the default, current viewer, always samples the active Viewer if there is more than
one. Selecting pixels in a Viewer or clicking on a Viewer’s tab causes the Viewer to become the
active or current Viewer. Alternatively, you can select a named Viewer from the dropdown when
more than one is open.
• layer - the default, current layer, displays the layer(s) specified in the Viewer channels
dropdown.
Note: If rgba channels are not present, the first four available channel values are written
into the Pixel Analyzer’s rgba controls.
Note: The color swatches may not update immediately if you select a layer from the sample
dropdown that is not currently visible in the Viewer, because the new layer must be
rendered in the background before the swatch is calculated.
Alternatively, you can select a named layer from the dropdown. When switching between
Viewers, the Pixel Analyzer attempts to match the selected layer in the new Viewer. If that layer
doesn't exist, the default current layer is displayed.
Tip: You can use the layer dropdown to sample a layer that is not displayed in the Viewer,
allowing you to compare position and color values between layers.
4. Ensure that the mode dropdown is set to pixel selection.

5. If you want to make multiple selections, without losing your previous selections, enable
accumulate.
6. There are three methods of selecting pixels for analysis:
• single - hold Ctrl/Cmd and click in the Viewer to select a pixel to add it to the swatches.
USER GUIDE
728
Using the Pixel Analyzer | Analyzing Pixel Selections
• multiple - hold Ctrl/Cmd and click-and-drag in the Viewer to select multiple pixels as you move
the pointer.
• region - hold Ctrl/Cmd+Shift and click-and-drag in the Viewer to create a Region of Interest.
Adjust or move the ROI using the Viewer handles.
Tip: To deselect all pixels, hold Ctrl/Cmd and right-click in the Viewer.
With multiple selections in the Viewer, the color swatches provide the following information:
• current - the color value of the last pixel selected.
• min/max - the darkest and brightest color values from the selection.
• average - the mean average color value from the selection.
• median - the mid-point color between the darkest and brightest colors in the selection.
You can enable live median update to calculate the median swatch dynamically, rather than at
pen-up.
Note: In single mode, all swatches show identical information and in region mode, the
current swatch is disabled.
7. Click on the swatches to display rgba and hsvl color values for the selected swatch.
The Pixel Analyzer also detects inf (infinite) and nan (not a number) color values. The ! point in
the swatch indicates inf/nan values.
8. You can reset the swatches at any time by clicking clear selection, except when full frame mode
is enabled, in which case the button is grayed out.
9. Use the range dropdown to select the color bit-depth you intend to display. For example,
selecting 8-bit limits your color values to 255.
10. Set the min/max channel you want to use to calculate the min/max color swatches.
USER GUIDE
729
Using the Pixel Analyzer | Analyzing Full Frames
This control defaults to luminance (l), but if you wanted to display the minimum values in only the
red channel, for example, you could set min/max to red (r).
Analyzing Full Frames

The analyzer’s full frame mode samples the visible region of the Viewer by default. As a result,
actions that change the visible area, such as zooming in and out, alter the available color values.
Note: Enabling full frame processing forces the Pixel Analyzer to analyze the entire frame,
regardless of the portion of the Viewer currently visible.
1. Connect a Viewer to the output you intend to analyze. You can connect multiple Viewers to a
single output, and vice-versa, for comparison.
2. Click the contents menu and select Pixel Analyzer.
The Pixel Analyzer panel displays.
3. Use the sample dropdowns to control which Viewer and layer provides the swatch information:
• viewer - the default, current viewer, always samples the active Viewer if there is more than
one. Selecting pixels in a Viewer or clicking on a Viewer’s tab causes the Viewer to become the
active or current Viewer. Alternatively, you can select a named Viewer from the dropdown when
more than one is open.
• layer - the default, current layer, displays the layer(s) specified in the Viewer channels
dropdown.
Note: If rgba channels are not present, the first four available channel values are written
into the Pixel Analyzer’s rgba controls.
Note: The color swatches may not update immediately if you select a layer from the sample
dropdown that is not currently visible in the Viewer, because the new layer must be
rendered in the background before the swatch is calculated.
Alternatively, you can select a named layer from the dropdown. When switching between
Viewers, the Pixel Analyzer attempts to match the selected layer in the new Viewer. If that layer
doesn't exist, the default current layer is displayed.
USER GUIDE
730
Using the Pixel Analyzer | Applying Analysis Data
Tip: You can use the layer dropdown to sample a layer that is not displayed in the Viewer,
allowing you to compare position and color values between layers.
4. Ensure that the mode dropdown is set to full frame.

5. Click on the swatches to display rgba and hsvl color values for the selected swatch.
The Pixel Analyzer also detects inf (infinite) and nan (not a number) color values. The ! point in
the swatch indicates inf/nan values.
6. Use the range dropdown to select the color bit-depth you intend to display. For example,
selecting 8-bit limits your color values to 255.
7. Set the min/max channel you want to use to calculate the min/max color swatches.
This control defaults to luminance (l), but if you wanted to display the minimum values in only the
red channel, for example, you could set min/max to red (r).
Applying Analysis Data

You can use color information stored in the Pixel Analyzer swatches in any other control that includes
a swatch. For example, min/max color swatches analyzed in another shot can be applied to the
current shot by dragging the swatch to the Grade nodes’s blackpoint and whitepoint controls.
USER GUIDE
731
Using the Pixel Analyzer | Applying Analysis Data
USER GUIDE
732
Making Tonal Adjustments | Sampling White and Black Points
Making Tonal Adjustments

Defining tonal range (the black point, white point, and neutral value) is typically the first step in color
correcting a clip. Tonal range adjustments often improve contrast, but more importantly, they set the
stage for subsequent color corrections by properly dividing the colorspace into shadow, midtone, and
highlight regions.
Before tonal adjustment. After tonal adjustment.
Several of Nuke’s color correction effects offer tonal adjustment tools. Of these, Grade and Histogram
are probably the most intuitive to operate.
Sampling White and Black Points

The Grade node lets you define white and black points by sampling pixels from a Viewer frame.
To Define Tonal Range with the Grade Node

1. Click Color > Grade to insert a Grade node at the appropriate place in your script.
2. Connect a Viewer to the output of the Grade node so you can see the effect of your changes.
3. In the Grade properties panel, use the channels dropdown menu to select the channels you wish
to process.
4. Click the blackpoint parameter’s color swatch.
The eye dropper icon appears.
5. In the Viewer, press Ctrl/Cmd+Shift while clicking on the pixel you want to define as the black
point (typically the darkest pixel).
6. Click the whitepoint parameter’s color swatch. The eye dropper icon appears.
7. In the Viewer, press Ctrl/Cmd+Shift while clicking on the pixel you want to define as the white
point (typically the lightest pixel).
USER GUIDE
733
Making Tonal Adjustments | Making Basic Corrections
Making Basic Corrections

Adjustments to contrast, gamma, gain, and offset often comprise the bulk of the work in color
correction. Some artists prefer to makes these adjustments via sliders; others prefer curves (which
represent the range of color values in an image.) Nuke’s ColorCorrect and ColorLookup nodes offer
tools to suit either preference.
See Using ColorCorrect Sliders or Using ColorLookup Curves
Original. Contrast boost.
Original. Gain boost.
Original. Gamma boost.
USER GUIDE
734
Making Tonal Adjustments | Using ColorCorrect Sliders
Original. Offset boost.
Using ColorCorrect Sliders

The ColorCorrect node is particularly convenient for making quick adjustments to contrast, gamma,
gain, and offset. A single window houses sliders for all these basic corrections and allows you to apply
these to a clip’s master (entire tonal range), shadows, midtones, or highlights.
To Adjust Contrast, Gain, Gamma or Offset with the ColorCorrect Node

1. Click Color > ColorCorrect (or press C) to insert a ColorCorrect node at the appropriate place in
your script.
2. Connect a Viewer to the output of the ColorCorrect node so you can see the effect of your
changes.
3. In the ColorCorrect properties panel, use the channels dropdown menu to select the channels
you wish to process.
4. Drag the slider appropriate to the region you want to affect an operation you want to apply.
For example, to brighten the images highlights, you would drag on the highlights gain slider.
Remember too that you can use the color sliders to apply any of the corrections on a per channel
basis.
5. On the Ranges tab, enable test to show what is considered to be in the shadows, midtones, or
highlights.
This overlays the output with black (for shadows), gray (for midtones), or white (for highlights).
Green and magenta indicate a mixture of ranges.
USER GUIDE
735
Making Tonal Adjustments | Using ColorCorrect Sliders
Original image. With test enabled.

6. Still on the Ranges tab, you can use the shadow and highlight lookup curves to edit the range of
the image that is considered to be in the shadows or highlights. You can also look up color
information for the current pixel in the Viewer.
To return a curve to its default values, select it and click reset.
Warning: Do not adjust the midtone curve. Midtones are always equal to 1 minus the other
two curves.
7. To control how much of the original luminance is preserved after the color correction, enable and
adjust mix luminance. A value of 0 means the altered luminance is used in the output image. A
value of 1 produces a luminance value close to that of the original input image.
Original image.
Mix luminance set to 0. Mix luminance set to 1.
Note: When mix luminance is set to 1, the resulting luminance value is close to the original
luminance, but not exactly the same. The difference may vary depending on the color
corrections applied to the source image.
USER GUIDE
736
Making Tonal Adjustments | Adjusting Black Levels with the Toe Node
Adjusting Black Levels with the Toe Node

Toe lifts the black level, in a similar way to gain controls, but with a roll-off so that whites are mostly
not affected.
1. Click Color > Toe to create a Toe node. Connect it to the image whose black levels need adjusting.
2. Adjust the lift slider to change the black values to the specified gray value, without affecting any
original white values of the image.
3. If necessary, you can limit the effect to a particular channel with the channels controls.
4. If you need to, you can pick a channel in the (un)premultby dropdown to divide the image first
with that channel and then multiply it again afterward. Doing this can sometimes improve your
color correction results on anti-aliased egdes.
5. You can also use the mix control to dissolve between the original input (value 0) and the full effect
of the Toe node (value 1). If you only want to use one channel for mixing, you can specify that
using the mask control.
Using ColorLookup Curves

If you prefer to work with color curves, you can use the ColorLookup node to make contrast, gamma,
gain, and offset adjustments (and, in fact, many others). Color curves refer to line graphs of a given
color channel’s brightness. The horizontal axis represents the channel’s original, or input, values, and
the vertical axis represents the channel’s new, or output values.
USER GUIDE
737
Making Tonal Adjustments | Using ColorLookup Curves
As Figure shows, you can edit the ColorLookup node’s color curves to make all of the types of
corrections that are possible through the ColorCorrect node - and you can generally make these
corrections with more flexibility and precision than is possible with sliders.
Corrections through Corrections through

the ColorLookup node’s the ColorLookup node’s
color color
curves: Contrast boost. curves: Gain boost.
Corrections through Corrections through

the ColorLookup node’s the ColorLookup node’s
color color
curves: Gamma boost. curves: Offset boost.
To Make Basic Corrections with the ColorLookup Node

1. Click Color > ColorLookup to insert a ColorLookup node at the appropriate place in your script.
2. Connect a Viewer to the output of the ColorLookup node so you can see the effect of your
changes.
3. In the ColorLookup properties panel, click red, green, blue, or alpha if you want to limit the
subsequent operations to a particular channel.
You can select multiple curves in order to edit one curve with reference to another. Otherwise,
select the master curve (which represents all channels).
USER GUIDE
738
Making Tonal Adjustments | Using ColorLookup Curves
4. To speed up the color calculation, the ColorLookup node uses a precomputed lookup table
between 0 and the value specified in the range field. You can adjust the range value, or uncheck
the use precomputed table box if necessary to tell ColorLookup to not use a precomputed table.
5. In the Viewer, drag the cursor over the pixels you want to sample for the correction. In the
ColorLookup properties panel, press Ctrl+Alt (Cmd+Alt on a Mac) while clicking on the curve to
set points at the places where the red, green, and blue lines intersect with the color curve.
Viewing values from a sampled color.

6. Edit the position of the points and adjust the tangent handles to adjust the curve shape for the
color correction.
As an alternative to steps 4 and 5, you can use the source control to pick a source color for
adding points. Then, use target to pick a destination color. Finally, do one of the following:
• Click Set RGB to add points on the red, green, and blue curves, mapping source to target.
• Click Set RGBA to add points on the red, green, blue, and alpha curves, mapping source to
target.
• Click Set A to add points on the alpha curve, mapping source to target.
You can use these controls to match shadow, midtone, and highlights on two plates, for example.
Set source to shadow rgb in one, target to shadow rgb in the other, then press Set RGB. Same
for midtone and highlight areas.
USER GUIDE
739
Making Hue, Saturation, and Value Adjustments | Using ColorLookup Curves
Making Hue, Saturation, and

Value Adjustments
For certain color correction tasks like spill suppression, you ideally want to influence only a very
narrow range of color values. For such tasks, it’s often helpful to use effects that employ the Hue,
Saturation, and Value (HSV) color model. As its name indicates, the HSV color model breaks color into
three components:
• Hue, which refers to the color’s location on the traditional color wheel.
• Saturation, which refers to the extent to which the color has “soaked up” its hue.
• Value, which refers to the brightness of the color.
Original. Hue shift.
Original. Saturation decrease.
Original. Value decrease.
Nuke offers effects that allow you to correct the hue, saturation, and value components individually
or collectively.
USER GUIDE
740
Making Hue, Saturation, and Value Adjustments | Correcting HSV
Correcting HSV
Nuke’s HSVTool node lets you simultaneously adjust hue, saturation, and value components from a
single properties panel. It also features a color replacement tool. The main strength of this node is
the precision it offers in limiting corrections to a narrow swath of colors.
Adjusting color within Adjusting color within

a specific range of pixel a specific range of pixel
values. values.
For example, suppose you wanted to add a bit more punch to the waterfront scene by diversifying the
rooftop hues. To do so, you could limit the correction to the rooftop’s ochre-colored hues by
sampling a few pixels, then shift their values. Because you limited the color range, the surrounding
image would be generally unaffected by the shift.
To make HSV corrections with the HSVTool node

1. Click Color > HSVTool to insert an HSVTool node at the appropriate place in your script.
2. Connect a Viewer to the output of the HSVTool node so you can see the effect of your changes.
3. Limit, as appropriate, the range of colors you want subsequent corrections to influence:
• In the HSVTool properties panel, click the srccolor color swatch. Ctrl/Cmd+click on the Viewer to
sample a single color displayed, or Ctrl/Cmd+Shift+drag to sample a range of colors. To sample
a single color from the node’s input while viewing its output, Ctrl/Cmd+Alt+click on the Viewer.
To sample a region from the input, Ctrl/Cmd+Alt+Shift+drag on the Viewer.
• The Range sliders on Hue, Saturation, and Brightness clamp to the sampled range.
• For any color component, drag on the Range sliders to expand the color range as necessary.
• For any color component, drag on the Range Rolloff slider to fine tune the color range. Doing
so, adjusts the amount of falloff allowed past the limits defined by the Range sliders.
4. Make the necessary HSV corrections:
USER GUIDE
741
Making Hue, Saturation, and Value Adjustments | Correcting Hue Only
• For hue corrections, drag on the Rotation slider to input color wheel value between 0 and 360.
• For saturation corrections, drag on the Saturation Adjustment slider to input values between -
1 (completely desaturated to some shade of gray) and 1 (completely saturated).
• For value corrections, drag on the Brightness Adjustment slider to input values between -1
(black) and 1 (white).
You can also make color replacements using the srccolor and dstcolor parameters: First sample the
color you wish to replace with the srccolor color swath, then sample the color which you wish to use
as the replacement with the dstcolor color swath. The color in dstcolor replaces the color in srccolor
throughout the image.
Also, keep in mind that the HSVTool node makes an excellent keyer. You can use its Hue, Saturation,
and Brightness range sliders to precisely select a range of colors, then use the channel output
dropdown at the bottom of the dialog to output this selection as a matte channel. This dropdown lets
you specify which color components (hue, saturation, value, etc.) are added to the matte.
Correcting Hue Only

Nuke’s HueCorrect node lets you make precision adjustments to the levels of saturation in a range of
hues. You do so via edits to a series of suppression curves.
Editing the suppression curve.
By selecting which curve you edit and how much of that curve you alter, you can precisely limit the
influence of the effect.
USER GUIDE
742
Making Hue, Saturation, and Value Adjustments | Correcting Hue Only
Suppressing spill
For the compositor, HueCorrect is obviously of greatest use in diminishing green, blue, or redscreen
spill.
To suppress spill with the HueCorrect node:
1. Click Color > HueCorrect to insert a node at the appropriate place in your script.
2. Connect a Viewer to the output of the HueCorrect node so you can see the effect of your changes.
3. In the HueCorrect properties panel, select the channels you want to influence:
• Click sat to influence all channels (red, green, blue, and alpha) equally.
• Click lum to influence all channels, but with luminance weighting in effect (meaning that the red
channel receives approximately 30% of the effect; the green, 60%; and the blue, 10%).
• Click red to apply the curve as a lookup on the red channel only, looking up the pixel's hue on
the curve and then multiplying the red value in the pixel by the lookup result.
• Click green to apply the curve as a lookup on the green channel only, looking up the pixel's hue
on the curve and then multiplying the green value in the pixel by the lookup result.
• Click blue to apply the curve as a lookup on the blue channel only, looking up the pixel's hue on
the curve and then multiplying the blue value in the pixel by the lookup result.
• Click r_sup to apply a suppression function to reduce the level of the red channel. While the red
curve is used to directly multiply the red channel by the curve value, the r_sup curve is used to
control the amount that the red channel is suppressed.
• Click g_sup to apply a suppression function to reduce the level of the green channel. While the
green curve is used to directly multiply the green channel by the curve value, the g_sup curve is
used to control the amount that the green channel is suppressed.
• Click b_sup apply a suppression function to reduce the level of the blue channel. While the blue
curve is used to directly multiply the blue channel by the curve value, the b_sup curve is used to
control the amount that the blue channel is suppressed.
Note that you can select multiple curves in order to edit one curve with reference to another.
4. If necessary, drag the cursor over the Viewer to sample the image pixels that are representative of
the part of the image you want to correct. Then, in the HueCorrect properties panel, press Ctrl+Alt
(Cmd+Alt on a Mac) while clicking on the curve to plot a particular pixel’s value on the curve. This
lets you see what portion of the curve you want to edit.
5. Edit the curve as necessary - typically this means dragging down on control points in the hue
region that you wish to suppress.
USER GUIDE
743
Making Hue, Saturation, and Value Adjustments | Correcting Saturation Only
6. To control how much of the original luminance is preserved after the color correction, enable and
adjust mix luminance. A value of 0 means the altered luminance is used in the output image. A
value of 1 produces a luminance value close to that of the original input image.
Original image.
Mix luminance set to 0. Mix luminance set to 1.
Note: When mix luminance is set to 1, the resulting luminance value is close to the original
luminance, but not exactly the same. The difference may vary depending on the color
corrections applied to the source image.
Correcting Saturation Only

For the times when you just want to correct the saturation component and don’t require limiting the
correction to any particular channel, you can use Nuke’s Saturation node. Its controls are bare bones -
basically, just a saturation slider.
To make saturation corrections with the Saturation node

1. Click Color > Saturation to insert a Saturation node at the appropriate place in your script.
2. Connect a Viewer to the output of the Saturation node so you can see the effect of your changes.
3. Drag the saturation slider to make the necessary corrections.
USER GUIDE
744
Masking Color Corrections | To Mask a Color Correction
Masking Color Corrections

Virtually all the color-correction effects in Nuke include mask parameters that lets you limit the
correction to the non-black pixel values of a matte image. For example, suppose you want to add a
blue cast to the following scene without affecting the buildings.
Original image. Color-corrected image.
Mask image. Color-corrected image.
You could create a garbage mask that covers the river, then boost the red channel’s gamma in the
area of the frame that underlies the mask.
Typically, mask controls are located toward the bottom of the properties panel. However, in the case
of multi-purpose effects like HSVTool, there may be multiple mask controls, so that you can limit
each type of correction with a different mask.
Selecting a mask channel.
To Mask a Color Correction

1. Open the node’s properties panel and locate the mask controls.
USER GUIDE
745
Masking Color Corrections | To Mask a Color Correction
2. Select the channel you wish to use as the mask from the dropdown menu.
3. If you check inject in the mask controls, the mask from the mask input is copied into the
predefined mask.a channel. This way, you can use the last mask input again downstream. You
can also set a stream of nodes to use mask.a as the mask, and then change the masking of all of
them by simply connecting a new mask into the mask input connector of the first node.
5. To blur the edges of the mask, check fringe.
dragging on the mix slider.
7. If you want to output only the portion of the frame underlying the mask, check the (un)premult
by box.
USER GUIDE
746
Applying Grain | Using Synthetic Grain
Applying Grain
Grain matching - ensuring that all of the elements in a composite, including those which were digitally
generated, look like they were shot on the same film stock - is often one of the final steps in achieving
a convincing integration of all of a composite’s elements. Nuke offers effects for synthetically creating
grain and for reading in practically-created grain (grain derived from actual film stock).
An example of applying An example of applying

grain grain
to an image: Grainless to an image: Grained
image. image.
Using Synthetic Grain

Nuke offers several nodes for creating synthetic grain: Dither, Grain, ScannedGrain, and if you have a
NukeX license, F_ReGrain. Of these, Dither is the crudest - it basically lets you specify the amount of
noise per channel.
Grain includes presets for matching film stock and a means for controlling the mix between the
generated grain and the backplate. ScannedGrain offers film stock presets, plus synthetic grain
controls for applying practical grain.
F_ReGrain, for NukeX, is designed to sample an area of grain from one image and then to generate
unlimited amounts of this grain with exactly the same statistics as the original. This new grain can
then be applied to another image. See Using F_ReGrain for more information.
To add synthetic grain with the Grain node

1. Click Draw > Grain to insert a Grain node at the appropriate place in your script.
2. Connect a Viewer to the output of the Grain node so you can see the effect of your changes.
USER GUIDE
747
Applying Grain | Using Practical Grain
3. From the presets dropdown menu, select one of the film stock you want to match.
4. Adjust the Size sliders for the red, green, and blue channels to shrink or enlarge the granules.
5. Adjust the Irregularity sliders to increase or decrease the random quality of the grain, according
to the different channels.
6. Adjust the Intensity sliders to increase or decrease the contrast of the grain against the original
image.
Using Practical Grain

Although Nuke’s ScannedGrain node offers controls for creating synthetic grain (ones comparable to
those just discussed), it’s main use is for reading in and applying scanned grain - that is, grain derived
from actual film stock. If your facility has such sequences available, you can read them in and apply
them using the ScannedGrain node. Creating grain files is described below, as well as using the
resulting grain files with the ScannedGrain node.
To Create Film Stock Sequences

1. Film a gray card. Only about 50 frames are needed.
2. Scan the film in.
3. Select Image > Read to load the scanned image into Nuke.
4. Add a Blur node (Filter > Blur) after the image to blur the image until you cannot see any grain.
Then, blur the image a bit more.
5. Select Merge > Merge to insert a Merge node in your script. Connect the A input of the Merge
node into the original image, and the B input into the Blur node. Then, open the Merge controls
and select minus from the operation dropdown menu. The blurred image is subtracted from the
original image.
The purpose of this and the previous step is to subtract any overall gray level from the grain so
that only the grain is left.
USER GUIDE
748
6. Select Color > Math > Add to insert an Add node after the minus node. In the Add node controls,
enter 0.5 in the value field. This adds a value of 0.5 to all channels.
This step is necessary, because the ScannedGrain node subtracts 0.5 from the channels when it
reads the grain file (the subtraction is needed to store negative numbers in most file formats).
7. Select Image > Write to insert a Write node after the Add node. Render the output. Any file
format will do (for example, we have used the .rgb extension in the grain files on our website).
To Add Scanned Grain with the ScannedGrain Node

1. Click Draw > ScannedGrain to insert a ScannedGrain node at the appropriate place in your
script.
2. Connect a Viewer to the output of the ScannedGrain node so you can see the effect of your
changes.
3. Click the folder icon of the grain field and navigate to the appropriate film stock sequence. Select
Open.
4. If necessary, check the resize box to scale the grain sequence up or down to match your working
resolution.
5. In the min. width field, define a minimum width (in pixels) that images have to have in order to
receive grain.
6. Enter values into the red, green, and blue amount fields to increase or decrease on a per-channel
basis the density of granules. (This is accomplished, crudely speaking, by boosting or reducing the
gain of the grain sequence.)
Now you’re ready to fine-tune the blending between the grain and backplate.
USER GUIDE
749
To Mix the Grain and Backplate

1. Drag on the saturation slider to increase or decrease the intensity of the grain’s hue across all
channels.
2. If necessary, you can also use the supplied curve editor to edit the grain sequence’s color curves.
In this manner, you can alter gain, gamma, contrast, etc. on a per channel basis. (These curves
function in the same manner as those described in Using ColorLookup Curves).
3. To set a low threshold, based on the input image, below which the grain is not subtracted, adjust
the low_clip slider.
USER GUIDE
750
Applying Mathematical Operations | Clamping Channel Values
Applying Mathematical
Operations
Nuke’s Color icon in the Toolbar houses a number of nodes which are designed to apply common
mathematical operations to channels. These operations include clamps, offsets, inversions,
multiplications, and expressions.
Clamping Channel Values

To clamp a channel’s values is to ensure that its blackest blacks and whitest whites are visible on an
intended display device. Nuke’s Clamp node lets you assign “legal” values to colors that are either too
light or dark for the intended display device.
Clamping black and Clamping black and

white pixels to “legal” white pixels to “legal”
values. values.
For this effect, you use Nuke’s Clamp node.

1. Click Color > Clamp to insert a Clamp node at the appropriate point in your script.
2. Connect a Viewer to the output of the Clamp node so you can see the effect of your changes.
3. In the Clamp properties panel, use the channels field to select the channel you wish to clamp.
4. Drag the minimum slider to the legal value. (This has the effect of causing black values to go
gray.)
5. Drag the maximum slider to the legal value. (This has the effect of causing white values to go
gray.)
USER GUIDE
751
Applying Mathematical Operations | Offsetting Channel Values
Offsetting Channel Values

To offset a channel’s values is to add a fixed value to them, which, in effect lightens the whole channel.
You can also add a negative value to a channel, in which case the channel gets darker.
Offsetting channel values. Offsetting channel values.
For this effect, you use Nuke’s Add node.

1. Click Color > Math > Add to insert a Add node at the appropriate point in your script.
2. Connect a Viewer to the output of the Add node so you can see the effect of your changes.
3. In the Add properties panel, use the channels field to select the channel you wish to offset.
4. Use the value slider to input the value you wish to add to the channel’s values.
5. If you are using premultiplied input images, you may want to check (un)premult by and select
rgba.alpha from the dropdown menu. This simulates doing the addition before the
premultiplication was done.
Inverting Channel Values

To invert a channel is to subtract its values from one, which causes its blacks to become white and its
whites to become black. In the course of building a script, you’ll have frequent need to invert mattes
in particular.
Inverting channel values. Inverting channel values.
To invert channels you use Nuke’s Invert node.

1. Click Color > Invert to insert an Invert node at the appropriate point in your script.
2. Connect a Viewer to the output of the Invert node so you can see the effect of your changes.
USER GUIDE
752
Applying Mathematical Operations | Multiplying Channel Values
3. In the Invert properties panel, use the channels field to select the channel you wish to invert.
Multiplying Channel Values

To multiply a channel’s values is to times them by a given factor, which has the effect of lightening the
channel while preserving the black point. (This operation is also knows as gain.)
Multiplying channel values. Multiplying channel values.
For this effect, you use Nuke’s Multiply node.

1. Click Color > Math > Multiply to insert a Multiply node at the appropriate point in your script.
2. Connect a Viewer to the output of the Multiply node so you can see the effect of your changes.
3. In the Multiply properties panel, use the channels field to select the channel whose values you
wish to multiply.
4. Use the value slider to input the factor by which to you want to times the channel’s values.
Applying Expressions to Channel Values

Up till now, the discussion has focused on how to apply simple mathematical formulae - additions,
subtractions, multiplications, etc. - to a channel’s values. Nuke’s Expression node, however allows you
to apply complex formulae to a channel’s values. The actual syntax for expressions is rather complex,
and thus must be deferred to Expressions. For now, you can read about the basics of how to operate
the Expression node.
1. Click Color > Math > Expression to insert an Expression node at the appropriate point in your
script.
2. Connect a Viewer to the output of the Expression node so you can see the effect of your changes.
3. In the Expression properties panel, use the channel dropdown menus and buttons to select the
channel to which you wish to apply an expression.
4. Type the actual expression in the = field next to the channel.
For example, to assign noise to the red channel, then boost the gain of that result by 20 you
would type (random*r)*20.
USER GUIDE
753
Applying Mathematical Operations | Applying Expressions to Channel Values
5. If necessary, you can apply different expressions to different sets of channels by repeating the
above steps for the other channel dropdown menus and buttons.
6. If you need to use a long expression in several fields, you can use the fields on top of the
properties panel for assigning the expression temporarily to a variable. Enter your variable on the
left side of the = sign, and the expression on the right. You can then use the variable to represent
the expression in the = fields next to the channels.
A checkerboard modified using an

Expression node.
USER GUIDE
754
Transforming the Color Space | Applying Expressions to Channel Values
Transforming the Color Space

Whenever you read a clip into a script, it is automatically converted to Nuke’s native color space,
which is 32-bit per channel RGB, a linear format. This conversion takes place even if the clip you read
in is in the Kodak Cineon format, which is a logarithmic format.
The reverse of this conversion, called a lin-to-log conversion, also automatically takes place when you
write the processed element back out of the script - that is, Nuke automatically converts it back into a
Cineon element.
USER GUIDE
755
Overriding the Default Cineon Conversion | To Override the Default Cineon Conversions Using Log2Lin
Overriding the Default Cineon

Conversion
Nuke uses the Kodak-recommended settings when making Cineon conversions in either direction. It’s
rare that you would want to override these settings, but if it becomes necessary you can use Nuke’s
Log2Lin or PLogLin nodes.
To Override the Default Cineon Conversions Using

Log2Lin
1. Double-click on the Read node of the Cineon element whose conversion you wish to override.
2. In the Read properties panel, set the colorspace dropdown menu to linear. This halts the
automatic log-to-lin conversion.
3. Click Color > Log2Lin to insert a log2lin node directly after the Read node.
4. In the Log2Lin properties panel, set the operation dropdown menu to log2lin.
5. Set black, white, and gamma to the appropriate values.
6. Copy the Log2Lin node and insert it just before the element’s Write node.
7. Open up the properties panel of the second Log2Lin node and set the operation dropdown menu
to lin2log. This gives you the reverse of the conversion you created above.
8. Double click on the element’s Write node.
9. In the Write properties panel, set the colorspace dropdown menu to linear. This halts the
automatic lin-to-log conversion and lets the one you create above have priority.
To Convert Between Logarithmic and Linear Color Spaces

You can also convert between logarithmic and linear color spaces with the PLogLin node. This
alternative method is better known as the Josh Pines log conversion, and it’s based on using a single
gray value, rather than a black and a white one, like in the Log2Lin node. To use the PLogLin node:
1. Much like with the Log2Lin node, when you’re using the PLogLin node you need to first make sure
your Read or Write nodes aren’t automatically converting your color space. To do this, click on
your Read or Write node, and select linear in the colorspace dropdown.
2. Create a PLogLin node by clicking Color > PLogLin. Connect it to either your Read node’s output
or the Write node’s input, depending on which one you’re converting the color space for.
USER GUIDE
756
Overriding the Default Cineon Conversion | To Create a 3D LUT in Log Color Space (for example
3. In the operation dropdown, select the operation you want PLogLin to perform. Select log to lin to
convert from logarithmic to linear, and lin to log to do the reverse.
4. Adjust the linear reference value slider to the linear value that corresponds with the logarithmic
reference value and the log reference value slider to the value that corresponds with the linear
reference value.
5. You can also adjust the film response gamma value in the negative gamma field, and use the
density per code value field to tell PLogLin what type of change occurs in the negative gamma
for each log space code value when converting.
6. If you need to, you can pick a channel in the (un)premult by dropdown to divide the image first
with that channel and then multiply it again afterward. Doing this can sometimes improve your
color correction results on anti-aliased edges.
7. You can also use the mix control to dissolve between the original input (value 0) and the full effect
of the PLogLin node (value 1). If you only want to use one channel for mixing, you can specify that
using the mask control.
To Create a 3D LUT in Log Color Space (for example

Cineon)
For good visual fidelity when using 3D LUTs, it is recommended that a log color space is used for 3D
LUT generation. To create a 3D LUT in a log colorspace (Cineon in this example):
1. Create a CMSTestPattern node at a required density.
2. Connect this to a Colorspace node to convert in: Cineon to out: Linear. This is a reverse log
conversion, and gives a higher density of samples in the shadows of your grading curve, with
reduced density in the highlights.
3. Connect the Colorspace node to your grading nodes.
4. Connect the grading nodes to another Colorspace node to convert in: Linear to out: Cineon. This
is the log conversion, and converts the grade samples back to the normalized range (0-1) for 3D
LUTs.
It is important to distribute the samples appropriately due to the low resolution of 3D LUTs.
5. Connect a GenerateLUT node to the last Colorspace node.
Tip: To reproduce the grade, set a Vectorfield node to Cineon for both the colorspace in
and colorspace out controls.
USER GUIDE
757
Making Other Colorspace Conversions | To Convert an Element in Nuke’s Native Color Space into
Making Other Colorspace

Conversions
You can also convert elements from Nuke’s native color space to other color spaces more appropriate
to a given process or intended display device. For conversions such as this, use Nuke’s Colorspace
node, which supports RGB, HSV, YUV, CIE, and CMS formats (and various sub-formats).
To Convert an Element in Nuke’s Native Color Space into

Another Color Space
1. Click Color > Colorspace to insert a Colorspace node into the appropriate place in your script.
2. In the Colorspace properties panel, set the rightmost dropdown menu in the out controls to the
appropriate standard.
3. Set the dropdown menu in the middle of the out controls to the appropriate standard.
4. Set the leftmost dropdown menu in the out controls to the color space of your choice.
5. If you wish to reverse this conversion later in the script:
• Copy the Colorspace node and insert it at the appropriate point in your script.
• Set the out controls to sRGB, D55, and RGB.
• Set the in controls to match the values you entered in steps 2, 3, and 4 above.
6. If you wish write out the element in the new color space:
• Double-click on the element’s Write node.
• In the Write properties panel, set the colorspace dropdown menu to linear. This halts the
automatic conversion and lets the one you create above have priority.
Converting Color Spaces with the OCIOColorSpace Node

Much like the ColorSpace node, you can use the OCIOColorSpace node for converting an image
sequence from one colorspace to another. Conversions with the OCIOColorSpace node are based on
the OpenColorIO library (for more information, see http://opencolorio.org). Using OCIOColorSpace is
pretty simple:
1. Select Color > OCIO > OCIO ColorSpace and connect it to your image sequence.
2. Select the channel or layer you want the conversion to affect using the channels controls.
3. In the in dropdown, select the colorspace of your input image.
4. In the out dropdown, select the colorspace you want to convert the image to.
USER GUIDE
758
Making Other Colorspace Conversions | Converting Color Spaces with the OCIOColorSpace Node
For more information on the OCIOColorSpace node, and other nodes based on the OpenColorIO
library, see Color Nodes in the Nuke Online Help for more information.
Note: Nuke is bundled with a pre-compiled version of the OpenColorIO library and a suite
of OCIO nodes. If you already have a pre-built version of the OCIO library on your system and
your environment is not set up correctly for Nuke, you may encounter problems. For OCIO
to work correctly, Nuke requires the compiled versions of PyOpenColorIO and
libOpenColorIO to match. To override Nuke's pre-packaged version with your own custom
version, follow the steps below:
• Linux:
Manually replace the Nuke packaged versions of libOpenColorIO.so and
PyOpenColorIO.so with your custom versions of these files. These can be found in the
<NukeInstall> and <NukeInstall>/plugins directories respectively.
• Mac:
Set your NUKE_PATH to the location of PyOpenColorIO.so and your DYLD_LIBRARY_PATH
to the location of OpenColorIO.so. For example:
export NUKE_PATH="/myOCIOLibraryLocation/"
export DYLD_LIBRARY_PATH="/myOCIOLibraryLocation/"
NOTE: Due to an Apple security update in Mac OS X 10.7 (Lion), DYLD variables can't be
loaded from the environment.plist. See http://support.apple.com/kb/TS4267 for more
information.
• Windows:
Both OpenColorIO.dll and PyOpenColorIO.pyd must be in the same directory. You then
need to set your NUKE_PATH to this directory. For example:
set NUKE_PATH=\myOCIOLibraryLocation\
• All Platforms:
In addition to the steps above, you need to set the OCIO environment variable to point to
your OCIO configuration file. This overrides the configuration file specified in Nuke’s
preferences (see Setting Preferences). For example:
export OCIO="/myOCIOConfigLocation/config.ocio"
The OCIO nodes in Nuke are compiled against a specific version of the OCIO libraries (for
the current version, see Third-Party Libraries and Fonts). If you’re using your own custom
libraries, recompile the OCIO nodes against your versions of the libraries. Failure to follow
these steps may result in errors when adding OCIO nodes or modifying OCIO options in the
preferences.
USER GUIDE
759
Changing the Viewer Color Space | Adding Context Viewer Processes
Changing the Viewer Color Space

By default, a script’s Viewers display images in Nuke’s native color space. You can, however, set a
script’s Viewers to display images in non-native color spaces. Changing the display color space in no
way affects your rendered output. You are applying a display-only lookup table.
Select the desired color space from the viewer process dropdown menu.
Adding Context Viewer Processes

You can add variables to register certain viewer processes in OCIODisplay nodes by creating a custom
config.ocio file and then specifying the variable in the to_reference file transform.
In this example, the variable is called SHOT. If you intend to use the same name, make sure that
SHOT is not assigned as an environment variable.
To edit the ocio.config file:

1. Open your ocio.config file in a text editor and review the colorspaces entries. The example
shows the sRGB entry.
- !<ColorSpace>
name: sRGB
family: ""
equalitygroup: ""
bitdepth: 32f
description: |
Standard RGB Display Space
isdata: false
allocation: uniform
allocationvars: [-0.125, 1.125]
USER GUIDE
760
Changing the Viewer Color Space | Adding Context Viewer Processes
to_reference: !<FileTransform> {src: "sRGB.spi1d", interpolation:

linear}
2. Edit the to_reference line to include the variable you intend to create.
to_reference: !<FileTransform> {src: "$SHOT.spi1d", interpolation: linear}
3. Save the file.
To register the viewer process:

1. Launch Nuke.
2. Press S over the Node Graph to open the Project Settings, or navigate to Edit > Project Settings.
3. Switch to the Color tab and use the color management dropdown to select OCIO.
4. Set OCIO config to custom and enter the file path to your ocio.config file.
5. Click the viewer process dropdown and select the process you want to register, in this case sRGB
(default).
An error message displays in the Viewer because Nuke can't find the specified reference.
6. Click the viewer process dropdown and select show panel to open the OCIODisplay node
properties.
7. Click the Context tab and enter the key1 and value1 pair as SHOT/sRGB.
The LUT specified by the variable is applied to the Viewer.
Note: Viewer processes are display-only LUTs and do not affect rendered output.
USER GUIDE
761
Filtering and Spatial Effects
This chapter explains how to create custom filter effects using the Convolve node (see Applying
Convolves) and how to simulate depth-of-field (DOF) blurring using the ZDefocus node (see
Simulating Depth-of-Field Blurring).
Applying Convolves
The Convolve node lets you create custom filter effects by supplying your own filter image. This image
is used as the convolution matrix. In other words, the new value of a pixel is calculated by centering
the filter image on the pixel, examining its neighbors, multiplying each pixel value by the
corresponding pixel values in the filter image, and then adding the results together. This allows you
to defocus a clip and create lens blur effects (bokeh) in the shape of the filter image (also see
Simulating Depth-of-Field Blurring).
The image input. The filter input.
The result.
Using the Convolve Node

To use the Convolve node:
USER GUIDE
762
Filtering and Spatial Effects |
1. Click Filter > Convolve to insert a Convolve node after the image you want to receive the
convolution filter effect (the image input).
2. Attach a Viewer to the output of the Convolve node.
An error appears in the Viewer because we haven’t connected the filter input yet.
3. Connect a filter image to the filter input. This image represents the shape and size of the camera
aperture used to shoot the image input. It can be any shape you like, for example, a pentagon or
a hexagon.
If you don’t have a filter image, you can create one using the Roto node (Draw > Roto) or the Flare
node (Draw > Flare).
A simple Convolve setup.

Note that you don’t necessarily need to crop the filter image to a smaller size, as fast Fourier
transforms are used to speed up convolutions with large filter images.
4. In the Convolve node controls, set channels to the channels of the Source image that you want
to affect. By default, the convolve effect is applied to all channels.
5. To select what channel to use from the filter input, do one of the following:
• Set the filter channel dropdown menu to the channel you want to use. By default, this menu is
set to rgba.alpha.
• To use the same channels from the filter input as the image input (that is, whatever channels is
set to), check use input channels.
6. In most cases, you can leave normalize checked. This means the filter input is divided by the sum
of all the pixels in it before using it, and ensures that the convolution doesn’t change the overall
brightness of the image.
7. Use the filter bounds dropdown menu in the Convolve controls to select whether you want to
limit the filter image to:
• shape - the filter input’s bounding box. In this case, Convolve only uses the bounding box area,
and the center of the filter is the center of the bounding box. This is the default value. You may
want to use it, for example, if your filter input is a roto shape with a small bounding box that
doesn’t fill the entire format area.
USER GUIDE
763
Filtering and Spatial Effects |
• format - the filter input’s format. In this case, Convolve uses the entire format area, allowing
you to offset the filter image within the format.
8. If you want to mask the convolve effect, check mask and select a mask channel using the controls
on the right. For example, you can select a depth channel as a mask to simulate depth-of-field
blurring.
A simple Ramp node used The result.

as
the mask.
Note that you can also use the ZDefocus node for more accurate depth-of-field blurring. For more
information, see Simulating Depth-of-Field Blurring.
9. To dissolve between the original image input and the full convolve effect, adjust the mix slider.
take effect.
USER GUIDE
764
Simulating Depth-of-Field Blurring |
Simulating Depth-of-Field
Blurring
The ZDefocus node blurs the image according to a depth map channel. This allows you to simulate
depth-of-field (DOF) blurring.
The original image.
Using ZDefocus to simulate a more narrow depth of field.

In this case, areas under the green sign remain in focus,
whereas areas in front of and behind it are blurred.
Any defocused highlights in the image bloom in the shape
of the filter image, creating a bokeh effect.
USER GUIDE
765
Simulating Depth-of-Field Blurring | Quick Start
In order to defocus the image, ZDefocus splits the image up into layers, each of which is assigned the
same depth value everywhere and processed with a single blur size. After ZDefocus has processed all
the layers, it blends them together from the back to the front of the image, with each new layer going
over the top of the previous ones. This allows it to preserve the ordering of objects in the image.
Quick Start
1. Create a ZDefocus node and connect it to your script.
See Connecting ZDefocus.
2. Adjust the blur settings.
See Adjusting Blur Settings.
3. Adjust the shape of any out-of-focus highlights.
See Adjusting the Shape of Out-of-Focus Highlights.
4. If necessary, enhance the highlights to make lens shape effects more visible.
See Enhancing Out-of-Focus Highlights.
5. If necessary, mask the blur effect.
See Masking Blur Effects.
take effect.
Connecting ZDefocus
To connect ZDefocus:
1. Create a ZDefocus node by clicking Filter > ZDefocus.
USER GUIDE
766
Simulating Depth-of-Field Blurring | Connecting ZDefocus
2. Connect the ZDefocus node’s image input to the image you want to blur.
Note that this image also needs to contain a depth map channel. If your depth channel and rgba
channels exist in two different files, you can use a ShuffleCopy node to combine them.
3. Use the channels dropdown menu to select the channels you want to blur.
4. Set depth channel to the channel in the image input that contains the depth map. By default, the
depth information is taken from depth.Z.
Note that the depth map should not be anti-aliased. If it is, pixels along an edge between two
objects can be assigned a depth that is in-between the depth of the front object and back objects.
This looks wrong, as it suggests that those edge pixels are floating somewhere between the
objects.
5. If you want to use your own filter image (rather than the predefined disc or bladed images),
connect that to the ZDefocus node’s filter input and set filter type to image in the node’s
properties.
The filter image represents the shape and size of the camera aperture used to shoot the input
footage. As the clip in the image input is blurred, any highlights in the clip bloom in the shape of
the filter image, creating a bokeh effect.
Note: Loading scripts from pre-Nuke 8.0v7 enables the legacy resize mode checkbox
automatically, for backward compatibilty, and uses the filter bounds dropdown to
determine how images used in filtering are resized.
Adding new ZDefocus nodes hides the legacy resize mode checkbox and allows you to use
the image filter dropdown to give you more flexibility when calculating blur.
See Using a Custom Filter Image for more information.
You can create a filter image using the Roto node (Draw > Roto) or the Flare node (Draw > Flare),
for example.
Note that you don’t necessarily need to crop the filter image to a smaller size, as fast Fourier
transforms are used to speed up convolutions with large filter images.
6. Attach a Viewer to the ZDefocus node.
USER GUIDE
767
Simulating Depth-of-Field Blurring | Adjusting Blur Settings
A ZDefocus script.
7. Proceed to Adjusting the Shape of Out-of-Focus Highlights below.
Adjusting Blur Settings

To adjust the Blur settings:
1. Use math to specify how you want to use the depth channel to calculate the distance between
the camera and an object. For example, some programs use higher values to denote further
away, while in others they mean closer to the camera:
• direct - The Z value in the depth channel directly controls blur. For example, if Z is 0.5, then the
blur size is 0.5 times the value of the size control (unless this is bigger than maximum, in which
case it is clamped to maximum).
• depth - The Z value in the depth channel is the distance between the camera and whatever is in
the image at that pixel.
• far = 0 - The Z value in the depth channel is equal to 1/distance. The values are expected to
decrease from large positive values close to the camera to zero at infinity. This is compatible
with depth maps generated by Nuke and RenderMan.
• far = 1 - Near plane = 0, far plane = 1. This is compatible with depth maps generated by OpenGL.
• -direct - As with the direct mode, the Z value in the depth channel directly controls blur. In
other words, each layer is blurred by the same amount as in the direct mode. However, in this
mode, the layers are interpreted as being in the opposite order, so a higher depth value places a
layer in front of another rather than behind it.
• -depth - The Z value in the depth channel is -distance in front of the camera. This is the same
as depth, but the distances are negative to start with.
• far = -0 - The Z value in the depth channel is equal to -1/distance. This is compatible with depth
maps generated by Maya.
• far = -1 - Near plane = 0, far plane = -1.
2. In the Viewer, drag the focal point widget on top of the area that you want to be entirely in focus.
This automatically updates the focal point coordinates in the ZDefocus properties and sets the
focus plane control to the Z depth value at those coordinates. Any layers with this Z depth value
are left in focus. You can also set a depth slice around the focus plane that is in focus (as
described in step 4), but any other areas of the image are blurred according to the depth map.
USER GUIDE
768
Focal point widget under the green sign.
Focal point widget near the cars in the distance.
Tip: The fill foreground control, enabled by default, attempts to compensate for missing
information by filling regions in the foreground which are revealed when the foreground
goes out of focus. However, because the true image information isn't available in these
regions, fill foreground can sometimes introduce undesirable artefacts by adding things
which aren't there.
If you see blurry artefacts in the foreground, rather than sharp edge artefacts, try disabling
this control.
To better see the effect of your changes, you can also set output to focal plane setup. ZDefocus
now displays depth-of-field (DOF) info in the rgb channels:
• red - Less than DOF (in front of the area that’s in focus).
USER GUIDE
769
• green - Inside DOF (in focus). Note that if depth of field is set to 0, nothing is displayed in
green.
• blue - Greater than DOF (behind the area that’s in focus).
When show image is enabled, this information is overlaid upon the input image.
Focal point widget under Focal point widget near

the green sign. the cars in the distance.
3. To widen the area that is entirely in focus, increase depth of field. This sets a depth slice around
the focus plane that is entirely in focus (and shown in green whenever output is set to focal
plane setup). True theoretical depth of field would set this to zero.
Depth of field set to 0. Depth of field set to 1.

4. To apply a small amount of blur to the in-focus region, make sure blur inside is enabled. This
gives a smoother transition between the in-focus region and the out-of-focus regions around it.
5. Set output back to result and adjust the amount of blur at infinite depth by setting the size value.
Note that the amount of blur nearer the camera than the focus plane may be larger.
If you have set math to direct, the size is multiplied by the depth to give the blur size at that
depth. Setting size to 1 allows you to use the values in the depth map as the blur size directly.
6. If you increased the size value in step 5, it’s a good idea to also increase the maximum value. No
blurring greater than this value is generated no matter where the object is in relation to the
camera.
For maximum processing speed, you may want to keep this value as low as possible.
7. By default, automatic layer spacing is enabled, which means ZDefocus automatically works out
how many depth layers to use, based on the maximum blur size (maximum). In this mode, the
layers are closer together near to the focal plane, where a small change in the blur amount is
more obvious, and increasingly more widely-spaced further away (this is equivalent to setting
layer curve to a value of 1 when controlling the layers manually; see step 8).
USER GUIDE
770
To visualize the layers, you can set output to layer setup. This is like focal plane setup, but
displays depth-of-field (DOF) information after the depth has been divided into layers.
Output set to layer setup.

The maximum number of blur sizes that are used between 0 and maximum is 256. This means
you can have up to 256 layers behind the focal plane, and up to 256 in front of it as well.
8. If you uncheck automatic layer spacing, you can control the number of layers manually using
depth layers. Note that the more layers you use, the longer it takes to process the blur.
Depth layers set to 10. Depth layers set to 50.

If necessary, you can also increase layer curve to concentrate the layers around the focal plane.
The default value of 0 produces evenly spaced layers.
Layer curve set to 0. Increasing the layer curve

This produces evenly value concentrates the
spaced layers
layers. around the focal plane.
9. Proceed to Adjusting the Shape of Out-of-Focus Highlights.
USER GUIDE
771
Simulating Depth-of-Field Blurring | Adjusting the Shape of Out-of-Focus Highlights
Adjusting the Shape of Out-of-Focus Highlights

As the clip in the image input is blurred, any out-of-focus highlights (bokeh) in the clip assume the
shape of the filter image (of the filter type selected).
The original image. Out-of-focus highlights

(bokeh) in the shape of
the filter image.
How you create the filter image is up to you. You can:

• Use a predefined disc shape as the filter image. See Using a Predefined Disc Image.
• Use a predefined bladed image (polygon) as the filter image. See Using a Predefined Bladed Image.
• Use your own custom image in the filter input as the filter image. See Using a Custom Filter Image.
Using a Predefined Disc Image

1. To see the effect of your changes, set output in the ZDefocus controls to filter shape setup.
2. Set filter type to disc.
3. Use the filter shape control to dissolve the filter shape between Gaussian at 0 and disc at 1.
Filter shape set Filter shape set

to 0 (Gaussian). to 1 (disc). This is
the default value.
4. Use the aspect ratio control to set the filter aspect ratio, which is 1:1 by default. Values less than
1 squeeze the filter on the x axis, and values larger than 1 squeeze it on the y axis.
This allows you to simulate the cat's eye effect, caused by vignetting inherent within some lens
designs.
USER GUIDE
772
Aspect ratio set Aspect ratio set

to 0.5. to 1.5.
5. Proceed to Enhancing Out-of-Focus Highlights.
Using a Predefined Bladed Image

1. To see the effect of your changes, set output in the ZDefocus controls to filter shape setup.
2. Set filter type to bladed.
3. Use the blades control to set the number of iris blades that make up the camera's diaphragm. A
value of 3 produces a triangle, 4 a square, 5 a pentagon, 6 a hexagon, and so on.
This field only accepts integers larger than 1.
Blades set to 5. Blades set to 8.

This is the default
value.
4. Adjust roundness to control the rounding of the filter polygon’s sides. A value of 0 equals no
rounding.
Roundness set Roundness set

to 0. to 8.
5. Use rotation to rotate the filter image in degrees. Positive values produce counter-clockwise
rotation, and vice-versa.
USER GUIDE
773
6. Use the aspect ratio control to set the filter aspect ratio, which is 1:1 by default. Values less than
1 squeeze the filter on the x axis, and values larger than 1 squeeze it on the y axis.
This allows you to simulate the cat's eye effect, caused by vignetting inherent within some lens
designs.
Aspect ratio set Aspect ratio set

to 0. to 2.
7. To adjust the distribution of light inside the out-of-focus highlights:
• Adjust inner size to control the size of the inner polygon, as a percentage of the outer polygon.
Inner size set Inner size set

to 0.2. to 0.8.
• Use inner feather to add outward or inward feathering around the inner polygon. With values
larger than 0.5, your feather effect is outward and, respectively, if your values are smaller than
0.5, the feather effect is inward. A value of 0.5 produces no feathering.
Inner feather set Inner feather set

to 0.2. to 0.8.
• Adjust inner brightness to control the brightness of the inner polygon, where 0 is equal to black
and 1 to white.
USER GUIDE
774
Inner brightness Inner brightness

set to 0.3. set to 0.7.
• If you want to simulate catadioptric lenses, check catadioptric. When using catadioptric lenses,
the defocused areas of the image are annular, producing donut-shaped bokeh. You can use
catadioptric size to adjust the size of the catadioptric hole in the filter.
Catadioptric size Catadioptric size

set to 0.3. set to 0.7.
Using a Custom Filter Image

1. Set filter type in the ZDefocus controls to image.
This tells ZDefocus to use the filter input rather than a predefined shape as the filter image.
Note that the filter image can be a color image. This can be useful, for example, if you want to
add color fringing to your out-of-focus highlights to simulate chromatic aberration.
The filter image.

2. If you want to display the filter image in the Viewer, set output to filter shape setup.
USER GUIDE
775
Simulating Depth-of-Field Blurring | Enhancing Out-of-Focus Highlights
Note: Loading scripts from pre-Nuke 9.0v2 enables the legacy resize mode checkbox
automatically, for backward compatibilty, and uses the filter bounds dropdown to
determine how images used in filtering are resized.
Adding new ZDefocus nodes hides the legacy resize mode checkbox and allows you to use
the image filter dropdown to give you more flexibility when calculating blur.
3. To select what channel to use from the filter input, do one of the following:
• Set the filter channel dropdown menu to the channel you want to use. By default, this menu is
set to rgba.alpha.
• To use the same channels from the filter input as the image input (that is, whatever channels is
set to), check use input channels.
4. In newer Nuke scripts, use the image filter dropdown to select the required filter. See Choosing a
Filtering Algorithm for more information on the available filters.
Note: When using filters that employ sharpening, such as Rifman and Lanczos, you may
see a haloing effect. If necessary, check clamp image filter to correct this problem.
In older Nuke scripts with legacy resize mode enabled, use the filter bounds dropdown to select
whether you want to limit the filter image to:
• shape - the filter input’s bounding box. In this case, ZDefocus only uses the bounding box area,
and the center of the filter is the center of the bounding box. This is the default value. You may
want to use it, for example, if your filter input is a roto shape with a small bounding box that
doesn’t fill the entire format area.
• format - the filter input’s format. In this case, ZDefocus uses the entire format area, allowing
you to offset the filter image within the format.
Enhancing Out-of-Focus Highlights

To enhance the out-of-focus highlights:
1. To make bokeh lens shape effects warmer and more visible, check gamma correction.
This means a gamma lookup curve of 2.2 is applied before blurring and then reversed for the final
output.
USER GUIDE
776
Simulating Depth-of-Field Blurring | Masking Blur Effects
Without gamma With gamma correction.

correction.
2. You can also make the lens shape effects more visible by enabling bloom.
Any highlights above the bloom threshold value are multiplied by the bloom gain value.
Highlights below the bloom threshold are not affected.
This allows you more control over the highlights than gamma correction; however, gamma
correction may bring out some of the highlights better.
Without bloom. With bloom.

3. Proceed to Masking Blur Effects.
Masking Blur Effects

To mask the blur effect:
• Make sure there is a mask channel in the ZDefocus node’s image input and nothing is
connected to the mask input.
• Connect a mask to the mask input of the ZDefocus node. If you cannot see the mask input,
open the node’s controls and make sure mask is set to none.
If you want the mask from the mask input copied into the predefined mask.a channel, also
check inject. This way, you can use the mask input again downstream.
2. If you don’t want to use the alpha channel as the matte, select the channel you want to use from
the mask dropdown menu.
By default, the blur is limited to the non-black areas of this channel.
3. If necessary, check invert to reverse the mask, so that the blur is limited to the non-white areas of
the mask.
4. To blur the edges of the mask, check fringe.
USER GUIDE
777
Simulating Depth-of-Field Blurring | Masking Blur Effects
5. To dissolve between the original image (at 0) and the full ZDefocus effect (at 1), adjust the mix
slider. A small light-gray square appears on the node in the Node Graph to indicate that the full
effect is not used.
USER GUIDE
778
Creating Effects
Several nodes in Nuke let you create various effects on your input images. In this chapter, we describe
three of these nodes: LightWrap, Glint, and Text.
Quick Start
1. To adjust the soft edges and light spills that occur in the border between your foreground and
background, you can use the LightWrap node (Draw > LightWrap). For more information, see
Background Reflections on Foreground Elements.
2. With the Glint node you can add star-shaped glints on your image. See Creating Star Filter Effects
on Image Highlights.
3. Using the Text node, you can add text elements, such as credits, to your footage. For more
information, see Creating Text Overlays.
USER GUIDE
779
Background Reflections on Foreground Elements |
Background Reflections on
Foreground Elements
You can use the LightWrap node to create background reflections on foreground elements. The node
creates a reflection of light around the edges of your foreground element by blending in whatever is
in the background.
The composite without the LightWrap The composite with the LightWrap effect.
effect.
You may have noticed that objects filmed in front of very bright backgrounds have the appearance of
softened edges as the light spills round from the background. When adding foreground layers onto a
background plate, using the LightWrap node can dramatically improve the quality of your composite.
The composite without the LightWrap The composite with the LightWrap effect.
effect. Note how the foreground fits better into
Note the hard edges around the the background.
trailing hand and the bottom of the skirt.
If you want to use LightWrap, you should apply it on your foreground element before you composite
the foreground over the background with the Merge node.
USER GUIDE
780
Background Reflections on Foreground Elements | Using the LightWrap Node
Using the LightWrap Node

To use the LightWrap node:
1. Select Draw > LightWrap to add a LightWrap node after your foreground and background
images.
2. Connect your foreground element to input A of the LightWrap node, and the background image
to input B.
3. Connect a Viewer to the output of the LightWrap node so you can see the effect of your changes.
4. Adjust the Diffuse and Intensity sliders to control both the spread and brightness of the
reflections on the foreground element. These sliders need to be balanced out together. You may
want to start by bringing Diffuse all the way down to better see what you are blending in from the
background. Then, adjust Intensity before going back to the Diffuse slider and, if necessary,
Intensity again until you are happy with the result.
Low Diffuse. High Diffuse.
Low Intensity. High Intensity.

5. If you want to create a uniform effect around the edges of the foreground rather than have the
effect adjust itself according to the background, check Disable luminance based wrap on the
LightWrap tab.
6. In case you don’t want to merge the LightWrap effect with the foreground element in order to
keep the LightWrap effect as a separate element, check Generate wrap only on the LightWrap
tab.
USER GUIDE
781
Background Reflections on Foreground Elements | Using the LightWrap Node
7. By default, the LightWrap effect is only applied inside the foreground element’s alpha. If you want
to extend the effect outside it, making the element seem to glow, check Enable Glow.
8. On the Tweaks tab, you can also adjust the following controls:
• FGBlur to determine how much the foreground matte is blurred. The more blur, the more of
the background is added to the foreground.
• BGBlur to control how much the background is blurred before it is merged with the foreground
element.
• Saturation to adjust the saturation of the effect.
• Luma Tolerance to increase or decrease the luminance values of the effect.
• Highlight Merge to control how the foreground element is merged with the background. The
default merge operation, called plus, adds the elements together, producing a glow effect.
• Check Use constant highlight to use a constant color of your choice rather than the
background in the LightWrap effect. Select the color using the controls next to Constant
Highlights Color.
9. On the CCorrect tab, you can color correct the LightWrap effect produced.
USER GUIDE
782
Creating Star Filter Effects on Image Highlights | Using the Glint Node
Creating Star Filter Effects on

Image Highlights
With the Glint node, you can create star-shaped rays around all the bright points in an image.
The original image with The image after using

bright points. the Glint node.
Using the Glint Node

To use the Glint node:
1. Select Draw > Glint to add a Glint node after the image you want to add star-shaped rays to.
2. From the channels dropdown menu and checkboxes, select the channels to which you want to
apply the effect.
3. In the no. of rays field, enter the number of rays you want coming out of the bright points in your
image. For example, if you want to create five-pointed stars, enter 5.
4. To change the threshold for how bright the highlights in the input image need to be to cause the
glint effect, adjust the tolerance slider. Only the pixels above the threshold bloom with the effect.
Low tolerance value. High tolerance value.

5. To determine the length of the rays, adjust the length slider. To give every other ray a different
length and determine that length, adjust the odd ray length slider.
USER GUIDE
783
6. To determine how many steps the rays are formed of, enter a value in the steps field. The more
steps you use and the shorter the rays are, the smoother the rays become.
The steps value set to 3. The steps value set to 5.

7. To rotate the star-shapes, adjust the rotation slider. Increasing the value rotates the rays
clockwise, whereas decreasing the value rotates them counter-clockwise.
8. To change the color in the beginning of the rays near the center point of the stars, adjust the from
color slider. To change the color in the end of the rays, adjust to color. By default the from color
is set to white, and the to color to black.
From color set to white From color set to pink

and to color set to black. and to color set to yellow.
9. If needed, you can also make the following adjustments:
• If you want to change the aspect ratio of the stars, adjust the aspect ratio slider.
• By default, the brightest image on the rays is used as the center point for the star. However, if
you prefer the images forming the rays to be added up in forming the center point, uncheck
max.
• To only output the Glint effect without merging it into the original input image used to create it,
check effect only.
USER GUIDE
784
Effect only disabled. Effect only enabled.

• To mask the shape that is used to create the rays, check w and select the mask channel from the
dropdown menu.
• To perform a gamma correction on the highlights that cause glint before the glint effect is
applied, adjust the gamma slider.
• To mask the glint effect, check mask and select a mask channel using the controls on the right.
• To dissolve between the original input image and the full glint effect, adjust the mix slider.
USER GUIDE
785
Creating Text Overlays | Preparing a Text Overlay
Creating Text Overlays

Using Nuke’s Text node, you can add text overlays on your images. You can simply type in the text
you want to have displayed, or use Tcl expressions (such as [metadata values]) or Tcl variables to
create a text overlay.
See Entering Text for a list of Tcl expressions, Tcl variables, HMTL named entities, hex entities, and
decimal entities you can use in the Text node.
Text overlays can also be animated using animation layers so that their properties, such as position,
size, and color, change over time. These features make the Text node useful, for example, for
creating slates or scrolling credits.
Preparing a Text Overlay

Creating text overlays involves some preparation, particularly which channels hold the text and the
positioning of the initial text box.
1. Select Draw > Text to create a Text node and connect it to a Viewer.
2. In the Text node properties, select the channels you want the text to appear in from the output
controls.
3. If you want to multiply any channels by the drawn text so that they are set to black outside the
text shape, select those channels using the premult controls.
4. From the clip to dropdown menu, select how you want to restrict the output image.
• no clip - do not restrict the output image.
• bbox - restrict the output image to the incoming bounding box.
• format - restrict the output image to the incoming format area.
• union bbox+format - restrict the output image to the combination of the incoming bounding
box and format area.
• intersect bbox+format - restrict the output image to the intersection of the incoming bounding
box and format area.
5. If you want to clear the affected channels to black before drawing on them, check replace.
By default, replace is not enabled and the text is drawn on top of the input image.
6. If you're happy with the text initially appearing in the top left corner of the format area, you can
start Entering Text using the message field.
7. If you'd like to have more control over the text's initial position or draw a text box to constrain
your text, ensure that edittext is enabled above the Viewer and then either:
• Click in the Viewer to place the cursor where you want the text to appear,
USER GUIDE
786
Creating Text Overlays | Preparing a Text Overlay
OR
• Draw a box in the Viewer to contain your text.
Tip: The cursor appears as in the Viewer when you initially add the Text node.
Text entered in a custom box is limited by the boxx, r, and t boundaries on the Transform tab,
but the y boundary can be overstepped.
Tip: You can adjust the box at any time using the boxxyrt controls or by dragging the Viewer
handles.
8. If you want to mask out part of the image, see Masking Regions of the Viewer
9. Once you're happy with the cursor position, proceed to Entering Text.
USER GUIDE
787
Entering Text | Example Variables and Entities
Entering Text
The message field in the properties panel is used to enter the text, Tcl expressions, Tcl variables, or a
combination of these that you want to display in your output. For some examples of these formats,
see the table below.
Text entry behaves much the same as regular text editors, but there are a few rules to observe:
• The edit text control above the Viewer must be enabled if you want to type directly in the Viewer,
though you can enter text in the message field at any time.
• Press Return to begin a new line in both the Viewer and the message field.
• Navigate around the Viewer or message field using the arrow keys.
• Tcl expressions must be placed in square brackets, such as [date].
• To display special Unicode characters, such as foreign language characters and copyright signs, you
can:
• Use HTML named entities, such as < to display <, or © to display ©.
• Use hex entities, such as < to display <, or © to display ©.
• Use decimal entities, such as < to display <, or © to display ©.
• Type Unicode characters, such as < or ©, on your keyboard or cut and paste them from other
applications. UTF-8 character encoding is used to store them in the control's value and in the
saved Nuke script.
These special characters only work if the font you are using supports the required character.
Note: We recommend using the above entities rather than typing <, for example. This is
because future versions of the Text node may interpret HTML mark-up. In HTML, some
characters, such as the greater than and less than signs, are reserved. If you used these signs
within your text now, future versions could mistake them for HTML mark-up.
Example Variables and Entities

The following table gives examples of Tcl expressions, Tcl variables, HMTL named entities, hex
entities, and decimal entities you can use in the message field of the Text node.
USER GUIDE
788
Message Prints
Tcl expressions
[date] Week day, day, month, hh:mm:ss, and time zone. For example, Thu Jan 15
14:22:20 GMT.
[date %a] Abbreviated week day name. For example, Thu.
[date %A] Full week day name. For example, Thursday.
[date %b] Abbreviated month name. For example, Jan.
[date %B] Full month name. For example, January.
[date %d] Day (01-31).
[date %D] Date (dd/mm/yy). For example, 15/01/10.
[date %H] Hour (00-23).
[date %I] Hour (01-12).
[date %m] Month (01-12).
[date %M] Minutes (00-59).
[date %p] AM or PM.
[date %r] Time (12-hour clock). For example, 11:04:07 AM.
[date %S] Seconds (00-59).
[date %T] Time (24-hour clock). For example, 14:06:54.
[date %y] Abbreviated year (00-99). For example, 10.
[date %Y] Full year. For example, 2010.
[date %z] Numeric time zone. For example, -0800.
[date %Z] Time zone. For example, GMT.
[frame] Frame number. For example, 23.
[metadata] List of all the keys in the incoming metadata.
USER GUIDE
789
Message Prints
[metadata values] List all of the keys and values in the incoming metadata.
[metadata key] Value of the key in the incoming metadata. Replace key with the name of the
key whose value you want to display. For example, you may be able to use
[metadata input/filename] to display the name and location of the image
file, or [metadata input/ctime] to display the timestamp for an input file.
[value root.name] Script directory path and script name. For example, Users/john/Nuke_
scripts/myscript.nk.
Tcl variables
$env(ENVIRONMENT_ The value of the environment variable specified. Replace ENVIRONMENT _

VARIABLE) VARIABLE with an environment variable you have set. For example, you can
use $env(USER) to display the user name (for example, john) on Mac and
Linux, or $env(USERNAME) to display it on Windows and Linux.
For a list of environment variables specific to Nuke, see Environment

Variables.
$version_long The full version number of Nuke. For example, 11.3v5.
$threads Number of render threads used to calculate images. This is in addition to a

main thread used to update the graphical user interface (GUI).
HTML named entities
& &
' '
Å Å
Á Á
Â Â
Æ Æ
À À
Ç Ç
USER GUIDE
790
Message Prints
© ©
é é
ê ê
è è
ë ë
€ €
> >
< <
Ñ Ñ
ø ø
õ õ
Ö Ö
ö ö
" "
® ®
ß ß
Ü Ü
ü ü
hex entities
# #
% %
& &
* *
USER GUIDE
791
Message Prints
@ @
™
œ œ
š š
< <
> >
© ©
é é
decimal entities
£ £
© ©
® ®
¿ ¿
ê ê
ß ß
à à
Tip: To get a list of all the Tcl expressions you can use with date, type x on the Node Graph,
set the script command dialog that opens to Tcl, enter date -h, and click OK.
USER GUIDE
792
Fonts and Font Properties | Example Variables and Entities
Fonts and Font Properties

The FreeType library used by the Text node supports a large number of fonts, including TrueType,
OpenType, and PostScript fonts. In order to display the available fonts faster in the font dropdown
menu, these fonts are cached as an XML file when you open a Text node's properties panel. The font
cache XML file is called fontmapping.fcache and its location is specified by the environment variable
NUKE_TEMP_DIR. See Updating the Font Cache for more information.
Note: See Environment Variables for more information on setting environment variables.
Note: The font cache file is not used when rendering and is not required on a render farm.
Nuke retrieves the fonts available in the Text node from various locations in a set order, before
caching them:
1. The location specified in the Project Settings > Font > project font path control.
2. The .nuke/fonts directory and all other plug-in folders.
3. The location specified by the NUKE_FONT_PATH environment variable.
Note: Locations 1 and 3 allow you to specify multiple paths using the OS standard syntax.
On Windows, for example, you could specify "c:\windows\font1;c:\windows\font2;..."
4. The local fonts folder.

5. The system fonts folder, assuming that Project Settings > Font > include system fonts is
enabled. For example, on Windows C:\Windows\Fonts\.
Note: If several locations contain the same font, Nuke uses the font from the first directory
it finds it in.
In general, there is no guarantee that a font family and style, such as {Arial : Regular}, will render
identically cross-platform. To avoid this, Nuke ships with several fonts common to all platforms. If
you disable Project Settings > Font > include system fonts, only the fonts that ship with Nuke are
available in the font controls.
See Third-Party Libraries and Fonts for a list of fonts that ship with Nuke.
USER GUIDE
793
Fonts and Font Properties | Selecting a Font
Note: Only fonts that are physically stored in a file are available in the font controls. As a
result, the list of system fonts available in Nuke may be different to that available in other
software applications on the same machine.
Selecting a Font
1. Highlight the text you want to affect in the message field or Viewer.
2. In the Text properties panel, click the font family dropdown to select the required family.
3. Select the required font style from the dropdown.
Note: The styles available depend on the family selected.
4. Click the information icon to display the location of the selected font. For example:
The font family and style used are saved in the Nuke script, along with some extra information, in
order to make sure the same family and style are used if the script is opened on a different machine
or operating system. If the font family and style are not the same, Nuke displays a warning message.
Adjusting Font Size and Spacing

1. Highlight the text you want to affect in the message field or Viewer.
2. To adjust the overall size of the font, use the font size slider. When leading is set to 0, this
parameter also controls the spacing between each line of text.
When rendering the font, font size controls the font hinting used. Font hinting adjusts which
pixels are interpolated to more clearly render a font. At small sizes and on low resolution output
devices, it has a big impact on the legibility of the font. For best results, you should use the font
size parameter (rather than the scale control on the Group tab) to control the size of the font and
keep scale set to 1.
USER GUIDE
794
Fonts and Font Properties | Adjusting Font Size and Spacing
Tip: Use the global font scale control to adjust the scale multiplier for all characters in the
current Text node, regardless of the font size specified for individual characters.
3. You can adjust fonts asymmetrically using the font width and font height controls.
You can also use the controls above the Viewer to adjust font size.
4. To increase or decrease the spacing between individual characters, adjust the kerning slider. By
using negative values, you can make the letters overlap.
Note: If you select more than one character or the last character in a group, the kerning
control is disabled.
You can also use the control above the Viewer to adjust kerning.
5. To increase or decrease the space between each character and the previous character adjust the
tracking control. Negative values move characters towards the preceding character.
You can also use the control above the Viewer to adjust tracking.
6. To adjust the selected text's height above the baseline, in screen space, use the baseline shift
control. The baseline is the imaginary line upon which most letters rest. This control allows you to
reliably line up the baseline of different fonts.
You can also use the control above the Viewer to adjust baseline.
7. If your text overlay contains several lines of text, you can adjust the spacing between each line by
using the leading slider. By using negative values, you can make the letters overlap.
USER GUIDE
795
Fonts and Font Properties | Justifying Fonts
Note: Unlike the baseline shift control, leading affects all text within the bounding box
regardless of selection.
Justifying Fonts
The justify controls affect the positioning of the text within the bounding box on the x and y axes.
1. Use the xjustify dropdown to control how you want to align the text horizontally:
• left - align the text along the left edge of the on-screen text box. This leaves the right edge of the
text ragged.
• center - align the text from the center of the on-screen text box. This leaves both edges of the
text ragged.
• right - align the text along the right edge of the on-screen text box. This leaves the left edge of
the text ragged.
USER GUIDE
796
• justify - align the text both along the left and the right edge of the on-screen text box. This
leaves no ragged edges. The justification is done by expanding the spaces between letters. If
there are no spaces or the spaces get more than about three times wider than they were, letters
are expanded.
2. Use the yjustify dropdown to control how you want to align the text vertically:
• top - align the text against the top edge of the on-screen text box.
• center - align the text from the center of the on-screen text box.
USER GUIDE
797
• bottom - align the text against the bottom edge of the on-screen text box.
USER GUIDE
798
Updating the Font Cache | Justifying Fonts
Updating the Font Cache

Once the font cache has been generated, Nuke accepts the content of the cache and no longer
generates it from scratch every time it is requested by the font controls. If you add or remove fonts,
the cache is not automatically updated and you need to update it manually. There are several ways to
do so:
• Navigate to Project Settings > Font and click Rescan font paths.
• Manually remove the fontmapping.fcache.xml font cache file from NUKE_TEMP_DIR and restart
Nuke.
• Open Nuke's Script Editor and execute nuke.rescanFontFolders()
Go to Help > Documentation > Python Developer's Guide for more information on Python
commands relating the font cache.
USER GUIDE
799
Transforming Text | Transforming Selections
Transforming Text
The Text node allows you to transform selections in the Viewer or create groups of selections in the
animation layers table and transform them together. Both methods support animation using
keyframes.
Transforming Selections
Selections in the Viewer are transformed by hand, that is, by manipulating handles drawn on your
selections.
1. Enable transform in the controls above the Viewer.
2. Select the required character(s) in the Viewer to enable the transform handles.
3. If you hover the cursor over a selection, the cursor changes to show the handle's function
depending on position.
4. Translate, rotate, and scale your selections by dragging the required handles.
If you want to animate your changes, see Animating Transforms for more information.
Tip: By default, scaling is done uniformly around the center of your selections. To scale
relative to the opposite corner or side instead, hold down Ctrl/Cmd while dragging the
transform handles.
USER GUIDE
800
Transforming Text | Transforming Groups
Transforming Groups
The Group tab saves group selections in the animation layer table allowing you to affect layers
separately using the transform controls.
1. Open the properties and click the Group tab to display the transform controls and animation
layers table.
2. Select the required character(s) in the Viewer, or message field, and then click the + button to
create an animation layer, or click create group above the Viewer.
You can create as many animation layers as required, each one is named according to the
selections you made in the Viewer.
3. Select an animation layer and use the transform controls on the Group tab to transform your
selection.
If you want to animate your changes, see Animating Transforms for more information.
4. Click the - button to remove an animation layer.
USER GUIDE
801
Animating Transforms | Animating Selections
Animating Transforms
The Text node supports animation in a similar way to most other Nuke nodes, in that you can
animate individual controls in the properties panel, but the Group tab animation layer table enables
you to visualize your animations more easily.
Animating Selections
1. Set the starting position for the selection(s) using the Viewer handles as described in
Transforming Text.
2. Click the button above the Viewer to add a starting keyframe.

3. Move the playhead to a new frame and transform the selection using the Viewer handles.
A keyframe is added automatically on the new frame.
4. Repeat the process to produce the required animation.
5. Click to remove a keyframe at the current frame.
Animating Groups
1. Set the starting position for the groups using the transform controls as described in Transforming
Text.
2. Select a group in the animation layers table and then click the button above the Viewer to add
a starting keyframe.
3. Move the playhead to a new frame and transform the group using the transform controls or the
handles in the Viewer.
5. Click to remove a keyframe at the current frame.
USER GUIDE
802
Changing the Text Color | Animating Groups
Changing the Text Color

To change the text color:
1. In the Text node properties, click the Color tab.
2. Adjust the color parameter or click on the color picker button to select a color for the text.
3. If you want to create a color gradient across the text, select anything other than none from the
ramp:
• linear - the ramp changes linearly from one color into another.
• smooth0 - the ramp color gradually eases into the point 0 end. This means colors in the point 0
end are spread wider than colors in the point 1 end.
• smooth1 - the ramp color eases into the point 1 end. This means colors in the point 1 end are
spread wider than colors in the point 0 end.
• smooth - the ramp color gradually eases into both ends. This means colors in the point 0 and
point 1 ends are spread wider than colors in the center of the ramp.
Linear ramp. smooth ramp.
smooth0 ramp. smooth1 ramp.

4. Use the color control to select a color for the ramp at the point 1 end (by default, the top end).
Then, use color 0 to select the color for the ramp at the point 0 end (by default, the bottom end).
5. To adjust the spread and angle of the ramp, drag the p0 and p1 Viewer handles to a new location.
You may need to press O twice on the Viewer to make these controls appear.
Alternatively, you can enter new x and y values for the point 1 and point 0 controls in the Text
node properties.
USER GUIDE
803
Masking Regions of the Viewer | Animating Groups
Masking Regions of the Viewer

If you want to mask the effect of the Text operation, the easiest way is to use a Roto node connected
to the mask input:
1. After connecting the Roto node, draw the mask in the Viewer. You can use any channel, but Roto
defaults to the alpha channel.
2. Select the mask channel from the Text node's mask dropdown menu, in this case rgba.alpha.
3. To invert the mask as shown, check invert.
Standard masking. Inverted masking.
USER GUIDE
804
Adding Shadows to Text | Animating Groups
Adding Shadows to Text

You can add shadows to your text using the Shadows tab in the Text node's properties.
To add a shadow to your text, simply select the Shadows tab and check enable drop shadows.
If you want the drop shadow to be the same color as the input image, enable inherit input color.
Otherwise, use the color controls to select a color for the drop shadow.
Use opacity to adjust the opacity of the drop shadow relative to the input's alpha channel.
Opacity set to 0.4. Opacity set to 0.8.
To adjust the direction of the shadow, use the angle slider. A value of 0 or 360 equals left direction.
USER GUIDE
805
Angle set to 0 (or 360). Angle set to 225.
Use distance to set the distance of the shadow from the input content.
Distance set to 15. Distance set to 45.
If you want to blur the shadow, increase the softness value.
Softness set to 0. Softness set to 20.
If necessary, set the shrink/expand slider to a negative value to erode the shadow or a positive value
to dilate it.
USER GUIDE
806
Shrink/expand set to -5. Shrink/expand set to 5.
USER GUIDE
807
Adding a Drop Shadow | Animating Groups
Adding a Drop Shadow

The DropShadow gizmo allows you to create a drop shadow for any input image that has an alpha
channel with values greater than 0. To add a drop shadow an image:
1. Select Filter > DropShadow to insert a DropShadow gizmo after your image Read node.
2. Connect your background image to the DropShadow gizmo's bg input. Your image and drop
shadow are merged over the background image.
3. In the DropShadow gizmo properties, check enable drop shadow to enable the drop shadow
effect.
with drop shadow with drop shadow

disabled enabled
4. If you want the drop shadow to be the same color as the input image, enable inherit input color.
Otherwise, use the color controls to select a color for the drop shadow.
with black selected with the inherited

input color
5. Use opacity to adjust the opacity of the drop shadow relative to the input's alpha channel.
Opacity set to 0.4 Opacity set to 0.8
USER GUIDE
808
6. To adjust the direction of the shadow, use the angle slider. A value of 0 or 360 equals left
direction.
Angle set to 0 Angle set to 225

(or 360)
7. Use distance to set the distance of the shadow from the input content.
Distance set to 5 Distance set to 15

8. If you want to blur the shadow, increase the softness value.
Softness set to 0 Softness set to 10

9. If necessary, set the shrink/expand slider to a negative value to erode the shadow or a positive
value to dilate it.
Shrink/expand Shrink/expand
set to 0 set to 3.5
USER GUIDE
809
10. Use the input [operation] bg dropdown menu to select how you want to combine the input and
bg inputs.
USER GUIDE
810
Analyzing and Matching Clips
This chapter concentrates on the CurveTool node. The node analyzes an aspect of a frame sequence
and creates an animation curve based on the analysis. You can then use the curve data to drive
effects elsewhere. For instance, you can add matching flicker to a CG render.
Introduction
You can use the CurveTool node to analyze four different aspects of your frame sequence, depending
on which curve type you select in the node controls:
• AutoCrop finds black regions (or any color you pick) around the edges of the frame sequence and
tracks their size and position over time. This is useful for running a Crop node to remove
unnecessary outer pixels and speed up the calculation. For more information, see Cropping Black
Edges .
• Avg Intensities is useful for obtaining the average pixel values in a frame sequence and then
matching that intensity elsewhere. It takes the first value in the frame range and the next value
selected, adds them together and divides by two, returning the average between the two. You might
want to use it to match the background plate’s fire flicker in the smoke in the foreground plate, for
example. For more information, see Analyzing the Intensity of a Frame Sequence and Removing
Flicker .
• Exposure Difference analyzes the exposure changes in the frame sequence. It takes the first value
in the frame range and the next value selected, and returns the difference between the two. You can
use the results to match the same exposure elsewhere. For more information, see Analyzing
Exposure Differences.
• Max Luma Pixel tracks the brightest and dimmest pixels in the frame sequence. This can be useful
in the following case, for example. Let’s say you have a night-time sequence depicting a person
moving inside a dark house holding a flashlight, and want to add lens flare on the moving flashlight.
Knowing where the brightest pixel is located over time allows you to match-move the lens flare and
position it correctly without having to manually animate it. For more information, see Tracking the
Brightest and Darkest Pixels .
USER GUIDE
811
Analyzing and Matching Clips |
Tip: If you are familiar with Shake, you may have used the PixelAnalyzer node. The
CurveTool node is the Nuke equivalent of PixelAnalyzer.
USER GUIDE
812
Cropping Black Edges |
Cropping Black Edges

You can crop black edges (or any color you choose) from your footage to eliminate unnecessary
computation:
1. Select Image > CurveTool to insert a CurveTool node after the image sequence you want to
analyze.
2. Make sure a Viewer is connected to the CurveTool node.
3. In the CurveTool controls, select AutoCrop from the Curve Type dropdown menu.
4. Using the color parameters, select the color you want to track.
5. To control how far the color can deviate from the selected color and still be cropped off, use the
Intensity Range slider.
6. From the channels dropdown menu and checkboxes, select the channels you want to analyze.
7. If you want to analyze an area in the frames rather than entire frames, define a region of interest
either by dragging the edges of the frames to a new position in the Viewer, or by defining the area
using parameters labeled ROI.
8. Click Go! to analyze the frames. This opens the Frames to Execute dialog.
9. In the dialog, define the frames to analyze. Enter the first frame, followed by a comma and the
last frame. Click OK. Nuke starts analyzing the frame sequence.
10. You’ll find the results of the analysis on the AutoCropData tab where the parameter values have
turned blue to indicate they are animated over time. To see the animation curve, right-click on a
parameter field and select Curve editor.
Once Nuke has created the animation curve, you can copy the animation or any of its values into a
Crop node, for example, to match the analyzed crop area there. Ctrl/Cmd+click on the animation
button and drag and drop it to another parameter to create an expression linking the two.
USER GUIDE
813
Analyzing the Intensity of a Frame Sequence |
Analyzing the Intensity of a Frame

Sequence
You can analyze your footage to find out the average intensity values in it:
1. Select Image > CurveTool to add a CurveTool node in an appropriate place after the image
sequence you want to analyze and match.
2. Connect a Viewer to the CurveTool.
3. In the node’s controls, select Avg Intensities from the Curve Type dropdown menu.
4. Select the channels you want to analyze from the channels dropdown menu and checkboxes.
5. By default, the region of interest that is to be analyzed covers the entire frame. If you want to
analyze a smaller area, resize and reposition the region of interest in the Viewer by dragging its
edges to a new position. You can also resize the region of interest using the ROI parameters in
the properties panel.
6. In the # frames for base average field, enter the range of frames that each frame being
analyzed is compared against. The frames are compared onwards from each frame analyzed.
Thus, a value of 1 would compare each frame to the frame following it, whereas a value of 5
would compare each frame to the following 5 frames.
The higher frame range you use, the more accurate and time-consuming the calculation becomes.
However, a high frame range is not always needed. For analyzing and matching fire flicker, you’d
probably want to go frame by frame, whereas removing flicker would require a wider frame range
to ensure a good average is obtained as the result.
7. To analyze the sequence, click Go!. This opens the Frames to execute dialog.
8. In the dialog, specify the frame range you want to analyze and match. Enter the first frame,
followed by a comma and the last frame. Click OK. Nuke now analyzes the frame sequence.
9. Move to the IntensityData tab where you’ll find the results of the analysis. You’ll notice that the
parameter input fields have turned blue. This indicates that they are animated. To see the
animation curve, right-click on the values and select Curve editor.
color correction node, for example, to match the analyzed intensity there. Ctrl/Cmd+click on the
animation button and drag and drop it to another parameter to create an expression linking the two.
USER GUIDE
814
Removing Flicker |
Removing Flicker
You can also use the CurveTool to stabilize flickering in your footage. To do this:
1. Connect a CurveTool node to your footage.
2. In the Curve Type dropdown, select Avg Intensities.
3. If necessary, select the channels you want analyzed in the channels controls, and adjust the region
of interest (ROI) box in the Viewer to cover your analysis area.
4. When you’re ready, click Go! and specify a frame range in the Frames to Execute dialog that
opens. CurveTool analyzes your footage, and the resulting values appear on the IntensityData
tab.
5. Create a Grade node (Color > Grade) and connect it to your footage.
6. From the CurveTool’s IntensityData tab, Ctrl/Cmd+drag the analysis result to the multiply field in
the Grade node controls.
7. Right-click on the multiply field, and select Edit expression, or press the equals (=) button with the
field selected.
8. In the dialog that appears, add 1/ in front of the Expression field entry. This inverts the brightness
values detected by the CurveTool node, and enables the Grade tool to stabilize brightness
changes causing flickering in your footage.
USER GUIDE
815
Analyzing Exposure Differences |
Analyzing Exposure Differences

You can analyze the differences in exposure in your frame sequence:
1. Select Image > CurveTool to add a CurveTool node after the image sequence you want to analyze.
2. Add a Viewer after the CurveTool.
3. Under Curve Type, select Exposure Difference.
4. From the channels dropdown menu and checkboxes, select the channels you want to analyze.
5. If you want to analyze an area in the frame rather than the entire frame, define a region of
interest either by dragging the edges of the frame box to a new position in the Viewer, or by
defining the area using parameters labeled ROI.
6. To analyze the sequence, click Go!. The Frames to Execute dialog opens.
7. Specify the frame range you want to analyze. Enter the first frame, followed by a comma and the
last frame. Click OK. Nuke now performs the analysis.
8. You can find the results of the analysis on the IntensityData tab where the parameter input fields
have turned blue to indicate they are animated. To see the animation curve, right-click on one of
the input fields and select Curve editor....
color correction node, for example, to match the analyzed exposure there. Ctrl/Cmd+click on the
animation button and drag and drop it to another parameter to create an expression linking the two.
USER GUIDE
816
Tracking the Brightest and Darkest Pixels |
Tracking the Brightest and

Darkest Pixels
You can track the brightest and the darkest pixels in your frame sequence:
1. Select Image > CurveTool to add a CurveTool node after the image sequence you want to analyze.
2. Connect a Viewer to the CurveTool.
3. From the Curve Type dropdown menu, select Max Luma Pixel.
4. Click Go! to analyze the frame sequence. This opens the Frames to Execute dialog.
5. Define the frame range you want to analyze. Enter the first frame, followed by a comma and the
last frame. Then, click OK. Nuke analyzes the frame sequence, tracking both the position and the
values of the brightest and darkest pixels.
6. You can find the results of the analysis on the MaxLumaData tab. You’ll notice that the input
fields have turned blue to indicate that they are animated over time. To see the animation curve,
right-click on an input field and select Curve editor.
Once Nuke has created the animation curve, you can copy the animation or any of its values into
another node to match that node’s effect with the brightest or darkest pixels in the frame sequence.
Ctrl/Cmd+click on the animation button and drag and drop it to another parameter to create an
expression linking the two.
USER GUIDE
817
3D Compositing
Nuke’s 3D workspace allows you to set up a 3D composite for camera moves, set replacement, and
other applications where you need to simulate a "real" dimensional environment.
Overview
This chapter explains how to set up a 3D scene in Nuke, and how to add objects and cameras in the
3D workspace. You’ll also see how to texture objects, transform objects and cameras, and render out
scenes for use in other areas of your script.
Although the 3D workspace has many potential uses, you’re most likely to use it - at least initially - to
create pan-and-tile scenes. These are scenes with 2D image planes arranged into a curved shape, and
then rendered out through an animated camera to give the illusion of a seamless environment.
Simple pan-and-tile scene.
The 3D objects in Nuke appear as round shapes to differentiate them from objects that perform 2D
operations. As shown above, you can mix 2D and 3D objects together in the node tree. For example,
you can texture a 3D object with a 2D clip, or take the rendered output from a 3D scene and use it as
a 2D background.
USER GUIDE
818
3D Compositing |
Script with 2D and 3D operators
USER GUIDE
819
Setting Up a Scene | The Scene Node
Setting Up a Scene
Each 3D scene includes the following objects: a Scene node, a Camera node, one or more geometry
nodes (i.e., card, sphere, obj), and a ScanlineRender node. Examples of 3D scenes are shown in
Overview and the figure below. In the example, the Scene node receives the output from the
geometry nodes and sends the composite of those objects to the ScanlineRender node, where the
output is converted back to 2D.
Core nodes for a 3D composite.
Your script may contain multiple Scene nodes, cameras, and 3D render nodes. All 3D objects loaded
in the Properties Bin appear in the 3D Viewer, regardless of whether they are connected to the same
Scene node.
The Scene Node

Regardless of its location in your script, the Scene node is the highest-level node in the scene
hierarchy because it references all the elements in a 3D workspace - all the geometric objects,
cameras, and materials.
To Add a Scene Node

Select 3D > Scene from the Toolbar.
USER GUIDE
820
Setting Up a Scene | The ScanlineRender Node
The ScanlineRender Node

Every Scene node in a script should be connected to a ScanlineRender node, which tells Nuke to
render the results of the scene. The ScanlineRender node also allows you to toggle between a 2D and
3D view of the scene.
To Add a ScanlineRender Node

1. Select the Scene node.
2. Select 3D > ScanlineRender from the Toolbar.
3. Connect the obj/scn input to a Scene or geometry node.
4. Connect the cam input to the main camera.
5. Connect the optional bg input to composite a background image into the scene.
6. Press Ctrl+I (Cmd+I on a Mac) to open a new Viewer to display the output of the ScanlineRender
node.
When an image is connected to the bg input, its resolution becomes the output resolution for the
ScanlineRender node.
The Camera Node

Cameras may be connected to either the Scene node or the ScanlineRender node. The camera
connected to the ScanlineRender node is the camera used for rendering.
1. Select 3D > Camera to insert a camera node.
2. Drag an output connector from the Camera node to a Scene node or connect the Camera node to
a ScanlineRender node’s cam input.
When connecting cameras for the 3D scene, the camera you want to use for rendering should be
connected to the ScanlineRender node, like this:
USER GUIDE
821
Setting Up a Scene | The Camera Node
Connecting cameras to the scene.
Any additional cameras should be connected to the Scene node. When you have multiple cameras
associated with a 3D scene, you can switch between them by selecting the viewing camera from the
dropdown menu at the top of the Viewer. See the next section, Using the 3D Viewer, for more
information.
USER GUIDE
822
Using the 3D Viewer | Switching to the 3D Viewer
Using the 3D Viewer

When you have a 3D setup in your script, any Viewer window can toggle between the 2D and 3D
display modes. The 2D mode shows the result of a rendered scene, the 3D mode shows the
perspective from cameras in the scene.
The 3D Viewer.
When you do not have a Camera node in your script, the 3D Viewer uses default views (see the figure
below for the list of options). These views are similar to different cameras that you can look though,
but they don’t appear as objects that you can manipulate in the scene.
Switching to the 3D Viewer

Open a Viewer and press Tab or V to toggle between the 2D and 3D modes - or select the view you
want from the dropdown menu at the top right corner of the Viewer window.
USER GUIDE
823
Using the 3D Viewer | To Navigate in the 3D Viewer
Switching to the 3D Viewer.
The “built-in” views give you different perspectives on your 3D scene. You can quickly switch between
the views by pressing the keyboard shortcuts for right view (X), left view (Shift+X), top view (C), bottom
(Shift+C), front (Z), back (Shift+Z), and three-quarter perspective (V).
To Navigate in the 3D Viewer

• Dolly: Press Alt and middle-mouse-button drag.
• Pan: Press Alt and left-mouse-button drag.
• Tilt: Press Ctrl/Cmd and left-mouse-button drag.
• Spin: Press Ctrl/Cmd and left-mouse-button drag.
• Roll: Press Ctrl/Cmd+Shift and left-mouse-button drag.
• Look through camera: Select a camera object, press H.
• Fit the scene: Press F to fit the entire 3D scene within the Viewer.
To Change the 3D Viewer Display Properties

1. Open the Preferences dialog (Shift+S), and select Panels > Viewer Handles.
USER GUIDE
824
Using the 3D Viewer | To Look Through a Camera
3D Viewer properties.
2. Make the desired changes to the 3D bg and fg colors.
3. From the 3D control type dropdown menu, select the navigation control scheme you want to use
(Nuke, Maya, Houdini, Lightwave, or Modo).
4. Click OK.
Note: The 3D control type also affects the mouse button assignments for panning and
zooming in the node graph and 2D Viewers.
To Look Through a Camera

1. Press V to make sure you are looking through the 3D perspective view, and not one of the
orthographic views.
2. From the 3D Viewer window, select the camera from the dropdown menu in the top right corner.
USER GUIDE
825
Using the 3D Viewer | To Lock the 3D Camera View
Note: This selection does not change the camera used for rendering. This changes only the
camera to “look through” for the current 3D Viewer.
Cameras in the current data stream automatically appear in the dropdown menu of cameras you can
select. To select a camera that doesn’t appear in the menu, double-click the camera node to open its
panel, and it is added to the menu.
To Lock the 3D Camera View

You can choose to lock the 3D view to the selected camera or light. You can toggle between the
unlocked and locked modes by clicking the 3D view lock button, or by pressing Ctrl/Cmd+L.
• unlocked: to freely move in the 3D view without restrictions. The 3D view lock button is gray.
• locked: to lock your movement to the camera or light you’ve selected in the dropdown menu on the
right side of the 3D view lock button. The 3D view lock button is red.
To Use the Interactive 3D Camera View Mode

With the interactive 3D camera view mode you can change the camera or light values according to
your movement in the Viewer. You can activate the interactive mode by Ctrl/Cmd+clicking the 3D
view lock button. When the interactive mode is on, the 3D view lock button turns green. In order to
activate the interactive mode, you need to have a Camera or a Light node selected in the dropdown
on the right side of the 3D view lock button.
When the interactive mode is on, you can use the plus (+) and the minus (-) keys to change the
translate values of the camera or light you’ve selected. When the interactive mode is off, these keys
zoom your view in and out.
USER GUIDE
826
3D Scene Geometry | To Use the Interactive 3D Camera View Mode
3D Scene Geometry
There are several options for inserting 3D geometry into your scenes. You can:
• create primitive shapes: cards, cubes, cylinders, and spheres. See Using Built-in Primitive Geometry.
• import models and point clouds created in other 3D applications and exported into OBJ
(Wavefront), FBX, or Alembic files. See Importing Geometry and Point Clouds from Other
Applications.
• create your own geometry and point clouds from scratch. See Creating Point Clouds and Geometry
from Scratch on page 841
Once you've inserted a 3D geometry object into your scene, you can:
• adjust how the object is displayed in the Viewer. See Object Display Properties on page 848.
• select entire objects, as well as vertices or faces on objects. See 3D Selection Tools on page 850.
• merge objects together. See Merging Objects on page 855.
• modify object shapes. See Modifying Object Shapes on page 856.
USER GUIDE
827
Using Built-in Primitive Geometry | Working with Cards
Using Built-in Primitive Geometry

This section describes the primitive shapes built in with Nuke: cards, cubes, cylinders, and spheres.
You can use them as the building blocks for other, more complex shapes, or place them in the
background of your 3D scene where extra detail isn't visible. If you modify them with transforms, they
can also be useful as foreground objects.
Working with Cards

A card is the simplest type of object you can add to a scene (and probably the type you will use most
often). It’s merely a plane onto which you can map a texture - typically a clip you are using as part of a
pan-and-tile setup.
A card object.
A card object may be deformed as a bilinear or bicubic object with controls contained in the card’s
parameters. You can also insert other 3D nodes, such as ProceduralNoise or RadialDistort, to change
the card geometry.
Card nodes have extended bicubics (bicubics with more control points). They allow you to subdivide a
card, giving you finer control for warping an area. You can subdivide the card into an evenly spaced
grid or pick a location to add a row, column, or both.
To Add a Card Object:

1. Click 3D > Geometry > Card to insert a Card node.
2. Drag the Card node’s img pipe to the Read node that has the image you want to apply to the card.
3. Connect the Card node to the appropriate Scene node to add it to the 3D scene.
USER GUIDE
828
4. Use the card object’s transform controls to manipulate the position, scale, and rotation of the
card in 3D space. For more information, see Transforming from the Node Properties Panel.
Deforming Card Objects

The Deform tab in the Card Properties panel lets you convert the card into a mesh surface that may
be pulled and reshaped.
A bicubic deformation offers the greatest degree of surface elasticity. You can add any number of
control points on the card and translate these points and their tangents in any direction. The control
point tangents exert a magnetic-like influence over the objects surface.
The Card node can have any number of

control points you can translate.
To deform a Card object:

1. Double-click the Card node to open its controls.
2. Go to the Deform tab, and select the mesh type for the deformation: bilinear or bicubic.
3. By default, the card has three control points on the x axis, and three on the y axis. To add more
control points, do any of the following:
• Enter new values in the x/y points fields and click the new shape button. For example, to create
a shape with 4 points on the x axis and 6 on the y axis, change the x points value to 4 and the y
points value to 6, and click new shape.
USER GUIDE
829
• To evenly subdivide the current shape in the x or y directions, click the x subdivide or y
subdivide buttons. This adds one control point between every existing control point in the
selected direction. The x/y points fields are also updated to reflect the current number of
control points.
• To add one row or column of control points, adjust the u or v slider. The u slider specifies the
position of new columns, and the v slider the position of rows. In the Viewer, a yellow cross
marker indicates the position of the new row or column. You can also move the cross marker by
dragging it to a new position in the Viewer. The u and v sliders’ values are updated as you move
the marker. When you are happy with the position, click the uv subdivide button. A row or
column is added in the position you specified. Clicking the button again has no effect, because
there is already a subdivision at the specified position.
4. If you selected bicubic under type, you can adjust the way control point tangents behave when
you are making your changes to the card. Do any of the following:
• To have the original tangents adjusted to create a more uniform subdivision when you are using
x subdivide, y subdivide, or uv subdivide, check uniform subdivision. If you do not check this,
Nuke maintains the original tangents.
• You can move the tangents in the Viewer by clicking and dragging. If you want to move a tangent
together with the opposite tangent so that the two tangents form a continuous line, check
smooth tangent. To break the tangent from the opposite tangent and move the tangent alone,
uncheck smooth tangent.
• To change the length of the opposite tangent to always match the length of the tangent you are
moving, check mirror tangent. If you do not check this, the opposite tangent length is not
changed.
5. Drag the points displayed in the mesh to deform the card.
To translate the control points and tangents:

1. If necessary, double-click on the Card node to display its controls, and go to the Deform tab.
2. Only the controls for the selected point are displayed in the bottom of the Card properties panel.
To translate another point, you can select a new point in the Viewer or use the arrow buttons in
the bottom of the Card controls to move to the next control point.
3. To translate the control points and their tangents:
USER GUIDE
830
Using Built-in Primitive Geometry | Working with Cubes
• Increment or decrement the numbered x, y, and z fields. For each control point, the controls for
translating the point itself are shown on top of the controls for translating the tangents.
• Or drag on any control point or tangent to translate it relative to the current angle of view.
Working with Cubes

A cube is the familiar six-sided polyhedron. You can transform any of its sides (and, of course, texture
it with clips).
A cube object.
To Add a Cube
1. Click 3D > Geometry > Cube to insert a Cube node.
2. Drag the Cube node’s img pipe to the Read node containing the clip you want to use as a texture.
3. Drag one of the Scene node’s numbered pipes to the Cube node to place the cube in the scene.
4. Use the cube object’s transform controls to manipulate the position, scale, and rotation of the
cube in 3D space. For more information, see Transforming from the Node Properties Panel.
5. Translate any of the cube’s sides to alter its shape.
To Translate a Cube’s Sides

1. If necessary, double click on the Cube node to display its parameters (and thereby select the
object in the scene).
2. Increment or decrement the cube fields. (Assuming a positive z view of the object, x refers to the
left side; y, the bottom side; n, the back side; r, the right side; t, the top side; and f, the front side.)
Or drag on any side order to translate it relative to the current angle of view.
USER GUIDE
831
Using Built-in Primitive Geometry | Working with Cylinders
Working with Cylinders

A cylinder is an object with two identical flat ends that are circular or elliptical and one curved side.
You can control its geometric resolution, or face count (and, of course, texture it with images).
A cylinder object.
To Add a Cylinder
1. Click 3D > Geometry > Cylinder to insert a Cylinder node.
2. Drag the Cylinder node’s img pipe to Read node containing the clip you want to use as a texture.
3. Drag one of the Scene node’s numbered pipes to the Cylinder node to place the cylinder in the
scene.
4. Use the cylinder object’s transform controls to manipulate the position, scale, and rotation of the
cylinder in 3D space. For more information, see Transforming from the Node Properties Panel.
Adjusting Geometric Resolution

By default, a cylinder has 30 rows and 30 columns. You can, however, increase or decrease either
number as appropriate. For example, the figure below shows a cylinder whose geometric resolution
has been decreased to 10 rows and 3 columns.
USER GUIDE
832
Using Built-in Primitive Geometry | Working with Spheres
A low-resolution cylinder.
To adjust a Cylinder’s geometric resolution:

1. If necessary, double click on the Cylinder node to display its parameters (and thereby select the
2. Increment or decrement rows field to adjust the number of latitudinal divisions on the cylinder.
3. Increment or decrement columns field to adjust the number of longitudinal divisions on the
cylinder.
Working with Spheres

A sphere is the familiar globe-shaped polyhedron. You can control its geometric resolution, or face
count (and, of course, texture it with images).
A sphere object.
To Add a Sphere
1. Click 3D > Geometry > Sphere to insert a Sphere node.
2. Drag the Sphere node’s img pipe to Read node containing the clip you want to use as a texture.
USER GUIDE
833
Using Built-in Primitive Geometry | Working with Spheres
3. Drag one of the Scene node’s numbered pipes to the Sphere node to place the sphere in the
scene.
4. Use the sphere object’s transform controls to manipulate the position, scale, and rotation of the
sphere in 3D space. For more information, see Transforming from the Node Properties Panel.
Adjusting Geometric Resolution

By default, a sphere has 30 rows and 30 columns. You can, however, increase or decrease either
number as appropriate. For example, the figure below shows a sphere whose geometric resolution
has been decreased to 2 rows and 4 columns (making it, in effect, an octahedron).
An octahedron generated with

a low-resolution sphere.
To adjust a Sphere’s geometric resolution:

1. If necessary, double click on the Sphere node to display its parameters (and thereby select the
2. Increment or decrement rows field to adjust the number of latitudinal divisions on the sphere.
3. Increment or decrement columns field to adjust the number of longitudinal divisions on the
sphere.
USER GUIDE
834
Importing Geometry and Point Clouds from Other Applications | Importing Geometry from OBJ Files
Importing Geometry and Point

Clouds from Other Applications
Sometimes, you may need to import files or objects created in 3D applications, such as Maya or
Boujou. Depending on what you want to import and from where, there are different ways of doing
the import:
• To import geometry from OBJ (.obj) files, see Importing Geometry from OBJ Files.
• To import geometry or point clouds from FBX (.fbx) files, see Importing Geometry and Point Clouds
from FBX Files. FBX is a standard file format many applications can export to, and .fbx files contain
3D scenes from which you can import cameras, lights, transforms, meshes, and point clouds into
Nuke.
• To import geometry or point clouds from Alembic (.abc) files, see Importing Geometry and Point
Clouds from Alembic Files. You can export to Alembic from most popular 3D applications.
Importing Geometry from OBJ Files

You can import into a Nuke scene 3D objects from other software programs that have been saved out
in the .obj (Wavefront) format. You cannot manipulate .obj objects at the vertex level from inside
Nuke, but you can texture and transform them.
An imported OBJ object.
To import an OBJ object

1. Click 3D > Geometry > ReadGeo to insert a ReadGeo node.
2. In the ReadGeo parameters, click the file field’s folder icon. The file navigation dialog appears.
USER GUIDE
835
Importing Geometry and Point Clouds from Other Applications | Importing Geometry and Point Clouds
3. Navigate to the OBJ file, then click Open. Nuke reads in the OBJ file.
4. Drag the ReadGeo node’s img pipe to the Read node containing the clip you want to use as a
texture.
5. Drag one of the Scene node’s numbered pipes to the ReadGeo node to place the OBJ object in the
scene.
Importing Geometry and Point Clouds from FBX Files

FBX is a standard 3D file format that gives you access to 3D scenes created in other applications
supporting the same format. What you generally have in an .fbx file is an entire 3D scene containing
cameras, lights, meshes, non-uniform rational B-spline (NURBS) curves, transformation, materials,
and so on. From this scene, you can extract cameras, lights, transforms, and meshes into Nuke. This
way, you can, for example, create a mesh in Maya, export it in a .fbx file, and use the same mesh
again in Nuke.
Note: For the FBX SDK version used in Nuke, see Third-Party Library Versions.
Tip: If you have trouble with .fbx files, it may be because they were written with an older
version of FBX. If they load very slowly, it is also possible that they are ASCII rather than
binary. To get around these problems, you can use the FBX converter on the Autodesk
website (http://usa.autodesk.com/fbx/download/). It converts between various different
formats, including older FBX versions, ASCII, and binary, and is available on Windows, Mac,
and Linux.
Importing Meshes from FBX Files

The ReadGeo node lets you import meshes (or NURBS curves/patch surfaces converted to meshes)
from FBX files. Using one ReadGeo node, you can read in a single mesh or all the meshes in a .fbx
file.
The mesh’s vertices, normals, UVs, and vertex colors are read on a per frame basis or at frame 0. If
there are any shape or cluster deformers, they are applied to the vertices. Materials or textures are
not read in.
To import a mesh from an .fbx file:

1. Select 3D > Geometry > ReadGeo to insert a ReadGeo node into your script.
USER GUIDE
836
2. In the ReadGeo controls, click the folder icon next to the file field and navigate to the .fbx file that
contains the mesh you want to import. Click Open.
3. From the animation stack dropdown menu, select the take you want to use. FBX files support
multiple takes. Usually, one of them is a default take that contains no animation.
4. From the node name dropdown menu, select the mesh you want to import from the .fbx file.
5. To adjust the frame rate used to sample the animation curves, enter a new value (frames per
second) in the frame rate field. The frame rate you enter is only used if you check use frame
rate. Otherwise, the frame rate from the .fbx file is used.
6. If you want to import all the meshes in the .fbx file rather than just one, check all objects. This
overrides whatever you have selected under node name. If the objects are animated, check read
on each frame. This bakes each object's transform into the mesh points and preserves the
animation.
7. If you want to modify the transform properties imported from the .fbx file, uncheck read
transform from file and make the necessary modifications. As long as read transform from file
is unchecked, your changes are kept.
8. To reload the transform properties from the .fbx file, click the Reload button.
Importing Point Clouds from FBX Files

The ReadGeo node also lets you import point clouds from FBX files.
To import a point cloud from an .fbx file:

1. Select 3D > Geometry > ReadGeo to insert a ReadGeo node into your script.
2. In the ReadGeo controls, click the folder icon next to the file field and navigate to the .fbx file that
contains the point cloud you want to import. Click Open.
3. From the animation stack dropdown menu, select the take you want to use. FBX files support
multiple takes. Usually, one of them is a default take that contains no animation.
4. In the objecttype dropdown, select PointCloud.
5. To adjust the frame rate used to sample the animation curves, enter a new value (frames per
second) in the frame rate field. The frame rate you enter is only used if you check use frame
rate. Otherwise, the frame rate from the .fbx file is used.
6. If you want to modify the transform properties imported from the .fbx file, uncheck read
transform from file and make the necessary modifications. As long as read transform from file
is unchecked, your changes are kept.
7. To reload the transform properties from the .fbx file, click the Reload button.
USER GUIDE
837
Importing Geometry and Point Clouds from Alembic Files

You can import meshes (or NURBS curves/patch surfaces converted to meshes) and point clouds
from Alembic files (.abc file format) into a Nuke scene. During the import, Nuke allows you to control
which nodes in the Alembic scene get loaded by using an import dialog. If there is only one item in
the Alembic file, it loads automatically. Below is an example of an imported octopus Alembic file,
courtesy of Sony Pictures Imageworks.
For more information on Alembic, see http://code.google.com/p/alembic/
Tip: In addition to meshes and point clouds, you can also import cameras and transforms
from Alembic files.
To learn how to export files in the Alembic (.abc) format, refer to Exporting Geometry,
Cameras, Lights, Axes, or Point Clouds for more information.
To Import Meshes and Point Clouds from an Alembic File

1. Click Image > Read or press R on the Node Graph.
The Read File(s) dialog displays.
2. Select the Alembic file you want to import from the file location, then click Open.
The Alembic import dialog displays. By default, all items are selected in the Scene Graph when the
import dialog is opened, as shown below:
USER GUIDE
838
Selected parent items are indicated by a yellow circle, and selected child items by a yellow bar
(these turn orange when you select them in the list). Unselected items do not have an indicator
next to them.
3. To import specific items, you must first deselect the root item by clicking on the yellow circle. This
de-selects the root and any child items. Then, select specific items in the scenegraph by clicking
on the blank space where the circles had been, as shown below:
Alternatively, you can right-click on an item and select:

• Select as parent - to select this item and make it a parent to other items. This allows you to
create a separate Nuke node for this item (and any child items underneath it) in the next step.
• Select as child - to select this item and make it a child to the nearest parent item up the tree.
• Deselect - to deselect this item (that is, not import it from the scene).
You can also select multiple items by pressing Ctrl/Cmd or Shift while clicking them.
• Click Create all-in-one node to create one Nuke node for everything that’s selected, regardless
of whether the items are selected as parent or child.
• Click Create parents as separate nodes to create one Nuke node for each parent item (yellow
circle) in the tree. This node contains all the child items (yellow bars) under the parent.
Tip: If you always want .abc files to import all-in-one without displaying the import
scenegraph, enable Preferences > Behaviors > File Handling > always load abc files as
all-in-one.
Nuke creates ReadGeo, Camera, and Axis nodes as necessary, depending on what you selected to
import from the scene.
5. On ReadGeo nodes, you can adjust the following:
• If the objects you imported are animated, make sure read on each frame is checked. This
preserves the animation. When read on each frame is disabled, use lock frame to select the
frame the object is loaded at.
• If you don’t want to render motion blur, disable sub frame for faster UI interactions.
• In the frame rate field, define a frame rate (frames per second) to sample the animation curves.
• Use render points as to determine how point primitives are rendered: either as Nuke point
clouds or Nuke particles.
USER GUIDE
839
• Enable use geometry colors to apply geometry color attributes read from .abc files and apply
them to the Nuke geometry.
Note: When disabled, this control can cause differences in rendered output when
compared to previous versions of Nuke. If this occurs, enable use geometry colors in the
ReadGeo properties panel.
• Any mesh items imported are listed on the Scenegraph tab of the ReadGeo node. To display all
items that exist in the Alembic file in the list, check view entire scenegraph. This allows you to
add items to the list or remove them from it by clicking on the yellow and blank indicators on
the right.
Tip: To load specific items from an Alembic file, you can also create a ReadGeo, Camera, or
Axis node, check read from file, and click the folder icon on the File tab to browse to an
.abc file.
USER GUIDE
840
Creating Point Clouds and Geometry from Scratch | Importing Geometry and Point Clouds from
Creating Point Clouds and

Geometry from Scratch
In addition to using the built-in primitives and imported geometry, Nuke also allows you to create
dense point clouds and geometry from scratch.
This section focuses on creating point clouds using the DepthToPosition, PositionToPoints, and
DepthToPoints nodes available in Nuke. However, if you have a NukeX license, you can also use the
CameraTracker and PointCloudGenerator nodes to create point clouds or the ModelBuilder node to
build 3D models for 2D shots. See Creating Dense Point Clouds and Using ModelBuilder for more
details.
USER GUIDE
841
Creating a Position Pass Using the DepthToPosition Node | To Create a Position Pass using
Creating a Position Pass Using the

DepthToPosition Node
DepthToPosition takes depth data contained in an image file and the camera data to create a 2D
position (xyz) pass. This pass is created by projecting the depth through the camera and recording the
xyz positions of each projected point. The DepthToPosition node can be used together with the
PositionToPoints node to create a point cloud, similar to the effect achieved with the DepthToPoints
node.
To Create a Position Pass using DepthToPosition

1. From the 3D > Geometry menu, select DepthToPosition to add the node to your script.
2. Read in an image with a depth pass and connect it to the image input of the node.
3. From the 3D menu, select a Camera and connect it to the camera input of the DepthToPosition
node.
4. In the DepthToPosition tab of the DepthToPosition node, select the depth channel from the
depth dropdown menu. You should now see a position pass in the Viewer.
USER GUIDE
842
Creating a Position Pass Using the DepthToPosition Node | To Create a Position Pass using
5. If you want to change the output channel, select the channel you would like, or create your own,
from the output dropdown menu.
6. Set the far control value to specify at what distance from the camera the depth values are
ignored. This prevents large depth values from creating unwanted banding caused by the
precision of the floating math calculations.
The DepthToPosition node allows you to create a position pass from an image’s depth pass, and feed
that into PositionToPoints to ultimately create a point cloud. In fact the three nodes,
DepthToPosition, PositionToPoints, and DepthToPoints can be used together, as DepthToPoints is a
gizmo that contains the DepthToPosition and PositionToPoints nodes.
To learn more about generating depth maps, please refer to Generating Depth Maps.
USER GUIDE
843
Creating a Dense Point Cloud Using the PositionToPoints Node | To Create a Point Cloud Using
Creating a Dense Point Cloud

Using the PositionToPoints Node
PositionToPoints takes position data contained in an image file (rendered from a 3D application) and
recreates the image as a dense 3D point cloud in Nuke. The x, y, and z vertices in the position channel
are used to define point positions in 3D space, the size and number of which can be adjusted using
the point size and point detail controls.
To Create a Point Cloud Using PositionToPoints

1. From the 3D > Geometry menu, select PositionToPoints to add the node to your script.
2. Read in a position pass and connect it to the node.
If you have multiple images containing separate position and normal information, connect them
to the pos and norm inputs, respectively.
Note that the pos and norm inputs only appear once you’ve connected the unnamed input.
The normals are contained in

the source image, while
a second image contains
the position pass.
3. Select the position channel from the surface point dropdown menu. If your image contains a
normal channel, select this from the surface normal dropdown menu. You should now see the
point cloud in the 3D Viewer.
If you supplied the node with separate position and normal information via the pos and norm
inputs, the surface point and surface normal dropdowns are disabled.
4. To change the number of points, adjust the point detail slider. A value of 0 means no points are
displayed. A value of 1 displays all available points.
5. To change the size of the points, adjust the point size slider.
USER GUIDE
844
Creating a Dense Point Cloud Using the PositionToPoints Node | To Create a Point Cloud Using
Point cloud created from

the PositionToPoints node.
USER GUIDE
845
Creating a Point Cloud Using the DepthToPoints Node | To Create a Point Cloud Using DepthToPoints
Creating a Point Cloud Using the

DepthToPoints Node
DepthToPoints is a gizmo containing the DepthToPosition and PositionToPoints nodes. It can be used
to generate a 3D point cloud from a depth pass and 3D camera. DepthToPoints takes the depth data
and color information contained in an image file and recreates the image as a 3D point cloud.
To Create a Point Cloud Using DepthToPoints

1. From the 3D > Geometry menu, select DepthToPoints to add the gizmo to your script.
2. Read in an image with a depth pass and connect it to the image input of the gizmo.
3. From the 3D menu, select a Camera and connect it to the camera input of DepthToPoints.
4. In the DepthToPoints User tab, select the depth channel from the depth dropdown menu. If your
channel contains a normal channel, select this from the surface normal dropdown menu, or
connect an image containing normal data to the norm input of the node. You should now see the
point cloud in the 3D Viewer.
5. To change the number of points, adjust the point detail slider. A value of 0 means no points are
displayed. A value of 1 displays all available points.
6. To change the size of the points, adjust the point size slider.
USER GUIDE
846
Creating a Point Cloud Using the DepthToPoints Node | To Create a Point Cloud Using DepthToPoints
Tip: If you are having trouble viewing the image and point cloud, ensure that your image’s
alpha is not set to black.
USER GUIDE
847
Object Display Properties | To Edit an Object’s Display Attributes
Object Display Properties

You can adjust the display characteristics of all geometric objects in a scene. These settings don’t
affect the render output of the scene; these are for display purposes only in the 3D Viewer.
To Edit an Object’s Display Attributes

1. Double click on the object’s node to display its parameters.
2. From the display dropdown menu, select the display type that you want for the object.
These are how each of the display options appear:

• wireframe displays only the outlines of the object’s geometry.
• solid displays all geometry with a solid color.
• solid+wireframe displays the geometry as solid color with the object’s geometry outlines.
USER GUIDE
848
Object Display Properties | To Edit an Object’s Display Attributes
• textured displays only the surface texture.
• textured+wireframe displays the wireframe plus the surface texture.
USER GUIDE
849
3D Selection Tools | Selection Modes
3D Selection Tools
Using the 3D selection tools in the top right corner of the Viewer, you can select nodes, vertices and
faces on a 3D object. Additionally, you can select individual 3D objects in the 3D view, which is handy
when you have more than one object in a node.
Selection Modes
Nuke features three selection modes, Rectangle, Ellipse, and Lasso. Selection modes work the same
in all Viewer contexts, whether you're selecting vertices, faces, or objects. The selection mode
dropdown is located above the Viewer at the top-right.
Rectangle Ellipse Lasso
Each mode uses the same selection modifiers:

• Shift - additive selection. Holding Shift and selecting vertices, faces, or objects adds the selection to
any existing selection.
USER GUIDE
850
3D Selection Tools | Selecting Nodes
• Alt + Shift - subtractive selection. Holding Alt + Shift and selecting vertices, faces, or objects
removes the selection from any existing selection.
Selecting Nodes
You can select nodes with the Node selection tool . This is the default selection tool, and
corresponds to the familiar way of selecting 3D object nodes in the Viewer.
Selecting Vertices on a 3D Object

You can select vertices on a 3D object using the Vertex selection tool. To save selections into your
script, you need to use the GeoSelect node. You can find the 3D selection tools in the top right corner
of the Viewer.
USER GUIDE
851
3D Selection Tools | To Save or Restore Selected Vertices Using the GeoSelect Node
Note: The selection mode is set by the Viewer - changing geometry nodes translates your
selections to the new object.
You can also use the Vertex selection tool to edit geometry using the EditGeo node. See Modifying
Objects Using the EditGeo Node.
To select vertices in the Viewer

1. Attach a 3D object to the Viewer and activate the Viewer’s 3D mode.
2. Click the Vertex selection tool to activate it. It is disabled by default.

3. Select vertices on the object by dragging a marquee over the vertices you want to select.
4. By default, a new selection replaces an existing one, but you can also add and remove from an
existing selection:
• To add to existing selection, hold down Shift while selecting.
• To remove from existing selection, hold down Shift+Alt while selecting.
5. You can also include hidden items by selecting the occlusion test button . Deselect it if you
don’t want to include occluded vertices in your selection. Select if you want to select both visible
and occluded vertices.
To Save or Restore Selected Vertices Using the

GeoSelect Node
With the default Viewer 3D selection tools you can only make a temporary selection. To save or
restore a selection, you can use the same 3D selection tools in the Viewer with the GeoSelect node:
1. Create a GeoSelect node (3D > Modify > GeoSelect) and attach a 3D object to its input. The
GeoSelect properties panel needs to be open in order for you to use the node for selecting
vertices.
2. Click the Vertex selection tool from the 3D selection panel.

3. Select vertices on the object by dragging a marquee over the vertices you want to select.
4. By default, a new selection replaces an existing one, but you can also add and remove from
existing selection:
USER GUIDE
852
3D Selection Tools | Selecting Faces on a 3D Object
5. In the GeoSelect properties panel, you can also uncheck the selectable box to freeze your
selection so it can’t be changed. Checking the box again enables selection again and you can
continue selecting vertices as normal.
6. With the display and render controls you can select the display option you want for your object.
For more information on these controls, see Object Display Properties.
7. When you’re happy with your selection, click save selection to store the vertices in the GeoSelect
node.
8. The vertices are saved by the GeoSelect node and can be restored by clicking restore selection.
Selecting Faces on a 3D Object

You can select faces on a 3D object using the Face selection tool . Each 3D object is divided into
polygonal faces by wireframe lines, and with the Face selection tool you can select one or many of
these.
You can also use the Face selection tool to edit geometry using the EditGeo node. See Modifying
Objects Using the EditGeo Node.
1. Click the Face selection tool to activate it. It is disabled by default.
2. Select faces on the object by dragging a marquee over the faces you want to select.
3. By default, a new selection replaces an existing one, but you can also add and remove from
existing selection:
Marquee-selecting faces on a cube.
USER GUIDE
853
3D Selection Tools | Selecting 3D Objects
Selecting 3D Objects
When you have more than one 3D object in one node (such as in a case when you’re reading in a .fbx
file with multiple objects), you can use the Object selection tool to select them. To select objects,
activate the Object selection tool and click on the object you want to select in the 3D view.
Matching Position, Orientation, and Size to 3D Selection

You can match the position, orientation, and size of a 3D object, such as a Card node, to your
selected vertices in another object. Do the following:
1. Click the snap dropdown menu on the properties panel of the 3D object you want to move
and select Match selection position. The object is now positioned to match the selected vertices.
2. If you select snap > Match selection position, orientation, your 3D object is positioned and
aligned according to the 3D vertex selection.
3. If you select snap > Match selection position, orientation, size, your 3D object is positioned,
aligned, and scaled according to the selected vertices.
USER GUIDE
854
Merging Objects | To Merge Your 3D Objects
Merging Objects
With the MergeGeo node, you can merge your 3D objects together to process all of them at the same
time. For example, after merging your objects, you can use a Transform node to move the objects
together, or add an ApplyMaterial node to apply a global material to them (note that this overrides
any individual materials applied to the geometry before it was merged).
To Merge Your 3D Objects

1. Select 3D > Modify > MergeGeo to insert a MergeGeo after the 3D objects in your script.
2. Connect the objects you want to merge to the MergeGeo node’s inputs.
You can now process all the objects you connected to the MergeGeo node together.
USER GUIDE
855
Modifying Object Shapes | Modifying Objects Using the EditGeo Node
Modifying Object Shapes

Many nodes under the Modify menu let you modify the shape of an object as a whole. Modifying only
selected portions of an object is also supported using the EditGeo node.
You can modify 3D objects by selecting vertices or faces in the Viewer using the EditGeo node or by
using lookup curves, power functions, images, a Perlin noise function, a distortion function, or
trilinear interpolation.
Modifying Objects Using the EditGeo Node

EditGeo allows you to directly modify an object’s vertices or faces, depending on the Viewer selection
mode currently active. You can also modify multiple objects simultaneously if the EditGeo is
downstream of a MergeGeo node with multiple geometry inputs.
To Modify Objects Using EditGeo

1. Select the 3D object or MergeGeo you want to modify in the Node Graph.
2. Navigate to 3D > Modify > EditGeo to insert an EditGeo node.
3. In the node’s controls, use the display dropdown menu to select how you want to view your object
in the Viewer while making changes to it.
Tip: If you don’t have an image or texture associated with your geometry, you may find that
enabling the headlamp control in the Viewer properties improves visibility.
Press S in the Viewer to display the Viewer properties, then enable the 3D > headlamp
checkbox.
4. Set the Viewer selection mode using the 3D selection tools in the top right corner of the Viewer:
• Vertex selection - allows you to select individual vertices on geometry, giving you fine control
while editing. You can also marquee select multiple vertices or use additive selection by holding
down Shift.
• Face selection - allows you to select individual faces on geometry. You can also marquee select
multiple faces or use additive selection by holding down Shift.
USER GUIDE
856
Note: You can turn off occlusion testing by clicking the disable icon in the Viewer to allow
you to select vertices and
faces that are hidden from the view point.
For example, selecting faces on the “front” of a cube with occlusion testing disabled also
selects faces on the opposite, hidden side of the cube.
See 3D Selection Tools for more information.

5. Set how the axis handles are aligned with your selection using the axisalignment dropdown:
• object - the position of the xyz axis is determined by the average position of all vertices in the
selection. The orientation of the axis is the same as the object's orientation.
• average normal - the position of the xyz axis is determined by the average position of all
vertices in the selection. The orientation of the axis is aligned to the average of the current
selection's normals.
In this mode, the z axis always points directly away from the current selection.
Object alignment Average normal alignment
Tip: Holding Ctrl/Cmd+Alt allows you to drag the center of the axis across the surface of the
geometry, with the orientation set to the nearest face's normal.
USER GUIDE
857
6. Make vertex or face selections in the Viewer as required. Hold Shift to add to a selection or
Alt+Shift to remove selections.
7. When you’re satisfied with your selections, drag the axis associated with the selection to a new
position to edit the geometry.
To Animate Edits
1. Set the Viewer selection mode using the 3D selection tools in the top right corner of the Viewer:
• Vertex selection - allows you to select individual vertices on geometry, giving you fine control
while editing. You can also marquee select multiple vertices or use additive selection by holding
down Shift.
• Face selection - allows you to select individual faces on geometry. You can also marquee select
multiple faces or use additive selection by holding down Shift.
2. Make selections in the Viewer and then click in the properties panel.
3. Scrub the playhead to the required frame and transform the geometry using the handles in the
Viewer.
Tip: Click to remove a keyframe at the current frame.
5. Click reset geometry to return the vertices to their original positions.
USER GUIDE
858
Modifying Objects Using Lookup Curves | Modifying Objects Using the EditGeo Node
Modifying Objects Using Lookup

Curves
The CrosstalkGeo and LookupGeo nodes offer you direct global control over each of the vertex x, y,
and z values respectively. You can, for example, only modify all the y values without touching the x
and z values.
You change the different vertex values (x, y, or z) by modifying their associated 2D curves in lookup
tables (LUTs). The x axis in the LUT represents the current vertex value, and the y axis the new vertex
value.
By default, the curve is a diagonal line where all the points in the curve have the same value on the y
axis (the new value) as they do on the x axis (the current value). Because both x and y values are the
same, there is no change in the object’s shape.
By modifying, for example, the CrosstalkGeo node’s y LUT the following way, you can set some of the
vertex y values of a sphere to 0 to squash its bottom half:
Modifying the CrosstalkGeo node’s LUT.
With the CrosstalkGeo node, you can also use one of the vertex x, y, and z values to evaluate the
lookup curve and then add the result to another vertex value. For example, you could modify the x->y
curve, using the vertex x value to find the new value on the curve, and then add that to the vertex y
value. This way, you can modulate the y values by another channel.
By default, these curves are horizontal lines at y=0. They produce no change, because the value added
to the vertex (the new value on the y axis) is 0.
USER GUIDE
859
Modifying Objects Using Lookup Curves | To Modify Objects Using Lookup Curves
To Modify Objects Using Lookup Curves

1. Select 3D > Modify > CrosstalkGeo or LookupGeo to insert a CrosstalkGeo or LookupGeo node
anywhere after the 3D object you want to modify.
2. Attach a Viewer to the node to see your changes.
4. From the list on the left, select the curve you want to modify. For example, you’d select z to only
modify the vertex z values.
In the case of the CrosstalkGeo node, you can also select y->x, for example, to use the vertex y
value to evaluate the curve and add the result to the vertex x value.
5. Adjust the curve as necessary. To insert points on the curve, Ctrl/Cmd+Alt+click on the curve.
USER GUIDE
860
Modifying Objects Using a Power Function | To Modify Objects Using a Power Function
Modifying Objects Using a Power

Function
The LogGeo node lets you modify the shape of your 3D objects using a power function. Using this
node, you can raise each of the vertex x, y, and z values to a power (Xx, Yy, Zz). This can have a
different effect depending on whether you are dealing with negative or positive values.
To Modify Objects Using a Power Function

1. Select 3D > Modify > LogGeo to insert a LogGeo node anywhere after the 3D object you want to
modify.
in the Viewer while making changes to it. See Object Display Properties.
4. Check swap. This swaps the values and the powers they are raised to around (for example,
7 5
changes 5 into 7 ).
5. In the log x, y, and z fields, enter the power you want to raise the respective vertex values to. For
example, if you want to raise the vertex z values to the power of 20, enter 20 in the z field.
Alternatively, you can adjust your 3D object in the Viewer by dragging the white control point to a
new location. You can find the control point just outside the object.
6. To clamp the negative x, y, and z values to 0.0, check clamp black. This option is only valid if you
have checked swap.
Tip: If you set the log x, y, and z values to 1 and check swap, the LogGeo node produces no
change in the incoming geometry. If you want to try out how the node works, this is a good
place to start as you can then gradually adjust the values from there.
The following images illustrate the effect of the LogGeo node on the default Nuke cylinder and sphere
when swap is checked in the LogGeo controls. Notice that if these objects were not positioned in the
default location (centered around 0,0,0), the results would be different.
USER GUIDE
861
Modifying Objects Using a Power Function | To Modify Objects Using a Power Function
The LogGeo node applied The LogGeo node applied

to to
the default cylinder and the default cylinder and
sphere: Log x, y and z set sphere: Log x, y, and z set
to 0.5. to 1 (no change).
The LogGeo node applied The LogGeo node applied

to to
the default cylinder and the default cylinder and
sphere: Log x, y, and z set sphere: Log x, y, and z set
to 2. to 2.
USER GUIDE
862
Modifying Objects Using an Image - Method 1 | To Modify Objects Using an Image
Modifying Objects Using an Image

- Method 1
With the DisplaceGeo node, you can modify geometry based on an image. When using the node,
each vertex is displaced along its normal with a value corresponding to the image pixel the vertex’s uv
attribute points to. The higher the pixel value, the greater the displacement.
The following image illustrates the principle behind the DisplaceGeo node. A Card node is modified
to resemble the pattern of a Checkerboard image.
Using the DisplaceGeo node to modify geometry.
To Modify Objects Using an Image

1. Select 3D > Modify > DisplaceGeo to insert a DisplaceGeo node anywhere after the 3D object you
want to modify.
USER GUIDE
863
Modifying Objects Using an Image - Method 1 | To Modify Objects Using an Image
4. Read in your image map and connect it to the DisplaceGeo node’s displace input.
5. Adjust the following controls:
• From the channels dropdown menu and check boxes, select the channels to use for the
displacement value.
• From the source dropdown menu, select the source for the displace value. For example, if you
selected rgb or rgba from the channels dropdown menu, you can use the red, green, blue, or
alpha channel or the pixel luminance as the source. You can also select rgb relative to move the
vertices on the x, y, and z axes by the amounts in rgb, or rgb absolute to move the vertices to
the values in rgb.
• To define the scale of the displacement, adjust the scale slider. The higher the value, the bigger
the displacement.
• To give x, y, and z different weightings, enter new weights the weight fields. By default, each
weighting is set to 1. If you don’t want to make changes to a value, set its weight to 0.
• To offset x, y, and z values, enter the value by which you want to offset them in the offset fields.
For example, if you enter 0.5 in the y offset field, 0.5 is added to the y value.
• To change the size of the filtering applied to the image before the displacement, adjust the filter
size slider.
• To select the filtering algorithm applied to the image before the displacement, select an
algorithm from the filter dropdown menu. For more information, see Choosing a Filtering
Algorithm.
• To change the name of the attribute that’s used as the vertex’s UV coordinates to find the image
pixel, enter a name in the attrib name field.
• Usually, the normals aren’t correct after the vertices have been moved. To recalculate them after
the displacement, check recalculate normals.
USER GUIDE
864
Modifying Objects Using an Image - Method 2 | Connecting the Displacement Node
Modifying Objects Using an Image

- Method 2
Like the DisplaceGeo node, the Displacement shader node also performs displacement mapping and
at first glance the nodes seem very similar. However, the approach they have on performing
displacement mapping is different.
Displacement mapping is a technique for adding geometric detail to object surfaces as you render
them. Unlike the DisplaceGeo node, The Displacement node does this on the fly, only displacing
those parts of the geometry that are visible at any given moment. Displacement considers the point
of view of camera to determine which parts of the displacement need rendering, thus saving render
time. It’s also possible to optimize the level of tessellation to be the level that you need for an object
at a certain distance.
Connecting the Displacement Node

1. Create the Displacement node by clicking 3D > Shader > Displacement.
2. Connect your geometry to the Displacement node’s output. If you want, you can connect a texture
to the Displacement node’s input.
3. Connect the image you want to create the displacement from in the displacement input.
4. Optionally, you can use a separate map for calculating the normals. Connect this to the normals
input.
5. Proceed to Adjusting the Displacement Controls below.
Adjusting the Displacement Controls

1. Use the displacement channel dropdown menu to select the channel from your displacement
input that you want to use as the displacement map.
2. If you’re using a normals input, uncheck build normals and set normal expansion to:
• none to use the normals as they are,
• XY to multiply them on x and y dimensions, and
• XYZ to multiply them in x, y and z dimensions.
3. Use scale to set the overall scale of the displacement.
4. Set filter size to the size of the filter you want to use when sampling the input image.
USER GUIDE
865
Modifying Objects Using an Image - Method 2 | Adjusting Displacement Controls for Rendering
5. Use the filter dropdown menu to select a filtering algorithm. For more information, see Choosing
a Filtering Algorithm.
6. If you want to automatically calculate normals after the displacement, check build normals.
Uncheck this, if you want the normals calculated from the normals input.
7. Proceed to Adjusting Displacement Controls for Rendering below.
Adjusting Displacement Controls for Rendering

Before rendering a scene where you’ve used Displacement, adjust the controls on the Tessellation
tab to gain speed and quality for the render process:
1. Use max subdivision to set the maximum number of iterations of polygon subdivision that
occurs in tessellation.
If you’re used to working with DisplaceGeo or displacement in other applications, it’s worth noting
that you can get a similar amount of detail with lower geometry subdivisions using the
Displacement node.
For example, a 30x30 Card is constructed from the tessellation of 1800 triangles (30x30x2), but
you can achieve a similar tessellation levels using the Displacement node with fewer card
subdivisions:
Card subdivisions max tessellation

subdivisions triangles
1x1 (2 triangles) 5 2048
Note: Using a high number of geometry and Displacement subdivisions is likely to slow
down rendering so far it becomes unusable - A 10x10 card with max subdivisions set to 4
generates 51200 triangles!
2. Set the mode dropdown menu to the mode used in polygon subdivision:
• uniform - uniform polygon tessellation. This is a good option mainly for testing your result, and
only in rare occasions the best option to actually use for a displacement render.
• screen - tessellation is determined by the screen size. This is the default and often the best
mode option. The tessellation is determined by the size of the tessellate polygons on the screen.
This mode ensures that no new polygons are created once a particular polygon edge screen
length is reached.
USER GUIDE
866
Modifying Objects Using an Image - Method 2 | Adjusting Displacement Controls for Rendering
• adaptive - tessellation is determined by the complexity of the displacement. This option

attempts to estimate flat areas in the image where tessellation is unnecessary. The calculation is
based on the threshold controls which are only active if the adaptive mode is selected.
3. If you set mode to screen or adaptive, set pixel edge length to the maximum size of polygons
used in tessellation.
4. If you set mode to adaptive, you can also adjust the following:
• edge threshold - edges larger than this threshold get divided, whereas edges smaller than this
are subdivided according to the normal and displace thresholds.
• normal threshold - detects normal orientations to determine if the surface is flat or not. If the
angle between adjacent normals is larger than this threshold, tessellation occurs.
• displace threshold - compares the degrees to which two points on the surface are displaced,
and if the results do not match, tessellation occurs.
Note: Keep in mind that applying the Displacement shader to a very complex high
resolution geometry can be very slow.
USER GUIDE
867
Modifying Objects Using a Perlin Noise Function | To Modify Objects Using a Perlin Noise Function
Modifying Objects Using a Perlin

Noise Function
The ProcGeo, or ProceduralNoise, node lets you modify your 3D objects using a Perlin noise function
that creates seemingly random noise. For example, you could use the ProcGeo node to generate
animated noise for rippling waves or clouds, or to create a terrain from a flat card, like in the
following image:
Using the ProcGeo node to create a terrain from

a card object.
You can select the type of noise and control its look in the ProcGeo node’s parameters.
To Modify Objects Using a Perlin Noise Function

1. Select 3D > Modify > ProceduralNoise to insert a ProcGeo node anywhere after the 3D object
you want to modify.
USER GUIDE
868
Modifying Objects Using a Perlin Noise Function | To Modify Objects Using a Perlin Noise Function

4. From the ProceduralNoise Method dropdown menu, select the type of noise you want to use:
Turbulence or fBm (Fractal Brownian Motion).
5. To select whether to modify the x, y, or z values or all of them, use the Orientation dropdown
menu.
6. To change the look of the noise, adjust the rest of the parameters. For example, to control the
amount of detail of the noise, adjust Octaves.
USER GUIDE
869
Modifying Objects Using a Distortion Function | To Modify Objects Using a Distortion Function
Modifying Objects Using a

Distortion Function
The RadialDistort node is a non-linear transformation of the vertices along directions from the object
center, giving either a barrel or pin-cushion distortion. In the following image, two cylinders have
been distorted using the RadialDistort node.
Barrel and pin-cushion distortions.
To Modify Objects Using a Distortion Function

1. Select 3D > Modify > Radial Distort to insert a RadialDistort node anywhere after the 3D object
you want to modify.
USER GUIDE
870
Modifying Objects Using a Distortion Function | To Modify Objects Using a Distortion Function
4. To select whether the distortion is a barrel or pin-cushion, adjust the distortion slider. Values
below 0 produce a barrel distortion, whereas values above 0 produce a pin-cushion distortion. If
you set the value to 0, the 3D object is not distorted.
5. To control the magnitude of the distortion, adjust the power bias slider. The higher the value, the
more distorted the object becomes.
6. To move the center point of the distortion, enter new coordinates in the rotation center fields.
7. To control the amount of distortion in each of the x, y, or z directions, adjust the values in the
scale fields.
8. To keep the object’s center in its original place in the 3D space, check preserve center.
USER GUIDE
871
Modifying Objects Using a Trilinear Interpolation | To Modify Objects Using a Trilinear Interpolation
Modifying Objects Using a

Trilinear Interpolation
With the Trilinear node, you can warp the object as a whole by using a trilinear interpolation to warp
the object’s bounding box. For example, you can use this node to create animated object
deformations, such as the squish/squash of a bouncing ball.
To Modify Objects Using a Trilinear Interpolation

1. Select 3D > Modify > Trilinear to insert a Trilinear node anywhere after the 3D object you want to
modify.
4. To move each corner of the bounding box, enter new coordinates in the p0, p1, p2...p7 fields. To
cancel your changes and reset the box, select reset shape to input.
5. To not use the object’s bounding box but define a box yourself, go to the Source box tab and
uncheck use incoming bounding box. Adjust the src0 and scr1 coordinates define the box. To
change the color of the box, click the box button.
USER GUIDE
872
Materials and Textures | To Modify Objects Using a Trilinear Interpolation
Materials and Textures

This section teaches you how to:
• use the nodes in the 3D > Shader menu to control what material your objects seem to be made of.
See Object Material Properties.
• merge two Shader nodes together. See Merging Two Shader Nodes.
• merge a material with the objects rendered behind it in the 3D scene. See Merging a Material with
the Objects Behind.
• replace selected material channels with a constant color to make one object hold out the others.
See Replacing Material Channels with a Constant Color.
• project texture images onto your 3D objects. See Projecting Textures onto Objects.
• import a set of texture patches following the UDIM scheme and apply them to the surface of a 3D
object. See Importing UDIM Patches.
USER GUIDE
873
Object Material Properties | To Modify Objects Using a Trilinear Interpolation
Object Material Properties

The nodes under the Shader menu let you define the material attributes of geometric objects in a
scene, including the quality of light reflected back to the camera from an object’s surface. Using these
nodes, you can control what material your objects seem to be made of.
You can also add several Shader nodes one after the other to produce more complex effects. For this,
you should use the unlabeled inputs on the Shader nodes.
The material property settings you apply affect the render output of the scene.
You can insert 3D shader nodes in the following places in your scripts:
• between the 2D image you’re using for the surface texture and the 3D object node that creates the
surface, or
• after the 3D object nodes using the ApplyMaterial node. This is a good way to apply a global
material to all objects. See Applying a Material Using the ApplyMaterial Node.
You can use the map connectors to input a mask image to limit the effect of the material change.
Diffuse and Specular nodes.
Note: You can only see the effect of your changes to an object’s material properties in the
2D view.
USER GUIDE
874
Object Material Properties | Applying a Material Using the ApplyMaterial Node
Applying a Material Using the ApplyMaterial Node

The ApplyMaterial node applies a material from the mat input to your 3D object(s).
1. Select 3D > Shader > ApplyMaterial to insert an ApplyMaterial node into your script.
2. Connect the unnamed input of the ApplyMaterial node to your geometry (for example, a Sphere,
ReadGeo, or ModelBuilder node).
Tip: If you want to apply a global material to several objects, you can also connect the
unnamed input to a MergeGeo node. This overrides any materials applied to the individual
geometry nodes before they were merged.
3. Connect your materials (for example, a 2D texture image, a BasicMaterial node, or a Wireframe
node) to the ApplyMaterial node’s mat input.
USER GUIDE
875
Object Material Properties | Applying a Material Using the ApplyMaterial Node
By default, ApplyMaterial applies the material from the mat input onto all incoming geometry
objects.
4. If you created your geometry using ModelBuilder or imported an Alembic file using ReadGeo, you
can choose to only apply the material onto a particular object in the incoming geometry. To do so,
open the ApplyMaterial properties and set filter to name. This allows you to tell ApplyMaterial to
ignore any geometry that doesn't match the filter on the right.
Note: You can also limit a material to a particular object if your geometry was created or
imported using a third-party plug-in that adds a name attribute for the geometry objects.
5. To set how to filter the incoming geometry objects, set the dropdown menu next to name to:
• equals - set the material on any objects whose name matches the string in the filter name field
exactly.
• doesn't equal - set the material on any objects whose name does not match the string in the
filter name field exactly.
• contains - set the material for any objects whose name contains the string in the filter name
field.
This can be useful when you have some structure to your object names. For example, if you
have objects like /Root/Chair/Seat, /Root/Chair/Back, and /Root/Table, you can select contains
and set the filter name field to Chair to apply the material to all parts of the chair while leaving
the table alone.
• doesn't contain - set the material for any objects whose name does not contain the string in
the filter name field.
USER GUIDE
876
Object Material Properties | Adjusting the Diffuse Color
6. To set the filter name, type the name directly into the text entry field or use the choose button to
open the Object Name Chooser dialog and select a filter name from a list of incoming geometry
objects.
Tip: You can also Ctrl/Cmd+click or Shift+click in the Object Name Chooser dialog to select
multiple objects.
Adjusting the Diffuse Color

The Diffuse node lets you adjust the color of the material when illuminated. The material appears
darker as the surface points away from the light, as the light is not falling on it.
1. Select 3D > Shader > Diffuse to insert a Diffuse node into your script.
2. Place the Diffuse node between your 2D texture image and your 3D object node, or connect it to
an ApplyMaterial node’s mat input.
3. In the Diffuse properties, use the channels dropdown menu to select the channels you wish to
process.
4. Adjust the white slider to control the diffuse color. By default, this is in grayscale, but you can
also adjust the individual r, g, and b values. The higher the value, the brighter the material.
USER GUIDE
877
Object Material Properties | Adjusting Specular Highlights
Diffuse: Low white value. Diffuse: High white value.
Adjusting Specular Highlights

You can use the Specular node to control how bright and wide the highlights on the material seem.
The location of the viewpoint is significant: the specular highlights are the brightest along the direct
angle of reflection.
1. Select 3D > Shader > Specular to insert a Specular node into your script.
2. Place the Specular node between your 2D texture image and your 3D object node, or connect it to
3. In the Specular properties, use the channels dropdown menu to select the channels you wish to
process.
4. Adjust the white slider to control the brightness of the specular highlight. The higher the value,
the shinier the material seems.
Low white value. High white value.

5. To control the width of the highlights, adjust the min shininess and max shininess sliders.
Low shininess value. High shininess value.
USER GUIDE
878
Object Material Properties | Simulating Materials That Emit Light
6. If necessary, adjust shininess channel to control how the input channels are used to map the
black and white values to the min shininess and max shininess parameters when a mapSh input
is connected. Select red to use the red channel for the mapping, green to use the green channel,
blue to use the blue channel, luminance to use the luminance, or average rgb to use the
average of the red, green, and blue channels.
Simulating Materials That Emit Light

You can use the Emission node to simulate lamps or other sources that emit light.
1. Select 3D > Shader > Emission to insert an Emission node into your script.
2. In the Emission properties, use the channels dropdown menu to select the channels you wish to
process.
3. Adjust the emission slider to change the brightness of non-illuminated areas for the surface.
The higher the value, the more light the material seems to emit and the brighter it appears.
Adjusting Diffuse, Specular, and Emission Using a Single

Node
The BasicMaterial node is a combination of the Diffuse, Specular, and Emission nodes, allowing you
to control all three aspects of the material from a single properties panel.
1. Select 3D > Shader > BasicMaterial to insert a BasicMaterial node into your script.
2. Place the BasicMaterial node between your 2D texture image and your 3D object node, or connect
it to an ApplyMaterial node’s mat input.
3. The BasicMaterial node has several map inputs you can use to mask the effect of the node. Use:
• mapD to modulate the diffuse component,
• mapS to modulate the specular component,
• mapE to modulate the emission component, and
• mapSh to modulate the shininess value.
4. In the BasicMaterial properties, use the channels dropdown menu to select the channels you
wish to process.
5. Adjust emission to change the color of the light the material emits. Note that when you have an
image connected to the unlabeled input of the BasicMaterial node and adjust this value, you need
to look at the rendered 2D image to see the effect of your changes. Changing the emission value
does not have any effect in the 3D Viewer.
6. Adjust diffuse to control the color of the material when illuminated.
7. Use specular to control how bright the highlights on the material seem.
USER GUIDE
879
Object Material Properties | Simulating Smooth, Regular Surfaces
8. Adjust min shininess and max shininess to set the minimum and maximum shininess values. If
you haven’t connected an image to the mapSh input of the node, the average of these values is
used as the shininess value for the material.
9. Select a shininess channel to control how the input channels are used to map the black and
white values to the minShininess and maxShininess parameters when a mapSh input is
connected. Select red to use the red channel for the mapping, green to use the green channel,
Simulating Smooth, Regular Surfaces

The Phong node uses the Phong algorithm to smooth edges between faces. It provides realistic
shading and highlights for smooth materials, such as skin and other organic surfaces.
1. Select 3D > Shader > Phong to insert a Phong node into your script.
2. Place the Phong node between your 2D texture image and your 3D object node, or connect it to
3. The Phong node has several map inputs you can use to mask the effect of the node. Use:
• mapD to modulate the diffuse component,
• mapS to modulate the specular component,
• mapE to modulate the emission component, and
• mapSh to modulate the shininess value.
4. In the Phong properties, use the channels dropdown menu to select the channels you wish to
process.
5. Adjust color to change the material color.
6. Adjust emission change the color of the light the material emits.
7. To control the color of the material when illuminated, adjust diffuse.
8. To control how bright the highlights on the material seem, adjust specular.
9. To control how shiny the material appears, adjust shininess.
10. To set the minimum and maximum shininess values, adjust min shininess and max shininess. If
you haven’t connected an image to the mapSh input of the node, the average of these values is
used as the shininess value for the material.
11. Select a shininess channel to control how the input channels are used to map the black and
white values to the minShininess and maxShininess parameters when a mapSh input is
connected. Select red to use the red channel for the mapping, green to use the green channel,
USER GUIDE
880
Object Material Properties | Rendering a Wireframe Overlay on Your Geometry
Rendering a Wireframe Overlay on Your Geometry

The Wireframe node allows you to render a wireframe overlay on the surface of your geometry object
or particle simulation. This can be useful, for example, if you want to:
• check that your texture projection correctly lines up with your geometry.
• create a quick render of your 3D scene to check the positioning of objects.
• create motion graphics.
• create "making of" videos.
Note: The Wireframe node currently only works if you are rendering your 3D scene using
ScanlineRender, RayRender and PrmanRender do not support the Wireframe shader.
1. Select 3D > Shader > Wireframe to insert a Wireframe node into your script.
2. Place the Wireframe node between your 2D texture image and your 3D object node, or connect it
to an ApplyMaterial node’s mat input.
USER GUIDE
881
3. In the Wireframe properties, use the channels dropdown menu to select the channels you wish
to process.
4. From the operation dropdown, select how to apply the wireframe overlay to your geometry:
• opaque - display the wireframe on fully opaque black input geometry.
• see through - display the wireframe on fully transparent geometry.
• over - display the wireframe on top of the input shader or texture.
USER GUIDE
882
• multiply - multiply the wireframe by the input shader or texture and display it on fully
transparent geometry.
• modulate - apply standard diffuse shading to the wireframe and display it on top of the input
shader or texture. This takes into account any lights in the scene.
5. To set the width of the wireframe lines (in pixels), adjust line width.
USER GUIDE
883
Line width set to 0.5. Line width set to 3.

6. To set the color and transparency of the wireframe lines, adjust line color.
Line color set to white. Line color set to cyan.
USER GUIDE
884
Merging Two Shader Nodes | To Merge Two Shaders
Merging Two Shader Nodes

With the Shader menu’s MergeMat node, you can combine two shader nodes together, using
compositing algorithms like none, replace, over, and stencil. The MergeMat node is particularly
useful for combining multiple Project3D nodes, allowing you to composite 2D images projected onto
the 3D geometry atop each other.
To Merge Two Shaders

1. Select 3D > Shader > MergeMat to add a MergeMat (over) node after the two shader nodes you
want to combine.
2. Connect the MergeMat node to the img input of the 3D object you want to project the images on.
3. Connect the shader nodes to the MergeMat node’s A and B inputs. A refers to the foreground
element, and B to the background element.
For example, if you wanted to combine two Project3D nodes and composite their results onto a
sphere, your node tree would look something like the following:
4. From the operation dropdown menu, select how you want to composite the results of the two
shader nodes together:
• to only use input B in the composite, select none.
• to only use input A in the composite, select replace.
USER GUIDE
885
Merging Two Shader Nodes | To Merge Two Shaders
• to composite input A over input B using a mask, select over.

• to use input B outside the mask area, select stencil.
• to use input B inside the mask area, select mask.
• to add input B to input A, select plus.
• to use input A if it is greater than input B or else use input B, select max.
• to use input A if it is less than input B or else use input B, select min.
5. For operations (such as over) that need an alpha channel (mask), select which channel to use for
the alpha from the alpha channels dropdown menu.
USER GUIDE
886
Merging a Material with the Objects Behind | To Merge a Material with the Objects Behind
Merging a Material with the

Objects Behind
The Shader menu’s BlendMat node sets how the pixels colored by the material is applied to combine
with the pixels from objects behind. It is like the MergeMat node, but instead of blending with
another material, it blends with whatever is rendered behind in the 3D scene.
The right-hand image shows the BlendMat node

applied to the checkered sphere (that has a
checkered alpha channel) and the BlendMat
operation set to stencil.
To Merge a Material with the Objects Behind

1. Select 3D > Shader > BlendMat to add a BlendMat node after the material you want to merge
with the background pixels.
2. Connect the BlendMat node to the img input of the 3D object you want to project the material on.
USER GUIDE
887
3. From the channels dropdown menu, select the channels you want to affect.
4. From the operation dropdown menu, select how you want to composite the BlendMat node’s
input material and the background pixels together:
• to set the material to black, select none.
• to show the material where the material and the background overlap, select replace.
• to composite the material over the background pixels according to the material’s alpha, select
over.
USER GUIDE
888
• to show the background pixels where the material’s alpha is black, select stencil. Where the
material’s alpha is white, the material is set to black.
For this to work, the BlendMat node needs to process the alpha channel, so set channels to
rgba.
This operation is the opposite of mask.
• to show the background pixels where the material’s alpha is white, select mask. Where the
material’s alpha is black, the material is also set to black.
For this to work, the BlendMat node needs to process the alpha channel, so set channels to
rgba.
This operation is the opposite of stencil.
• to add the background pixels to the material, select plus.
USER GUIDE
889
• to use the material if its pixel values are greater than the background pixels or else use the
background pixels, select max.
• to use the material if its pixel values are less than the background pixels or else use the
background pixels, select min.
USER GUIDE
890
Replacing Material Channels with a Constant Color | To Merge a Material with the Objects Behind
Replacing Material Channels with

a Constant Color
The FillMat node lets you replace selected material channels with a constant color. Typically, you
would use this node to make one object hold out the others. When you set the FillMat color to 0, it
acts as a “3D cookie cutter” and makes a black hole where the material would otherwise be.
A sphere in front of a cube. The same scene with the

sphere material’s rgba The alpha channel after
channels set to black applying the FillMat node.
using the FillMat node.
This is similar to using a black Constant node as the input texture. However, the advantage of using
the FillMat node is that you can easily apply it to the alpha channel in addition to the rgb channels.
Another advantage is that the FillMat node doesn’t break the shading sequence, so you can insert it
after other material nodes in your node tree.
1. Select 3D > Shader > FillMat to insert a FillMat node between the 2D image you’re using for the
surface texture and the 3D object node that creates the surface.
USER GUIDE
891
Replacing Material Channels with a Constant Color | To Merge a Material with the Objects Behind
2. In the FillMat controls, use the channels controls to select the channels you want to replace with
a constant color.
3. Use the color control to select the constant color. By default, this is set to black (0).
USER GUIDE
892
Projecting Textures onto Objects | Projecting Textures with the UVProject Node
Projecting Textures onto Objects

You can use the UVProject and the Project3D nodes to project texture images onto your 3D objects.
This way, you can add detail, surface texture, or color to your geometry, making the geometry more
realistic and interesting.
The UVProject node changes the uv values of the vertices whereas the Project3D node is a material
shader.
Projecting Textures with the UVProject Node

The UVProject node sets the uv coordinates for the object, allowing you to project a texture image
onto the object. If the object already has uv coordinates, this node replaces them.
1. Select 3D > Modify > UVProject to insert a UVProject node anywhere after the 3D object you want
to modify.
4. Connect an Axis or a Camera node to the UVProject node’s axis/cam input. If you connect an Axis
node, project the texture UV coordinates onto the object using the axis transform values (that is,
translation, rotation, scale, etc.). If you connect a Camera node, do a similar projection as with the
axis but also use the camera lens information, such as the aperture.
5. Adjust the following parameters:
• From the projection dropdown menu, select the projection type. Usually, it’s best to select a
type that’s close to the object’s surface shape. For example, if your object is a sphere, like a
football or a planet, select spherical.
• From the plane dropdown menu, select the projection direction: XY, YZ, or ZX to project the
texture image along the z, x, or y axis. This dropdown menu is only available if you selected
planar as the projection type.
• From the projecton dropdown menu, select both, front or back depending on whether you
want to project the texture on the front face of the object, its back face or both. The front face of
an object is the one facing the camera and similarly the back face is the one furthest away from
the camera.
• Check view frustum culling if you want the UVProject node to affect only the vertices inside the
camera view frustum. Any vertices outside the view frustum are not affected and they still keep
their original UV coordinates. Uncheck if you want the node to affect all vertices.
• To mirror the texture UV coordinates in the horizontal direction, check invert u. To mirror them
in the vertical direction, check invert v.
USER GUIDE
893
Projecting Textures onto Objects | Projecting Textures with the Project3D Node
• To scale (stretch or squash) the texture UV coordinates in the horizontal direction, adjust the u
scale slider. To scale them in the vertical direction, adjust the v scale slider. The higher the
value, the more the texture is stretched.
• To change the name of the attribute that’s used as the vertex’s UV coordinates to find the image
pixel, enter a name in the attrib name field.
Projecting Textures with the Project3D Node

The Project3D node projects an input image through a camera onto the 3D object.
1. Select 3D > Shader > Project3D to insert a Project3D node after the image you want to project.
Connect a Camera node to the Project3D node’s cam input.
2. Insert a 3D geometry node (a Sphere for example) after the Project3D node.
3. Attach a Viewer to the 3D geometry node to see your changes.
5. From the project on dropdown menu, select to project the image on either the front facing, back
facing, or both polygons.
6. To extend the input image at its edges with black, check crop. To extend the image with the edge
colors, uncheck crop.
7. If you want to use ray casting to test the projection and find out which parts of it are occluded,
you can use the occlusion mode dropdown. Select:
• none - to disable occlusion testing.
• self - to tell Project3D that only the geometry connected to it can cause occlusion.
• world - to tell Project3D that other objects in the scene can cause occlusion.
USER GUIDE
894
Importing UDIM Patches | Projecting Textures with the Project3D Node
Importing UDIM Patches

When applying textures to models that use regions of UV space outside the standard (0,0) - (1,1)
range, it’s common to use one texture for each 1x1 square. These textures can be numbered in a
variety of ways. UDIM is a numbering scheme that identifies the first texture that’s applied to the (0,0)
- (1,1) region as 1001, with numbers increasing by one for each texture in the U direction, and by ten
for each texture in the V direction.
For example, the texture for the region (1,0) - (2,1) would have UDIM 1003, and the texture for (0,1) -
(1,2) would have UDIM 1011. A set of patches following the UDIM scheme might look like this:
• color.1001.tif
• color.1002.tif
• color.1003.tif
• color.1004.tif
• color.1005.tif
Where color is the channel name and 1001, 1002, ... is the UDIM value for each patch in UDIM space.
In Nuke, you can import a set of patches following the UDIM scheme and quickly apply them to the
surface of a 3D object. To do this:
1. Select Image > UDIM Import from the Toolbar.
2. Select the UDIM image or sequence to import and click Open.
3. To ignore an individual patch, disable the checkbox next to it.

4. To add additional patches, click Add Files.
USER GUIDE
895
Note: If there’s a conflict between the patches that you’ve added, a notification appears
underneath the Add Files button describing the conflict.
A typical conflict would be if you attempted to import multiple files sharing the same UDIM
value.
5. Check the postage stamp checkbox to enable a thumbnail view of the patch on the Read node.
6. Check the group nodes checkbox to place the Read nodes for each individual patch into a Group
node. The name of the Group node is based on the name of the image or sequence.
7. Click Ok to import the files.
This creates a Read node for every patch in the sequence and appends a UVTile node (which allows
you to modify a patch’s coordinates in UV space) to each one.
Note: If you checked group nodes in the UDIM import dialog, highlight the Group node
and press Ctrl+Return (Cmd+Return on a Mac) to view the contents.
The UDIM value for each patch is inserted into the udim field in the UVTile node. This value is
important as it offsets every Read node (and therefore, patch) to the correct UDIM space. A number of
MergeMat nodes are used to combine the Read/UVTile node combinations.
Note: The MultiTexture node at the end of the tree optimizes rendering when multiple
MergeMat nodes are chained together. It uses the default Iop vertex shader and only
handles multi-texturing. This node has no fields or knobs and is intended for use only with
the UDIM Import operation.
To apply these patches to a 3D object in your scene, simply connect the output of the last node in the
tree to the img input of a ReadGeo node.
To move a patch to a different point in UV space, in the UVTile control panel enter a different UDIM
value or disable the udim field and enter values in the u and v fields.
USER GUIDE
896
Note: By default the UDIM Import operation parses the file name based on the UDIM
number. It is possible, however, to create a custom parsing function to work with your own
naming conventions. Please refer to the Python Developers Guide for details on how to do
this.
USER GUIDE
897
Lighting | Projecting Textures with the Project3D Node
Lighting
The nodes under the Lights menu let you control the lighting in your scene. Using these nodes, you
can bring objects out or push them back, create an illusion of depth, simulate the conditions in the
real world, or simply alter the feeling of the scene.
Nuke features four types of light you can use in your 3D scenes: direct light, point light, spot light, and
environment light. You can add these using the DirectLight, Point, Spotlight, and Environment nodes.
In addition to these, there is a Light node, which lets you create direct, point, and spot lights, as well
as read in lights from .fbx files. For more information, see Working with Lights.
The Light, DirectLight, Point, and Spotlight nodes all have controls that you can use to adjust how the
lights cast shadows in your 3D scene. See Casting Shadows.
Lighting is also affected by the 3D object normals, which are used to determine how the light should
bounce off a surface at any particular point. In Nuke, the Normals node allows you to manipulate
object normals in order to control the diffuse and specular light contributions. See Manipulating
Object Normals.
There’s also a Relight node, which takes a 2D image containing normal and point position passes and
lets you relight it using 3D lights. See Relighting a 2D Image Using 3D Lights.
USER GUIDE
898
Working with Lights | Inserting Direct Lights
Working with Lights

There are several types of lights that you can add and manipulate in your 3D scene to create the
required effect.
Inserting Direct Lights

A direct light is a light that emits parallel light in one direction. It appears to illuminate all objects with
equal intensity, as if it was coming from a far away source. Being at an infinite distance from the
objects, direct light has orientation, but no position. A real world example of a direct light is the sun.
You can use direct light to simulate sunlight and moonlight, for example.
1. Select 3D > Lights > Direct to insert a DirectLight node in your script.
2. Connect the DirectLight node to the Scene node.
3. In the DirectLight node’s controls, adjust the following:
• Drag the color slider to change the light color.
• Drag the intensity slider to change the brightness of the light.
• To control the direction of the light, enter values in the rotate fields.
• To adjust the settings for shadows, change values for the controls on the Shadows tab. For
more information on these controls, see Casting Shadows.
Inserting Point Lights

A point light is a point in 3D space that emits light in every direction. A real world example of a point
light is a light bulb. You can use point light to simulate light bulbs, lamps, and candles, for example.
1. Select 3D > Lights > Point to insert a Point node in your script.
2. Connect the Point node to the Scene node.
USER GUIDE
899
Working with Lights | Inserting Spot Lights
3. In the Point node’s controls, adjust the following:

• To control how much light the object gets from the light source (based on the distance between
the object and the light source), use the falloff type dropdown menu. A Linear type diminishes
the light at a fixed rate as it travels from the object, whereas Quadratic and Cubic types
diminish the light at an exponential rate. If you select No Falloff, the distance between the light
source and the object does not affect the lighting.
• To control the position of the light in the 3D space, enter values in the translate fields.
• To adjust the settings for shadows, change values for the controls on the Shadows tab. Note,
however, that the Point light doesn’t cast shadows if you’re using ScanlineRender. For more
information on these controls, see Casting Shadows.
Inserting Spot Lights

A spot light is a point in 3D space that emits a cone-shaped light in a given direction. A real world
example of a spot light is a desk lamp.
1. Select 3D > Lights > Spot to insert a Spotlight node in your script.
2. In the node’s controls, adjust the following:
• Drag the cone angle slider to control the spread of the light (how wide or narrow the beam is)
in degrees from 0 to 180.
• Drag the cone penumbra angle slider to control the softness along the edge of the area of
illumination. A negative value fades inward from the circle's edge. A positive value fades outward
from the circle's edge. The cone falloff should be set to zero or a low value in order to see the
softness. This feature is only visible in the rendered objects and not in the 3D OpenGL Viewer.
• Drag the cone falloff slider to control how concentrated the light is (that is, how much the light
diminishes from the center of the circular region out to the edge). The higher the value, the
more focused the light becomes. The falloff is independent of the falloff type.
• To control how much light the object gets from the light source (based on the distance between
the object and the light source), use the falloff type dropdown menu. A Linear type diminishes
the light at a fixed rate as it travels from the object, whereas Quadratic and Cubic types
diminish the light at an exponential rate. If you select No Falloff, the distance between the light
source and the object does not affect the lighting.
• To control the direction of the light, enter values in the rotate fields.
• To control the position of the light in the 3D space, enter values in the translate fields.
USER GUIDE
900
Working with Lights | Inserting Environment Lights
• To adjust the settings for shadows, change values for the controls on the Shadows tab. For
more information on these controls, see Casting Shadows.
Inserting Environment Lights

An environment light is a light that illuminates the objects using an image of light from a real-world
environment. This image-based lighting is generated using High Dynamic Range Images (HDRI). When
HDR images are created, several differently exposed images are combined to produce a single image
of the surrounding environment. As a result, HDR images have a wide range of values between light
and dark areas, and represent the lighting conditions of the real world more accurately.
To use environment light, you first need to shoot a real life environment as an HDR image. Using the
SphericalTransform node, you then convert this image into a spherical mapped image. The sphere is
used to surround the 3D objects, so that the mapped image color illuminates them.
Environment light only works with shiny object materials that can reflect the mapped image. It results
in a very realistic lighting that makes it easier to integrate the objects into the environment.
1. Read an HDR image of the environment into your script.
2. Select Transform > SphericalTransform to insert a SphericalTransform node after the HDR
image. You use this node to convert the HDR image into a spherical mapped image. In the node’s
controls, select the Input Type and the Output Type (in this case, Sphere).
3. Select 3D > Lights > Environment to insert an Environment node in your script. Connect the
SphericalTransform node to the Environment node’s map input, and the Environment node to the
Scene node.
USER GUIDE
901
Working with Lights | Inserting Environment Lights
4. In the Environment node’s controls, adjust the following:

• From the filter dropdown menu, select a filtering algorithm for the map image. For more
information, see Choosing a Filtering Algorithm.
• To change the blur size of the map image, adjust the blur size slider.
USER GUIDE
902
Working with Lights | Inserting Direct Lights, Point Lights, or Spot Lights
Inserting Direct Lights, Point Lights, or Spot Lights

The Light node includes the DirectLight, Point, and Spotlight nodes, so you can set it to act as any of
these three nodes. The advantage of using a Light node in this way is that if you want to change the
light type later, you can do so without setting up a new node. For example, you might insert a direct
light, but then realize that what you actually need is a spot light. If you inserted the direct light using a
DirectLight node, you need to delete this node and insert a SpotLight node instead. However, if you
inserted the direct light using a Light node, you can simply change the light type from directional to
spot in the Light controls.
Tip: The node can also be used to import lights from .fbx files. This is described under
Importing Lights from FBX Files.
1. Select 3D > Lights > Light to insert a Light node into your script.
2. In the Light controls, select the light type you want to use: point, directional, or spot. The
controls are enabled and disabled according to the light type you select. For example, if you chose
directional light, you get the same controls that appear on the DirectLight node.
3. Adjust the controls as necessary. For information on the functions of the controls, refer to the
following:
• If you selected point as the light type, see Inserting Point Lights.
• If you selected directional as the light type, see Working with Lights.
• If you selected spot as the light type, see Inserting Spot Lights.
Using the Look Input

You can use the optional look input of the Light node, so that the light automatically rotates to face
towards the connected input. You can attach a Camera, Light, or Axis node to the look input. For
example, you can connect an Axis node to the look input so that the light rotates to face the axis,
wherever it is moved.
USER GUIDE
903
If you animate a card to move along the x axis, you can attach a Camera and a Light node with the
look inputs so that they automatically rotate and face the card as it moves. To do this, complete the
following steps:
1. After animating your card, insert an Axis node.
2. Expression link the Axis node to the Card node by holding Ctrl, clicking on the translate
animation button in the Card node's properties, and dragging it to the translate animation
button in the Axis node's properties. The expression link is displayed as a green line with an arrow
denoting the direction of the expression. See Linking Expressions for more information.
3. Insert a Camera node and either drag the camera to the required position in the Viewer, or use
the Camera node properties to adjust the position of the camera.
USER GUIDE
904
4. Connect the Camera's look input to the Axis node.
5. Insert a Light node and drag the light to required position in the Viewer, or use the Light node
properties to adjust the position of the camera.
USER GUIDE
905
6. Connect the Light node's look input to the Axis node
USER GUIDE
906
Working with Lights | Importing Lights from FBX Files
7. Playback your animated card and notice that the Camera and Light follow the animated card.
Importing Lights from FBX Files

way, you can, for example, create a light in Maya, export it in a .fbx file, and use the same light again
in Nuke.
and Linux.
You can use the Light node to read in directional, point, and spot lights from FBX scene files (for
more information on these three light types, refer to Lighting). One Light node only reads in one light.
Therefore, if your .fbx file contains three lights and you want to import all of them into Nuke, you
need to use three Light nodes.
To import a light from an .fbx file:

1. Select 3D > Lights > Light to insert a Light node in the place where you want to add the light in
your script.
USER GUIDE
907
Working with Lights | Importing Lights from FBX Files
2. In the Light controls, check read from file. This enables the controls on the File tab, allowing you
to read in lights from an .fbx file. It also disables all controls whose values are filled from the .fbx
file. You can still view these values and use them in expressions, but you cannot modify them,
because they are read from the .fbx file. Any changes you make in the .fbx file are reflected in
these values of the Light node.
3. On the File tab, click the folder icon and browse to the .fbx file that contains the light you want to
use. Click Open.
4. From the animation stack dropdown menu, select the take you want to use from the .fbx file.
FBX files support multiple takes in the same file. One of the takes is usually a default take without
any animation.
5. From the node name dropdown menu, select the light node you want to import from the .fbx
file.
6. If you want to override the frame rate used in the .fbx file to sample the animation curves, enter a
new frame rate (frames per second) in the frame rate field. Check use frame rate to use the rate
you entered (rather than the one in the .fbx file).
7. To scale the intensity channel values read from the .fbx file, adjust the intensity scale slider. If
the light is too dark, increase this value.
8. If you want to modify the light properties imported from the .fbx file, uncheck read from file on
the Light tab and make the necessary modifications. As long as read from file is unchecked, your
changes are kept.
9. To reload the light properties from the .fbx file, make sure read from file is checked and click the
reload button on the File tab.
USER GUIDE
908
Casting Shadows | Importing Lights from FBX Files
Casting Shadows
The Light, Point, DirectLight, and Spotlight nodes all have controls that you can use to adjust how the
lights cast shadows in your 3D scene. The following geometry nodes also have controls that allow you
to select whether they receive shadows cast by the lights and whether they themselves cast shadows
on other objects:
• Card
• Cube
• Cylinder
• Sphere
• MergeGeo
• ModelBuilder
• PointCloudGenerator
• PositionToPoints
• ReadGeo
• Scene
Note: The method used to create shadows varies between different render nodes:
• ScanlineRender uses depth mapping to create shadows. It first renders a depth map for each
light that casts a shadow. The depth map is rendered from the light’s point of view, and each pixel in
the depth map represents the distance from the light to the nearest surface the light illuminates in a
specific direction. The depth map is then compared to the render from the point of view of the
camera. If a point is farther away in the image that the camera sees than in the depth map, then that
point is considered to be in shadow.
Depth map shadows are often faster to render than raytraced shadows, but may not appear as
realistic.
USER GUIDE
909
Casting Shadows | To Cast Shadows:
• PrmanRender creates shadows through raytracing. It fires individual light rays from the camera
into the scene for each pixel. When a light ray hits a surface in the scene, PrmanRender traces so
called shadow rays between the intersection point and every light source in the scene. If there are
obstacles between the intersection point and the light sources, the intersection point is considered
to be in shadow.
The advantage of raytracing is that it can be used to create more accurate shadows and shadows
that have soft edges, much like those in the real world. However, compared to creating shadows
using depth mapping, raytracing can take much longer to render.
To Cast Shadows:
1. Open the properties panels for the 3D objects in your scene.
2. Check the cast shadow or receive shadow box, or both.
Cast shadow tells objects in a light’s path to cast shadows onto other objects in the scene.
Receive shadow tells objects to display shadows cast by other objects in the scene. This is useful
for showing occlusion in a scene.
You can also use a Scene node to define these settings for all 3D objects connected to it. In the
Scene properties, set shadow to override inputs and use the cast shadow and receive shadow
controls to override the corresponding controls on the individual 3D objects.
USER GUIDE
910
Casting Shadows | Adjusting Shadows When Using ScanlineRender
3. Make sure you’ve got a Shader node attached to any 3D objects that you want to receive shadows.
4. Attach a light to your scene, and check the cast shadows box on the Shadows tab.
In our example node tree below, we are using a ScanlineRender node, but you could use a
PrmanRender node instead.
Note: If you are using a point light, casting shadows is not supported in the ScanlineRender
node.
5. Depending on the render node you are using, proceed to either:

• Adjusting Shadows When Using ScanlineRender below, or
• Adjusting Shadows When Using PrmanRender.
Adjusting Shadows When Using ScanlineRender

1. On the Shadows tab of the light node properties, set shadow mode to:
• solid - Objects seen from the light that cast shadows are considered completely solid.
• clipped alpha - objects that cast shadows are considered to be transparent if the object’s alpha
is below the light’s clipping threshold control. All other alpha values fully occlude the light.
• full alpha - Shadows are calculated based on how light is reduced when it passes through non-
opaque occluders.
This affects shadows cast by objects, based on the objects’ opacity.
USER GUIDE
911
Solid. (The shadow is Clipped alpha.

rectangular because the
Full alpha.
leaf image is applied on
a Card object.)
2. If you want the light node to to output the shadow map, set output mask to the channel where
you want to store this. You can enable this even if cast shadows is disabled.
3. If you set shadow mode to solid or clipped alpha, you can adjust the following:
• depthmap resolution - this sets the resolution of the depth map. Larger values result in
shadows with less crunchy edges and less artifacts, but require more time to process.
Depthmap resolution set Depthmap resolution set

to 800. to 7400.
Note: You can also fix crunchy edges by increasing the number of samples instead of
increasing the depth map resolution.
• samples - this sets the number of samples for the light when generating soft shadows. If soft
shadows in your scene appear dotty or noisy, try increasing this value. The higher the value, the
smoother the soft the shadows become.
Samples set to 1. Samples set to 50.
USER GUIDE
912
• jitter scale - the amount of jitter used when doing percentage-closer filtering (PCF) for soft
shadows. A larger jitter scale value results in softer, more perceptually accurate shadows.
PCF works by sampling the depth map at many different positions around the same spot. The
final shadow value for that spot is an average of how many of the samples were occluded or
visible from point of view of the light.
• bias - this is a constant offset that moves the surface sample point away from surface, towards
the light that is casting the shadow. If self-shadowing artifacts appear in the image, you may
want to increase this value. Note, however, that if you increase the value too much, some
shadows may start moving away from the base of the objects that cast them.
Shadow artifacts. Increasing bias can help

reduce artifacts.
• slope bias - this is like bias, but the offset is proportional to the slope of the depth map. This
allows you to give a different offset to each value in the depth map, depending on the surface’s
slope relative to the light. For example, if the surface’s slope relative to the light is shallow, the
value in the depth map may be a correct approximation and no offset (or only a very small
offset) is needed. If the surface’s slope relative to the light is steep, the values in the depth map
are less likely to be a correct approximation and a larger offset is needed.
If increasing bias reduced the existing self-shadowing artifacts but introduced more artifacts in
other areas of the image, you may want to bring bias down a little and increase slope bias
instead. Then, tweak both values until you’re happy with the result.
In the clipped alpha mode, you can also adjust:
• filter - the type of filter to use. For more information on the available filtering algorithms, see
• clipping threshold - any surface samples with alpha values below this threshold are considered
transparent. The higher the value, the more areas on objects that cast shadows are considered
transparent and pass through some light.
USER GUIDE
913
Casting Shadows | Adjusting Shadows When Using PrmanRender
Clipping threshold set to 0.1. Clipping threshold set to 0.9.

4. If you set shadow mode to full alpha, you can adjust the following:
• filter - the type of filter to use. For more information on the available filtering algorithms, see
• scene epsilon - an offset that moves the sampling point away from the geometry surface,
towards the light that is casting the shadow. Increasing this value can reduce self-shadowing
artifacts.
Tip: To generate accurate shadows from a Direct light, view the scene through the light
(using the Viewer's camera dropdown menu) and adjust the Direct light's scale control so
that the part of the scene that should cast shadows fits within the view. This ensures that
none of the shadow-casting geometry is missed by the depth map.
Adjusting Shadows When Using PrmanRender

1. Open the PrmanRender properties and make sure shadows is enabled.
2. Open the light node properties and go to the Shadows tab.
3. Make sure shadow mode is set to solid. The other two modes aren’t relevant when using
PrmanRender.
4. If you want to cast soft shadows from the light, increase sample width. This value multiplies the
width of the soft area around the egde of a shadow. The higher the value, the larger the soft area.
Sample width set to 1. Sample width set to 15.

5. If you increased sample width in the previous step and the resulting soft shadows in your scene
appear dotty or noisy, increase samples. This sets the number of samples for the light when
generating soft shadows. The higher the value, the smoother the soft shadows become.
USER GUIDE
914
Casting Shadows | Adjusting Shadows When Using PrmanRender

6. If self-shadowing artifacts appear in the image, increase the bias value. This moves the surface
sample point away from surface. Note, however, that if you increase the value too much, some
shadows may start moving away from the base of the objects that cast them.
Shadow artifacts. Increasing bias can help

reduce artifacts.
USER GUIDE
915
Manipulating Object Normals | To Manipulate Object Normals
Manipulating Object Normals

Object normals are vectors that are perpendicular to the surface. They are used in lighting
calculations to determine how the light should bounce off a surface at any particular point. By
manipulating them, you can control the diffuse and specular light contributions.
To Manipulate Object Normals

1. Select 3D > Modify > Normals to insert a Normals node anywhere after the 3D object whose
lighting you want to adjust.
2. Connect a Camera, Axis, or light node to the Normals node’s lookat input.
3. In the Normals controls, open the action dropdown menu and select:
• unchanged to make no changes.
• set to assign the normals value to the normal x, y, and z fields.
• build to rebuild each normal based on the surrounding vertices. Adjust the threshold angle
slider to determine the break angle where two faces no longer constitute a smooth surface. An
angle of 0 means all faces are flat, whereas 180 means all faces are smooth. A good average
setting is 60.
• lookat to point all normals towards the Normals node’s lookat input.
• delete to remove the named attribute from the object. For example, if you remove the N
attribute, the object has no normals.
USER GUIDE
916
Relighting a 2D Image Using 3D Lights | To Relight a 2D Image Using Relight
Relighting a 2D Image Using 3D

Lights
The Relight node takes a 2D image containing normal and point position passes and lets you relight it
using 3D point lights. Essentially bypassing the need to return to a 3D application and re-render the
lighting, Relight provides a quick and interactive way to relight a 3D scene in a 2D environment.
Relight works by applying a 3D shader to a 2D image using the normal and point position passes
stored in separate image channels, and lets you attach and manipulate a 3D point light (or multiple
lights via a Scene node).
Note: Relight only works with Light nodes set to light type > point.
To Relight a 2D Image Using Relight

1. From the 3D > Lights menu, select Relight to add the node to your script.
2. Read in a 2D image containing normal and point position passes and connect it to the color input
of the node.
Tip: If the position pass and normal vectors are contained in separate images, they can be
combined using a ShuffleCopy node, which connects to Relight through the color input.
You can create normal and point position passes using the DepthGenerator node in NukeX, for
example. See Generating Depth Maps.
3. In the Relight properties panel, select the channel containing the normal data from the normal
vectors dropdown menu.
4. Select the channel containing the point position information from the point positions dropdown
menu.
USER GUIDE
917
Relighting a 2D Image Using 3D Lights | To Relight a 2D Image Using Relight
5. Connect a Light node with light type > point to the lights input, or multiple lights via a Scene
node.
6. Connect the camera that was used to render the original scene to the cam input.
7. Attach a shader (a Phong node, for example) to the material input. Depending on the type of
shader you attach, ensure that you've defined the necessary properties for it. For information on
defining material properties, refer to Object Material Properties.
Note: The camera input does not appear until the light inputs have been connected to a
light or scene node, and the material inputs do not appear until the camera input has been
connected.
8. Switch to the 3D Viewer and position your lights.

9. If the image in the color input contains an alpha channel and you'd like to use this as a mask to
limit the effects of Relight, check use alpha.
10. If required, adjust the ambient slider to set the global ambient light level for the scene.
11. To combine the lighting information from the Relight node with the original 2D image, use a
Merge node with operation set to multiply. Connect the Relight node to the Merge node’s A input
and your 2D image to its B input.
USER GUIDE
918
Cameras | To Relight a 2D Image Using Relight
Cameras
Nuke supports multiple cameras in a scene, with each providing a unique perspective.
Cameras in a scene.
For details on how to add a camera, look through it, and edit its lens characteristics, see Working with
Cameras.
In addition to using cameras to view and render 3D scenes, you can also set up cameras that project
2D images onto geometry in your scene. For more information, see Projection Cameras.
If you have created a camera in a third-party 3D application (such as Maya) and want to use it in Nuke,
you can export it from your 3D application in the FBX or Alembic format and then import it into
Nuke. See Importing Cameras from FBX Files and Importing Cameras from Alembic Files.
If your camera was created in Boujou, you can also use the import_boujou.tcl script to import it into
Nuke. See Importing Cameras from Boujou.
You can extrapolate a 3D point's position from imported camera data using the PointsTo3D node,
which can help you position geometry in a scene. See <xref> for more information.
Tip: NukeX and Nuke Studio include CameraTracker which creates animated cameras
without resorting to third-party applications. See Camera Tracking for more information.
USER GUIDE
919
Working with Cameras | To Add a Camera
Working with Cameras

This section explains how to add a camera to your script, look through it, lock it, and edit its lens
characteristics.
To Add a Camera
1. Click 3D > Camera to insert a Camera node.
OR
In the 3D Viewer, select create camera to place a new camera at the current position and
orientation in 3D space.
2. To setup the rendering camera, drag a connector from the new Camera node to the
ScanlineRender node.
OR
To setup an additional scene camera for viewing, drag a connector from the new Camera node to
the Scene node.
To Look Through a Camera

1. Press V to make sure you are looking through the 3D perspective view, and not one of the
orthographic views.
2. From the 3D Viewer window, select the camera from the dropdown menu in the top right corner.
USER GUIDE
920
Working with Cameras | To Lock the 3D Camera View
Note: This selection does not change the camera used for rendering. This changes only the
camera to “look through” for the current 3D Viewer.
Cameras in the current data stream automatically appear in the dropdown menu of cameras you can
select. To select a camera that doesn’t appear in menu, double-click the Camera node to open its
panel and add it to the menu.
To Lock the 3D Camera View

You can lock the 3D view to the selected camera or light. You can toggle between the unlocked and
locked modes by clicking the 3D view lock button, or by pressing Ctrl/Cmd+L.
• unlocked - move freely in the 3D view without restrictions. The 3D view lock button is gray.
• locked - lock your movement to the camera or light you’ve selected in the dropdown menu on the
right side of the 3D view lock button. The 3D view lock button is red.
To Use the Interactive 3D Camera View Mode

With the interactive 3D camera view mode you can change the camera or light values according to
your movement in the Viewer. You can activate the interactive mode by Ctrl/Cmd+clicking the 3D
view lock button. When the interactive mode is on, the 3D view lock button turns green. In order to
activate the interactive mode, you need to have a Camera or a Light node selected in the dropdown
on the right side of the 3D view lock button.
When the interactive mode is on, you can use the plus (+) and minus (-) keys to change the translate
values of the camera or light you’ve selected. When the interactive mode is off, these keys zoom your
view in and out.
To Edit a Camera’s Lens Characteristics

1. If necessary, double-click on the Camera node to display its parameters.
2. Click the Projection tab.
3. Drag the focal length slider to adjust the camera’s level of magnification.
4. Drag the near slide to edit the position of the camera’s forward clipping plane. Objects closer to
the camera than this plane are not rendered.
USER GUIDE
921
Working with Cameras | Using the Look Input
Note: The value for the near clipping plane must always be positive to produce a sensible
result.
5. Drag the far slider to edit the position of the camera’s rearward clipping plane. Objects farther
from the camera than this plane are not rendered.
6. Increment the window translate u (horizontal axis) and v (vertical axis) sliders to translate the
camera’s output along either axis.
7. Increment the window scale u (horizontal axis) and v (vertical axis) sliders to scale the camera’s
output on either axis.
8. Drag the window roll slider to rotate the camera’s output on z.

You can use the optional look input of the Camera node, so that the camera automatically rotates to
face towards the connected input. You can attach a Camera, Light, or Axis node to the look input. For
example, you can connect an Axis node to the look input so that the camera rotates to face the axis,
wherever it is moved.
If you animate a card to move along the x axis, you can attach a Camera and a Light node with the
look inputs so that they automatically rotate and face the card as it moves. To do this, complete the
following steps:
1. After animating your card, insert an Axis node.
2. Expression link the Axis node to the Card node by holding Ctrl, clicking on the translate
animation button in the Card node's properties, and dragging it to the translate animation
button in the Axis node's properties. The expression link is displayed as a green line with an arrow
denoting the direction of the expression. See Linking Expressions for more information.
3. Insert a Camera node and either drag the camera to the required position in the Viewer, or use
the Camera node properties to adjust the position of the camera.
USER GUIDE
922
4. Connect the Camera's look input to the Axis node.
5. Insert a Light node and drag the light to required position in the Viewer, or use the Light node
properties to adjust the position of the camera.
USER GUIDE
923
6. Connect the Light node's look input to the Axis node
7. Playback your animated card and notice that the Camera and Light follow the animated card.
USER GUIDE
924
USER GUIDE
925
Projection Cameras | First a Little Math...
Projection Cameras
In addition to viewing and rendering a 3D scene, cameras can also project a 2D still image or image
sequence onto geometry in the scene. This is similar to the front-projection systems used in practical
photography, where a background image or other element is projected onto the stage and
photographed with other elements.
In Nuke, a projection camera can receive camera data tracked from the original shot - or another shot
- to setup a projection that is match-moved to another source.
This setup requires these nodes: a projection camera, a Scene node, a Project3D node, a geometry
object node (what you’ll be projecting onto), and a 2D node with the image that you want to project.
First a Little Math...

When you create a projection camera, you need to gather some information and do a few small
calculations to make sure the projection works. Here are the bits of information you need:
• Focal length of the lens that photographed the projection image.
• Resolution of scanned image.
• Scanner pitch of the film scanning device.
After you have this information, you need to do these calculations to get the horizontal and vertical
aperture settings for the projection setup:
horiz. res. / scanner pitch = horizontal aperture
vertical res. / scanner pitch = vertical aperture
So, for example, if your image resolution is 720 x 486 and the scanner pitch is 20, then these are the
results:
720 / 20 = horizontal aperture = 36
486 / 20 = vertical aperture = 24.3
Generally, for most professional projects, you can get the lens focal length from the camera report
for the shot(s). If that is not available, you may be able to extrapolate lens information by running the
shot through a 3D tracking application, such as Boujou, Syntheyes, or RealViz.
USER GUIDE
926
Projection Cameras | Setting Up the Projection Camera Script
Setting Up the Projection Camera Script

Once you have the horizontal and vertical aperture and the lens focal length for the image you want
to project, you can complete the projection camera setup.
To Add a Projection Camera

1. Select 3D > Camera to add a new camera to your script and rename the node to identify it as a
projection camera.
2. Select 3D > Shader > Project3D to add a Project3D node to the script.
3. Connect the 2D image (i.e., Read node) to the Project3D node.
4. Connect the projection camera to the Project3D node.
5. Connect the Project3D node to the geometry node that should receive the 3D projection.
6. Double-click the projection camera node to load its parameters.
7. Click the Projection tab in the camera’s panel and then enter the information you gathered for
focal length, horiz aperture, and vert aperture.
USER GUIDE
927
Projection Cameras | To View a 3D Scene over a 2D Background Image
When you are finished, view the 3D scene to check the placement of the projection. The next section
explains how to preview the 2D and 3D elements together to check the results of the composite.
To View a 3D Scene over a 2D Background Image

1. Select the Scene node and press 1 to display its output to the Viewer.
2. If necessary, press Tab to toggle the Viewer to 3D mode.
3. Select the rendering camera object or node and press H to look through it.
4. Select the node with the 2D image you want to see in the Viewer, and then press Shift+2.
The Shift+2 keystroke connects the image to the Viewer (assigning the next available connection,
number 2), and also sets up the compare wipe.
5. Select the desired option from the Viewer composite dropdown menu, such as over, under,
minus, and wipe).
This last step superimposes the two elements in the Viewer. The crosshair (shown below) is the
control that lets you adjust the location and angle of the wipe for the comparison.
USER GUIDE
928
Projection Cameras | To View a 3D Scene over a 2D Background Image
Comparing a 3D scene over a 2D image.
USER GUIDE
929
Importing Cameras from FBX Files | Importing Cameras from an FBX File
Importing Cameras from FBX

Files
way, you can, for example, create a camera in Maya, export it in a .fbx file, and use the same camera
again in Nuke.
and Linux.
Importing Cameras from an FBX File

The Camera node lets you read in the standard FBX cameras (Producer Perspective, Producer Top,
Producer Bottom, Producer Right, Producer Left, Producer Front, Producer Back) and any other
cameras.
Using one Camera node, you can only import one camera from a .fbx file. If you need to import
several cameras, you need to use one Camera node per camera.
1. Select 3D > Camera to insert a Camera node in the place where you want to add the camera in
your script.
2. In the Camera controls, check read from file. When this is checked, the controls on the File tab
are enabled, and you can use them to read in a camera from an .fbx file. Any controls whose
values are read in from the .fbx file are disabled. You can still view these values and use them in
expressions but, as long as read from file is checked, you cannot modify them. Modifying the
USER GUIDE
930
Importing Cameras from FBX Files | Importing Cameras from an FBX File
values in the .fbx file, however, affect the disabled values in the Camera controls, because these
are reloaded from the .fbx file every time the node is instantiated.
3. To read in a camera from an .fbx file, click the folder icon on the File tab. Navigate to the .fbx file
and select Open.
FBX files support multiple takes in one file. Usually, one of the takes is a default take with no
animation.
5. From the node name dropdown menu, select the camera node you want to import from the .fbx
file.
6. In the frame rate field, define a frame rate (frames per second) to sample the animation curves.
To use this rate rather than the one defined in the .fbx file, check use frame rate.
7. To have the camera rotation values calculated using the look up vector and look at position, check
compute rotation. If you don’t check this, Nuke uses the rotation channel from the .fbx file
instead of computing a new one. The rotation values are always computed when there is a look at
target.
8. If you want to modify the camera properties imported from the .fbx file, uncheck read from file
on the Camera tab and make the necessary modifications. As long as read from file is
unchecked, your changes are kept.
9. To reload the camera properties from the .fbx file, make sure read from file is checked and click
the reload button on the File tab.
USER GUIDE
931
Importing Cameras from Alembic Files | To Import Cameras from an Alembic File:
Importing Cameras from Alembic

Files
You can import cameras from Alembic files (.abc file format) into a Nuke scene. During the import,
Nuke allows you to control which nodes in the Alembic scene get loaded by using an import dialog. If
there is only one item in the Alembic file, it loads automatically.
Tip: In addition to cameras, you can also import meshes (or NURBS curves/patch surfaces
converted to meshes), point clouds, and transforms from Alembic files.
To Import Cameras from an Alembic File:

next to them.
USER GUIDE
932
Importing Cameras from Alembic Files | To Import Cameras from an Alembic File:

5. On the Camera nodes, you can adjust the following:
• From the animation stack dropdown menu, select the take you want to use from the .abc file.
Alembic files support multiple takes in one file.
• From the node name dropdown menu, select the camera you want to import from the .abc file.
• In the frame rate field, define a frame rate (frames per second) to sample the animation. To
use this rate rather than the one defined in the .abc file, check use frame rate.
• If you want to modify the camera properties imported from the .abc file, uncheck read from
file on the Camera tab and make the necessary modifications. As long as read from file is
• To reload the camera properties from the .abc file, make sure read from file is checked and
click the reload button on the File tab.
• To have the camera rotation values calculated using the look up vector and look at position,
check compute rotation. If you don’t check this, Nuke uses the rotation channel from the .abc
file instead of computing a new one. The rotation values are always computed when there is a
look at target.
.abc file.
USER GUIDE
933
Importing Cameras from Boujou | To Import a Camera from Boujou:
Importing Cameras from Boujou

Nuke is shipped with a script called import_boujou.tcl, which lets you load in cameras created with
Boujou.
To Import a Camera from Boujou:

1. Save the Boujou camera solve as a .txt file.
2. In Nuke, click on a content menu button and select Script Editor. The Script Editor opens in the
pane whose content menu you used.
3. In the input pane of the Script Editor (that is, the lower pane), enter nuke.tcl("import_boujou").
Click the Run the current script button on the top of the Editor, or press Ctrl+Return
(Cmd+Return on a Mac).
4. In the File Browser that opens, navigate to the .txt file you saved in step 1.
A Camera, a ScanlineRender, and a Group node are loaded into Nuke. The Group node contains
cylinders to represent points from Boujou.
Tip: You can also open Nuke’s Boujou Text File Browser by doing the following:
1. Press x on the Node Graph to open the Nuke script command dialog.
2. In the dialog, check Tcl (if it’s not already checked).
3. In the command field, enter import_boujou.
4. Click OK.
These steps can be used to replace the first three steps in the above instructions.
USER GUIDE
934
Locating a 3D Point from an Animated Camera | Connecting the PointsTo3D Node
Locating a 3D Point from an

Animated Camera
Animated cameras created in third-party applications, such as Maya, or Nuke's CameraTracker
contain enough parallax information to locate a 3D point in a 2D scene using the PointsTo3D node.
PointsTo3D uses three reference frames in an image sequence to extrapolate the location of a 2D
point in 3D space.
Connecting the PointsTo3D Node

1. Read in your image sequence. See Loading Image Sequences for more information.
2. Import your animated camera as described under Cameras.
Tip: NukeX and Nuke Studio include CameraTracker, which creates animated cameras
without resorting to third-party applications. See Camera Tracking for more information.
3. In the toolbar on the left of the interface, navigate to Transform > PointsTo3D to add a
PointsTo3D node.
4. Connect the image sequence and Camera to the PointsTo3D node and then connect the Viewer.
Setting Reference Frames

1. In the Properties panel, set Camera type to free move.
USER GUIDE
935
Locating a 3D Point from an Animated Camera | Calculating the 3D Position
2. Scrub through the sequence and pick out three frames where the feature is clearly visible.
3. Scrub to the first reference frame where your feature is visible and then drag the pointA Viewer
widget from the bottom-left of the Viewer to the feature's location.
4. Click set frame.
The 2D coordinates and reference frame number are inserted into the Point A controls.
Tip: You can adjust the 2D point using the x and y controls in the properties for pixel-
perfect positioning.
5. Scrub to the second reference frame where your feature is visible and then drag the pointB
Viewer widget from the bottom-left of the Viewer to the feature's location.
6. Click set frame.
7. Scrub to the third reference frame where your feature is visible and then drag the pointC Viewer
widget from the bottom-left of the Viewer to the feature's location.
8. Click set frame.
You can now see coordinate and frame information for all three reference frames in the
properties.
Calculating the 3D Position

1. Click Calculate to display the range dialog.
2. Click the Frame range dropdown to select the range bounds and the frames to calculate, if
required.
3. To calculate the position from the image sequence using a lower resolution, enable Use proxy.
This can speed up the calculation, but the results may not be as accurate.
4. To stop the calculation if an error occurs during the process, disable Continue on error.
5. Click OK.
USER GUIDE
936
Locating a 3D Point from an Animated Camera | Calculating the 3D Position
Nuke runs through the sequence, tracking the 2D position against the three reference frames to
determine the corresponding 3D point.
Once the calculation is complete, the point2d Viewer widget is placed on the tracked point in 2D
space.
6. Press Tab in the Viewer to switch to the 3D perspective view. The calculated 3D point is
highlighted in 3D space.
The 2D points used to calculate the The resulting position in the 3D Viewer.
3D position.
You can use the calculated 3D point to help you place elements in the scene, or click generate axis to
add an Axis node to control all elements in the scene together. See Parenting to Axis Objects for more
information on axes.
USER GUIDE
937
Transforming Geometry, Cameras, and Lights | Calculating the 3D Position
Transforming Geometry,
Cameras, and Lights
Transform operations include moving, scaling, rotating the objects in your 3D scene. When an object
node is active, you can enter specific transform settings in the node parameters (see Transforming
from the Node Properties Panel), or directly manipulate the object with the transform handles
displayed in the 3D Viewer (see Using the Transform Handles). Transformations occur from the
location of the object’s pivot point (see Transformations and the Pivot Point).
Nuke allows you to transform several objects together or have one object always face another (see
Parenting to Axis Objects and Using the TransformGeo Node).
You can also link transform parameters to imported track or camera data (see Applying Tracks to an
Object), or control the transforms with animation curves.
Using FBX or Alembic files, you can import transforms from third-party 3D applications, such as Maya
(see Importing Transforms from FBX Files and Importing Transforms from Alembic Files).
USER GUIDE
938
Using the Transform Handles | To Move an Object with the Transform Handles
Using the Transform Handles

Transform handles appear when a 3D object with transform capabilities is loaded into the Properties
Bin. The colors of the handles correspond to the axes available in 3D space: red transforms the x-
axis, green transforms the y-axis, and blue transforms the z-axis.
To Move an Object with the Transform Handles

• Drag an object to move it on any axis.
• Shift+drag to constrain movement to one axis.
To Rotate an Object with the Transform Handles

• Ctrl+click (Mac users Cmd+click) between the rotation rings and drag to rotate the object on any
axis.
• Click on a specific axis ring and drag to constrain the rotation to one axis.
USER GUIDE
939
Using the Transform Handles | To Scale an Object with the Transform Handles
To Scale an Object with the Transform Handles

• Ctrl+Shift+click (Mac users Cmd+Shift+click) and drag on the central yellow square to scale
uniformly on all three axes.
• Ctrl+Shift+click (Mac users Cmd+Shift+click) and drag on a single axis square to scale on that
axis.
USER GUIDE
940
Transforming from the Node Properties Panel | To Set Transformation Options
Transforming from the Node

Properties Panel
The transform handles (see Using the Transform Handles) are a convenient way to move objects
around in the 3D workspace, but when you want more precision, you should enter values directly into
the object’s node panel. The panel also includes transform and rotation order options, which are not
available within the 3D Viewer.
The following assumes you’ve already loaded the object’s parameters into the Properties Bin.
To Set Transformation Options

• From the transform order dropdown menu, select an option to define the order by which
transformations are executed (s signifies scale, r, rotation; and t, translation).
• From the rotation order dropdown menu, select an option to define the axial order by which
rotations are executed.
To Transform an Object from Its Panel

• To move the object along one or more axes, increment or decrement the translate x, y, and z
fields.
• To rotate the object, increment or decrement the rotate x, y, and z fields.
• To scale the object on all axes simultaneously, increment or decrement the uniform scale field.
• To scale the object asymmetrically (on x, y, or z), increment or decrement the scale x, y, and z fields.
• To skew the object (warp it by rotating its local axes), increment or decrement the skew x, y,and z
fields to rotate the corresponding axis (and associated object vertices).
USER GUIDE
941
Transformations and the Pivot Point | To Move the Pivot Point
Transformations and the Pivot

Point
When you make changes to an object’s position, scaling and rotation, these occur from the location
of the object’s origin point or pivot. By default, the pivot point is located at the intersection of the
object’s local axes.
You can offset the pivot point and move it anywhere you like - you can even move it outside of the
object. Subsequent local transformations then occur relative to the new pivot point location.
To Move the Pivot Point

1. Double-click on the object node to display its parameters.
2. Change the values of the pivot x, y, and z fields to move the local axis in any direction.
Pivot point location and coordinates.
The object’s graphical overlay points to the location of the pivot point with a line. All subsequent local
transformations occur relative to this pivot point.
Once you’ve defined the location of an object’s pivot point, you can use the object’s transform
parameters to translate, rotate, scale, and skew the object relative to the pivot point.
USER GUIDE
942
Parenting to Axis Objects | To Add an Axis Object
Parenting to Axis Objects

An axis object works as a null object by adding a new transformational axis to which other objects
may be parented. Even when objects already have their own internal axes, it’s sometimes useful to
parent in a separate axis.
For example, the Axis node has been parented to the other objects in the scenes (the two image
planes and the camera). The result is an axis object that globally controls the scene. Rotating it, for
example, rotates all objects in the scene, as the figure below shows.
Rotating an entire scene with an axis object.
Tip: To move several objects together, you can also merge them using a MergeGeo node
(see Merging Objects) and then control them using a TransformGeo node (see Using the
TransformGeo Node).
To Add an Axis Object

1. Click 3D > Axis to insert an Axis node.
2. To create the parent relationships, connect the Axis node to all object nodes you wish to control
with the Axis transformation controls.
USER GUIDE
943
Parenting to Axis Objects | To Add an Axis Object
To create a nested transformational hierarchy, chain additional Axis nodes to the first one you
inserted. For example, you could create a hierarchy of three Axis nodes to control rotation, scale, and
transform.
Creating a nested transformational

hierarchy.
In the above example, Axis3 (rotation) is connected to Axis2 (scale). Axis2 is connected to Axis1
(transform), and Axis1 is connected to the TransformGeo node(s) that you want to affect. With the
Axis nodes connected in this manner, their transformation data ripples down the chain and is added
to the settings of the TransformGeo node.
USER GUIDE
944
Parenting to Axis Objects | Using the Look Input

You can use the optional look input of the Axis node, so that the axis automatically rotates to face
towards the connected input. You can attach a Camera, Light, or Axis node to the look input. For
example, you can connect an Camera node to the look input so that the axis rotates to face the
camera, wherever it is moved. See Working with Lights or Working with Cameras for an example.
USER GUIDE
945
Using the TransformGeo Node | To Have One 3D Object Always Face Another
Using the TransformGeo Node

The TransformGeo node allows you to move, rotate, scale, and perform other transformations on
several objects merged together with a MergeGeo node. It also lets you connect geometry objects to
an Axis node. By doing so, you can move all the connected objects together by using the Axis
transformation controls. All you need to do is insert a TransformGeo after each geometry object,
connect the Axis node to the TransformGeo nodes’ axis input, and adjust the transform controls of
the Axis node. For more information, see Parenting to Axis Objects.
Another use of the TransformGeo node is to have the rotation of one object depend on the position
of another so that the first object is always rotated to face or “look at” the second one. For example,
you can have a sphere object always facing a cylinder object, regardless of the cylinder’s position.
When the cylinder is moved to a new position, the sphere is automatically rotated to face it.
To Have One 3D Object Always Face Another

1. Select the 3D object node (for example, a sphere) that you want to face another object.
2. Select 3D > Modify > TransformGeo to insert a TransformGeo node.
3. Select the object you want the first object to face (for example, a cylinder), and insert a
TransformGeo after this node, too.
4. Connect the first TransformGeo node into the look input of the second TransformGeo node.
USER GUIDE
946
5. Open the controls of the first TransformGeo node and go to the Look tab.
6. From the look axis dropdown menu, select the axis around which the object rotates to face the
other object:
+Z +Y
+X -Z
USER GUIDE
947
-Y -X
7. Use the rotate X, rotate Y, and rotate Z check boxes to select the axes the object rotates around.
For the first object to truly face the second, you need to check all three check boxes.
8. Adjust the look strength slider to define the extend of the rotation. The smaller the value, the
less the object is rotated. Setting the value to 0 produces no rotation.
9. If you want to use an alternate scheme to calculate the rotation, check use quaternions. This may
be useful for smoothing out erratic rotations along the selected look axis.
If you now adjust the second TransformGeo node’s transform controls, you’ll notice that the first
object automatically rotates to face the second object. For more information on how to adjust the
transform controls, see Using the Transform Handles and Transforming from the Node Properties
Panel.
USER GUIDE
948
Applying Tracks to an Object | To apply a channel file to an object
Applying Tracks to an Object

Nuke can import channel files and apply the motion data to the transformation parameters of any
camera or object. The most common purpose for this is to simulate a practical camera move or move
objects along a defined path.
Channel files contain a set of Cartesian coordinates for every frame of animation in a given shot. This
information is calculated by 3D tracking software, such as 3D-Equalizer, Maya, or Boujou, and then
exported as channel files.
To apply a channel file to an object

1. Double-click on an object or camera node to display its parameters.
2. Click import chan file. The file navigation dialog appears.
3. Navigate to the channel file, then click OK.
4. Nuke reads in the channel data and displays a status message about the number of data frames
imported. You’ll also notice the object’s translation parameters turn green to indicate these
parameters are now controlled by animation data. Scrub the Viewer and you’ll notice the object or
camera now moves according to the transformation data imported from the channel file.
Note: You can use the export chan file button to export as a chan file any animated
translation parameters which you’ve applied to given object. This is a useful method of
sharing setups between artists.
Tip: You can also use channel files to import cameras created in other applications into
Nuke. However, as the chan file format is not a standard file format, you may need a file
format converter to export chan files from other applications.
USER GUIDE
949
Importing Transforms from FBX Files | Importing Transforms from an FBX File
Importing Transforms from FBX

Files
way, you can, for example, create a transform in Maya, export it in a .fbx file, and use the same
transform again in Nuke.
Note: For the FBX SDK version used in Nuke, see Third-Party Library Versions on page 2021.
and Linux.
Importing Transforms from an FBX File

The Axis node reads in transforms, markers and nulls (locators) from FBX files. You can use it to
import one transform, marker, or null per Axis node.
1. Select 3D > Axis to insert an Axis node in your script. Connect the Axis node to a Scene node.
2. In the Axis controls, check read from file. This enables the controls on the File tab, allowing you
to import transforms from an .fbx file. It also disables controls whose values are filled in from the
.fbx file. As long as read from file is checked, you cannot modify these values. You can, however,
view them and use them in expressions. The values are reloaded from the .fbx file every time the
node is instantiated, so any changes you make in the .fbx file’s values are reflected in the Axis
controls.
USER GUIDE
950
Importing Transforms from FBX Files | Importing Transforms from an FBX File
3. On the File tab, click the folder icon to open the File Browser. Navigate to the .fbx file that
contains the transform you want to use. Click Open.
FBX files support multiple takes, one of which is usually a default take with no animation.
5. From the node name dropdown menu, select the transform, marker, or null you want to import
from the .fbx file.
6. If you do not want to use the frame rate from the .fbx file for sampling the animation curves, in
the frame rate field, enter a new value (frames per second). To override the frame rate defined in
the .fbx file and use the one you defined here, check use frame rate.
7. If you want to modify the transform properties imported from the .fbx file, uncheck read from
file on the Axis tab and make the necessary modifications. As long as read from file is
8. To reload the transform properties from the .fbx file, make sure read from file is checked and
USER GUIDE
951
Importing Transforms from Alembic Files | To Import Transforms from an Alembic file:
Importing Transforms from

Alembic Files
You can import transforms from Alembic files (.abc file format) into a Nuke scene. During the import,
Nuke allows you to control which nodes in the Alembic scene get loaded by using an import dialog. If
there is only one item in the Alembic file, it loads automatically.
Tip: In addition to transforms, you can also import meshes (or NURBS curves/patch surfaces
converted to meshes), point clouds, and cameras from Alembic files.
To Import Transforms from an Alembic file:

next to them.
USER GUIDE
952
Importing Transforms from Alembic Files | To Import Transforms from an Alembic file:

5. On Axis nodes, you can adjust the following:
• From the animation stack dropdown menu, select the take you want to use from the .abc file.
Alembic files support multiple takes in one file.
• From the node name dropdown menu, select the transform, marker, or null you want to
import from the .abc file.
• In the frame rate field, define a frame rate (frames per second) to sample the animation. To
use this rate rather than the one defined in the .abc file, check use frame rate.
• If you want to modify the transform properties imported from the .abc file, uncheck read from
file on the Axis tab and make the necessary modifications. As long as read from file is
• To reload the transform properties from the .abc file, make sure read from file is checked and
.abc file.
USER GUIDE
953
Adding Motion Blur to the 3D Scene | To Import Transforms from an Alembic file:
Adding Motion Blur to the 3D

Scene
To create more realism for a 3D scene, you’ll want to add motion blur to it based on the movement of
your 3D camera. This can be done in two ways:
• Adjust the samples value in the ScanlineRender or RayRender node’s Properties panel. The image is
sampled multiple times over the shutter period. This way is the most accurate, but also the slowest,
because the full rendering calculation is done multiple times for each pixel. See Adding Motion Blur
Using a Renderer.
• Use the VectorBlur node to generate motion blur based on motion vectors. This way is faster to
render, as it is only affected by the pixel resolution of the final image rather than the complexity of
your 3D scene. See Adding Motion Blur Using VectorBlur.
USER GUIDE
954
Adding Motion Blur Using a Renderer | To Import Transforms from an Alembic file:
Adding Motion Blur Using a

Renderer
To add motion blur using ScanlineRender or RayRender:
1. In the node’s controls, go to the MultiSample tab.
2. Increase the samples value to sample the image multiple times over the shutter time. The higher
the value, the smoother the result, but the longer the render time.

Shutter set to 20. Shutter set to 40.

USER GUIDE
955
current frame, enter a negative value. For example, a value of -0.5 would open the shutter half a
5. If you're using ScanlineRender, you can add randomness to the distribution of samples in time so
they don’t produce regularly spaced images, by adjusting randomize time. This affects all
samples except the first and the last one. The larger the value, the larger the time difference
between the samples.
6. If you're using ScanlineRender, set sample diameter to the diameter of the circle that the
samples for each pixel are placed in for antialiasing. The larger the value, the more pixels are
jittered.
7. If you're using ScanlineRender, adjust the focus diameter to randomly orbit the camera about a
point at the focal distance in front of it for each sample to produce depth-of-field effects from
multiple samples.
Note: The focal distance is set in the Camera node's controls, in the Projection tab.
8. If rendering has become very slow, you can approximate multi-sample rendering and reduce
render times by using stochastic samples. This sets the number of samples, per pixel, to use in
stochastic estimation (0 is disabled). Lower values result in faster renders, while higher values
improve the quality of the final image.
Note that rather than setting stochastic samples to a very large value, you may want to leave it at
a relatively low value and increase samples instead.
Stochastic sampling is based on Robert L. Cook’s Stochastic Sampling in Computer Graphics,
available in ACM Transactions on Graphics, Volume 6, Number 1, January 1996.
Stochastic samples set to 0. Stochastic samples set to 5.
USER GUIDE
956
9. To distribute the samples uniformly over the shutter time, enable uniform distribution. This
gives the same importance to all samples over the shutter time, and produces more accurate
results for stochastic multi-sampling.
For more information on the ScanlineRender and RayRender nodes, see Rendering a 3D Scene.
USER GUIDE
957
Adding Motion Blur Using VectorBlur | To use VectorBlur with ScanlineRender
Adding Motion Blur Using

VectorBlur
Nuke’s VectorBlur node generates motion blur by blurring each pixel into a straight line, using the
values from the motion vector channels (u and v channels) to determine the direction of the blur.
Compared to generating motion blur using the ScanlineRender node’s MultiSample controls, this is
less accurate but faster to render.
You can create the necessary motion vectors for use with VectorBlur in several ways:
• Use the ScanlineRender or RayRender nodes. This is usually more accurate than using a
MotionBlur3D node (described below), and works well with both static and moving cameras as well
as both linear and non-linear camera movement. See To use VectorBlur with ScanlineRender or To
use VectorBlur with RayRender.
• Alternatively, if your 3D scene is static or nearly so and the camera movement over the shutter time
is nearly linear, you can also use the Filter > MotionBlur3D node. However, note that MotionBlur3D
only uses camera information to produce motion vectors. If you have moving objects in your scene
but the camera is static, MotionBlur3D isn’t able to produce any output. See To use VectorBlur with
MotionBlur3D.
• Several third-party 3D applications can also produce motion vector information as two-channel,
floating point images that you can use with VectorBlur. If possible, you should unpremultiply these
images. See To use VectorBlur with Third-Party Motion Vectors.
To use VectorBlur with ScanlineRender

1. In your 3D scene, open the ScanlineRender properties and go to the Shader tab.
2. Make sure motion vectors is set to one of the following (not off):
• classic - Render motion vectors the classic (pre-Nuke 6.1) way. This option is only provided for
backwards compatibility, and isn't always accurate.
• velocity - Store the velocity of every single pixel in the motion vector channels (pre-Nuke 7.0
way). This option is only provided for backwards compatibility. In order to have the same
behavior as Nuke 6.3, set samples to 1.
• distance - For every pixel, store the distance (in pixels) between samples in the motion vector
channels. This is the recommended option that usually produces the best results. It also allows
the VectorBlur node to produce curved vector blur where interpolation between two frames is
made according to a curve rather than linearly.
3. Use motion vector channels to select where to store the generated motion vectors.
USER GUIDE
958
Adding Motion Blur Using VectorBlur | To use VectorBlur with ScanlineRender
If you view the channels you selected, you should see they now contain motion vectors.
A render of a 3D scene. Motion vectors produced by

ScanlineRender.
4. On the MultiSample tab, leave samples set to 1. You can increase this value later if non-linear
movement cannot be approximated sufficiently well using VectorBlur’s linear approach.
5. Select Filter > VectorBlur to insert a VectorBlur node after ScanlineRender.
A node tree with ScanlineRender and VectorBlur.

6. In the VectorBlur properties, use the channels dropdown menu to select the channels to blur.
7. Select the motion vector channels from the uv channels dropdown menu. These should be the
same channels as the channels you chose to create in step 3.
8. Set the mv presets dropdown to set which renderer format to apply to the motion vectors. Each
preset contains a scale and offset specific to the renderer, so that the vectors are in the format
expected by Nuke.
9. Use the blur uv dropdown to blur the motion vectors themselves, before blurring the image, to
smooth the results between regions with very different motion. Choose either:
• uniform - applies a small uniform blur to soften edges in the UV map, or
USER GUIDE
959
Adding Motion Blur Using VectorBlur | To use VectorBlur with RayRender
• linear - applies a linearly-weighted blur to blend between regions with different motion. This
has a stronger effect than the uniform option.
Tip: The uniform option can be used to create blur similar to pre-Nuke 10 results.
10. Set shutter offset to 0. This means the shutter opens at the current frame.
11. Select the blur type using the dropdown.
12. To adjust the amount of blur, adjust the motion blur controls.
The original image. Motion blur produced by

VectorBlur.
To use VectorBlur with RayRender

1. In your 3D scene, open the RayRender properties and go to the AOVs tab.
2. Enable output AOV and then use the motion vector dropdown to select where to store the
generated motion vectors.
A render of a 3D scene. Motion vectors produced by

ScanlineRender.
3. On the MultiSample tab, leave samples set to 1. You can increase this value later if non-linear
movement cannot be approximated sufficiently well using VectorBlur’s linear approach.
USER GUIDE
960
Adding Motion Blur Using VectorBlur | To use VectorBlur with RayRender
4. Select Filter > VectorBlur to insert a VectorBlur node after RayRender.
A node tree with RayRender and VectorBlur.

6. Select the motion vector channels from the uv channels dropdown menu. These should be the
same channels as the channels you chose to create in step 2.
7. Set the mv presets dropdown to set which renderer format to apply to the motion vectors. Each
preset contains a scale and offset specific to the renderer, so that the vectors are in the format
expected by Nuke.
9. Set shutter offset to 0. This means the shutter opens at the current frame.
USER GUIDE
961
Adding Motion Blur Using VectorBlur | To use VectorBlur with MotionBlur3D
The original image. Motion blur produced by

VectorBlur.
To use VectorBlur with MotionBlur3D

1. In your 3D scene (which should be static or nearly static), open the ScanlineRender properties, go
to the Shader tab, and set motion vectors to off.
This tells ScanlineRender not to produce motion vectors, which is what you want if you are going
to use MotionBlur3D to generate the vectors.
2. Select Filter > MotionBlur3D to insert this node and connect it to the ScanlineRender node’s
output.
3. Connect the rendering camera to the cam input of the MotionBlur3D node. Note that this has to
be a moving camera. The camera movement over the shutter time should be linear or nearly
linear.
4. In the MotionBlur3D properties, use Output UV to select where to store the generated motion
vectors.
5. Select Filter > VectorBlur to insert this node and connect it to the MotionBlur3D node.
USER GUIDE
962
Adding Motion Blur Using VectorBlur | To use VectorBlur with MotionBlur3D
A node tree with MotionBlur3D and VectorBlur.

6. In the VectorBlur properties, select the motion layer from the uv channels dropdown menu.

10. To adjust the length of the blur, adjust the Shutter setting in the MotionBlur3D Properties panel.
Low Shutter value. High Shutter value.
USER GUIDE
963
Adding Motion Blur Using VectorBlur | To use VectorBlur with Third-Party Motion Vectors
To use VectorBlur with Third-Party Motion Vectors

1. Select Filter > VectorBlur to insert a VectorBlur node after a 2D image that contains motion
vector channels.
3. Select the motion vector channels from the uv channels dropdown menu.
4. Set the mv presets dropdown to apply the presets associated with the selected renderer's default
settings, if present.
5. If your motion vectors have been premultiplied, check uv alpha and specify the channel used to
premultiply the image.
6. Select the blur type and the required amount of blur using the motion blur controls.
USER GUIDE
964
Exporting Geometry, Cameras, Lights, Axes, or Point Clouds | To use VectorBlur with Third-Party
Exporting Geometry, Cameras,

Lights, Axes, or Point Clouds
You can export geometry, cameras, light, axes and point clouds into an FBX (.fbx) or Alembic (.abc)
file using the WriteGeo node.
1. Create a WriteGeo node and attach it to a Scene node.
2. In the WriteGeo controls, select fbx or abc in the file type dropdown, and check the objects to
export:
• geometries - to write the scene geometries into the file.
• cameras - to write the scene cameras to the file.
• lights - to write the scene lights into the file (.fbx only).
• axes - to write the scene axes into the file.
• point clouds - to write the scene point clouds into the file.
3. If you’re writing FBX files:
• Check ascii file format if you don’t want to write out a binary .fbx file.
• Check animate mesh vertices if you want the mesh vertices animated and keyframes created
at every frame. The animated meshes use vertex point caches for the data. A directory with _fpc
appended to the file name is created to contain the point caches. By default the animated
meshes check box is off so the point cache directory is not created.
4. If you're writing Alembic files, select the storage format to use when writing the file:
• HDF - A storage format that maintains backwards compatibility.
• Ogawa - A storage format that offers faster file reading and smaller files.
5. Browse to the destination you want to save the .fbx or .abc file in the file field and give it a name.
6. Check the limit to range box if you want to disable the node when executed outside of a specified
frame range. In the frame range fields, enter the range of frames you want to make executable
with this Write node.
7. Select the views to export, such as left and right for stereoscopic projects.
8. Click Execute.
9. If necessary, you can change the frame range you want to include in the file using the pop-up
dialog and then click OK.
A progress bar displays, and the selected file is saved to the specified directory.
USER GUIDE
965
Rendering a 3D Scene | Choosing a Render Node
Rendering a 3D Scene
The 3D Viewer displays the scene using an OpenGL hardware render. When you build a scene, Nuke
renders high-quality output from the perspective of the camera connected to the render node. The
rendered 2D image is then passed along to the next node in the compositing tree, and you can use
the result as an input to other nodes in the script.
Choosing a Render Node

Nuke ships with two render nodes, ScanlineRender and RayRender, which are suited to different
tasks. ScanlineRender, as the name suggests, renders row-by-row advancing down the picture.
RayRender traces a path from the camera, or virtual eye, to the light source pixel-by-pixel.
Note: NukeX and Nuke Studio also include Pixar’s PhotoRealistic RenderMan®, another ray
renderer. See PrmanRender for more information.
The ScanlineRender and RayRender nodes have the same inputs and share some of the same
controls in their respective Properties panels, but they both have different strengths and
weaknesses:
• ScanlineRender generally produces faster results, but is less accurate with reflection and refraction.
• ScanlineRender supports Deep workflows downstream by injecting a deep channel into the node
tree. See Using ScanlineRender to Generate Deep Data.
• RayRender generally produces very accurate reflection, but at the cost of processing time.
• RayRender does not currently support Deep workflows or render sprite-based particles from
NukeX's Particles system.
As a general rule, if you can afford to wait for a render and don't need Deep data or particles that rely
on sprites, use RayRender.
Rendering Out a Scene

To render a 3D scene into 2D space you need three things: your scene, a render camera, and a render
node. When set up correctly, the output of the render node is passed down the node tree for
compositing as normal.
To render out a scene:

1. Set up your 3D scene as normal, including geometry, materials, lights, etc.
USER GUIDE
966
Rendering a 3D Scene | To Add Motion Blur to the 3D Scene
2. Add a render node and connect the scene to the obj/scn input.
3. Make sure the render camera is connected to the render node's cam input. See Cameras for more
information.
4. Add a background to the bg input, if required. You can use the bg input to composite a
background image into the scene and to determine the output resolution.
Note: If the bg input is not used, the render node output defaults to the root.format or
root.proxy_format defined in the Project Settings.
5. Toggle the Viewer back to 2D.

6. Connect the output of the render node to the appropriate 2D nodes in your script.
See the Nuke Reference Guide for more information on the different render node controls.
To Add Motion Blur to the 3D Scene

You can use the properties on the MultiSample tab to add motion blur to your 3D scene. For more
information, see Adding Motion Blur to the 3D Scene.
USER GUIDE
967
Stereoscopic Scripts
The title of this chapter is slightly misleading, as Nuke isn’t actually limited to stereoscopic views, but
rather provides multi-view support for as many views as you need. The views do not have to be stereo
pairs, but since that is the most obvious application, this chapter mainly deals with stereoscopic
projects.
Quick Start
In many ways, Nuke lets you work on stereoscopic material just like you would on any other images.
However, there are also a few stereo-specific settings and nodes that you need to be aware of when
compositing stereoscopic material. The following teaches you how to set up your stereo project, read
in and view your images, use the stereo nodes, and render the final output.

1. The first step in working on stereo footage in Nuke is to set up views for them in your project
settings (you can open up the project settings by pressing S over the Node Graph). Press Set up
views for stereo on the Views tab to do this. For more information, see Setting Up Views for the
Script.
2. You can then load your stereo footage into Nuke, either by entering a variable in the file name
field or by using a JoinViews node. For more information, see Loading Multi-View Images.
3. In the Viewer, you can select which view to display with the views buttons. You can also display
the views side by side or mix them together with the SideBySide and MixViews nodes. For more
information, see Displaying Views in the Viewer.
4. Sometimes you need to make changes to one view, while the other one remains as it is. In these
cases you need to either split off a view using the View menu or, in case of RotoPaint, use the
right-click options in the stroke/shape list to split off views. For more information, see Selecting
Which Views Display Changes.
If you need to perform totally different actions on the two views, you can separate one view for
processing with the OneView node. See Performing Different Actions on Different Views.
5. If you have a disparity field available, you can also use the correlate options in the RotoPaint node
and the View menu to have changes made to one view automatically reproduced in the other.
For more information, see Reproducing Changes Made to One View.
6. With the ShuffleViews node you can rearrange your views, and with the Anaglyph node you can
convert your footage into a red and cyan anaglyph footage. For more information, see Swapping
Views and Converting Images into Anaglyph.
USER GUIDE
968
Stereoscopic Scripts |
7. Sometimes you need to re-adjust the convergence, or the inward rotation of your left and right
view cameras. This changes the point in the image that appears at screen depth when viewed with
3D glasses. For more information, see Changing Convergence.
8. Finally, you can preview a view of your stereo project by flipbooking it and rendering it out. For
more information, see Previewing Stereoscopic Images.
USER GUIDE
969
Setting Up Views for the Script | Creating Views Automatically
Setting Up Views for the Script

You can import your footage and let Nuke create the views automatically or set up views in advance in
the project settings. This allows you to process the individual views separately or both views together,
and see the effect of your changes on each view.
If you are likely to need the same views in several projects, you may want to save the views you
created in a template.nk script file. For more information on how to do this, see Template Scripts.
Creating Views Automatically
Note: Automatic view creation is not implemented for multi-view .mov files. See Creating
and Managing Views Manually for information on how to create the views.
1. Read in your multi-view files as normal.

A warning dialog is displayed.
2. Click Add Views, Replace Views, or No:

• Add Views - add the views in the incoming clip to those that exist in the project.
• Replace Views - replace all existing project views with those in the incoming clip.
• No - import the clip and display only the first view in the file, retaining any existing views in the
project.
You can now access the views in your project from the view dropdown menu of certain nodes’
controls. You’ll also notice that each view has its own button in the Viewer controls.
Creating and Managing Views Manually

1. Select Edit > Project Settings or pres S in the Node Graph.
2. Go to the Views tab. The exiting views are listed in the views field.
USER GUIDE
970
Setting Up Views for the Script | Creating and Managing Views Manually
3. If you want to remove the view called main and add views called left and right, click the Set up
views for stereo button.
Note: If the script is already set up for multiple views, Set up view for mono is displayed.
Clicking Set up view for mono deletes all existing views and adds a single main view.
The two new views are assigned colors, red and green by default. To change the colors, double-
click on the color field and select another color from the color picker.
If you check Use colors in UI?, these colors are used in Node Graph connections, split views
indicators on nodes, and Viewer and ShuffleViews node controls to make it easier to differentiate
between views.
USER GUIDE
971
4. If you want to add other new views, click the + button.

A new view is added automatically.
5. To rename the view, double-click its name in the view list and enter the new name.
6. Repeat these steps as necessary until you’ve got the views you want. You can assign colors to all
views by double-clicking the area on the right of the view name.
7. To delete an unnecessary view, select the view from the list and click the - button.
8. To move a view around in the list of views, click the up and down arrows above the views panel.
9. To select the view that’s displayed whenever you load the project, set the hero dropdown to the
appropriate view.
You can now access the views in your project from the view dropdown menu of certain nodes’
controls. You’ll also notice that each view has its own button in the Viewer controls.
USER GUIDE
972
If you created many views, you may want them to appear in a dropdown menu rather than as
separate buttons in the Viewer and node controls. To switch to using dropdown menus, disable View
selection uses buttons? on the Views tab of the Project Settings panel.
USER GUIDE
973
Loading Multi-View Images | To Read Images in
Loading Multi-View Images

Once you have set up the views, you are ready to read your images into Nuke. To make things easier,
the images you read in should have the view name or the first letter of the view name in the filename,
for example filename.left.0001.exr, filename.l.exr, or lefteyefilename.0001.cin.
If you are using .exr files, your files can contain both the input for the left eye and the input for the
right eye, as .exr files support multiple views in a single file. With any other file types, you need to
have separate files for the left and right inputs.
To Read Images in
1. Select Image > Read.
2. Navigate to the files containing the images intended for either the left or right eye (or in the case
of .exr and .mov images, both eyes), and select Open.
• If the images you want to read in contain a view name or the initial letter of one (for example,
left, right, l or r) in their file names, replace this with the variable %V or %v in the file field of
the Read node’s controls. Use %V to replace an entire view name (for example, left or right),
and %v to replace an initial letter (for example, l or r). When a variable is used, Nuke reads in
the missing inputs and combines all inputs into a single output.
For example, if you read in image.left.cin and changed the name to image.%V.cin, Nuke would
read in both image.left.cin and image.right.cin with the same Read node, provided that views
called left and right existed in your project settings. Both input images would be combined into
a single output.
Note: Mac and Linux operating systems can be case-sensitive or case-insensitive. If your OS
is case-sensitive, you'll need to make sure you use the correct case when naming your left
and right views, as the %v variable can only retrieve the case used in the view name.
You can also use the %V and %v variables at a directory level. For example, let’s say you have set
up views called testleft, testmiddle and testright, and you have the following directories and
files:
mydirectory/testleft/image.testleft.cin
mydirectory/testmiddle/image.testmiddle.cin
mydirectory/testright/image.testright.cin
If you now read in image.testleft.cin and changed the pathname to
mydirectory/%V/image.%V.cin, all three inputs would be read in with the same Read node.
USER GUIDE
974
Loading Multi-View Images | Combining Views from Different Files into a Single Output
• If the images you want to read in do NOT contain a view name or the initial letter of one (for
example, left, right, l or r) in the file names and are not stereo .exr files, insert a Read node for
each input and combine them into a single output using the JoinViews node (see below for
instructions on how to do that).
• If the images you want to read in are in the stereo .exr file format, you do not need to do
anything. However, remember that not all .exr files contain multiple views. If you are using files
that are not, follow the instructions in the first two points.
• If the images you want to read in are in the multi-view .mov file format, you need to open the
Read node's properties and disable First track only.
You'll notice that the .mov Read node in the Node Graph is now marked with to denote
multiple views. You can now switch between views using the buttons above the compositing
Viewer.
Combining Views from Different Files into a Single Output

1. Select Image > Read to read in your image sequences containing the different views.
2. To insert a JoinViews node, select Views > JoinViews.
3. Connect the inputs of the JoinViews node into the appropriate Read nodes. There should be an
input for each view you have created in the project settings. The inputs are labeled with the name
of the view.
If you have assigned colors to the views and checked Use colors in UI? on the Views tab of your
project settings, the connecting arrows reflect the view colors. If this does not happen and the
arrows are black, you may have connected the inputs the wrong way around. Check that you have
connected each Read node to the correct input of the JoinViews node.
USER GUIDE
975
Loading Multi-View Images | Combining Views from Different Files into a Single Output
Nuke combines the inputs into a single output.
USER GUIDE
976
Displaying Views in the Viewer | To Display a Particular View
Displaying Views in the Viewer

You can only display the views that exist in your project settings. To see a list of these views, or to add
or delete views, select Edit > Project Settings and go to the Views tab. For more information, see
Setting Up Views for the Script.
To Display a Particular View

1. Add a Viewer into your script if you haven’t already done so.
2. On top of the Viewer controls, do one of the following:
• If you have checked View selection uses buttons? in the project settings, click the button of the
view you want to display. For example, click the right button (assuming you have a view called
right in your script).
• If you haven’t checked View selection uses buttons? in the project settings, select the view you
want to display from the dropdown menu.
Note: Nuke lists the views in .exr files in the order they appear in the clip's header, so a
view named 'left' may not always be the first view displayed above the Viewer.
If your views do not appear in the correct order, you can rearrange them in the Project
Settings > Views tab. See Setting Up Views for the Script for more information.
Tip: You can also press the ; (semicolon) and ’ (forward single quote) keys to move between
different views in the Viewer.
USER GUIDE
977
Displaying Views in the Viewer | To Display Two Views Next to Each Other
To Display Two Views Next to Each Other

2. If necessary, combine your views into a single output using the JoinViews node. For more
information on how to do this, see Loading Multi-View Images.
3. Select Views > Stereo > SideBySide to insert a SideBySide node in an appropriate place in your
script.
4. In the SideBySide node’s controls, select the two views you want to display from the view1 and
view2 dropdown menus. View1 is displayed on the left and view2 on the right.
5. If you want to display one view on top of another rather than next to it, check vertical. View1 is
displayed above view2.
6. If you want to swap the views around in the Viewer, click the swap button.
The Viewer displays the two selected views simultaneously, so you can easily compare them.
To Display a Blend Between Two Views

2. If necessary, combine your views into a single output using the JoinViews node. For more
information on how to do this, see Loading Multi-View Images.
3. Select Views > Stereo > MixViews to insert a MixViews node into your script. This node displays a
blend between two views in the Viewer, allowing you to check how elements in these views are
aligned.
4. In the MixViews controls, use the views buttons or dropdown menus to select the two views to
blend between.
USER GUIDE
978
Displaying Views in the Viewer | To Display a Blend Between Two Views
5. To control the blend between the views, adjust the mix slider. Setting the slider to 0 or 1 displays
only one of the views. Values between 0 and 1 produce different blends between the views.
USER GUIDE
979
Selecting Which Views Display Changes | Splitting Views Off
Selecting Which Views Display

Changes
By default, Nuke applies any changes you make to all views of the processed node. To apply changes
to a particular view only (for example, the left view but not the right), you must first do one of the
following:
• In the case of most nodes, split the view off in the node’s controls.
• In the case of RotoPaint nodes, select the view you want to process from the view dropdown menu
in the node’s controls.
These methods are useful, for example, when you want to perform the same operation on both views
but use different values for each.
Splitting Views Off

Nuke allows you to split views off in order to apply changes to the existing views separately.
1. Insert a process node (for example, ColorCorrect) in the appropriate place in your script.
2. If you haven’t already done so, attach a Viewer to the node. From the Viewer’s controls, select the
view you want to make changes to.
3. Open the node’s controls.
4. Click the view button next to the control you want to adjust. From the menu that opens,
select Split off [view name]. For example, to apply changes to a view called left, select Split off
left. You can also split all the node’s controls by selecting Split all knobs from the right-click
menu.
USER GUIDE
980
Selecting Which Views Display Changes | To Show Separate Values for Each View
An eye appears on the view button and the node gets a small green dot on it in the Node
Graph to indicate that views have been split off.
If you have assigned colors to the views and checked Use colors in UI? in your project settings,
dots also appear on the node to indicate which views have been split off. For example, if you are
using red for the left view and split off that view, a red dot appears on the node.
Any changes you now make using the control in question are only applied to the view you chose to
split off. Changes to controls that have not been split off are still applied to all views.
To Show Separate Values for Each View

Once you have split off a view, you can apply changes to the existing views separately. Simply click on
the small arrow on the left side of a control you have split off. This divides the control so that you can
see the values for each view.
Adjusting the split control for only the current view and for
all views separately.
USER GUIDE
981
Selecting Which Views Display Changes | To Unsplit Views
To Unsplit Views
1. In the node’s controls, click the view button .
2. From the menu that opens, select Unsplit [view]. For example, to unsplit a view called left, you’d
select Unsplit left.
3. Repeat step 2 for all views you want to unsplit.
Selecting the View to Process When Using the RotoPaint

Node
To select the view to process:
1. Open the RotoPaint node’s controls.
2. From the view dropdown menu, select the view you want to process. To apply changes to all
views at the same time, select all the views separately.
3. If you selected to process just one view, make sure you are viewing the selected view in the Viewer
when making your changes.
USER GUIDE
982
Performing Different Actions on Different Views | To Extract a View for Processing
Performing Different Actions on

Different Views
In case you need to perform totally different actions on the two views, you can add a OneView node
to separate one view for processing.
To Extract a View for Processing

1. Select Views > OneView to insert a OneView node in an appropriate place in your script.
2. In the OneView node’s controls, select the view you want to make changes to from the view
dropdown menu.
All views are extracted, and any changes you make are only applied to the view you selected
(regardless of which view you are displaying in the Viewer).
To make changes to a different view, select it from the OneView node’s view dropdown menu.
To merge views from two separate streams, select Views > JoinViews to combine the views (or delete
the OneView node from your script).
If you need to extract all views, process them individually, and then merge them together, use the
Split and Join menu item. This menu item is actually a combination of the OneView and JoinViews
nodes. It first extracts all the views you have set up in your project settings and then merges them
back together. It’s no different to using several OneView nodes together with a JoinViews node, but
speeds up work, because you don’t need to add each node in a separate go. To use the menu item,
select Views > Split and Join.
For example, if you have created views called left and right in your project settings and use a Split
and Join menu item after your Read node, you get the following node tree:
USER GUIDE
983
Performing Different Actions on Different Views | To Extract a View for Processing
You can then add any necessary nodes, such as color corrections, between the OneView and
JoinViews nodes.
USER GUIDE
984
Reproducing Changes Made to One View | Reproducing X and Y Values
Reproducing Changes Made to

One View
When rotoscoping, creating paint effects, or doing other operations dependent on image locality, you
can have changes made to one view automatically reproduced in the other. This applies to the
RotoPaint node, Roto node and any nodes, groups, or gizmos that have controls for x and y
coordinates.
To reproduce changes made with the above nodes, groups, or gizmos, you need a disparity field that
maps the location of a pixel in one view to the location of its corresponding pixel in the other view.
You can create a disparity field using the O_DisparityGenerator plug-in, which is included in the Ocula
plug-in set, or a 3D application. Once you have the disparity field, you can store it in the channels of
an .exr file or use the ShuffleCopy node to add the disparity channels in the data stream where you
need them.
For more information on reproducing paint strokes, Bezier shapes, or B-spline shapes, see
Reproducing Strokes/Shapes in Other Views.
Reproducing X and Y Values

Whenever there are values in any x or y control in Nuke for one view, you can automatically generate
the corresponding values for the other view. This is true for both nodes and gizmos. For example,
you can use a Tracker node to track something in one view, and then have the track’s x and y position
generated for the other view automatically.
1. Make sure there is a disparity field upstream from the image sequence you are manipulating. If
the image sequence is an .exr file, the disparity field can be included in its channels. Otherwise,
you can use a ShuffleCopy node or Ocula’s O_DisparityGenerator plug-in to add it in the data
stream.
2. Insert a node that has an x and y control after the image sequence you are manipulating.
3. Attach a Viewer to the node you added in the previous step, and make your changes in one view.
4. From the View menu next to the x and y controls, select Correlate [view] from [view] using
disparity, for example Correlate right from left using disparity. This generates the
corresponding x and y values for the other view.
USER GUIDE
985
Reproducing Changes Made to One View | Reproducing X and Y Values
If you have Foundry’s Ocula plug-ins installed, you can also select Correlate [view] from [view]
with Ocula. This way, extra refinements are made when creating the corresponding x and y
values, and the results may be more accurate.
5. If you want to adjust the x and y values further, you need to adjust both views independently.
Adjustments you make to one view are not automatically generated for the other.
USER GUIDE
986
Swapping Views | To Rearrange Views
Swapping Views
You can rearrange the views in your script using the ShuffleViews node. For example, you can swap
the left and right views around in the pipeline, so that Nuke uses the left input for the right eye and
vice versa.
To Rearrange Views
1. Select Views > ShuffleViews to insert a ShuffleViews node in an appropriate place in your script.
2. In the ShuffleViews controls, click add as necessary.
3. Use the buttons or dropdown menus to select which view to replace with which. For example, to
swap the left and right views around, you need to make the following selections:
• On one row, select left under get, and right under from (“get left from right”). The left view is
now replaced with the right view.
• On another row, select right under get, and left under from (“get right from left”).
The right view is replaced with the left view.
If there aren’t enough rows of buttons or dropdown menus on the ShuffleViews node’s properties
panel, click the add button to add a row.
To remove unnecessary rows in the ShuffleViews node’s controls, click the delete button next to
the row you want to remove.
USER GUIDE
987
Converting Images into Anaglyph | To Convert Your Images into Anaglyph
Converting Images into Anaglyph

You can use the Anaglyph node to convert your inputs into anaglyph images, which produce a 3D
effect when viewed with 2-color anaglyph glasses.
To Convert Your Images into Anaglyph

1. Select Views > Stereo > Anaglyph to insert an Anaglyph node in an appropriate place in your
script.
2. Use the views controls in the Anaglyph properties panel to select which views you want to use for
the left and the right eye.
Nuke converts the input images into grayscale anaglyph images. The left input is filtered to
remove blue and green, and the right view to remove red.
3. To add color into the images, drag right on the amtcolor slider, or insert a value between 0
(grayscale) and 1 (colored) into the amtcolor input field.
If the images include areas that are very red, green, or blue, adding more color into them may not
produce the best possible results.
USER GUIDE
988
Converting Images into Anaglyph | To Convert Your Images into Anaglyph
4. To invert the colors and use the red channel from the right input and the blue and green channels
from the left, check the (right=red) box.
5. To control where the images appear in relation to the screen when viewed with anaglyph glasses,
enter a value in the horizontal offset input field. To have the images appear in front of the
screen, you would usually enter a negative value. To have the images appear further away, you
would usually enter a positive value. (This is not the case if you have swapped the left and right
views around.)
Tip: If you like, you can register the Anaglyph node as a Viewer Process. This way, you always
have it as a viewing option in the Viewer’s Viewer Process dropdown menu and can apply it
to the current Viewer without having to insert the node in the Node Graph. Do the following:
2. To register the Anaglyph node as a Viewer Process, save the following in your menu.py:
nuke.ViewerProcess.register("Anaglyph", nuke.createNode, ("Anaglyph",
""))
3. Restart Nuke.
4. To apply the Anaglyph Viewer Process, select it from the Viewer Process dropdown menu
in the Viewer controls.
5. To adjust the Anaglyph Viewer Process controls, select show panel from the Viewer
Process dropdown menu.
For more information on Viewer Processes, see Using the Viewer Controls.
USER GUIDE
989
Changing Convergence | To Convert Your Images into Anaglyph
Changing Convergence
The ReConverge node lets you shift convergence (the inward rotation of the eyes or cameras) so that
any selected point in the image appears at screen depth when viewed with 3D glasses. This point is
called the convergence point. It is the point where the lines of sight from the two cameras meet.
Changing convergence moves the point where the lines of sight from
the two cameras meet.
At the convergence point, the different views in the image are aligned and appear at screen depth
when viewed with 3D glasses. Anything behind the convergence point appears behind the screen,
while anything in front of it seems to pop out of the screen. This is illustrated in the figure below.
Convergence controls where elements in the image appear

in relation to the screen when viewed with 3D glasses.
USER GUIDE
990
Changing Convergence | To Change the Convergence Point of a Stereo Image
Changing convergence changes the perceived depth of the images. It moves all the elements in the
image backwards or forwards a fixed distance while keeping the distance between them the same.
This is illustrated in the figure below, where the gray rectangles represent elements depicted in a
stereo image.
Changing convergence changes the perceived depth of the images.
Often, the element of an image that appears closest to the audience is used as the convergence point.
However, to make an element in your image jump out of the screen, you need to converge on
something behind this element.
To calculate the convergence shift, the ReConverge node needs a disparity field that maps the
location of a pixel in one view to the location of its corresponding pixel in the other view. To create
the disparity field, you can use the O_DisparityGenerator plug-in, which is part of the Ocula plug-in
set. Alternatively, you can create the disparity field in a 3D application. Once you have the disparity
field, you can store it in the channels of an .exr file or use the ShuffleCopy node to add the disparity
channels in the data stream where you need them.
Note that the ReConverge node only shifts views horizontally, not vertically.
To Change the Convergence Point of a Stereo Image

1. Make sure there is a disparity field upstream from the image sequence whose convergence you
want to change. If the image sequence is an .exr file, the disparity field can be included in its
channels. Otherwise, you can use a ShuffleCopy node or Ocula’s O_DisparityGenerator plug-in to
add it in the data stream.
2. From the Toolbar, select Views > Stereo > ReConverge to insert a ReConverge node after the
image sequence whose convergence you want to adjust.
3. Attach a Viewer to the ReConverge node.
4. To better view the effect of the ReConverge node, insert an Anaglyph node (Views > Stereo >
Anaglyph) between the ReConverge node and the Viewer.
USER GUIDE
991
Changing Convergence | Using the Same Convergence Point Throughout a Sequence
5. Make sure the ReConverge properties panel is open. You should see the convergence point
overlay in the Viewer. Drag the point on top of the point you want to appear at screen level when
viewed with 3D glasses. The convergence shifts to this location.
You can also move the convergence point by entering the point’s x and y coordinates in the
Convergence upon fields.
6. By default, the ReConverge node moves the right view to achieve the convergence shift. However,
if you like, you can use the Mode dropdown menu in the ReConverge controls to move the left
view instead (select shift left) or move both views equally (select shift both).
7. If necessary, adjust the offset for convergence (in pixels) in the ReConverge controls. To bring all
elements of your image forward from the screen level, enter a positive value in the Convergence
offset field. To move all elements further away, enter a negative value.
It is also possible to use the same element as the convergence point throughout the image sequence.
You can, for example, have the same actor always appear at screen depth. To converge on the same
element throughout the sequence, link the ReConverge node with a Tracker node.
Using the Same Convergence Point Throughout a

Sequence
1. Insert a Tracker node after the image sequence whose convergence you want to adjust.
USER GUIDE
992
Changing Convergence | Using the Same Convergence Point Throughout a Sequence
2. Track the point that you want to appear at screen level throughout the sequence. For more
information on how to use the Tracker node, refer to Tracking and Stabilizing .
3. When you have the track animation data, apply it to the ReConverge node’s Converge upon
control via a linking expression. The easiest way to do this is to Ctrl/Cmd+drag the animation
button next to the tracks list to the animation button next to the Converge upon control.
USER GUIDE
993
Previewing Stereoscopic Images | Previewing Using the Viewer Stereo Modes
Previewing Stereoscopic Images

You can preview stereo images using the Viewer stereo modes or by flipbooking the sequence using
Nuke's default flipbook.
Previewing Using the Viewer Stereo Modes

The Viewer stereo modes allow you to see both views at once, in either anaglyph or interlaced per
scanline. To enable of Viewer stereo mode, do the following:
1. Right-click in the Viewer to display the context-sensitive menu.
2. Navigate to Stereo Modes and select the required mode:
• Single - the default mode, shows only the view selected using the buttons above the Viewer.
• Anaglyph - produces a 3D effect when viewed with 2-color anaglyph glasses.
• OpenGL Stereo - displays both views at once on a 3D monitor for review purposes. See
Displaying OpenGL Stereo in Comp Viewers for more information.
• Interlace - shows both views, but alternately per scanline.
Anaglyph Stereo Mode Interlaced Stereo Mode
Displaying OpenGL Stereo in Comp Viewers

The Viewer OpenGL Stereo mode allows you to see both views at once on a 3D monitor for review
purposes. OpenGL Stereo is only supported on NVIDIA Quadro series GPUs and AMD Radeon Pro
series GPUs on Windows and Linux OS.
Note: OpenGL Stereo mode is not supported on Mac due to limitations in macOS and not
supported with AMD GPUs on Linux.
USER GUIDE
994
Previewing Stereoscopic Images | Displaying OpenGL Stereo in Comp Viewers
Enabling OpenGL Stereo Output
Windows
To enable NVIDIA GPU stereo output:
1. Right-click on the desktop and select NVIDIA Control Panel.
2. Navigate to 3D Settings > Manage 3D Settings > Stereo - Enable.
3. Proceed to Switching to OpenGL Stereo Output.
To enable AMD GPU stereo output:

1. Double-click the AMD taskbar icon and select Advanced Settings.
2. Navigate to the AMD Pro Settings and check Enable Quad Buffer Stereo.
USER GUIDE
995
Previewing Stereoscopic Images | Displaying OpenGL Stereo in Comp Viewers
3. Select either Auto-Stereo (Horizontal Interleaved) or Auto-Stereo (Vertical Interleaved) and

click Apply.
Linux
1. Open a command prompt and enter:
nvidia-xconfig --stereo=3
Tip: For more information on the nvidia-xconfig utility, please see the man page: man
nvidia-xconfig
Note: OpenGL Stereo mode is not supported with AMD GPUs on Linux.
Switching to OpenGL Stereo Output

2. Navigate to Stereo Modes and select OpenGL Stereo.
If you select OpenGL Stereo mode before enabling your GPU settings, a warning is displayed.
USER GUIDE
996
Previewing Stereoscopic Images | Flipbooking Stereo Sequences
The first time you select OpenGL Stereo, a warning message is displayed.
Tip: You can disable the warning by enabling Don't show again and clicking OK.
OpenGL Stereo is displayed in the Comp Viewer.
Flipbooking Stereo Sequences

To flipbook stereo images, do the following:
1. Select the node that you want to flipbook.
2. Select Render > Flipbook Selected from the main menu bar (or press Alt+F).
You can also press the Flipbook this Viewer button at the bottom-right of the Viewer. This
flipbooks the nodes that are connected to the Viewer.
A dialog opens.
USER GUIDE
997
Previewing Stereoscopic Images | Flipbooking Stereo Sequences
3. Check that your settings are correct in the dialog. The settings in the dialog are almost identical to
the settings when you flipbook non-stereo sequences. For more information about these settings,
see Flipbooking Sequences.
4. In the frame range field, enter the frame range you want to preview (for example, 1-35 or 1-8 10
12-15).
5. Using the View control, select the view you want to flipbook, either left or right.
6. Click OK.
Nuke renders the selected view as a temporary sequence using the frame range and resolution
defined in the script’s settings. This may take a few moments.
7. After the render is complete, Nuke launches Flipbook Viewer and loads in the temporary
sequence.
You can play it back and view it using Flipbook Viewer controls.
USER GUIDE
998
Rendering Stereoscopic Images | Rendering EXR Files
Rendering Stereoscopic Images

You can render several views using a single Write node. When using the stereo extensions for the .exr
file format, Nuke writes the output of both views into a single file. With any other file types, the views
are written into their respective files.
Rendering EXR Files

To render .exr files:
1. Select Image > Write to insert a Write node in an appropriate place in your script.
2. In the Write node’s controls, select exr from the file type dropdown menu.
3. From the views dropdown menu, select the view(s) you want to render, for example left, right.
4. Adjust any other Write controls as necessary and click Render. Nuke prompts you for the frames
to render.
Nuke writes the selected views into a single file.
Rendering Other File Formats

To render files that are not in the .exr file format:
1. Select Image > Write to insert a Write node in an appropriate place in your script.
2. In the Write nodes’ controls, select the file type of your images from the file type dropdown
menu.
3. When entering names for the rendered image sequences, you can use the variable %V (with a
capital V) to represent the words left and right (or any other full view names) in the file names,
for example filename.%V.####.exr. To represent the letters l and r (or the first letters of any
views), use the variable %v (with a lower-case v) instead. When rendering, Nuke then fills this in
with left, right, l, or r, and renders all views you specify in the next step.
4. Adjust any other Write controls as necessary and click Render. Nuke prompts you for the frames
to render as well as the views to execute (assuming you have set up several views in the project
settings).
Nuke renders several views, but writes them into separate files. If you did not specify a view in the file
names (using either the name of the view, its first letter, or a variable), you can only render one view.
USER GUIDE
999
Rendering Stereoscopic Images | Rendering Other File Formats
Note: For command line renders, you can pass the -view argument with a list of view
names to render, separated by a comma. If you do not specify a -view argument, Nuke
renders all views.
USER GUIDE
1000
Deep Compositing
Deep compositing is a way of compositing digital images using data in a different format to standard
"flat" compositing. As the name suggests, deep compositing uses additional depth data. This reduces
need for re-rendering, produces high image quality, and helps you solve problems with artifacts
around the edges of objects.
About Deep Compositing

A standard 2D image contains a single value for each channel of each pixel. In contrast, deep images
contain multiple samples per pixel at varying depths and each sample contains per-pixel information
such as color, opacity, and camera-relative depth.
For example, creating holdouts of objects that have moved in the scene has previously required re-
rendering the background, and problems frequently occurred with transparent pixels and anti-
aliasing. With Nuke’s Deep compositing node set, you can render the background once, and later
move your objects to different places and depths, without having to re-render the background. Any
transparent pixels, with motion blur for example, are also represented without flaw, so working with
deep compositing is not only faster, but you also get higher image quality.
Deep composite with ball objects among blue buildings.
Quick Start
With Nuke’s deep compositing node set, you can:
USER GUIDE
1001
Deep Compositing |
• Read in your deep image with the DeepRead node. See Reading in Deep Footage.
• Merge deep data with the DeepMerge, see Merging Deep Images.
• Generate holdout mattes from a pair of deep images using the DeepHoldout node. See Creating
Holdouts.
• Flatten deep images to regular 2D or create point clouds out of them. See Creating Deep Data.
• Sample information at a given pixel using the DeepSample node. See Sampling Deep Images.
• Crop, reformat, and transform deep images much in the same way as you would a regular image,
using the DeepCrop, DeepReformat and DeepTransform nodes. See Cropping, Reformatting, and
Transforming Deep Images.
• Create deep images using the DeepFromFrames, DeepFromImage, DeepRecolor, and
ScanlineRender nodes. See Creating Deep Data.
USER GUIDE
1002
Reading in Deep Footage |
Reading in Deep Footage

You read in deep images to Nuke with a DeepRead node, which is rather like reading in any other
images with the Read node. Nuke allows you to import deep images in two formats:
• DTEX (generated from Pixar’s PhotoRealistic RenderMan® Pro Server).
• Scanline OpenEXR 2.2, or above (tiled OpenEXR files are not supported).
USER GUIDE
1003
Importing DTEX Files | Importing Scanline OpenEXR Files
Importing DTEX Files

Before importing DTEX files, you need to set up Pixar's RenderMan Pro Server 20, or earlier, on your
computer.
Do the following:
1. Install RenderMan Pro Server on your computer, and set up the necessary environment variables
that enable Nuke to work with it. For details, have a look at Setting Up RenderMan Pro Server and
PrmanRender. Note that you don’t need a RenderMan Pro Server license to work with deep
images in Nuke, just installing the software is enough.
2. Create a DeepRead by clicking Deep > DeepRead.
3. Navigate to your .dtex image, and click Open. For more information about the Read node
controls, see the Managing Scripts chapter.
4. By default, Nuke tries to automatically detect the .dtex file type by looking at the subimage name.
If the name is either Deep Shadow or ends with (or is) .deepopacity, Nuke treats the file as a
deep opacity file. However, if you have manually changed the subimage name when rendering the
file, Nuke may not be able to detect the file type correctly. If this is the case, set the type
dropdown menu to one of the following:
• deepopacity - This forces Nuke to treat the file as an accumulated deep opacity file,
corresponding to a RenderMan Display Driver configuration of:
Display "Filename.dtex" "deepshad" "deepopacity"
• alpha - This forces Nuke to treat the file as the newer point-sampled alpha or color,
corresponding to a RenderMan Display Driver configuration of either:
Display "Filename.dtex" "deepshad" "a" or
Display "Filename.dtex" "deepshad" "rgba"
Importing Scanline OpenEXR Files

To import scanline OpenEXR files:
1. Create a DeepRead by clicking Deep > DeepRead.
2. Navigate to your deep .exr image, and click Open.For more information about the Read node
controls, see the Managing Scripts chapter.
3. By default, the exr prefix is attached to metadata keys to make them distinct from other
metadata in the tree. If you’d rather read metadata in "as is" without attaching a prefix, enable do
not attach prefix.
USER GUIDE
1004
Creating Deep Data | Converting a 2D Image Sequence to a Deep Frame Using Input Frames
Creating Deep Data

You can create deep data in Nuke by:
• sampling a regular 2D image sequence at multiple frames to create several samples for each pixel
in a single deep frame. See Converting a 2D Image Sequence to a Deep Frame Using Input Frames.
• converting a regular 2D image to a deep image with a single sample for each pixel at the depth
defined by the depth.Z channel. See Converting a 2D Image to a Deep Image.
• recoloring depth samples using a regular 2D color image. See Recoloring Depth Data.
• adding a ScanlineRender node to a 3D scene and connecting a Deep node downstream. See Using
ScanlineRender to Generate Deep Data.
Converting a 2D Image Sequence to a Deep Frame Using

Input Frames
You can use the DeepFromFrames node to create depth samples from input frames.
1. Connect the DeepFromFrames node to your footage. The deep image is created by placing each
frame at increasing depths.
2. To adjust the results, use the controls in the properties panel:
• samples - the number of samples to create per pixel in the output deep image.
• frame range - the range of frames to use for one deep image. For example, with the default
samples value (5) and frame range value (1-9) DeepFromFrames sample at times 1, 3, 5, 7 and 9.
• premult - check to premultiply the samples.
• split alpha mode - select additive to perform a straight division by the number of samples or
multiplicative to split the alpha so that it can be returned to its original value if flattened later
on (using the DeepToImage node, for example). If you select additive, the alpha can’t be
returned to its original value.
• zmin - the depth to assign to the first sample of each deep pixel output, corresponding to the
first frame in the range.
• zmax - the depth to assign to the last sample of each deep pixel output, corresponding to the
last frame in the range.
USER GUIDE
1005
Creating Deep Data | Converting a 2D Image to a Deep Image
A simple setup for creating a deep fog element.
Converting a 2D Image to a Deep Image

Using the DeepFromImage node, you can convert a regular 2D image to a deep image with a single
sample for each pixel at the depth defined by the depth.Z channel.
1. Connect DeepFromImage to the footage you want to convert to a deep image.
2. Use the premult input box in the properties panel to select whether you want the input channels
to be premultiplied or not.
3. Uncheck the keep zero alpha box if you want to drop any samples with a zero alpha value from
the resulting deep image. By default, the box is checked and the resulting deep image contains
the zero alpha samples.
4. You can also specify the depth using the z control in the properties panel. In that case, check the
specify z box to indicate you don’t want to use a depth channel from the input.
Recoloring Depth Data

Use the DeepRecolor node to merge deep buffer files that only contain opacity for each sample with a
standard 2D color image. DeepRecolor spreads the color at each pixel of the input 2D image across
all the samples of the corresponding pixel in the deep input.
1. Connect your deep source to the depth input of the DeepRecolor node, and your 2D image to the
color input. You might want to add an unpremultiply node between your color input and the
DeepRecolor if your 2D image is premultiplied.
2. In the properties panel, you can select which channels you want to use from the color input
image.
USER GUIDE
1006
Creating Deep Data | Using ScanlineRender to Generate Deep Data
In the example below, DeepRecolor takes an unpremultiplied .exr image and uses it to color the
.dtex file’s deep samples.
3. If at this point the alpha from the final high-quality flat render doesn’t match the alpha
represented by the deep samples (for example, as a result of the compression that usually
happens to deep files on disk or some change to the shader), you can check target input alpha.
This means the color input’s alpha is distributed amongst the deep samples, so that the final
resulting alpha after flattening of the deep data matches the color input’s alpha.
If you leave target input alpha unchecked, Nuke distributes the color to each sample by
unpremultiplying by the color image’s alpha and then remultiplying by the alpha of each sample.
In this case, the alpha from DeepRecolor may not match the alpha from its color input.
Using ScanlineRender to Generate Deep Data

The ScanlineRender node outputs deep data if there is a Deep node downstream.
1. Create a 3D scene and attach a ScanlineRender node to it to render the scene as a 2D image.
Note: Deep compositing only supports the over blend mode. As a result, if there is a
BlendMat node in the 3D scene, its operation always appears to be set to over when
converted to Deep.
2. Add a node from the Deep menu downstream from ScanlineRender.
USER GUIDE
1007
3. If you don't want deep samples with an alpha value of 0 to contribute to the output, open the
ScanlineRender properties and make sure that drop zero alpha samples is enabled.
4. Adjust the rest of the ScanlineRender properties as usual. For example:
• If you see any aliasing artifacts in the render, go to the MultiSample tab and increase samples.
This increases the number of deep samples per pixel.
Alternatively, you can set antialiasing to low, medium, or high on the ScanlineRender tab.
Samples set to a low value. Samples set to a high value.

• If you want to add motion blur to your 3D scene, increase the samples value to sample the
image multiple times over the shutter time.
In the shutter field, enter the number of frames the shutter stays open when motion blurring. If
rendering becomes very slow, you can approximate multi-sample rendering and reduce render
times by increasing stochastic samples.
USER GUIDE
1008
Without motion blur. With motion blur.

For more information on the ScanlineRender properties, see Rendering a 3D Scene.
Tip: You can use a DeepToPoints node after ScanlineRender to create a 3D point cloud that
represents the motion in the scene. For more information on DeepToPoints, see Creating 2D
and 3D Elements from Deep Images.
USER GUIDE
1009
Merging Deep Images | Using ScanlineRender to Generate Deep Data
Merging Deep Images

Use the DeepMerge node to merge the samples from multiple deep images, so that each output pixel
contains all the samples from the same pixel of each input.
1. Connect the data you want to merge to the DeepMerge node’s numbered inputs.
2. In the DeepMerge properties, make sure operation is set to combine.
3. You can check the drop hidden samples box in the properties panel to not include samples that
are completely occluded by nearer samples whose alpha value is one.
Merging two DeepRecolor results.

4. You can filter out samples using the drop zero threshold control. Increasing the value removes
more samples with very small alpha values, such as those caused by floating point inaccuracy.
5. The metadata from control allows you to control which input's metadata is passed down the
node tree.
6. Enable compute occluded samples if you want Nuke to calculate occlusion using the values of
the holdout samples in front of samples from main. This is a more accurate representation of
occlusion at depth, but can take longer to process.
For example:
M = main sample
USER GUIDE
1010
Merging Deep Images | Using ScanlineRender to Generate Deep Data
H = holdout sample
M0 remains unchanged since there are no holdout samples before it. M2 is affected by the
combined H0, H1, and H2 holdout samples and M4 is affected by all holdout samples.
USER GUIDE
1011
Creating Holdouts | Creating a Flattened Holdout with the DeepHoldout Node
Creating Holdouts
You can create holdouts using either the DeepHoldout or the DeepMerge nodes. The primary
difference is whether you want the node to produce a flattened output image or a deep output image
after the holdout.
Creating a Flattened Holdout with the DeepHoldout Node

The DeepHoldout node removes or fades out samples in the main input that are occluded by
samples in the holdout input. The result is a flattened image of your main input image, with holes
where objects in the holdout input image have occluded parts of it. If you need a deep output image,
then using the holdout operation in the DeepMerge node may be a preferable option.
To create a holdout using the DeepHoldout node:

1. Connect the deep image you want to remove or fade parts from to the main input.
2. Connect the deep image with the occluding parts to the holdout input.
3. You can now view the result, which is a holdout with red, green, blue and alpha channels. Note
that the output image is flattened.
A holdout of blue buildings with ball

shapes held out.
4. Enable compute occluded samples if you want Nuke calculate occlusion using the values of the
holdout samples in front of samples from main. This is a more accurate representation of
For example:
USER GUIDE
1012
Creating Holdouts | Creating a Deep Holdout with the DeepMerge Node
M = main sample
H = holdout sample
more samples with very small alpha values such as those caused by floating point inaccuracy.
Creating a Deep Holdout with the DeepMerge Node

Like the DeepHoldout node, the holdout operation of the DeepMerge node removes or fades out
samples in input B that are occluded by samples in input A. However, unlike the DeepHoldout node,
DeepMerge doesn’t flatten the data but produces a deep output image.
To create a holdout using the DeepMerge node:

1. Connect the deep image you want to remove or fade parts from to input B.
2. Connect the deep image with the occluding parts to input A.
3. In the DeepMerge properties, set operation to holdout.
more samples with very small alpha values, such as those caused by floating point inaccuracy.
5. The metadata from control allows you to control which input's metadata is passed down the
node tree.
6. Enable compute occluded samples if you want Nuke to calculate occlusion using the values of
the holdout samples in front of samples from main. This is a more accurate representation of
For example:
M = main sample
H = holdout sample
USER GUIDE
1013
Creating Holdouts | Creating a Deep Holdout with the DeepMerge Node
You can now view the result, which is a holdout with red, green, blue, and alpha channels. Note
that the output image is still a deep image.
USER GUIDE
1014
Creating 2D and 3D Elements from Deep Images | Creating a 2D Image from a Deep Image
Creating 2D and 3D Elements

from Deep Images
You can create a 2D image or a 3D point cloud from a deep image.
Creating a 2D Image from a Deep Image

You can use the DeepToImage node to flatten an image, in other words merge all the samples in a
deep image into a regular 2D image.
1. Connect the node to a deep image (or a DeepMerge with merged deep data) you want to flatten.
2. In the properties panel, the volumetric composition box is checked by default, but if you
uncheck it, Nuke only calculates the front depth of each sample and assumes the samples do not
overlap. If you uncheck this, the calculation takes less time, but if you have overlapping samples
in your deep image, the resulting image might not represent every pixel as expected.
Creating a Point Cloud from a Deep Image

You can use the DeepToPoints node to transform the deep pixel samples into points in 3D space that
you can see in Nuke's 3D view, much like a point cloud. This node is useful for position reference.
1. Connect the DeepToPoints node’s deep input to the deep image you want to view in 3D. If you
have a camera that you want to look at the point cloud through, connect it to the camera input.
2. Change to 3D view (by pressing Tab) to view the results.
3. In the properties panel, you can use the Point detail slider to adjust the density of the cloud.
Point detail set to 0.005. Point detail set to 0.05.

4. Adjust Point size to change the size of the points. You can also use 3D object selection on the
point cloud, and for example snap on your DeepToPoints results. (For more information on 3D
selection, see 3D Selection Tools.)
USER GUIDE
1015
Creating 2D and 3D Elements from Deep Images | Creating a Point Cloud from a Deep Image
Point size set to 2. Point size set to 6.
USER GUIDE
1016
Modifying Deep Data | Color Correcting Deep Images
Modifying Deep Data

Nuke allows you to color correct deep images as well as modify them using expressions.
Color Correcting Deep Images

The DeepColorCorrect node applies the color correction to each sample at each pixel.
There are control sets for adjusting shadows, midtones and highlights, as well as a master set for
adjusting all these at once. You can use the lookup curves on the Ranges tab to control the range of
the image that is considered to be in the shadows, midtones, and highlights. For more information
about the basics of color correcting, see Working with Color.
Tip: Make mattes by setting the gain value for your alpha channel to 0 and setting the offset
value for the alpha channel to 1 in the range you want a matte for.
Adjusting the Effect of Deep Color Correction

On the Masking tab you can set the point among the deep samples where the effect of your color
correction starts and finishes.
1. Check the limit_z box to activate the zmap tool.
2. Adjust the trapezoid so that the A delimiter marks the depth where you want the color correction
to start, B and C mark the length of the full effect and delimiter D indicates where the effect stops.
Zmap tool’s y axis, therefore, indicates the amount of the effect, and the x axis is the range of
your depth samples.
3. Use the mix control to adjust the overall mixing between the color corrected result and the
original image. Zero value is the original image, and value 1 is the full color correction result.
Modifying Your Deep Images with Expressions

You can use the DeepExpression node to run Nuke expressions on deep data. Use the controls in the
properties panel:
1. There are four fields for temporary expressions, just like in the normal Expression node. These
can be useful if you need to use a long expression in several fields and want to assign that
USER GUIDE
1017
Modifying Deep Data | Modifying Your Deep Images with Expressions
expression temporarily to a variable. Enter the variable name to the left of the equals (=) sign, and
the expression to the right. You can then use the variable name to represent the entire expression
in the expression fields for the channels.
2. In the chans0 - chans3 dropdowns you can then specify which channels you want to create
expressions for. This adds or removes expression fields below.
3. You can then enter your expressions for the different channels in the channel expression fields.
For more information about expressions, have a look at Expressions.
USER GUIDE
1018
Cropping, Reformatting, and Transforming Deep Images | Cropping Deep Images
Cropping, Reformatting, and

Transforming Deep Images
You can crop, reformat and transform deep images much in the same way as you would a regular
image, using the corresponding deep nodes.
Note: Remember that since the samples at each pixel can be located at arbitrary depths,
resampling during transforming may produce unexpected results since there may not be
samples at the same depth in adjacent pixels.
Cropping Deep Images

You can use the DeepCrop node to clip your deep image, much like the normal Crop node:
1. Connect the DeepCrop node to the deep image you want to crop.
2. Adjust the crop box in the Viewer in X and Y directions to define your crop area. Alternatively,
define your crop area using the bbox fields in the properties panel. If you want to keep the depth
samples outside the crop box, you can check the keep outside bbox box.
3. Use the znear and zfar controls in the properties panel to crop samples in depth. If you don’t
want to use either of these controls, you can disable them by unchecking the use box next to
them. If you want to keep your depth samples outside of the z range defined by these controls,
you should check the keep outside zrange box.
Reformatting Deep Images

DeepReformat is the Reformat node for deep data. You can use it to set your deep image’s
dimensions, scale, and so on. To reformat your deep image:
1. Connect the DeepReformat node to the deep image you want to resize.
2. In the type dropdown, select:
• to format - sets the output width and height to the selected format. Select the format in the
output format dropdown. If the format does not yet exist, you can select new to create a new
format from scratch. The default setting, root.format, resizes the image to the format indicated
on the Project Settings dialog.
USER GUIDE
1019
Cropping, Reformatting, and Transforming Deep Images | Transforming Deep Samples
• to box - sets the output width and height to dimensions you define in pixels. Enter values in the
width, height and pixel aspect fields to specify the dimensions.
• scale - sets the output width and height to a multiple of the input size. Use the scale slider to
define the factor. The scale factor is rounded slightly, so that the output image is an integer
number of pixels in the direction selected under resize type.
3. You can specify what kind of resize you want in the resize type dropdown. Select:
• none - to not resize the original.
• width - to scale the original until its width matches the output width. Height is then scaled in
• height - to scale the original so that it fills the output height. Width is then scaled in such a
• fit - to scale the original so that its smallest side fills the output width or height. The longest side
• fill - to scale the original so that its longest side fills the output width or height. The smallest side
• distort - to scale the original so that both sides fill the output dimensions. This option does not
preserve the original aspect ratio, so distortions may occur.
4. Check the center box to define whether the input pixels should be resampled to the new size
or centered in the output. If you don’t center, the lower left corners of the input and output are
aligned.
5. To further adjust your image’s layout, you can check the respective boxes for:
• flip - to swap the top and bottom of the image.
• flop - to swap the left and right of the image.
• turn - to turn the image 90 degrees.
• black outside - to set pixels outside the format black.
• preserve bounding box - to preserve pixels outside the output format rather than clipping
them off.
Transforming Deep Samples

You can use the DeepTransform node to reposition the deep samples.
1. Connect the node to the deep footage you want to transform.
2. Use the translate x, y, and z controls to translate your samples.
3. Scale the samples’ z depth using the zscale control. Values above 1 increase the depth, whereas
values below 1 decrease it.
4. If you connect a mask to the node’s mask input, you can use it to regulate how much of an effect
the depth transformation has in different parts of the frame.
USER GUIDE
1020
Sampling Deep Images | Transforming Deep Samples
Sampling Deep Images

You can use the DeepSample node to sample any given pixel in a deep image. The Deep Sample node
gives you the depth data as figures.
1. Connect the DeepSample node to another Deep node.
2. Position the pos indicator over the pixels you want to sample in the Viewer.
3. View the deep sample information in the sample table on the DeepSample properties panel.
4. You can also toggle the accumulate box to select whether you want to see the individual sample
values of the sample pixel (unchecked), or the final composited value (checked).
USER GUIDE
1021
Writing Deep Data | Transforming Deep Samples
Writing Deep Data

You can write out deep images in the scanline OpenEXR 2.2, or above, format using the DeepWrite
node, which shares a lot of controls with the standard Write node. Do the following:
1. Select Deep > DeepWrite to insert a DeepWrite node into your script.
2. In the properties panel, click the file or proxy field’s folder icon and browse to the directory where
you want to store the deep image.
3. After the path, type a name for the deep image, including the .exr extension, and then click OK. If
you’re rendering an image sequence, include the frame number variable (for example, ####) in
the name.
4. Use the datatype dropdown menu to select the bit depth for the rendered file: 16 bit half or 32
bit float.
5. Set compression to the compression type to apply to the rendered file.
6. From the metadata dropdown menu, select what metadata is included with the rendered file:
• no metadata - No custom attributes are created, and only metadata that fills required header
fields is written out.
• default metadata - The optional time code, edge code, frame rate, and exposure header fields
are also filled using metadata values.
• default metadata and exr/*
• all metadata except input/*
• all metadata
7. By default, unknown metadata keys have the prefix nuke attached to them when they are written
into the file. If you’d rather have them written into the file "as is", without the prefix, check do not
attach prefix.
8. Adjust the rest of the controls as necessary. For more information on them, see Output (Write)
Nodes.
9. Click the Render button.
Nuke prompts for a frame range, defaulting to the range in the frame range fields.
10. If necessary, change the start and end frames, and then click OK.
Nuke writes the deep data to a scanline OpenEXR 2.2, or above, file (tiled OpenEXR files are not
supported).
USER GUIDE
1022
Working with File Metadata
The Read node's Metadata tab and the nodes in the Metadata menu of the Toolbar let you work
with information embedded in your images. This section gives instructions on their usage, teaching
you to view, compare, edit, and render metadata.
Metadata
Metadata is a set of information about an image embedded in the image file. This information may
include the image’s original bit depth, width, and height, for example. It can be attached to the file by
the camera used to shoot the images, and/or edited later.
When Nuke loads an image, it reads in the metadata embedded in the image. The metadata is then
passed down the node tree so you can view and use it at any point in your script. For example, you
can reference metadata via expressions. You can also edit or delete existing metadata, add new
metadata to a file, and write the resulting metadata out to files.
Note: Metadata for QuickTime files does not show gamma or bit depth.
Note: The Read and Write node's timecode and edge code fields have been removed from
the Properties panel, but the metadata is still available using the input/timecode and
input/edgecode keys. See Viewing Metadata for more information.
Tip: When using a Merge node, you can choose which input’s metadata to pass down the
tree. In the Merge controls, set metadata from to either A or B.
As well as the Metadata tab in the Read node, the MetaData menu of Nuke's Toolbar features five
nodes that help you work with file metadata:
• ViewMetaData lets you inspect the metadata passed down by the input node. See Viewing Metadata.
• CompareMetaData lets you compare the metadata between two inputs and view the differences. See
Comparing Metadata Between Inputs.
• ModifyMetaData lets you edit existing metadata in the input stream, add metadata into the stream,
and delete metadata from the stream. See Modifying Metadata.
USER GUIDE
1023
Working with File Metadata |
• CopyMetaData lets you copy metadata from one input to another and filter metadata to exclude
some of it. See Copying and Filtering Metadata Between Inputs.
• AddTimeCode lets you add a timecode to the metadata passed down by the input node. See Adding
a Time Code to Metadata.
USER GUIDE
1024
Viewing Metadata |
Viewing Metadata
The simplest way to view file metadata is by clicking the Metadata tab in the Properties panel of a
standard Read node. All the available metadata is displayed, along with a simple search function.
To filter the lists of metadata, use the search metadata for field. For example, if you enter f in the
search metadata for field, only the keys and values that include the letter f are displayed. By default,
the search is done within both keys and values. If you want to limit the search to either keys or values
only, set within to keys only or values only. For example, you can view metadata specific to the
input’s file format by entering the file format (for instance, dpx/) in the search metadata for field
and setting within to keys only.
Note: When observing the creation time (input/ctime) of an image, Windows generally
differs from Linux and Mac. This is due to the different way in which the operating systems
deal with file creation times.
Once you know which keys exist on the input, you can reference them in expressions. See Accessing
Metadata Using Tcl Expressions.
You can also view metadata using the ViewMetaData node:

1. Select MetaData > ViewMetaData to insert a ViewMetaData node after the node whose
metadata you want to inspect.
2. In the ViewMetaData properties, you can see a list of the metadata embedded in the image. This is
divided into keys and their values.
USER GUIDE
1025
Viewing Metadata | Comparing Metadata Between Inputs
Comparing Metadata Between Inputs

To compare metadata between two inputs:
1. From the Toolbar, select MetaData > CompareMetaData to add a CompareMetaData node after
the two nodes whose metadata you want to compare.
2. Connect the nodes you want to compare to the A and B inputs of the CompareMetaData node.
A list of keys where there is a difference between the two inputs is shown in the
CompareMetaData properties.
USER GUIDE
1026
Modifying Metadata | To Add Metadata
Modifying Metadata
There are several ways to modify metadata in Nuke.
To Add Metadata
1. Select MetaData > ModifyMetaData to insert a ModifyMetaData node after the node whose
metadata you want to add a new key to.
2. In the ModifyMetaData controls, click on the plus (+) button. A placeholder appears in the
metadata box.
3. Double-click on the placeholder under key.
The Pick metadata key dialog opens.

4. In the field at the bottom of the dialog, enter a name for the new key you want to add to the
metadata. Click OK.
5. Double-click on the placeholder under data and enter a value for the new key.
The new key and its value are added to the metadata being passed through.
To Edit Metadata
metadata you want to edit.
2. In the ModifyMetaData controls, click on the plus (+) button.
USER GUIDE
1027
Modifying Metadata | To Remove Metadata
A placeholder appears in the metadata box.

4. Pick the key whose name or value you want to edit and click OK.
The key is added to the ModifyMetaData properties.
5. In the ModifyMetaData properties, double-click on the key or its value and edit the information as
required.
To Remove Metadata
metadata you want to edit.
2. In the ModifyMetaData properties, click on the plus (+) button. A placeholder is added to the
metadata list.
3. Double-click on the placeholder under action and select remove from the menu that opens.

5. From the list of existing keys, select the key you want to remove and click OK.
USER GUIDE
1028
Modifying Metadata | To Edit the List of Actions in the ModifyMetaData Properties
The node now removes the selected key from the metadata, as you can see if you view the
metadata in the output.
To Edit the List of Actions in the ModifyMetaData

Properties
• To perform a new action, click on the plus (+) button.
• To cancel an existing action, select it from the list and click on the minus (-) button. Note that this
only affects the ModifyMetaData actions, and does NOT delete keys from the metadata embedded
in the input image.
• To move an item up in the list, select it and click on the up arrow button.
• To move an item down in the list, select it and click on the down arrow button.
USER GUIDE
1029
Copying and Filtering Metadata Between Inputs | To Edit the List of Actions in the ModifyMetaData
Copying and Filtering Metadata

Between Inputs
To copy metadata from one input to another and/or filter metadata:
1. Select MetaData > CopyMetaData to insert a CopyMetaData node into your script.
2. Connect:
• the Image input to the node whose image you want to pass down the tree.
• the Meta input to the node whose metadata you want to copy to the output.
3. Set metadata from to one of the following:
• Image+Meta - to add the metadata from the Meta input to the metadata from the Image input.
If the inputs share any common metadata keys, the values taken from the Meta input override
those taken from the Image input.
• Meta only - to only use the metadata from the Meta input.
• Meta+Image - to add the metadata from the Image input to the metadata from the Meta input.
If the inputs share any common metadata keys, the values taken from the Image input override
those taken from the Meta input.
• Image only - to only use the metadata from the Image input. This produces the same result as
not using a CopyMetaData at all: both the image and metadata are taken from the Image input.
However, this option can be useful if you want to filter the metadata to exclude some of it (see
the next step for how to do this).
4. To filter the metadata taken from the inputs, use the copy only fields under Meta filtering
and/or Image filtering. For example, if you enter f in the copy only field under Meta filtering,
only the keys and values that include the letter f are copied from the Meta input. By default, the
search is done within both keys and values. If you want to limit the search to either keys or values
only, set within to keys only or values only. For example, you can copy metadata specific to the
input’s file format by entering the file format (for instance, dpx/) in the copy only field and setting
within to keys only.
USER GUIDE
1030
Adding a Time Code to Metadata | To Edit the List of Actions in the ModifyMetaData Properties
Adding a Time Code to Metadata

1. Select MetaData > AddTimeCode to insert an AddTimeCode node into your node tree.
A time code is added to the metadata being passed through. By default, the time code is
01:00:00:00 on the first frame. It is updated throughout the frame range according to the input
clip’s playback speed, which in turn is controlled by the fps (frames per second) parameter in the
Project Settings. If you change the fps value in the Project Settings, the time code in the metadata
is updated to reflect the change.
Nuke can also deal with drop frames, such as when a clip’s frame rate is 29.97 or 59.94. Instead of
the default HH:MM:SS:FF time code format, use the format HH;MM;SS;FF, delimited by ;
(semicolon).
Note: Using semicolon delimiters with non-drop frame time codes displays an error in the
Viewer.
2. If you don’t want the time code on the start frame to be 01:00:00:00, enter a new time code in the
startcode field.
3. If you want to specify the playback speed manually rather than get it from the metadata and
project settings, uncheck get FPS from metadata and enter a new value in the fps field.
4. If you want to specify a different start frame than the first frame, check use start frame? and
enter a new value in the start frame field.
If you want to display the time code on the image, insert a Text node after the AddTimeCode node
and enter [timecode] in the message field. For more information on referencing metadata via
expressions, see Accessing Metadata Using Tcl Expressions.
USER GUIDE
1031
Rendering Metadata | To Edit the List of Actions in the ModifyMetaData Properties
Rendering Metadata
When rendering with the Write node, Nuke lets you write out metadata into the following file formats:
.exr, .cin, .dpx, and .jpg. You cannot write out metadata into any other formats.
When rendering metadata into an .exr file, you can use the metadata dropdown menu in the Write
node controls to specify what to write out:
• no metadata - Do not write out any metadata, except for metadata that fills required header fields
(for example, file name and bbox).
• default metadata - Write out the time code, edge code, frame rate, and exposure.
• default metadata and exr/* - Write out the time code, edge code, frame rate, exposure, and
anything in exr/.
• all metadata except input/* - Write out all metadata, except anything in input/.
• all metadata - Write out all metadata.
When rendering any other file format than .exr, Nuke writes whatever metadata the file format
header is known to support. Therefore, what is written out varies according to the file format.
USER GUIDE
1032
Accessing Metadata Using Tcl Expressions | To Edit the List of Actions in the ModifyMetaData
Accessing Metadata Using Tcl

Expressions
You can access metadata via Tcl expressions in the following ways:
• To get a list of all keys in the incoming metadata, use the expression [metadata]. For example, if
you add a Text node after an image and enter [metadata] in the message field, a list of all the keys
in the incoming metadata appears on the image. The values of the keys are not displayed.
To get a list of all keys and values, use the expression [metadata values].
• To get the value of a particular key in the incoming metadata, use the expression [metadata key].
Replace key with the name of the key whose value you want to use. For example, to display the
name and location of the image file on the image, add a Text node after the image and enter
[metadata input/filename] in the message field.
• To get a filtered list of keys in the incoming metadata, use the expression [metadata keys filter].
Replace filter with whatever you want to use to filter the list. You can use asterisks (*) as wildcards in
your filter to substitute zero or more characters in the key names. For example, to get a list of all
the keys with the letter f in them, use the expression [metadata keys *f*]. To get a list of all the
keys starting with input/, use the expression [metadata keys input/*].
By default, the keys are listed on separate lines. To change this, you can use -s "separator" to have
the keys separated by a separator of your choice. Replace separator with whatever you want to
appear between the different keys. For example, to get a list of all the keys starting with input/ and
separated by spaces, you can use [metadata -s " " keys input/*]. To get the same list separated by
commas, use [metadata -s ", " keys input/*].
By default, if you attempt to access metadata that does not exist in the stream, Nuke returns an
empty string. To make this error instead, use the -e flag before other parameters.
For more information on using expressions, see the Expressions chapter.
USER GUIDE
1033
Accessing Metadata Using Python | To Edit the List of Actions in the ModifyMetaData Properties
Accessing Metadata Using Python

You can also access metadata using the Python programming language. For more information, see
the Nuke Python documentation (Help > Documentation).
USER GUIDE
1034
Audio in Nuke
In many compositing projects it’s vital to be able to key visual changes to cues on the audio track that
goes with the picture. You can use Nuke’s AudioRead node to read in an audio file, view it in the Curve
Editor and Dope Sheet in order to line up keyframes of your composition with the waveform of the
sound. You can then flipbook the audio with your footage to preview your comp with sound.
Quick Start
You can load audio files into Nuke using the AudioRead node, in much the same way as you read in
images with the Read node. You can read in uncompressed .wav and .aiff files and flipbook them
with your footage for playback.

1. Read in an audio file. See Reading Audio Files into the Node Graph.
2. Display an audio waveform for your audio clip and access its animation curve in the Curve Editor
or the Dope Sheet. See Creating and Editing Audio Curves.
3. When you’re done, you can flipbook your script to view and listen to the results. See Flipbooking
the Audio Track.
USER GUIDE
1035
Reading Audio Files into the Node Graph |
Reading Audio Files into the Node

Graph
You can drag and drop audio clips from the Project tab to the Node Graph, if the clip is already in
Nuke.
Otherwise, use the AudioRead node to read in an audio file:

1. To create an AudioRead node, click Other > AudioRead in the Nuke Toolbar.
The AudioRead node doesn’t have to be connected to other nodes.
Simple AudioRead node setup.
USER GUIDE
1036
Reading Audio Files into the Node Graph |
Tip: You can also load an audio file by creating a normal Read node and navigating to a
supported audio file.
2. In the AudioRead properties, use the file control to navigate to the audio file you want to read in.
You can read in uncompressed .wav and .aiff files.
3. Use the time range fields to enter the start and end times in seconds for the audio in Nuke.
4. In the file time range fields, enter the start and end times in seconds of the audio file read in.
These are automatically set to the values in the file, but you can change them to trim the data
used in Nuke.
5. If you want to discard your changes and reload the audio file, click reload.
6. Use the ratesource menu to select the source for the sample rate:
• file - reads the rate from the audio file.
• custom - lets you specify a custom sample rate in the rate field.
USER GUIDE
1037
Creating and Editing Audio Curves | Creating a Keyframe Curve
Creating and Editing Audio Curves

Once you have read in an audio file (see Reading Audio Files into the Node Graph), you can display an
audio waveform for your audio clip and access its animation curve in the Curve Editor or the Dope
Sheet.
Creating a Keyframe Curve

In the curves section of the AudioRead properties panel, you can generate curves out of the audio
data:
1. Set the keyframe interval you want to use when creating the curves in the key interval field. For
example, if you enter 3, keyframes are created to every third frame of the input footage.
2. Click generate to generate the audio data as a curve that you can use in the Curve Editor and
Dope Sheet.
3. View the left and right stereo levels on the current frame in the left and right fields and adjust if
necessary. Any changes are reflected on the curve automatically.
Modifying the Audio Curve in the Curve Editor and Dope

Sheet
When you’re working with your audio curve in the Curve Editor or Dope Sheet, there are a few right-
click options that you can use to adjust how the clip’s waveform displays:
1. Right-click in the Curve Editor or Dope Sheet and select View > Audio.
2. Then select Source and check the box for either ProjectDefault or an AudioRead node
depending on which one you want to view. If you’ve only got one AudioRead, it is the project
default.
3. If you’re working with a stereo clip with more than one audio channel, you can select your audio
channel by ticking the appropriate box under Channel.
4. Select a style in which you want your waveform to be drawn by selecting one of the DrawStyle
options:
• Off - to draw no audio waveform.
• Behind - to draw a waveform behind the animation curves.
• Below - to draw a waveform below the animation curves.
USER GUIDE
1038
Creating and Editing Audio Curves | Modifying the Audio Curve in the Curve Editor and Dope Sheet
Audio waveform in the Curve Editor.
USER GUIDE
1039
Flipbooking the Audio Track | Modifying the Audio Curve in the Curve Editor and Dope Sheet
Flipbooking the Audio Track

When you’re done, you can proceed to flipbooking your results:
1. Click the flipbook this Viewer button in the Comp Viewer.
2. In the Flipbook dialog, select the AudioRead file you want to use in the Audio dropdown.
3. Click OK. View and listen to your clip in Flipbook.
USER GUIDE
1040
Previews and Rendering
Nuke supports a fast, high-quality internal renderer, with superior color resolution and dynamic
range without a slowdown in the workflow.
About Rendering in Nuke

These are some of the key features of Nuke’s rendering engine:
• Multi-threaded rendering to take advantage of multiple processors in its calculations.
• Scanline (as opposed to buffer-based) rendering allows you to immediately see portions of render
output.
• Calculations performed with 32-bit precision, using linear light levels.
This chapter teaches you how to use the renderer’s various features to preview a script’s output and
generate its final elements. You’ll also learn how to preview using Flipbooking and Capture, and check
output on an external broadcast video monitor.
Quick Start
1. With the Viewer, you can preview your footage and use the ROI button to focus on a
particular part of it. For more information, see Previewing Output.
2. You can then flipbook your clip. A quick way of doing this is to click the Flipbook button in the
Viewer, set the frame range and other settings in the Flipbook dialog, and click OK to flipbook
and automatically launch the Flipbook Viewer. See Flipbooking Sequences.
3. Save out low resolution .jpg sequences using Capture to share your work, such as for peer review
purposes. Click the Capture button in the Viewer, set the frame range and other settings in
the Capture dialog, and click OK.
4. If you’ve read in an audio clip with the AudioRead node, you can flipbook that with your footage
just by selecting the right AudioRead node in the Audio dropdown. For more information, see
Audio in Nuke.
5. To check the final result in correct video colorspace and pixel aspect ratio, preview your footage
on an external broadcast video monitor. See Previewing on an External Broadcast Video Monitor.
USER GUIDE
1041
Previews and Rendering |
6. When you’re happy with your preview results, you can render out your clip. To do this, you need
to connect at least one Write node to your clip, and then set the render properties in the
properties panel. You can specify your render format in the filename field, and use the frame
control to offset your frame numbers if necessary. For more information, see Output (Write)
Nodes.
7. If you have more than one Write node connected to your node tree, you can render out all of
them, or select the ones you want to render. You can then click Render > Render all or Render
selected in the menu bar to start the renders. For more information, see Output (Write) Nodes.
USER GUIDE
1042
Previewing Output |
Previewing Output
This section explains how to preview individual frames in a Nuke Viewer window (see Previewing in a
Nuke Viewer), how to render a flipbook for a sequence of frames (see Flipbooking Sequences), and
how to preview output on an external broadcast video monitor (see Previewing on an External
Broadcast Video Monitor).
USER GUIDE
1043
Previewing in a Nuke Viewer | To Enable the ROI Render Feature
Previewing in a Nuke Viewer

When you connect a Viewer to a given node’s output (by selecting the node and pressing a number
key), Nuke immediately starts rendering the output in the Viewer using all available local processors.
Keep in mind the following tips in order to speed up this type of preview rendering:
• First, if you don’t need to evaluate the whole image, zoom into the area of interest. Nuke then
renders only the portion of scan lines visible within the Viewer.
• Alternatively, you can use the Viewer’s region of interest (ROI) feature to render only a portion of the
image, while seeing that result in the context of the whole image.
To Enable the ROI Render Feature

1. Press Alt+W over the Viewer. The Viewer’s ROI button turns red , indicating that the feature is
enabled.
2. Drag on the Viewer to draw the region of interest. The Viewer now renders only the pixels within
the region.
To Edit the Position or Size of Current ROI

1. Click the ROI button so that it turns red . The overlay for the current ROI appears in the
Viewer.
2. To reposition the ROI:
Using the crosshair in the middle of the ROI, drag the ROI to the desired location.
3. To resize the ROI:
Drag any corner or side of the ROI until you achieve the desired size.
To Disable the ROI Render Feature

Click the Viewer’s ROI button. It turns gray , signaling that it is off. The Viewer now renders all of
the visible image.
USER GUIDE
1044
Flipbooking Sequences | Setting the Viewer Cache Location and Size
Flipbooking Sequences
Flipbooking a sequence refers to rendering a range of images (typically at proxy resolution), then
playing them back in order to accurately access the motion characteristics of added effects.
You have a few options for flipbooking within Nuke:

• You can enable automatic disk caching of rendered frames, then play these frames back using
Nuke’s native Viewer. This option does not let you define a specific playback rate.
• You can render out temporary image sequences using the default flipbooking tool, a RAM-buffering
playback utility, which is displayed in its own Viewer and plays back sequences at the defined frame
rate.
If you have purchased a license for HieroPlayer, you can flipbook using HieroPlayer rather than the
default flipbook that ships with Nuke.
• You can also set up an external flipbooking application in Nuke using Python. For more
information, see the Nuke Python documentation (Help > Documentation).
The Nuke Viewer automatically saves to disk a version of every frame it displays. When you play
through sequences in the Viewer, it reads, where possible, from this cache of pre-rendered images,
making real-time play back possible (depending, of course, on image resolution and your hardware
configuration). You can define the location and size of the Viewer cache in the Preferences.
Depending on what viewer buffer bit depth has been set to in the Viewer settings, the cache can
contain 8-bit (byte), 16-bit (half-float), or 32-bit (float) image data. This offers a trade-off between
speed and quality. Half-float and float modes provide higher precision than byte, but are also slower
to process.
Setting the Viewer Cache Location and Size

1. Click Edit > Preferences to display the Preferences dialog.
2. In the Performance > Caching section, there is a Disk Caching – temp directory field. Use this
to enter the path name of the directory in which you want to store the flipbook images (for
example, c:/temp).
3. Using the comp disk cache size control, specify the number of gigabytes you want to allow the
image cache to consume.
4. Click OK in the bottom-right corner of the Preferences dialog to update preferences and then
restart Nuke.
The Viewer now caches each frame it displays in the directory specified. When you click the playback
buttons on the Viewer, or drag on the scrub bar, Nuke reads in images from this cache.
USER GUIDE
1045
Flipbooking Sequences | To Use HieroPlayer as Nuke's Flipbook
Note that the cached images have unique names reflecting their point of output location in the script.
This means that you can cache images from multiple nodes in the script without overwriting
previously cached images. For more information on caching, see Image Caching.
To Use HieroPlayer as Nuke's Flipbook

If you have purchased a license for HieroPlayer, you can flipbook using HieroPlayer rather than the
default flipbook that ships with Nuke:
1. Copy the flipbook_hiero_launch.py and hiero_flipbook.py file from:
<installation_folder>/pythonextensions/site-packages/hiero/examples/
to your .nuke folder.
2. Launch Nuke and open the Script Editor.
3. Type:
import hiero_flipbook
and press Ctrl/Cmd+Return.
4. Open the required comp and launch the flipbook by clicking the icon under the Viewer or by
pressing Alt+F on a selected node.

5. Select HieroPlayer from the Flipbook dropdown and proceed as described below.
Flipbooking a Sequence
To flipbook an image sequence, do the following:
1. Select the node that you want to flipbook the output of.
Note: If you select a Write node in the step above, you must first click its Render button in
order to manually render its output to the destination defined in the file field. This step is
necessary only in the case of Write nodes.
2. Select Render > Flipbook selected (or press Alt+F).

Alternatively, you can click the Flipbook this Viewer button at the bottom-right of the Viewer.
This flipbooks the nodes that are connected to the Viewer.

A Flipbook dialog opens.
3. Check that your settings are correct in the dialog. The default values are copied from the Viewer
you currently have active. You can change them if necessary:
• Flipbook - set the flipbooking application you want to use.
• Take settings from - set which Viewer should be used to draw default values from.
USER GUIDE
1046
Flipbooking Sequences | Flipbooking a Sequence
• Enable ROI - Check to define your region of interest.

• Channels - select which layer to display in the flipbook result.
• Frame range - set the frame range you want to flipbook.
• Use proxy - check to use proxy mode.
• Render in background - check to render in the background. If you check this, you can also set
Thread limit and Memory limit controls. The former limits the number of threads that Nuke
uses in the background and the latter limits the amount of cache memory that Nuke uses.
Note: If you’re rendering multiple sequences in the background, this can take up more than
the total RAM on your machine. When running background renders of any type, you need to
make sure they don't require more RAM all together than what's available on the machine,
otherwise you may experience problems such as hanging.
• Delete existing temporary files - Check to delete any existing temporary files with the same file
name before flipbooking.
• LUT - select the LUT appropriate for viewing. By default, the flipbook renders your files with a
linear colorspace and attempt to pass a LUT file to the flipbook.
• Burn in the LUT - If you check this box the flipbook files are rendered with the LUT applied. If
you uncheck it, the flipbook is displayed using it's equivalent LUT (based on the LUT’s name). If
you have an equivalent LUT available in the flipbook program, then it's better not to check the
Burn in the LUT box. This way, when you measure pixel values in the flipbook application they
match what you get in the Nuke Viewer.
• Audio - if you want to flipbook an audio file with your clip, select the AudioRead node you need
in this dropdown. For more information on audio files in Nuke, see Audio in Nuke.
• Continue on error - check to keep rendering even if an error occurs during the process.
• Views - check the view(s) to output, if you're working in a multi-view comp.
Note: The Views control is only available in multi-view comps. See Stereoscopic Scripts for
more information.
• Buffer - set which buffer you want include.

4. Click OK.
Nuke renders as a temporary sequence the output of the selected node using the frame range
and resolution defined in the script’s settings. This may take a few moments.
After the render is complete, Nuke launches Flipbook Viewer and loads in the temporary
sequence. You can play it back and view it using Flipbook Viewer controls.
USER GUIDE
1047
Flipbooking Sequences | Capturing the Viewer
If you flipbooked a stereo comp, you can right-click in the Flipbook and choose Stereo Modes to
view your flipbook in various configurations:
• Side by Side - displays the views side by side at the correct aspect ratio, and adds selection
controls above the Viewer.
• Squeezed Side by Side - displays the views side by side and squeezed to fit the format
horizontally, and adds selection controls above the Viewer.
• Squeezed Above by Below - displays the views above and below each other and squeezed to fit
the format vertically, and adds selection controls above the Viewer.
• Interlace H - displays the views interlaced horizontally, and adds selection controls above the
Viewer.
• Interlace V - displays the views interlaced vertically, and adds selection controls above the
Viewer.
• Checkerboard - displays the views using an alternating checkerboard pattern (one pixel from
left and one pixel from right), and adds selection controls above the Viewer.
• Anaglyph - displays the views simultaneously using a red hue for left and green hue for right,
and adds selection controls above the Viewer.
• Flicker - displays both views alternately, and adds selection controls above the Viewer.
Capturing the Viewer

You can capture the contents of the Viewer for a quick real-time flipbook, also known as a playblast,
and save the content out to .jpg for review. This is very useful for quickly checking the animation in
your 3D scene in real-time without having to do a full scanline render. Playblast flipbooks the Viewer
'as is', including 2D and 3D scenes, handle and transform overlays (such as roto shape outlines),
wipes, and so on.
USER GUIDE
1048
The contents of a Viewer containing 3D and 2D information, including

the wipe handle, captured in .jpg format.
To capture the Viewer contents:

1. Click the Capture button under the Viewer, to the right of the flipbook button.
The Capture dialog displays.
2. Select the required Flipbook tool using the dropdown.

3. If you have more than one Viewer in the script, select the required Viewer.
4. Check that your settings are correct in the dialog. The default values are copied from the Viewer
you currently have active. You can change them if necessary:
• Frame range - set which Viewer you want to draw default values from, and set the frame range
that you want to flipbook.
• Render in background - check to render in the background. If you check this, you can also set
Thread limit and Memory limit controls. The former limits the number of threads that Nuke
uses in the background and the latter limits the amount of cache memory that Nuke uses.
• Delete existing temporary files - disable this if you want to retain previously cached files.
• Buffer - select the buffer you want to capture.
• Customise write path - Check this to manually set the location (in the Write path field below)
where the .jpg files are stored.
• No flipbook - Enable this if you don't require the selected flipbooking tool to display the frame
range.
USER GUIDE
1049
Note: Nuke only captures sequences using the .jpg format. Don't forget to include printf or
hash frame padding, such as %4d or ####.
5. Click OK to capture the contents of the Viewer.

The frame range is cached as .jpgs to either the default location set in the Preferences dialog, or
– if enabled – the directory specified in the Write path control.
USER GUIDE
1050
Previewing on an External Broadcast Video Monitor | Capturing the Viewer
Previewing on an External
Broadcast Video Monitor
To check the final result in correct video colorspace and pixel aspect ratio, you can preview the
current Viewer image on an external broadcast video monitor. This option requires additional
hardware, such as a monitor output card or a FireWire port.
Our monitor out architecture interfaces directly with the AJA and BlackMagic device drivers, which are
unified across their respective hardware lines, meaning all current supported cards for the versions
detailed in Third-Party Library Versions should work. We've tested the following AJA and Blackmagic
hardware:
Formats
SD
HD
2K
UHD
4K
BNC
HDMI
No Yes Yes No
Platforms
Win, Mac, Linux Win, Mac, Linux Win, Linux Mac 10.9 and
10.10
Drivers
USER GUIDE
1051
• Windows - 12.5.1 • Windows - 12.5.1 • Windows - 12.5.1 Mac - 12.5.1

• Mac - 12.5.1 • Mac - 12.5.1 • Linux - 12.5.2.7
• Linux - 12.5.2.7 • Linux - 12.5.2.7
Note: The following should be taken into account when using the Monitor Out functionality
with AJA cards:
If you're running AJA cards on Linux, you can contact www.aja.com/support to obtain the
correct drivers.
12-bit monitor output is only supported with dual connection cards, that is cards with two
physical connections, not dual links combining two separate streams of data.
Hiero is unable to send out the right eye separately using the 2nd output cable of KONA 3G
cards. Instead, both views are sent through the 1st output and can be viewed using the side-
by-side, anaglyph, and interlacing options.

Formats
SD
HD
2K
UHD
4K
BNC
HDMI
No No Yes No
Platforms
USER GUIDE
1052

Win, Mac, Linux Win, Mac, Linux Win, Mac, Linux Win, Mac, Linux
Drivers
Desktop Video Desktop Video Desktop Video Desktop Video

10.7 to 10.8.4 10.7 to 10.8.4 10.7 to 10.8.4 10.7 to 10.8.4

12G
Formats
SD
HD
2K
UHD
4K
BNC
HDMI
Yes Yes Yes
(Both views through

one output, so the Full
Resolution option is
not available.)
Platforms
Win, Mac, Linux Win, Mac, Linux Win, Mac, Linux
Drivers
USER GUIDE
1053
Previewing on an External Broadcast Video Monitor | To Preview Output on an External Broadcast

12G
Desktop Video 10.7 to Desktop Video 10.7 to Desktop Video 10.7 to

10.8.4 10.8.4 10.8.4
Nuke can output images to external broadcast monitors in either 8- or 10-bit RGB color modes. 10-bit
color is automatically selected when the Viewer's gl buffer depth setting is half-float or float,
provided it is supported by the monitor output card. In all other cases, 8-bit color is used.
Note: Selecting a monitor output mode with the phrase 10-bit in its description only
outputs true 10-bit color if a gl buffer depth of either half-float or float is selected. Half-
float and float modes are considerably slower to process, so it is recommended to stick with
byte mode for monitor output unless 10-bit color is specifically required.
To Preview Output on an External Broadcast Video

Monitor
1. Press S on the Viewer to open the Viewer settings.
2. From the monitor output device dropdown menu, select the external device you want to use
and check enable monitor output. All available devices are automatically detected and listed in
this menu.
3. From the monitor output mode dropdown menu, select the display mode for the device you
selected in the previous step. The available options depend on the device you are using. By
default, the most recently used display mode is selected.
4. Press Ctrl/Cmd+U, navigate to Viewer > Toggle Monitor Out, or click on the monitor output
button in the Viewer overflow dropdown.
From now on, any output connected to the Viewer is sent to the monitor output device you selected.
USER GUIDE
1054
Previewing on an External Broadcast Video Monitor | To Disable Viewing on an External Broadcast
The monitor output image is always full frame, unzoomed (1:1 ratio), and unpanned, regardless of
what the Viewer it is linked to is set to. This means that slightly mismatching formats (for example,
640x512 / 640x448 for PAL/NTSC) are not rescaled to fit the monitor.
If you save your Nuke script with monitor output enabled, this setting is saved with the script. The
next time you open the same script, monitor output is enabled.
To Disable Viewing on an External Broadcast Video

Monitor
• Click on the monitor output button in the Viewer controls.
• Select Viewer > Toggle Monitor Output from the menu bar.
• Press Ctrl/Cmd+U. This can be particularly useful if you are using a monitor output device that can
draw things on the screen you are using to display the Nuke window (such as the Digital Cinema
Desktop Preview device installed with Apple Final Cut Pro).
• Press S on the Viewer to open the Viewer settings, and uncheck enable monitor output.
USER GUIDE
1055
Rendering Output | To Disable Viewing on an External Broadcast Video Monitor
Rendering Output
Nuke can render images locally on your workstation (see Output (Write) Nodes) or it can be setup to
render images on a network render farm (see Using the Frame Server on External Machines). Before
rendering, make sure that you are using the appropriate file name syntax (see File Name Conventions
for Rendered Images) and verify that your project settings have the correct output format and proxy
format selected (see Render Resolution and Format).
By default, if you choose to render frame 15, the resulting file is numbered accordingly, for example
image.0015.rgb. However, you can change this behavior via expressions, specified start frames, and
constant offsets (see Changing the Numbering of Rendered Frames).
Sometimes, you may want to render an image just to read the rendered image back in (see Using a
Write Node to Read in the Rendered Image). Because reading the output from a file is faster than
calculating its output by processing the node tree upstream, this can speed up large projects.
Note: Scripts that require Nuke to load a large number of files concurrently (for example,
by having hundreds of Read nodes followed by TimeBlurs) may exceed the number of
available file handles per process, causing problems when rendering scripts.
Nuke itself supports up to 2048 file handles on all systems; however, you may need to
increase the file handle limit on your system.
On Mac, you can increase the default limit of 256 (depending on your version) by entering
the following command from the Terminal and then running Nuke from the same Terminal
session:
ulimit -Sn 2048
USER GUIDE
1056
Render Resolution and Format | To Disable Viewing on an External Broadcast Video Monitor
Render Resolution and Format

Before rendering a script, it’s important to check what is the currently active mode: the full-size or the
proxy mode. Nuke executes all renders at the currently active scale. Thus, when a script is rendered in
proxy mode, processing is done at the proxy scale and image output goes to the file name in the
Write node’s proxy field. If you do not specify a proxy file name, the render fails with an error. It
never resizes the proxy image, and it does not write the proxy image over the full-size one.
To view and change the proxy resolution for the current script file, select Edit > Project Settings from
the menu bar, or press S with the mouse pointer over the Node Graph or the Properties Bin.
Changing the output resolution under

project settings.
From the Project Settings properties panel, you can select a new render format from the dropdown
menu of predefined resolutions, and toggle proxy rendering. You can also select the new option
under either full size format or proxy format or use the proxy scale fields to define custom render
resolutions for the composite. When rendering in proxy mode, use the dropdown menu on the right
to select whether to use the resolution defined under proxy format or proxy scale. Also check that
you have set read proxy file to what you want - this setting controls how Read nodes select the file to
read (full res file or proxy) in the proxy mode. For more information on these settings, refer to Setting
Up Your Script .
USER GUIDE
1057
Output (Write) Nodes | To Render a Single Write Node
Output (Write) Nodes

With the correct resolution and format selected, you then insert Write nodes to indicate where you
want to render images from the script.
Inserting Write nodes for rendering.
One Write node is usually placed at the bottom of the compositing tree to render the final output.
However, Write nodes have both input and output connectors, so they may be embedded anywhere
in the compositing tree.
You can execute renders for a single Write node or all Write nodes in your compositing script.
To Render a Single Write Node

1. Select the node in the script from which you want to render an image.
2. Select Image > Write (or press W over the Node Graph). Nuke attaches a Write node and opens
its properties panel.
3. Connect a Viewer to the Write node you want to render and verify that the correct resolution is
displayed for output. If necessary, press Ctrl/Cmd+P to toggle between full-res and proxy
resolution. The displayed output resolution is used for rendering.
USER GUIDE
1058
4. In the properties panel, click the file or proxy field’s folder icon (depending on whether you want
to render high res or low res images) and browse to the directory where you want to store the
rendered sequence. For instructions, see Using the File Browser.
5. After the path, type a name for the rendered image and then click OK. If you’re rendering an
image sequence, include the frame number variable (for example, ####) in the name.
See To Render Selected or All Write Nodes in the Script below for examples of valid file names
with the frame number variable.
6. If necessary, adjust the following controls:
• Using the channels dropdown menu and checkboxes, select the channels you want to render.
• Using the frame dropdown menu and input field, set the relation between the currently
processed frame and the numbering of the frame written out. For more information, see
Changing the Numbering of Rendered Frames.
• check read file if you want the output of the Write node to be produced by reading the
rendered file back in rather than by processing the upstream node tree. For more information
on this and the missing frames control, see Using a Write Node to Read in the Rendered Image.
• From the colorspace dropdown menu, select which lookup table to use when converting
between the images’ color space and Nuke’s internal color space.
• From the file type dropdown menu, select the file format for the rendered images. If you don’t
specify a file format, Nuke uses the extension in the file name to figure out the format.
USER GUIDE
1059
• Check the limit to range box if you want to disable the node when executed outside of a
specified frame range. In the frame range fields, enter the range of frames you want to make
executable with this Write node.
7. In the Write node properties, click the Render button.
8. In the Render dialog, adjust the render settings if necessary:
• Frame range - set the frame range you want to render.
• Render using frame server - enable this checkbox if you want to use Nuke's frame server to
render the frames specified in the current render task. When the frame server is disabled, Nuke
can only perform one render task at a time. Executing another render displays a warning and
the current render is paused until you acknowledge the message:
See Rendering Using the Frame Server for more information.

• Views - set which views to include in the render. See Selecting Which Views to Render for more
information.
9. Click OK.
Tip: When specifying the frame range to render, you can enter complex frame ranges into
the frame range prompt dialog. For example, if you enter 1-5 8 10 15 22-25, it only renders
those frames. You can also increment rendered frames using something like 10-50x10,
which resolves to only frames 10, 20, 30, 40, and 50 from the range.
Likewise, you can specify multiple ranges on the command line, for example:
nuke -F 1-5 -F 8 -F 10 -F 15 -F 20-40x5 -x myscript.nk
Tip: When rendering with the Write node, you can force a certain data type by adding the
data type and a colon before the file path. For example, you can enter ftiff:C:\Temp\test.tif
as the file path to render a file whose data type is ftiff and extension .tif.
USER GUIDE
1060
Output (Write) Nodes | To Render Selected or All Write Nodes in the Script
To Render Selected or All Write Nodes in the Script

1. Connect a Viewer to a Write node you want to render and verify that the correct resolution is
displayed for output.
2. If necessary, press Ctrl/Cmd+P to toggle between full-res and proxy resolution. The displayed
output resolution is used for rendering.
3. If you want, you can change the order in which your Write nodes are rendered by giving them
custom render order numbers in the render order field.
• With the desired Write node selected, select Render > Render selected (or press F7).
• Select Render > Render all (or press F5).
5. In the Render dialog, adjust the render settings if necessary. The default values are drawn from
the Viewer you have active.
• Frame range - set the frame range you want to render.
• Render using frame server - disable this checkbox if you don't want to use Nuke's frame
server to render the frames specified in the current render task. If you disable the frame server,
Nuke can only perform one render task at a time. Executing another render displays a warning
and the current render is paused until you acknowledge the message:
See Rendering Using the Frame Server for more information.

• Views - set which views to include in the render. See Selecting Which Views to Render for more
information.
6. Click OK.
Tip: When specifying the frame range to render, you can enter complex frame ranges into
the frame range prompt dialog. For example, if you enter "1-5 8 10 15 22-25", it only renders
those frames. Likewise, you can specify multiple ranges on the command line, for example:
nuke -F 1-5 -F 8 -F 10 -F 15 -F 22-25 -x myscript.nk
USER GUIDE
1061
Output (Write) Nodes | Selecting Which Views to Render
You can see the progress of your render in the status window that appears. When the render
is complete, the rendered images are added to the directory you specified.
Tip: When rendering with the Write node, you can force a certain data type by adding the
data type and a colon before the file path. For example, you can enter ftiff:C:\Temp\test.tif as
the file path to render a file whose data type is ftiff and extension .tif.
Selecting Which Views to Render

The Write nodes in a script determine which views are available at render time, but the Render dialog
can be used as a filter to limit the views rendered from those available. Write nodes and the Render
dialog default to all available views.
For example, if your script contained two Write nodes calling for the main view and left and right
views, but the Render dialog is set to render main, then only the main view is rendered.
Conversely, if the Write nodes are calling for the left and right views, but the Render dialog is set to
render main, then no views are rendered.
A slightly more complex example is shown in the image.
In the example above, the left-hand dialog is set to render only main and the right-hand dialog is set
to render only cam1 and cam2.
USER GUIDE
1062
Output (Write) Nodes | Notes on Rendering QuickTime Files
So, in the case of the first render, Write1 renders main, Write2 renders nothing, and Write3 renders
main.
In the case of the second render, Write1 renders cam1 and cam2, Write2 renders cam1, and Write3
renders nothing.
Notes on Rendering QuickTime Files

If you are rendering .mov files, you can:
• choose the QuickTime codec from the codec dropdown menu.
Note: If you're using the Avid DNxHD codec, Avid AVDn, avoid setting the pixel format
control to r408 as there is a known issue within the codec causing frames to darken with
each frame progression in the sequence.
• Set the encoder library used to write the file:
Note: Depending on the codec in use, this control may be read only. For example, Apple
ProRes 4444 always uses mov64, but Animation allows you to choose mov32 or mov64.
• mov32 - uses the full range of QuickTime codecs, but can be slow to process due to extra
complexity during decode.
• mov64 - uses its own packing and unpacking and streams decode/encode for extra processing
speed, but only supports a sub-set of QuickTime codecs.
Note: Nuke defaults to the fastest decoder for the codec used in the file - if you're reading
in a type supported by the mov64 sub-set, Nuke defaults to that reader. Otherwise, the
fallback mov32 reader is used.
• fps - set the playback frames per second for the output file.
• audio file - Allows you to specify a separate audio file to include in the output. Either enter the
filepath manually or click the browse button to locate the audio file.
• audio offset - Sets the start time of any audio file specified in the audio file control. The unit of
measure is specified using the units control. Negative values cause the audio to start before the
video and vice versa.
• write timecode - When enabled, Nuke writes the timecode into the .mov metadata, where
available.
USER GUIDE
1063
Note: The timecode is read from the input/timecode metadata key pair. If this field is
blank, the timecode is not written into the file.
Advanced QuickTimes Options

You can adjust advanced codec options by opening the Advanced dropdown. The options available
vary, depending on whether you're using the mov32 or mov64 encoder.
Control Description
mov32 encoder
codec options Click to display an advanced Compression Settings dialog.
fast start When enabled, MOVs are playable while still down loading.
use format When enabled, the rendered .mov uses the same pixel ratio as the input.
aspect
When disabled, the codec determines the pixel aspect to use.
Note: Codecs writing PAL and NTSC should be allowed to determine

the ratio during render, but formats that otherwise expect 1:1 pixel
ratios may require this override.
ycbcr matrix Sets the way RGB is converted to Y’CbCr. Rec 601 and Rec 709 follow the ITU.BC
specifications, whilst Nuke Legacy, Nuke Legacy Mpeg, and Nuke Legacy
YUVS are retained for backwards compatibility. Format-based sets the color
matrix to Rec 601 for formats with a width below 840 pixels and Rec 709 for
formats with a width of 840 pixels or above.
This setting is only available when you’re working with a Y’CbCr-based pixel type.
pixel format Lists pixel formats supported by the current codec. The pixel format defines the
type and layout Nuke requests from QuickTime:
• Pixel colorspace - either RGB(A) or YCbCr(A). This defines whether QuickTime
or Nuke’s QuickTime reader does the conversion between colorspaces. For a
Y’CbCr pixel type, choosing an RGB(A) colorspace means Nuke relies on
QuickTime to do the RGB to Y’CbCr conversion. Choosing a YCbCr(A)
colorspace means that Nuke is responsible for the conversion, and so a
USER GUIDE
1064
Control Description
specific ycbcr matrix can be used (this is recommended).

• Pixel bit depth - 8-bit, 16-bit, and so on. This sets the encoding depth used
when decompressing the frames. A large bit depth gives higher accuracy at the
cost of speed and memory usage.
• Pixel layout - 422, 444, 4444, and so on. This defines how the chroma channels
in the buffer are arranged. 444 buffers have lower spatial chroma sampling
than 422, so they are generally preferred when available. For all cases, Nuke
unpacks the sub-sampled buffer to full resolution.
• Range - either Biased or empty. For RGB(A) types, the values are full range
(from 0 to 1). For YCbCr(A) types, the values are in video range by default,
offering headroom at both ends of the scale. If this is set to Biased, then
headroom is only available at the top end.
• (4cc). This is the pixel type 4cc, as defined by the QuickTime API.
This setting defaults to the best format accepted by the codec.
write nclc When enabled, write the nclc data in the colr atom of the video sample.
write gamma When enabled, write the gama data in the gama atom of the video sample.
write prores When enabled, write the prores data in the prores header of the video sample.
mov64 encoder
bitrate Sets the target bitrate that the codec attempts to reach, within the limits set by
the bitrate tolerance and quality min/max controls.
Note: The bitrate control is only enabled for certain codecs, such as
MPEG-4 - Video.
bitrate tolerance Sets the amount that the bitrate can vary from the bitrate setting. Setting this
tolerance too low will result in renders failing.
Note: The bitratetolerance control is only enabled for certain codecs,

such as MPEG-4 - Video.
USER GUIDE
1065
Control Description
quality min Sets the quality range within which the codec can vary the image to achieve the
specified bitrate. Higher ranges can introduce image degradation.
quality max Note: The quality min/max controls are only enabled for certain
codecs, such as MPEG-4 - Video.
gop size Sets how many frames can be placed together to form a compression GOP
(group of pictures).
Note: Use caution with this control as large alterations can stop other
applications reading the rendered file.
Note: The gop sizecontrol is only enabled for certain codecs, such as
MPEG-4 - Video.
b frames Sets the maximum number of B frames that can be consecutive in the rendered
file.
The default, 0, does not impose any maximum number of B frames in the
output.
Note: The b frames control is only enabled for certain codecs, such as
MPEG-4 - Video.
write nclc When enabled, write the nclc data in the colr atom of the video sample.
Nuke writes the selected colorspace and pixel format, along with some other information, into the
metadata of the file. If you then read the rendered file in, Nuke reads that metadata and is able to
pick the correct defaults for the file. To see the file metadata yourself, use a ViewMetaData node.
When writing QuickTime files, some users have reported receiving the following error at the end of
rendering:
Failed to flatten movie data: the movie is open in another application.
This is because the output file has been opened by some other process. This can be the Finder on
Mac trying to show a preview, an OS file system search trying to index the file, or a virus checker, for
USER GUIDE
1066
Output (Write) Nodes | Notes on Rendering OpenEXR Files
example. The workaround is to turn off Fast Start in the Write node controls to skip the flattening
process that is affected.
Notes on Rendering OpenEXR Files

Nuke supports multi-part OpenEXR 2.2 files, which allow you to store your channels, layers, and views
in separate parts of the file. Storing the data this way can make loading .exr files faster, as Nuke only
has to access the part of the file that is requested rather than all parts. However, for backwards
compatibility, you also have the option to render your .exr files as single-part images.
To set how the data is stored in your rendered .exr file, open the Write properties and set interleave
to:
• channels, layers and views - Write channels, layers, and views into the same part of the rendered
.exr file. This creates a single-part file to ensure backwards compatibility with earlier versions of
Nuke and other applications using an older OpenEXR library.
• channels and layers - Write channels and layers into the same part of the rendered .exr file, but
separate views into their own part. This creates a multi-part file and can speed up Read
performance, as Nuke only has to access the part of the file that is requested rather than all parts.
• channels - Separate channels, layers, and views into their own parts of the rendered .exr file. This
creates a multi-part file and can speed up Read performance if you work with only a few layers at a
time.
USER GUIDE
1067
Rendering Using the Frame Server | Notes on Rendering OpenEXR Files
Rendering Using the Frame

Server
The Frame Server reduces render times by sharing work over the number of render processes
specified in the Preferences. The Frame Server logs renders in the Background Renders panel,
including information such as script name, Write node, frame range, progress, and whether or not
the render produced errors. The Frame Server does not display a progress bar for render tasks, but
you can disable Render using frame server in the Render dialog to use the process from older
versions of Nuke.
The Frame Server is disabled by default, but you can enable it in the Preferences > Performance >
Threads/Processes or on a per render basis in the Write node's Render dialog.
Note: Local Frame Server processes use ports 5558-5662.
Rendering Write nodes, either from a node's Properties or from the Render menu, displays the
Render dialog. See Output (Write) Nodes for more information on initiating renders. If the Frame
Server is enabled, clicking OK in the Render dialog starts the selected renders in the background,
allowing you to keep working.
The Background Renders tab is not displayed as part of the workspaces that ship with Nuke, but you
can enable Performance > Threads/Processes > focus background renders in the Preferences to
open the panel automatically when the Frame Server is in use.
Tip: You can also add the panel manually by right-clicking the button in the top-left of
any pane, and then selecting Windows > Background Renders.
USER GUIDE
1068
Rendering Using the Frame Server | Frame Server Preferences
The Background Renders panel includes information such as script name, Write node, frame range,
progress, and whether or not the render produced errors.
Frame Server Preferences

You can control how the Frame Server behaves using the Performance > Threads/Processes tab in
the Preferences:
• render using frame server (Nuke) - when enabled, the Frame Server is always used for rendering.
• focus background renders - when enabled, rendering using the Frame Server automatically opens
the Background Renders panel, or if it is already open, shifts focus to the panel.
• frame server processes to run - sets the number of render processes available to the Frame
Server.
USER GUIDE
1069
Rendering Using the Frame Server | Using the Frame Server on External Machines
Note: You must restart Nuke if you change the number of processes available.
• export renders - allows you to set limitations on the resources available for render processes:
• limit renderer (more responsive ui) – select this option to make the user interface more
responsive during rendering. It tells Nuke to use 2 threads to transcode and to use 25% of RAM
to cache. Using this option is likely to result in slower renders.
• no renderer limits (fastest transcoding) – select this option to ensure that renders happen as
quickly as possible. This option may result in a less responsive user interface during rendering.
• customize render limits – select this option to manually configure the number of threads
used and cache memory available during renders.
Using the Frame Server on External Machines

Although Nuke is capable of rendering frames internally, running the Frame Server on an external
machine can accelerate the process considerably by sharing work across a network of machines.
Note: The Frame Server requires a Nuke license (nuke_i) on the main workstation, but only
a Nuke render license (nuke_r) on the slave machines.
If you want to use an interactive license (nuke_i) on the slave machines, add the --
useInteractiveLicense argument to the runframeserver.py command described below.
Configuring the Frame Server on External Machines

Nuke's Frame Server can be set up on an external machine (or a number of machines) to render from
your Nuke Studio session. To do this, you need to run the runframeserver.py script on the external
machines, found inside the Python site-packages, with specific command line arguments.
Warning: In order for everything to work smoothly, you need to ensure that both your
external slave machines and main Nuke session can read and write files to a shared
location, such as an NFS share.
Depending on platform this can be done by manipulating your default umask setting, but
be aware that this alters the permissions of the created files.
USER GUIDE
1070
Rendering Using the Frame Server | Configuring the Frame Server on External Machines
Additionally, Macs and certain Linux distributions, such as RHEL, can not function as the
main workstation if the firewall is blocking the communication port 5560. You can configure
the firewall to allow certain ports through the firewall using the iptables command, but use
caution when doing so. For example:
sudo iptables -I INPUT -p tcp --dport 5560 --syn -j ACCEPT
Please refer to the documentation on firewalls for your particular platform for more
information.
The Frame Server uses a number of worker processes on the external machine, each of which
requires allocated resources, such as threads, memory, and so on. There are a number of arguments
that you must pass to runframeserver.py for the server to work correctly:
• --numworkers - this is the number of concurrent Nuke processes that are launched when you run
this server render node.
• --nukeworkerthreads - the number of threads that each worker is allocated. This is similar to
setting the -m argument when running Nuke from the command line.
• --nukeworkermemory - the amount of memory, in MB, allocated to each frame server worker.
• --workerconnecturl - the TCP port address of the main workstation you want to serve. For
example:
tcp://bob:5560
where bob is the resolved hostname of a machine you wish to serve. You can also use an IP address.
Tip: To ensure that you're entering a valid URL, try using the ping command to see if you get
a response.
• --nukepath - the path to the Nuke application on the slave workstation.
Tip: On Windows, if there are spaces in the file path, remember to place the path in quotes.
For example, --nukepath="C:\Program Files\Nuke11.3v5\Nuke11.3.exe"
On a Linux slave machine, an example command prompt entry running from the install directory
might look like this:
./python ./pythonextensions/site-
packages/foundry/frameserver/nuke/runframeserver.py --numworkers=2 --
nukeworkerthreads=4 --nukeworkermemory=8096 --workerconnecturl=tcp://bob:5560
--nukepath=./Nuke11.3
USER GUIDE
1071
Rendering Using the Frame Server | Frame Server Logs
On a Windows slave machine, an example command prompt entry running from the install directory
python.exe pythonextensions\site-
packages\foundry\frameserver\nuke\runframeserver.py --numworkers=2 --
--nukepath=Nuke11.3.exe
In the examples, we specify that the slave uses two Nuke workers, with four threads and 8 GB RAM
each, and are slaved to the main Nuke workstation running on bob.
Tip: If your slave machines run a different OS than your main Nuke machine, you can use
the --remap command line argument to convert file paths between them. The host file path
is read first followed by the slave file path. Nuke expects all file paths to use / (forward slash)
between directories. For example:
--remap "P:/,/mnt/renders/"
converts host paths beginning with P:/ (Windows style) to slave paths beginning with
/mnt/renders/ (Linux style).
You can check that the Frame Server and workers are connected by running the following lines in the
Script Editor on the main workstation:
from hiero.ui.nuke_bridge.FnNsFrameServer import frameServer
print [worker.address for worker in frameServer.getStatus(1).workerStatus]
Successful connections should report something similar to the following in the output panel:
['Worker 0 - henry.local - 192.168.1.11', 'Worker 0 - bob.local -
192.168.1.111', 'Worker 1 - henry.local - 192.168.1.11']
Where henry.local is the name of the remote slave, and bob.local is the name of the main Nuke
session.
Note: If the workers cannot contact the Frame Server, an exception is printed in the Script
Editor's output panel.
Frame Server Logs

Broker and Worker logging can to help diagnose Frame Server issues. The logs are written to NUKE_
TEMP_DIR/logs by default, and take the form:
broker.log
worker-0.log
USER GUIDE
1072
Rendering Using the Frame Server | Frame Server Logs
worker-1.log
worker-2.log
Note: Running the Frame Server using Python, as described above, always writes log files to
the specific OS temporary directory. For example, on Windows C:\temp is used.
Tip: You can use the FRAMESERVER_LOG_DIR environment variable to force Frame Server
logs into a different location. See for more information.
USER GUIDE
1073
File Name Conventions for Rendered Images | Writing Versions of Rendered Images
File Name Conventions for

Rendered Images
There is no parameter in the Write node to specify output format. Instead, format is indicated by a
prefix or an extension when you type the file name. Here is the appropriate syntax:
<prefix>:/<path>/<name>.<frame number variable>.<extension>
The optional <prefix>: can be any valid extension. <path> is the full path name to the directory where
you want to render. The <frame number variable> is usually entered as ####, with #### indicating
frame numbers padded to four digits.
You can change the padding by substituting #### with any number of hash marks. For example, two-
digit padding would be ##, three-digit would be ###, and five-digit would be #####.
With these conventions in mind, suppose you want to save an image sequence called “final_comp_
v01” to the TIFF16 format. Here are examples of names that work in the Write node:
tiff16:/<path>/final_comp_v01.####.tiff
/<path>/final_comp_v01.####.tiff16
tiff16:/<path>/final_comp_v01.####
All extensions supported by Nuke may be used for the prefix. See Appendix C: Supported File
Formats on page 2007 for a complete list of recognized extensions.
When a prefix is used, it takes precedence over the format represented by an extension. In fact, the
prefix makes the extension unnecessary.
You could, for example, enter exr:/<path>/### as the file name, and this would create an OpenEXR
image sequence with frame numbers only, padded to three digits.
Writing Versions of Rendered Images

You can write out versions of a file using the Alt+Up/Down arrow keys. Versions must written in the
following format in order for Nuke recognize them:
<path>/final_comp_v01.####.tiff
USER GUIDE
1074
Changing the Numbering of Rendered Frames | Using Expressions
Changing the Numbering of

Rendered Frames
By default, when you render an image sequence, Nuke assumes an exact relation between the
currently processed frame and the numbering of the frame written out. Therefore, if you choose to
render frame 15, the resulting file is numbered accordingly, for example image.0015.rgb. However,
the frame parameter on the Write node lets you change this behavior. For instance, if you have a
sequence that runs from image.0500.rgb to image.1000.rgb, you may want to render it so that the
frame numbering in the resulting files runs from image.0001.rgb to image.0501.rgb. You can do so via
expressions, specified start frames, and constant offsets. Each method is described below.
Using Expressions
1. Select Image > Write to insert a Write node into your script.
2. In the Write properties panel, click the file folder icon , then navigate to the directory path
where you want to save the rendered image sequence. Enter a name for the image sequence.
3. Set frame to expression. Enter an expression in the field on the right. The expression changes
the relation between the currently processed frame and the numbering of the frame written out.
The resulting file name for the current frame is displayed on the Write node in the Node Graph.
For example, if your clip begins from frame 500 and you want to name that frame image.0001.rgb
rather than image.0500.rgb, you can use the expression frame-499. This way, 499 frames are
subtracted from the current frame to get the number for the frame written out. Frame 500 is
written out as image.0001.rgb, frame 501 is written out as image.0002.rgb, and so on.
USER GUIDE
1075
Changing the Numbering of Rendered Frames | Specifying a Frame Number for the First Frame in the
two to get the number of the frame that’s written out. At frame 1, image.0002.rgb is written out; at
frame 2, image.0004.rgb is written out; at frame 3, image.0006.rgb is written out; and so on.
Specifying a Frame Number for the First Frame in the Clip

3. Select start at from the frame dropdown menu. Enter a start frame number in the field on the
right. This specifies the frame number given to the first frame in the sequence. The numbering of
the rest of the frames is offset accordingly.
For example, if your sequence begins from frame 500 and you enter 1 in the field, frame 500 is
written out as image.0001.rgb, frame 501 as image.0002.rgb, and so on. Similarly, if you enter 100
in the field, frame 500 is written out as image.0100.rgb.
Offsetting All Frame Numbers by a Constant Value

3. From the frame dropdown menu, select offset. Enter a constant offset in the field on the right.
This constant value is added to the current frame to get the number for the frame that’s written
out.
USER GUIDE
1076
Changing the Numbering of Rendered Frames | Offsetting All Frame Numbers by a Constant Value
For example, if your clip begins from frame 500 and you want to render this first frame as
image.0001.rgb rather than image.0500.rgb, you can use -499 as the constant offset. This way, 499
is subtracted from the current frame to get the number for the frame that’s written out. At frame
500, image.0001.rgb is written out; at frame 501, image.0002 is written out, and so on.
USER GUIDE
1077
Using a Write Node to Read in the Rendered Image | To Use a Write Node to Read in the Rendered
Using a Write Node to Read in the

Rendered Image
You can use a Write node to both render an image and read the rendered image back in. Because
reading the output of a Write node from a file is faster than calculating its output by processing the
node tree upstream, this can be particularly useful on large comps. When you have finished working
on a branch of the node tree, you can insert a Write node after it, render the output, and use the
same Write node to read the rendered image in. If you later need to edit the nodes upstream, simply
make your changes and render the Write node again to update the image being read in.
To Use a Write Node to Read in the Rendered Image

1. Render an image as described in Output (Write) Nodes. We recommend rendering the image as
an .exr. This way, Nuke writes the hash value of the incoming node tree into the rendered file. If
the node tree changes and the rendered file gets out of date, the hashes won’t match and Nuke
notifies you of the problem.
2. In the Write node properties, check read file. When this is on, Nuke ignores the upstream node
tree and uses the rendered image as the output of the Write node.
3. To check whether the input file is up to date with the input tree connected to the Write node,
check check file matches input. This only works with .exr files written by Nuke and when the
proxy mode and down-rez are disabled. If the input file cannot be checked, Nuke displays the
word unchecked on the Write node in the Node Graph.
4. If there is an error when loading the rendered file, select what to do from the missing frames
dropdown menu:
• error - display an error message on any missing frames.
• black - replace any missing frames with black.
• checkerboard - replace any missing frames with a checkerboard image.
• read input - display the result of the input tree rather than the rendered file on any missing
frames.
Tip: You can also use the Precomp node (Other > Precomp) to reduce portions of the node
tree to pre-rendered image inputs. For more information, see Using the Precomp Node.
USER GUIDE
1078
Using a Write Node to Read in the Rendered Image | What is the Hash Value?
What is the Hash Value?

The hash value is a unique number (for example, b1c9c0aff2012a8) calculated from a node and the
entire tree of nodes connected to its input. The class of the node and all the current control settings
contribute to the hash value.
You can display the hash value at any point in the node tree by selecting a node in the Node Graph
and pressing I. The hash is different at different points in the tree.
USER GUIDE
1079
Render Farms | What is the Hash Value?
Render Farms
Nuke is supported by virtually all third-party and proprietary render-queuing software. By integrating
Nuke with such a system, the render load can be distributed across all the Nuke-licensed machines
on your network, whether Windows, Mac, or Linux-based.
Note: Instead of setting up a render farm, you can take advantage of the internal Frame
Server, which allows you to setup external slave machines to process renders faster. See
Using the Frame Server on External Machines for more information.
Your installation of Nuke may be configured to send jobs to a network render farm, which is usually
made available under the Render menu (i.e., Render > Render). However, because this option must
be customized for each studio, you should check with your system administrator for instructions on
how to send a Nuke script for network rendering.
Tip: If you’re attempting to force Nuke to retry an operation rather than failing, such as for
license issues during rendering, you may find Nuke’s exit codes helpful:
0 = success
1 = render error
100 = license failure
USER GUIDE
1080
Organizing Scripts
As scripts grow in complexity and are worked on by several people, it becomes increasingly important
to organize them in a clear and meaningful way. These pages teach you how to:
• display information about your Nuke script. See Displaying Script Information.
• find and replace all or part of file names or file paths in any node with file or proxy controls. See
File Name Search and Replace.
• group nodes in the Node Graph using the Backdrop node or the Group node. See Grouping Nodes
in the Node Graph.
• add notes to the Node Graph. See Adding Notes to the Node Graph.
• use the Precomp node to save a subset of the node tree as a separate .nk script, render the output
of this saved script, and read the rendered output back into the main comp as a single image input.
See Using the Precomp Node.
• use the LiveGroup node, which combines the power of groups and precomps with the utility of
gizmos. See Using the LiveGroup Node.
USER GUIDE
1081
Displaying Script Information |
Displaying Script Information

To display script information, such as the node count, channel count, cache usage, and whether the
script is in full-resolution or proxy mode, do the following:
1. Select File > Comp Info (or press Alt+I).
The script information window opens.
2. If you make changes to your script while the window is open, click update to update the
information.
3. To close the information window, click close.
USER GUIDE
1082
Using Visual Diagnostics | Enabling Visual Diagnostics
Using Visual Diagnostics

Nuke can display accurate script profile data onscreen or output it to .csv or .xml file to help you
troubleshoot bottlenecks in slow scripts. When visual diagnostics are calculating, timing information
is displayed in the Node Graph, and the nodes themselves are colored according to the proportion of
the total processing time spent in each one. The data is then displayed in the Profile tab as a bar or
pie chart, timeline, or as a table.
Enabling Visual Diagnostics

Nuke's visual diagnostics use Profile nodes in the node tree to gather data. You can add as many
Profile nodes as required, but one strategy might be to start at the foot of the Node Graph and work
your way up until CPU or memory load decreases significantly.
Note: Profile nodes add processing overhead, but the data gathered still provides an
accurate picture of where processing time is spent in the script.
1. Navigate to Image > Profile to add a Profile node to the Node Graph.
2. Connect the Profile node to the point in the node tree where you want to collect profiling data.
Note: You cannot connect Profile nodes directly to 3D, Particle, or Deep nodes. To profile
these nodes, you'll need to render their output through a render node, such as
ScanlineRender, RayRender, or a third-party renderer like PRManRender.
3. Click open profile panel in the node properties or click the content menu and select Windows
> Profile.
The Profile tab opens.
4. Click the profile node dropdown and then select the a Profile node in the Node Graph by name,
OR
Select the a Profile node in the Node Graph and then choose selected.
USER GUIDE
1083
5. Set the frame range to collect data from and the required data type using the checkboxes.
Tip: The initial frame range is determined by the Project Settings > frame range control
and all data types are collected by default. The frame range control accepts complex
ranges, such as 1-10, 15-25 and 1-200x10. See Defining Frame Ranges for more information.
6. Click profile to start data collection.
Note: If you've imported existing profiling data from a .csv or .xml file, you can't profile the
current script. See Exporting and Importing Profile Data for more information.
As the specified frame range is calculated, timing information is displayed and the nodes in the
Node Graph are colored according to the proportion of the total processing time spent in each
one.
USER GUIDE
1084
The profiling data is stored in the Profile tab's DATA panel when complete. See Filtering Profile Data
for information on how to control what profiling data is displayed.
USER GUIDE
1085
USER GUIDE
1086
Filtering Profile Data | Enabling Visual Diagnostics
Filtering Profile Data

The Profile tab's FILTER dropdown allows you to output only the profiling data you need, but the full
data type set selected in the PROFILE dropdown remains available if you change your mind.
The following controls determine what output is available on-screen and for export to .csv or .xml
file:
• minimum threshold - allows you to hide display data that doesn't reach a certain minimum value.
For example, setting the threshold to 15% can hide any nodes that don't use at least 15% of the
available memory.
0% minimum threshold. 15% minimum threshold.
Tip: You can zoom in and out of the DATA pane using the buttons.
• display data - determines what data is displayed from CPU, wall, ops, and memory. The default,
all, displays all the available profiling data:
• CPU - the time that the CPU spent executing the processing code, in microseconds, aggregated
over all CPU threads. For example, with multi-threaded processing this is typically much larger
than the wall time. If the average CPU time per thread (CPU divided by the number of threads
used) is much shorter than the wall time, this suggests the CPU threads have spent a lot of time
not executing code and perhaps waiting on locks, which could indicate a performance problem.
Note: On Mac and Windows, the CPU time is not currently accurate. On Mac, the CPU value
is always similar to the wall time.
USER GUIDE
1087
• wall - the time taken as it would be measured by a clock on the wall - the actual time you have
to wait for the processing to complete. The wall time is also measured in microseconds.
• ops - the number of operators called in the node. Operators are Nuke's building blocks that
perform certain tasks. Nodes can contain one or more ops. For example, when a node needs to
resize something it would use a Transform op rather than an implementation of its own to do
the same thing.
See https://learn.foundry.com/nuke/developers/110/ndkdevguide/intro/terminology.html#op for
more information.
• memory - the total amount of system memory used by the node.
Tip: Bear in mind that some node classes also contain numerical values to differentiate
between current and deprecated classes, such as Text and Text2.
• display nodes - for large scripts, it is impractical to display all the nodes in the Node Graph in the
Profile tab. In bar, pie and timeline mode, only the top 15 results by resource usage are listed. In
table mode, all results are listed.
The display nodes control allows you to be more selective with your output:
• all - displays data for all the nodes in the script.
• list - enter the explicit node names to display in the DATA panel. You can type the names
manually or select nodes in the Node Graph and then click add selected.
You can also find nodes listed in the FILTER panel by clicking select to highlight them in the
node tree.
With display nodes set to all. With display nodes set to a list.
You can display data by node class, such as CameraTracker, or explicitly by node name, such as
Read2.
USER GUIDE
1088
Tip: You can zoom in and out of the DATA pane using the buttons.
• display frames - allows you to view a sub-set of the profiled frames:

• average - the mean average of all the frames specified by the display frames control.
• per frame - allows you to step through all the frames in the display frames range using the
buttons.
• chart type - sets the form of the output produced from the profiling data:
• bar or pie chart, and timeline - displays the top 15 results by resource usage.
• table - displays all results in table form, a visual representation of the .csv or .xml file output.
USER GUIDE
1089
Exporting and Importing Profile Data | Exporting Profile Data
Exporting and Importing Profile

Data
You can write the onscreen profile data to .csv or .xml file for analysis or comparison using the
export data profile control on the Profile tab. You can also import existing profile data from .csv or
.xml files for analysis using the import data profile control on the Profile tab.
Exporting Profile Data

To export profile data to a file:
1. Click open profile panel in the Profile node's properties or click the content menu and select
Windows > Profile.
2. Enter a valid file path in the export profile data control and then click export.
Tip: You can also browse to the export location using the button.
The profile data is saved to the specified location as a .csv or .xml file.
USER GUIDE
1090
Exporting and Importing Profile Data | Importing Profile Data
Importing Profile Data

To import profile data from a file:
1. Click open profile panel in the Profile node's properties or click the content menu and select
Windows > Profile.
2. Enter a valid file path in the import profile data control and then click reload.
Tip: You can also browse to the location of the file using the button.
The profile data is imported from the specified location and displayed in the DATA panel.
USER GUIDE
1091
Using Performance Timing | Enabling Performance Timings
Using Performance Timing

Nuke can display accurate performance timing data onscreen or output it to XML file to help you
troubleshoot bottlenecks in slow scripts. When performance timing is enabled, timing information is
displayed in the Node Graph, and the nodes themselves are colored according to the proportion of
the total processing time spent in each one, from green (fast nodes) through to red (slow nodes).
Note: You can also access timing information for individual nodes using Python. See
https://learn.foundry.com/nuke/developers/113/pythondevguide/performance.html for more
information.
Enabling Performance Timings

You can enable profiling from the Script Editor or from the command line using the -P argument. You
can also output the timing data to an XML file by adding the -Pf argument and a file name, though the
output is only produced at render time.
Note: Enabling performance timings interactively or from the command line adds an extra
processing overhead, but the data still provides an accurate picture of where processing
time is spent in the script.
Outputting Performance Timings Onscreen

Enabling performance timings onscreen allows you to view cumulative data, as you work, on a per
node basis.
From the Script Editor, enter:

nuke.startPerformanceTimers()
From the command line, enter:

./Nuke<version>/Nuke<version>.exe -P
USER GUIDE
1092
Using Performance Timing | Outputting Performance Timings Onscreen
Note: You can reset the timing data using nuke.resetPerformanceTimers() from the
Script Editor.
When the performance timers are active, the following information is available onscreen:
• cpu - the time that the CPU spent executing the processing code, in microseconds, aggregated over
all CPU threads. For example, with multi-threaded processing this is typically much larger than the
wall time. If the average CPU time per thread (cpu divided by the number of threads used) is much
shorter than the wall time, this suggests the CPU threads have spent a lot of time not executing
code and perhaps waiting on locks, which could indicate a performance problem.
Note: On Mac and Windows, the CPU time is not currently accurate. On Mac, the cpu value
is always similar to the wall time.
• wall - the time taken as it would be measured by a clock on the wall - the actual time you have to
wait for the processing to complete. The wall time is also measured in microseconds.
• ops - the number of operators called in the node. Operators are Nuke's building blocks that
perform certain tasks. Nodes can contain one or more ops. For example, when a node needs to
resize something it would use a Transform op rather than an implementation of its own to do the
same thing.
See https://learn.foundry.com/nuke/developers/113/ndkdevguide/intro/terminology.html#op for

more information.
• memory - the total amount of system memory used by the node.
In addition to the timing information, nodes are color-coded according to their profiling, green
through red, where red is a slow node. You can see from the example script that Defocus is a slow
node, whereas Merge is doing no work at all.
USER GUIDE
1093
Using Performance Timing | Writing Performance Timings to File
Note: You can stop displaying timing data using nuke.stopPerformanceTimers() in the
Script Editor.
Writing Performance Timings to File

You can output the timing data to an XML file using the -Pf argument and a file name. The output is
only produced at render time, either from a Write node or from a command line render.
On Windows, for example, the following command writes profile data to file when you render from a
Write node in the Node Graph:
Nuke<version>\Nuke<version>.exe -Pf <file path and name>.xml
Rendering from the command line, the following command renders 10 frames from a script called
profileTest.nk and writes the profile data to an XML file:
Nuke<version>\Nuke<version>.exe -x -Pf C:\temp\profileTest.xml
C:\temp\profileTest.nk 1-10
USER GUIDE
1094
File Name Search and Replace | To Search for a File Name or File Path and Replace It
File Name Search and Replace

With the Search and Replace function, you can quickly replace all or part of file names or file paths in
any node with file or proxy controls (for example in Read and Write nodes).
To Search for a File Name or File Path and Replace It

1. Select the node(s) where you want to replace all or part of a file name or file path.
2. Select Edit > Node > Filename > Search and Replace.
OR
Press Ctrl+Shift+/ (Mac users press Cmd+Shift+/).
3. In the dialog that opens, enter the string you want to search for and the string you want to replace
it with. Click OK.
Nuke searches for the string in the selected nodes and replaces it with the new string.
Note: You can also enter expressions into the Search and Replace dialog. Just remember
that the search field in the dialog only takes regular expressions. Any characters that have
specific meanings in regular expressions, such as [ and ], need to be preceded by the \
character. For example, [getenv HOME] would need to be entered as \[getenv HOME\].
You can also pass flags alongside the expression itself to control how the expression
behaves. For example, to perform case-insensitive searches, you can enter (?i) in the
beginning of the expression or after one or more whitespace characters.
USER GUIDE
1095
Grouping Nodes in the Node Graph | Grouping Nodes with the Backdrop Node
Grouping Nodes in the Node

Graph
You can group nodes in the Node Graph using the Backdrop node or the Group node. The Backdrop
node adds a background box behind the nodes, separating the nodes visually from the rest of the
node tree. A Group node, instead, combines a set of nodes into a single node, acting as a nesting
container for those nodes.
Grouping Nodes with the Backdrop Node

You can use the Backdrop node to visually group nodes in the Node Graph. Inserting a Backdrop
node creates a box behind the nodes. When you move the box, all the nodes that overlap the box are
moved, too. By inserting several backdrop nodes, you can group the nodes in your node tree onto
boxes of different colors and titles. This makes it easier to find a particular node in a large node tree,
for example.
You can also use the Z Order control in the Properties panel to layer-up Backdrop nodes. Backdrops
with lower Z Order values appear underneath those with a higher value.
Nodes grouped with Backdrop nodes.
To Group Nodes with a Backdrop Node

1. Select Other > Backdrop. A Backdrop node box appears in the Node Graph.
USER GUIDE
1096
Grouping Nodes in the Node Graph | Grouping Nodes with the Backdrop Node
2. Drag the triangle in the lower right corner of the box to resize the box as necessary.
3. Click on the box title bar and drag it to move the box behind the nodes you want to group
together. If there are any nodes on the box, they move together with the box.
4. To change the color of the box, open the Backdrop node’s controls by double-clicking on the title
bar, then click the left color button and pick a new color with the color picker that appears.
5. To change the title of the box, enter a new name for the Backdrop node in the node’s controls.
6. To layer-up Backdrop nodes, enter a value in the Z Order control in the Properties panel.
Backdrops with lower Z Order values appear underneath those with a higher value.
USER GUIDE
1097
Grouping Nodes in the Node Graph | Grouping Nodes with the Group Node
7. If you later want to remove both the box and the nodes inside it, click on the title bar and press
Delete. Ctrl/Cmd+clicking on the title bar and pressing Delete only removes the box and leaves
the nodes untouched. To remove the nodes and leave the box untouched, click on the triangle in
the lower right corner of the box and press Delete.
Tip: You may want to use Backdrop nodes as jump-to points throughout a project. For more
information, see Navigating Inside the Node Graph.
Grouping Nodes with the Group Node

You can use the Group node to nest multiple nodes inside a single node.
To Group Nodes with a Group Node

1. Select all the nodes you want to nest inside the Group node.
2. If you want to replace the original nodes with the Group node, right-click and select Edit > Node >
Group > Collapse To Group (or press Ctrl/Cmd+G on the Node Graph).
If you want to keep the original nodes in the layout in addition to the Group node, right-click and
select Edit > Node > Group > Copy Nodes To Group (or press Ctrl/Cmd+Alt+Shift+G on the Node
Graph).
USER GUIDE
1098
The selected nodes are nested into a group. The internal structure of the Group node is shown on a
separate tab that opens.
Tip: As an alternative to Edit > Node > Group > Collapse to Group, you can also select
Other > Group from the Toolbar or the Node Graph right-click menu.
To View the Nodes Nested Inside a Group Node

In the Group node’s controls, click the S button (short for Show) in the top right corner.
You can also select the node and then press Ctrl/Cmd+Enter to open the Group.
A new tab that contains the nested nodes opens.
To Ungroup Nodes
1. Select the Group node in the Node Graph.
2. Select Edit > Node > Group > Expand Group (or press Ctrl/Cmd+Alt+G).
The Group node is replaced with the nodes that were nested inside it.
OR
1. In the Group node’s controls, click the S button in the top right corner.
USER GUIDE
1099
A new tab that contains the nested nodes opens.

2. Copy the nodes from the new tab into your script. If you want to lock the connections between the
grouped nodes so that they cannot be accidentally disconnected during the copy-paste operation,
check lock all connections in the Group node’s controls.
3. Delete the unnecessary Group node from your script.
USER GUIDE
1100
Adding Notes to the Node Graph | To Add a Note to the Node Graph
Adding Notes to the Node Graph

Using the StickyNote node, you can add notes to the Node Graph. The notes can be any text or HTML
mark-up. Usually, they are made as annotations to the elements in the node tree.
To Add a Note to the Node Graph

1. Click on the part of the Node Graph where you want to add a note.
2. Select Other > StickyNote. A note box appears in the Node Graph.
3. In the StickyNote controls, enter your note in the label field. If you like, you can use HTML mark-
up. For example,
• to have a note appear in bold, you can use <b>my note</b>. This would appear as my note.
• to have a note appear in italics, you can use <i>my note</i>. This would appear as my note.
• to add an icon to your note, you can use <img src="Colorwheel.png"/>. This adds the Nuke
color wheel icon. You can also use your own icons in the same way as long as you save them in
your plug-in path directory. Most common image formats work, but we recommend using .png
files.
USER GUIDE
1101
Using the Precomp Node | To Add a Note to the Node Graph
Using the Precomp Node

The Precomp node is like a Group node, but its content is stored in an independent .nk file. This
allows you to save a subset of the node tree as a separate .nk script, render the output of this saved
script, and read the rendered output back into the main comp as a single image input.
Note: Precomp nodes are unsupported in Nuke Assist. You cannot create a Precomp in
Nuke Assist; however, if one already exists in the script, you can view, but not modify its
output. To indicate this, Precomp nodes are outlined in red in the Node Graph, and their
controls are grayed out. See Nuke Products for more information.
Precomp nodes can be useful in at least two ways. Firstly, they can be used to reduce portions of the
node tree to pre-rendered image inputs. This speeds up render time, as Nuke only has to process the
single image input instead of all the nodes that were used to create it. Because the original nodes are
saved in a separate .nk script, you also maintain access to them and can adjust them later if
necessary.
Secondly, Precomp nodes enable a collaborative workflow. While one artist works on the main comp,
others can work on the sections that have been exported using the Precomp node. These sections
can be edited, versioned, and managed independent of the main comp. For example, say you have a
comp that involves a complex, multi-layered CG render. A 3D artist can produce this as a separate
script that the compositor finishing the shot then reads in using a Precomp node. This way, the 3D
artist can modify and re-render the CG element portion of the comp without having to share the main
comp with the compositor.
USER GUIDE
1102
Creating Precomp Nodes | To Create a Precomp Node
Creating Precomp Nodes

Whether you want to use the Precomp node to speed up rendering or to enable a collaborative
workflow, the process of creating the Precomp is the same.
To Create a Precomp Node

1. Select the nodes you want to include in the separate precomp script. If you select a Group node,
the nodes nested inside it are copied into the precomp script.
2. Select Other > Precomp (or press Ctrl/Cmd+Shift+P on the Node Graph).
The Precomp Nodes dialog opens.
Note: If you don't select any nodes when you create a Precomp node, the Precomp Nodes
dialog is not displayed and the node is left blank. Using the file parameter in the Precomp
controls, you can then browse to an existing script to load into the Precomp node.
If you have several Write nodes in the existing script and want to control which of them is
used for the output of the Precomp node, you can select Other > Output to insert an
Output node after the Write node you want to use.
If the Precomp node cannot find an Output node in the script, it looks for a Write node and
sets Output node in the Precomp controls to the name of that node. The Output node field
can also be used later to override what is set as the Output node in the precomped script. To
do so, make sure you check enable. This check box allows you to toggle between the output
node chosen by default and the output node you specified in the Output node field.
3. Click the file browser icon next to Precomp script path, and browse to the directory where
you want to save the precomp .nk script. After the directory path, enter a name for the precomp
script, for example Precomp1_v01.nk. By default, the precomp script is saved next to the main
script. If the main script has not been saved, the precomp script is saved in the current directory.
4. Click the file browser icon next to Precomp render path, and browse to the directory where
you want to save the rendered output of the precomped nodes. After the directory path, enter a
USER GUIDE
1103
Creating Precomp Nodes | About the Hash Value
name for the rendered image, for example Precomp1_####.exr. If you like, you can also use
version numbers in the name, for example Precomp1_v01_####.exr.
Warning: We recommend rendering the image as an .exr because that way Nuke can write
the hash value of the incoming node tree into the rendered file. If the precomp script
changes so that the hashes won’t match, Nuke can then notify you, and you can update the
resulting image. If you use a file format other than .exr, you do not get this notification and
the rendered file is likely to become out of date. For more information on hash values, see
About the Hash Value.
5. From the Channels dropdown menu, select the channels you want to include in the rendered
output of the precomped nodes.
If you later need to adjust this selection, you can do so by setting the channels on the appropriate
Write node (by default, Write1) in the precomp .nk script.
6. From the Original nodes dropdown menu, select:
• add backdrop to create a backdrop behind the precomped nodes in the Node Graph.
• delete to delete the precomped nodes.
• no change to do nothing to the precomped nodes.
7. Click OK.
Nuke saves the selected nodes in the .nk script specified. This script also includes input and
output nodes, a Write node, and the current project settings.
In the Node Graph, the selected nodes are replaced with the Precomp node. Nuke opens the
properties of the Precomp node.
About the Hash Value

The hash value is a unique number (for example, b1c9c0aff2012a8) calculated from a node and the
entire tree of nodes connected to its input. The class of the node and all the current control settings
contribute to the hash value.
You can display the hash value at any point in the node tree by selecting a node in the Node Graph
and pressing I. The hash can be different at different points in the tree.
Precomp Nodes and Project Settings

When you create a Precomp node, the precomp .nk script gets its project settings from the main
script. The main script’s project settings are also used whenever the precomp script is loaded into the
main script. Therefore, if you open the precomp script in a separate instance of Nuke and change its
USER GUIDE
1104
Creating Precomp Nodes | Precomp Nodes and Project Settings
project settings so that they no longer match the main script’s settings, your changes do NOT have an
effect when the precomp script is loaded into the main script. If you want to change the project
settings, you should always do so in the main script rather than the precomp script.
USER GUIDE
1105
Using a Precomp Node to Speed-up Rendering | To Render a Precomp Node
Using a Precomp Node to Speed-

up Rendering
If you want to use a Precomp node to speed up rendering, you need to have the Precomp node read
in the output of the precomp script.
To Render a Precomp Node

1. Create a Precomp node. For more information on how to do this, see Creating Precomp Nodes.
2. In the Precomp node controls, click Render. Enter a frame range (for example, 1-100) in the
dialog that opens and click OK.
Nuke renders the output of the precomp script and saves the resulting image in the Precomp
render path directory you specified when you created the Precomp node.
If you have activated the proxy mode and have specified a proxy file name in the output node, the
rendered image is a proxy image.
3. If the render finishes successfully, read file for output is automatically turned on in the Precomp
properties. When this is checked, the Precomp node turns green and reads in the rendered
precomp image rather than calculating the output of the precomp script. The word (Read) and
the name of the image read in are also displayed on the Precomp node in the Node Graph to
indicate this.
4. In case there is an error opening the rendered output of the precomped nodes, select what to do
from the missing frames dropdown menu in the Precomp properties:
• error - display an error message on any missing frames.
• black - replace any missing frames with black.
• checkerboard - replace any missing frames with a checkerboard image.
• read input - display the result of the input tree rather than the rendered file on any missing
frames.
When several Precomp nodes are used like this to replace sections in a large, complex node tree, the
render times may become significantly faster.
Note: When rendering the output of the precomp script, Nuke automatically selects the
Write node to use by first looking for an Output node. If it can't find any Output nodes, it
tries to find a Write node, and sets output node in the Precomp node controls to the name
of the Write node. If it can't find any Output or Write nodes, it produces an error.
USER GUIDE
1106
Using a Precomp Node to Speed-up Rendering | To Render a Precomp Node
At any point, the Precomp node properties can be used to override what is set as the Output
node in the precomped script. To do so, open the Precomp properties and the advanced
controls. In the output node field, enter the name of the Write node whose output you’d
like to use. Make sure you also check enable. This check box allows you to toggle between
the output node chosen by default and the output node you specified in the output node
field.
Tip: You can also use Write nodes in a similar manner and have their output read in from
the rendered file rather than calculated using the nodes upstream. For more information,
see Using a Write Node to Read in the Rendered Image.
USER GUIDE
1107
Precomp Revisions | To View and Edit a Precomp Script
Precomp Revisions
The following describes how to open, edit, and reload a precomp script.
To View and Edit a Precomp Script

1. In the Precomp node properties, click Open.
This starts a new Nuke session and loads the precomp script.
2. Edit the precomp script as necessary.
3. If you are using version numbers in the Write node file name, select the Write node in the Node
Graph and choose Edit > Node > Filename > VersionUp from the menu bar. This changes the
version number in the file that’s rendered, for example, from Precomp1_v01_####.exr to
Precomp1_v02_####.exr.
4. Render the Write node. Note that if the proxy mode is on and you have entered a proxy file name
for the output node, Nuke renders a proxy image.
5. Select File > Save New Comp Version (or File > Save Comp As) to save the script with a new
version number, for example, Precomp1_v02.nk instead of Precomp1_v01.nk.
To Reload a Revised Precomp Script

1. Load the main comp.
2. Make sure the filename in the Precomp node’s file field matches the current name of the
precomp script.
If the precomp script name has been versioned up, you can simply select the Precomp node and
choose Edit > Node > Filename > VersionUp from the menu bar. This changes the name of the
script that is read in, for example, from Precomp1_v01.nk to Precomp1_v02.nk.
If the precomp script has been saved with a different name or you want to use a different script as
the precomp script, edit the filename in the file field.
3. In the Precomp properties, click Reload.
USER GUIDE
1108
Collaborative Workflow Example | To Enable a Collaborative Workflow
Collaborative Workflow Example

In this example, a compositor is responsible for the main comp that is delivered to the client. A
lighting TD is responsible for providing a multi-layered CG render for use in the main comp. The
following describes how the compositor and lighting TD could use the Precomp node to enable a
collaborative workflow.
To Enable a Collaborative Workflow

1. The compositor starts building the main comp.
2. The lighting TD does a render of the CG element and a comp of the layers, saving the script as cg_
v01.nk.
Warning: Note that we recommend rendering precomp images as .exr files. This way, Nuke
writes the hash value of the incoming node tree into the rendered file. If the precomp script
changes so that the hashes won’t match, Nuke can notify the user who can then update the
resulting image. If you use a file format other than .exr, you don't get this notification and
the rendered file is likely to become out of date.
For more information on hash values, see Creating Precomp Nodes.
3. The compositor creates a Precomp node reading from the file cg_v01.nk, and continues working
on the main comp.
4. The lighting TD continues to revise the CG render and the comp, versioning up the Write node
and the precomp script .nk name. (See Precomp Revisions.)
When better results are achieved, the lighting TD notifies the compositor of the new and
improved version of the precomp script.
5. The compositor versions up the Precomp node to the new, improved version and reloads the
precomp script. (See Precomp Revisions.)
USER GUIDE
1109
Using the LiveGroup Node | To Enable a Collaborative Workflow
Using the LiveGroup Node

LiveGroups are a type of container node that can be used in conjunction with LiveInput nodes so that
multiple artists can work on different parts of the same shot as separate scripts, without the need for
rendering. LiveGroups also offer all the functionality of Precomps, Groups, and Gizmos combining all
the functionality that they lack individually.
Function Group/Gizmo Precomp LiveGroup
Open source file
Version up and down without recreating
Expose controls from within the node
View internal nodes in the same Nuke session
Accepts inputs in the same Nuke session
Just like Precomps, LiveGroups can store independent, open source .nk files, allowing you to save a
subset of the node tree as a separate .nk script, render the output of this saved script, and read the
rendered output back into the master comp as a single image input.
You can also use LiveGroups like Group nodes to nest multiple nodes inside a single node. The
original nodes are replaced with the LiveGroup node. When you create a LiveGroup node, its internal
structure is shown in a separate Node Graph tab.
USER GUIDE
1110
LiveGroups can also be used as shared toolsets, similar to Gizmos, but without having to export
them. You can import a LiveGroup script into any other Nuke script, as long as Project Settings > is
live group is enabled and the script contains an Output node.
USER GUIDE
1111
Collaborative Scripts
The LiveGroup node is different to Precomps and Groups in that it allows multiple artists to work on
the same script at the same time, drawing work from disparate scripts into the master script. A
simple example is shown below, where the Keying, CameraTrack, and VFX LiveGroups contain the
node trees for those operations in separate scripts.
In this example, the artists can continually improve the expression linked Camera and green-screen
key while the 3D artist works on the shot. See Collaborative Workflow Examples for more
information.
USER GUIDE
1112
Collaborative Workflow Examples | Creating a Master Script
Collaborative Workflow Examples

LiveGroups allow several artists to work on the same master script at the same time, without having
to render the separate components like Precomps. You can set up a master script before beginning
or start work on individual sessions and then import them into another script, depending on how you
like to work.
Creating a Master Script

As a Supervisor, you might want to set up a master script to hold the work of various artists working
on a shot. In this example, we'll assume that there are Keying, Tracking, and 3D artists working
simultaneously.
To create a master script:

1. Create a Nuke script with the required Project Settings and read in the required assets.
See Setting Up Your Script for more information.
2. Add the basic nodes required for the Keying artist and place them in a LiveGroup. See Creating
LiveGroups for more information on creating LiveGroups.
Tip: You can give the LiveGroup node a more descriptive name to make things easier to
read.
3. Publish the LiveGroup to a network location accessible by the Keying artist. See Editing and
Publishing LiveGroups for more information. Published LiveGroups are locked and cannot be
edited unless you click Make Editable in the script.
4. Repeat steps 2 and 3 for the Tracking and 3D artists. An example is shown in the image.
USER GUIDE
1113
Collaborative Workflow Examples | Creating a Master Script
5. As the sub-script work is completed, you can then add the required nodes to complete the master
script.
For example, you can expression link the Camera data from the CameraTracking LiveGroup to a
Camera in the master script. See Managing LiveGroup Controls for more information.
6. As the sub-script artists version up their scripts, you can version up your LiveGroups in the master
script to read in the latest work without the need to re-render in the master script. See Versioning
LiveGroup Scripts for more information.
USER GUIDE
1114
Collaborative Workflow Examples | Combining Existing Scripts
Combining Existing Scripts

As an artist, you might be working on a Nuke script in isolation that needs to be incorporated into a
master script at a later date. LiveGroups allow you to point to a sub-script in the same way as a
Precomp, but with the added functionality of LiveGroups.
To import a sub-script using a LiveGroup:

1. Open up the master script and add a LiveGroup node. The VFX LiveGroup referencing particles_
v02.nk in the example.
2. Double-click the LiveGroup to open its Properties panel.
3. Enter the location of the sub-script in the file control.
4. Click Reload.
This example differs from creating a master script first because the VFX LiveGroup is self-
contained, meaning that the assets are inside the LiveGroup rather than in the master script.
You can also work the other way, with LiveInputs referencing LiveGroups, but there are a few
caveats to working this way. See Referencing a Master Script Using LiveInputs for more
information.
USER GUIDE
1115
Creating LiveGroups | Importing an Existing Script into a LiveGroup
Creating LiveGroups
You can create empty LiveGroups to read in existing .nk scripts or select a set of nodes in the Node
Graph and create a LiveGroup to contain those nodes, with the option to write them out to a new .nk
script.
Note: If you create and publish a LiveGroup Pythonically, you need to add xpos and ypos
values to the nodes in the Node Graph to position them correctly before you publish the
LiveGroup. See Nuke's Python Developer's Guide for more information.
Importing an Existing Script into a LiveGroup

1. In Nuke's Toolbar, select Other > LiveGroup.
A new Node Graph tab opens showing the empty LiveGroup.
2. Enter the location of the .nk script in the file control or browse to its location and click Open.
A warning displays, informing you that there are differences between the local script and the
saved script.
3. Click Yes to overwrite the local script with the saved script.
The LiveGroup icon turns gray and the LiveGroup tab updates to show the contents of the
imported .nk script. The LiveGroup is locked as indicated by the padlock on the LiveGroup's tab
and it behaves like regular Nuke Precomp node.
USER GUIDE
1116
Creating LiveGroups | Adding Nodes to a LiveGroup
Note: Click Open in the LiveGroup's Properties to start a new Nuke session containing just
the LiveGroup's contents.
Adding Nodes to a LiveGroup

1. Select all the nodes you want to nest inside the LiveGroup.
• Press Tab in the Node Graph, type LiveGroup and press enter,
• Press Ctrl/Cmd+L, or
• Right-click and select Edit > Node > Group > Collapse To LiveGroup.
The LiveGroup icon turns yellow and a LiveGroup tab shows the contents of the node. The
LiveGroup is not locked and it behaves like regular Nuke Group node.
USER GUIDE
1117
Creating LiveGroups | Viewing Nodes Inside a LiveGroup
Viewing Nodes Inside a LiveGroup

In the LiveGroup's Properties panel, click the S button in the top-right corner.
A new Node Graph tab opens containing the nested nodes. LiveGroups contain a LiveInput and
Output node by default, which points to the master script so that artists can collaborate without
depending on Read nodes in the LiveGroup.
See Referencing a Master Script Using LiveInputs below for more information.
Referencing a Master Script Using LiveInputs

The LiveInput node can reference any LiveGroup within a script, eliminating the need to include the
required assets in the sub-script. The only requirements are that the sub-script and master script
have been saved and the LiveGroups you intend to connect to are pointing at the new sub-script.
1. Create a LiveInput node in the sub-script and then save the script.
2. In the master script, create a LiveGroup node and enter the sub-script name in the file control.
3. Save the master script.
USER GUIDE
1118
Creating LiveGroups | Referencing a Master Script Using LiveInputs
4. In the sub-script, enter the master script name in the file control and click Reload.
5. Select the LiveGroup you want to reference from the liveGroup dropdown.
USER GUIDE
1119
Managing LiveGroup Controls | Picking and Adding Controls
Managing LiveGroup Controls

You can expose controls from within your LiveGroup by picking existing knobs from nested nodes or
by adding a new control, and even expression link these controls so that updates in the LiveGroup
are passed to the master script.
Note: You cannot add or remove controls when a LiveGroup is locked. To unlock a
LiveGroup, open the Properties panel LiveGroup tab and click Make Editable.
Picking and Adding Controls

LiveGroups support the same method as gizmos for exposing controls in the Properties panel. Any
control from a nested node within the LiveGroup can be exposed or you can add your own.
In Nuke's Properties panel, you can add controls using the Edit button or by right-clicking in the
panel and selecting Adding Knobs Manually. Drag-and-drop knobs significantly reduce the time spent
managing user knobs in a node's Properties panel when compared to the legacy Manage User
Knobs method within Nuke.
To add knobs to a LiveGroup:

1. Double-click the LiveGroup to open its Properties panel.
2. Double-click the node containing the knobs you want to expose to open its Properties panel.
3. Click the edit button to enable drag-and-drop functionality.
4. Drag the target knob or knobs from the source node's Properties and drop them onto the
LiveGroup's Properties.
USER GUIDE
1120
Managing LiveGroup Controls | Expression Linking Controls
5. Click the edit button to disable drag-and-drop functionality when you've added the required
knobs.
You can also expression link User knobs just like any other knob in Nuke, but with LiveGroups, you
can expression link knobs from sub-scripts as well. See Expression Linking Controls for more
information.
Expression Linking Controls

LiveGroups support expression links between the master and sub-scripts, allowing you to easily
update exposed control values that you know are likely to change, such as Camera transforms.
To expression link a sub-script control:

1. Double-click the LiveGroup containing the controls you want to link and navigate to the User tab.
2. Double-click the node you want to link the controls to, in this case a Camera.
USER GUIDE
1121
Managing LiveGroup Controls | Expression Linking Controls
3. Hold Ctrl/Cmd and drag the animation button from the LiveGroup to the node's control
properties.
4. Repeat the process for all the controls you want to link from the sub-script.
The Node Graph updates to show a green arrow representing the expression link and the
direction of information flow.
Now, whenever the CameraTrack LiveGroup is reloaded, any updates to the camera translation
and rotation are automatically applied in the master script.
USER GUIDE
1122
Editing and Publishing LiveGroups | Editing LiveGroups
Editing and Publishing LiveGroups

The state of a LiveGroup in the Node Graph controls how the master script and sub-scripts behave.
LiveGroups have five states:
The Editable state indicates that the LiveGroup is unlocked, but no edits
have been made to the LiveGroup.
The Edited state indicates that the LiveGroup has been edited in the current
Nuke session.
The Edited with Overrides state indicates that the LiveGroup has been
edited in the current Nuke session and that the edits are overriding controls
in the saved version of the LiveGroup.
See Overriding LiveGroup Controls for more information.
The Published state indicates that the LiveGroup has been saved to the
location specified in the file control. LiveGroups in this state cannot be
edited, but you can still adjust any controls exposed in the Properties.
The Published with Overrides state indicates that the LiveGroup has been
saved to the location specified in the file control and that one or more
controls exposed in the Properties are overriding controls in the saved
version of the LiveGroup.
See Overriding LiveGroup Controls for more information.
Editing LiveGroups
The Editable and Edited states apply to new LiveGroups and LiveGroups that you have edited in the
current Nuke session.
Note: You cannot add or remove controls when a LiveGroup is locked. To unlock a
LiveGroup, open the Properties panel LiveGroup tab and click Make Editable.
USER GUIDE
1123
Editing and Publishing LiveGroups | Versioning LiveGroup Scripts
A LiveGroup in the Edited state always writes its contents into the current script on save. If its file
control points to a sub-script, no changes are made in the sub-script until the LiveGroup is Published
to that location.
To change the state of a published LiveGroup to Editable without editing it:

1. Double-click the LiveGroup node to open its Properties.
2. Click Make Editable.
The LiveGroup's state changes to Editable in the Node Graph and the contents of the group are
added into the current script.
Versioning LiveGroup Scripts

Versions allow you to record the different stages of your workflow as you progress, quickly swapping
the contents of LiveGroups in and out without overwriting existing work. You can publish different
versions of LiveGroups in the same way as Nuke scripts.
To load a new version of a LiveGroup:

1. Select the required LiveGroup in the Node Graph,
OR
Double-click the LiveGroup to open its Properties.
2. Press Alt+up arrow or Alt+down arrow to load the next version.
If no version of the current LiveGroup exists, Nuke displays a warning.
To save a new version of a LiveGroup:

1. Double-click the LiveGroup to open its Properties.
2. Enter the new version number in the file control,
USER GUIDE
1124
Editing and Publishing LiveGroups | Publishing LiveGroups
OR
Press Alt+up arrow to increment the existing version number.
Note: Nuke displays a warning that the new version doesn't exist.
3. Click Publish to save the LiveGroup to the new version.
Publishing LiveGroups
The Published state only applies to LiveGroups that you have published to the location specified in
the file control.
LiveGroups in the Published state never write their contents into the current script on save. To make
changes to a Published LiveGroup, click Make Editable. It becomes Editable until you publish it
again by clicking the Publish button.
Note: If you create and publish a LiveGroup Pythonically, you need to add xpos and ypos
values to the nodes in the Node Graph to position them correctly before you publish the
LiveGroup. See Nuke's Python Developer's Guide for more information.
To Publish a LiveGroup:
1. Double-click the LiveGroup node to open its Properties.
2. Check that the file control points to the required location and version up the file path if required.
3. Click Publish.
The LiveGroup's state changes to Published in the Node Graph and the sub-script specified in the
file control is updated.
USER GUIDE
1125
Editing and Publishing LiveGroups | Publishing LiveGroups
Note: When you Publish a LiveGroup, the resulting .nk script retains the project settings
from the master script. The master script’s Project Settings are also used whenever the sub-
script is loaded into the master script.
If you open the LiveGroup script in a separate instance of Nuke and change its Project
Settings so that they no longer match the master script’s settings, your changes DO NOT
apply when the LiveGroup script is loaded into the master script. If you want to change the
Project Settings, you should always do so in the master script rather than the sub-script.
USER GUIDE
1126
Overriding LiveGroup Controls | Publishing LiveGroups
Overriding LiveGroup Controls

Published LiveGroups cannot be edited, unless you click Make Editable in the LiveGroups
Properties, but the exposed controls can be overridden in scripts that include the LiveGroup.
Override controls are marked with a yellow chip in the Properties panel, so you can quickly see which
controls are affecting the output.
Note: Override controls are also indicated on the LiveGroup node in the Node Graph
In the example, the current color and channels controls are overrides. The double yellow chip
indicates that one of the controls on the top line is an override, in this case the current color control.
Click the yellow chip to list the overridden controls on that line.
USER GUIDE
1127
Overrides can be reverted or applied to the LiveGroup controls, if the LiveGroup is in the editable
state.
To revert overrides to the values stored in the LiveGroup:

1. Open the LiveGroup Properties panel by double-clicking the node.
2. Click the custom tab on which the override controls are stored. In the example, the User tab.
3. Right-click the panel, making sure you're not hovering over a control, and select Revert overrides
to LiveGroup values.
The controls revert to the values stored in the LiveGroup and the overrides are removed.
USER GUIDE
1128
To apply the overrides to the LiveGroup controls:

1. Open the LiveGroup Properties panel by double-clicking the node.
2. Click Make Editable to unlock the LiveGroup.
3. Click the custom tab on which the override controls are stored. In the example, the User tab.
4. Right-click the panel, making sure you're not hovering over a control, and select Apply override
values to LiveGroup.
The override values are applied to the LiveGroup's controls and the overrides are removed.
5. Click Publish to overwrite the LiveGroup on disk or save a new version of the LiveGroup.
USER GUIDE
1129
Configuring Nuke
These pages show visual effects supervisors how to configure Nuke for multiple artists, prior to the
start of a project. These are the common application settings discussed:
• Command line operations
• Environment variables
• Gizmo, NDK plug-in, and Tcl script directories

• Python script directories
• OFX plug-in directories
• Favorite directories
• Cross-platform file paths
• Menu and Toolbar options
• Image formats
• Gizmos (Nuke group nodes or subscripts that allow only select modifications)
• Custom plug-ins (binary plug-ins made via the Nuke software developers kit)
• Generic Tcl ("Tickle") scripts
• Template scripts
• Common preferences
• Script's lookup tables (LUTs)
• Custom Viewer Processes
Note: If you copy and paste Python example scripts from this user guide into a text editor,
line indentations may not be preserved. If this is the case, correct the indentations manually.
USER GUIDE
1130
What Is a Terminal and How Do I Use One? |
What Is a Terminal and How Do I

Use One?
Many tasks in this chapter tell you to enter commands from a terminal or shell. This refers to a
window where you can enter commands directly rather than making selections through a user
interface.
The following describes how to open such a window for your operating system.
• Linux: Click the right mouse button over the desktop and select New Terminal (or Open Terminal)
from the right-click menu.
• Windows: From the Start menu, select All Programs > Accessories > Command Prompt.
• Mac: Click on the Terminal dock icon.
OR
Browse to the Applications > Utilities folder on your system hard drive, and double-click the
Terminal icon.
Inside the terminal or shell, you’ll see a command prompt, which looks similar to this:
Terminal prompt on Linux and Mac Command prompt for Windows
Once you see the command prompt, you can enter commands to perform various tasks like listing
the files in a directory or running programs. Here are some specific examples:
• On Linux or Mac, type pwd and press Enter to view the path of the current directory. On Windows,
the equivalent command would be cd.
• On Linux or Mac, type ls and press Enter to view a list of files in the current directory. On Windows,
the equivalent command would be dir.
• On Linux, Mac, and Windows, type cd followed by a full path name and press Enter to change
directories.
USER GUIDE
1131
Command Line Operations | On Windows
Command Line Operations

Command line flags activate various options when you launch Nuke from a command line or
Terminal, and provide additional functionality to Nuke. First let’s discuss how to launch Nuke from a
shell.
On Windows
Note: If you're using a terminal emulator, follow the steps under Mac or Linux, substituting
the file paths as required.
Open a command line prompt and change directory as follows:

cd C:\Program Files\Nuke11.3v5\
To launch Nuke, type this command:

Nuke11.3.exe
Alternatively, you can set a doskey to point to Nuke and then you can launch Nuke from any
directory:
doskey nuke="C:\Program Files\Nuke11.3v5\Nuke11.3.exe"
If you want to doskey NukeX, enter:

doskey nukex="C:\Program Files\Nuke11.3v5\Nuke11.3.exe" --nukex
If you want to doskey Nuke Studio, enter:

doskey nukes="C:\Program Files\Nuke11.3v5\Nuke11.3.exe" --studio
On Mac
cd /Applications/Nuke11.3v5/Nuke11.3v5.app/Contents/MacOS/

./Nuke11.3v5
Alternatively, you can set an alias to point to Nuke and then you can launch Nuke from any directory.
The procedure for this depends on what your default shell is. To get the name of the shell you are
using, launch Terminal and enter echo $SHELL.
USER GUIDE
1132
Command Line Operations | On Linux
If you are using a bash shell, enter:

alias
nuke='/Applications/Nuke11.3v5/Nuke11.3v5.app/Contents/MacOS/Nuke11.3v5'
Alternatively, if you are using a tcsh shell, enter:

alias nuke /Applications/Nuke11.3v5/Nuke11.3v5.app/Contents/MacOS/Nuke11.3v5
If you want to alias NukeX, enter:

alias nukex /Applications/Nuke11.3v5/Nuke11.3v5.app/Contents/MacOS/Nuke11.3v5
--nukex
If you want to alias Nuke Studio, enter:

alias nukes /Applications/Nuke11.3v5/Nuke11.3v5.app/Contents/MacOS/Nuke11.3v5
--studio
Tip: You can add aliases to a .cshrc or .bashrc file in your home directory so that they are
activated each time you open a shell. See your Systems Administrator for help setting this
up.
On Linux
cd /usr/local/Nuke11.3v5/

./Nuke11.3
Alternatively, you can set an alias to point to Nuke and then you can launch Nuke from any directory.
The procedure for this depends on what your default shell is. To get the name of the shell you are
using, launch Terminal and enter echo $SHELL.
If you are using a bash shell, enter:

alias nuke='/usr/local/Nuke11.3v5/Nuke11.3'
Alternatively, if you are using a tcsh shell, enter:

alias nuke=/usr/local/Nuke11.3v5/Nuke11.3
If you want to alias NukeX, enter:

alias nukex=/usr/local/Nuke11.3v5/Nuke11.3 --nukex
If you want to alias Nuke Studio, enter:

alias nukes=/usr/local/Nuke11.3v5/Nuke11.3 --studio
USER GUIDE
1133
Command Line Operations | Using command line Flags
Tip: You can add aliases to a .cshrc or .bashrc file in your home directory so that they are
activated each time you open a shell. See your Systems Administrator for help setting this
up.
Using command line Flags

Now you can start experimenting with command line flags on launching Nuke. Here’s one that
displays the version number and build date.
nuke -version
If you have an .nk script, you can render it on the command line without opening the GUI version.
Here’s an example that renders a hundred frames of a Nuke script:
nuke -F 1-100 -x myscript.nk
Note how you can use the -F switch on the command line to indicate a frame range, separating the
starting and ending frames with a dash.
Note: We recommend that you use the -F switch whenever defining a frame range on the
command line, which must precede the script name argument.
However, for backwards compatibility, you can also use the old syntax. To do so, place the
frame range at the end of the command and use a comma to separate the starting and
ending frames. For example:
nuke -x myscript.nk 1,100
To display a list of command line flags (switches) available to you, use the following command:
nuke -help
Here’s that list of command line flags in a table:
Switch/Flag Action
-a Formats default to anamorphic.
-c size (k, M, or G) Limit the cache memory usage, where size equals a number in bytes. You
can specify a different unit by appending k (kilobytes), M (megabytes), or G
(gigabytes) after size.
--cont Nuke attempts to render subsequent frames in the specified range after
USER GUIDE
1134
Switch/Flag Action
an error. When --cont is not specified, rendering stops when an error is

encountered.
--crashhandling 1 Breakpad crash reporting allows you to submit crash dumps to Foundry in
the unlikely event of a crash. By default, crash reporting is enabled in GUI
--crashhandling 0 mode and disabled in terminal mode.
Use --crashhandling 1 to enable crash reporting in both GUI and terminal

mode.
Use --crashhandling 0 to disable crash reporting in both GUI and

terminal mode.
Tip: You can also use the NUKE_CRASH_HANDLING environment

variable to control crash handling. See Environment Variables for
more information.
-d <x server name> This allows Nuke to be viewed on one machine while run on another.
(Linux only and requires some setting up to allow remote access to the X
Server on the target machine).
-f Open Nuke script at full resolution. Scripts that have been saved
displaying proxy images can be opened to show the full resolution image
using this flag. See also -p.
-F Frame numbers to execute the script for. All -F arguments must precede
the script name argument. Here are some examples:
• -F 3 indicates frame 3.
• -F 1-10 indicates frames 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10.
• -F 1-10x2 indicates frames 1, 3, 5, 7, and 9.
You can also use multiple frame ranges:

nuke -F 1-5 -F 10 -F 30-50x2 -x myscript.nk
--gpu ARG When set, enables GPU usage in terminal mode with an optional GPU
index argument, which defaults to 0. Use --gpulist to display the selectable
GPUs.
USER GUIDE
1135
Switch/Flag Action
Note: Overrides the GPU set in Preferences > Performance >

Hardware when run in interactive mode.
--gpulist Prints the selectable GPUs and their corresponding index for use with the
--gpu ARG option.
-h Display command line help.
-help Display command line help.
-i Use an interactive (nuke_i) RLM license key. This flag is used in conjunction
with background rendering scripts using -x. By default -x uses a nuke_r
license key, but -xi background renders using a nuke_i license key.
Note: If you still use FLEXlm licenses and you're interested in

making a move to RLM licensing, please contact
sales@foundry.com to obtain a replacement license.
-l New read or write nodes have the colorspace set to linear rather than
default.
-m # Set the number of threads to the value specified by #.
--multigpu If have multiple GPUs of the same type installed, you can enable this
preference to share work between the available GPUs for extra processing
speed. This is a global preference and is applied to all GPU enabled nodes.
See Windows, Mac OS X and macOS, or Linux for more information on the
GPUs Nuke supports.
-n Open script without postage stamps on nodes.
--nocrashprompt When crash handling is enabled in GUI mode, submit crash reports
automatically without displaying a crash reporter dialog.
Tip: You can also use the NUKE_NO_CRASH_PROMPT

environment variable to control the crash prompt. See
USER GUIDE
1136
Switch/Flag Action
Environment Variables for more information.
--nukeassist Launch Nuke Assist, which is licensed as part of a NukeX Maintenance

package and is intended for use as a workstation for artists performing
painting, rotoscoping, and tracking. Two complimentary licenses are
included with every NukeX license.
-p Open Nuke script at proxy resolution. Scripts that have been saved
displaying full resolution images can be opened to show the proxy
resolution image using this flag. See also -f.
-P Measure your nodes’ performance metrics and show them in the Node
Graph. See Using Performance Timing for more information.
--pause Initial Viewers in the script specified on the command line should be
paused.
-Pf <filename> Measure your nodes’ performance metrics and write them to an XML file
at render time. See Using Performance Timing for more information.
--nc Runs Nuke in Nuke Non-commercial mode. See About Nuke Non-
commercial for more information.
--priority p Runs Nuke with a different priority, you can choose from:
• high (only available to the super user on Linux/OS X)
• medium
• low
--python-no-root- Prevents the application of knob defaults to the root node when executing
knobdefaults a Python script.
-q Quiet mode. This stops all printing to the shell.
--remap Allows you to remap file paths in order to easily share Nuke projects
across different operating systems. This is the command line equivalent of
setting the Path Remaps control in the Preferences dialog.
The --remap flag takes a comma-separated list of paths as an argument.

The paths are arranged in pairs where the first path of each pair maps to
USER GUIDE
1137
Switch/Flag Action
the second path of each pair. For example, if you use:
nuke -t --remap "X:/path,Y:,A:,B:/anotherpath"

• Any paths starting with X:/path are converted to start with Y:
• Any paths starting with A: are converted to start with B:/anotherpath
The --remap flag throws an error if:

• it is defined when starting GUI mode, that is, without -x or -t
• the paths do not pair up. For example, if you use:
nuke -t --remap "X:/path,Y:,A:"
A: does not map to anything, and an error is produced.
The --remap flag gives a warning (but does not error) if you give it no
paths. For example:
nuke -t --remap ""
Note: Note that the mappings are only applied to the Nuke
session that is being started. They do not affect the
Preferences.nk file used by the GUI.
-s # Sets the stack size per thread, in bytes. The default value is 16777216 (16
MB). The smallest allowed value is 1048576 (1 MB).
None of Nuke's default nodes require more than the default memory
stack value, but if you have written a custom node that requests a large
stack from the memory buffer, increasing the stack size can prevent stack
overflow errors.
--safe Running Nuke in safe mode stops the following loading at startup:
• Any scripts or plug-ins in ~/.nuke
• Any scripts or plug-ins in $NUKE_PATH or %NUKE_PATH%
• Any OFX plug-in (including FurnaceCore)
--sro Forces Nuke to obey the render order of Write nodes so that Read nodes
can use files created by earlier Write nodes.
USER GUIDE
1138
Switch/Flag Action
-t Terminal mode (without GUI). This allows you to enter Python commands
without launching the GUI. A >>> command prompt is displayed during
this mode. Enter quit() to exit this mode and return to the shell prompt.
This mode uses a nuke_r license key by default, but you can get it to use a
nuke_i key by using the -ti flag combo.
--tg Terminal Mode. This also starts a QApplication so that Pyside/PyQt can be
used. This mode uses an interactive license, and on Linux requires an X
Windows display session.
-V level Verbose mode. In the terminal, you’ll see explicit commands as each
action is performed in Nuke. Specify the level to print more in the
Terminal, select from:
• 0 (not verbose)
• 1 (outputs Nuke script load and save)
• 2 (outputs loading plug-ins, Python, Tcl, Nuke scripts, progress and
buffer reports)
-v This command displays an image file inside a Nuke Viewer. Here’s an

example:
nuke -v image.tif
--view v Only execute the specified views. For multiple views, use a comma
separated list:
left,right
--version Display the version information in the shell.
-x eXecute mode. Takes a Nuke script and renders all active Write nodes.
Note: Nuke Non-commercial is restricted to encrypted .nknc

scripts with -x from the command line, that is, using the syntax:
nuke -x myscript.nknc
This mode uses a nuke_r license key. To use a nuke_i license key, use -xi.
This is the syntax:
nuke -xi myscript.nk
USER GUIDE
1139
Command Line Operations | General syntax
Switch/Flag Action
On Windows, you can press Ctrl+Break to cancel a render without exiting

if a render is active, or exit if not. Ctrl/Cmd+C exits immediately.
On Mac and Linux, Ctrl/Cmd+C always exits.

-X node Render only the Write node specified by node.
-- End switches, allowing script to start with a dash or be just - to read from
stdin
General syntax
This is the general syntax for using these options when launching Nuke at the command prompt:
nuke <switches> <script> <argv> <ranges>
<switches> - modifies the behavior of Nuke when run from the command line. A list of switches is
given in the table above. These are sometimes called flags.
<script> - the name of the Nuke script.
<argv> - an optional argument that can be used in Nuke. See the example below.
<ranges> - this is the frame range you want rendering.
Examples
Let’s consider some practical examples.
To launch Nuke and open a script.

nuke myscript.nk
Crazy I know, but I’ve called my script, -myscript.nk, and the hyphen at the start of the filename has
confused Nuke. To get round this if you don’t want to rename your file use the double hyphen syntax:
nuke -- -myscript.nk
USER GUIDE
1140
Command Line Operations | Examples
To display an image:
nuke -v polarbear.tif
To display several images:

nuke -v polarbear.tif whiteflower.psd mountains.cin
To display an image sequence (taxi.0001.tif, taxi.0002.tif,...,taxi.0050.tif):

nuke -v taxi.####.tif 1-50
To render frame 5 of a Nuke script:

nuke -F 5 -x myscript.nk
To render frames 30 to 50 of a Nuke script:

nuke -F 30-50 -x myscript.nk
To render two frame ranges, 10-20 and 34-60, of a Nuke script:

nuke -F 10-20 -F 34-60 -x myscript.nk
To render every tenth frame of a 50 frame sequence of a Nuke script:

nuke -F 1-50x10 -x myscript.nk
This renders frames 1, 11, 21, 31, 41.
In a script with two write nodes called WriteBlur and WriteInvert this command just renders frames 1
to 20 from the WriteBlur node:
nuke -X WriteBlur myscript.nk 1-20
Using [argv 0]
Let’s use [argv] to vary the output file name. Launch the GUI version of Nuke and create a node tree
that puts a checker into a Write node. Open the write node property panel by double clicking on it
and in the file text field enter this filename:
[argv 0].####.tif
Save the script and quit Nuke. On the command line type:
nuke -x myscript.nk mychecker 1-5
This renders 5 frames (mychecker.0001.tif, mychecker.0002.tif, etc.).
You can add another variable to control the output image file type. The file text field needs this:
[argv 0].####.[argv 1]
and then render the script using this command:
USER GUIDE
1141
Command Line Operations | Examples
nuke -x myscript.nk mychecker cin 1-5
to get mychecker.0001.cin, mychecker.0002.cin, etc.
The <argv> string can be any [argv n] expression to provide variable arguments to the script. These
must be placed between the <script> and the <ranges> on the command line. You can include multiple
expressions, but each must begin with a non-numeric character to avoid confusion with the frame
range control. For more information on expressions, see Expressions.
Using Python to convert TIFFs to JPEGs

This command line method converts 5 TIFF frames to JPEG.
nuke -t
>>> r = nuke.nodes.Read(file = ”myimage.####.tif”)
>>> w = nuke.nodes.Write(file = ”myimage.####.jpg”)
>>> w.setInput( 0, r )
>>> nuke.execute(“Write1”, 1,5)
>>> quit()
It’s a bit tedious typing these commands in line by line. So let’s put them in a text file called
imageconvert.py and get Nuke to execute the Python script.
cat imageconvert.py
r = nuke.nodes.Read(file = ”myimage.####.tif”)
w = nuke.nodes.Write(file = ”myimage.####.jpg”)
w.setInput( 0, r )
nuke.execute(“Write1”, 1,5)
nuke -t < imageconvert.py
You can also pass in the Python script as a command line parameter. Doing this allows you to enter
additional parameters after the script name to pass into your script. When you do so, note that
sys.argv[0] is the name of the Python script being executed, and argv[1:] are the other parameters you
passed in. One example of this is below. See the standard Python module optparse for other ways to
parse parameters.
cat imageconvertwithargs.py
import sys
r = nuke.nodes.Read(file = sys.argv[1])
w = nuke.nodes.Write(file = sys.argv[2])
w.setInput(0, r)
nuke.execute("Write1", 1, 5)
nuke -t imageconvertwithargs.py myimage.####.tif myimage.####.jpg
USER GUIDE
1142
Environment Variables | Setting Environment Variables
Environment Variables
Environment variables are named variables used to store a value, such as a specific file path. They can
be used to influence Nuke’s behavior. For example, Nuke uses the information stored in them to
define where to place certain files.
Setting Environment Variables

This section teaches you how to set environment variables, check if a particular environment variable
exists, and display a list of set environment variables.
On Windows
1. Right-click on My Computer and select Properties.
2. Go to the Advanced tab.
3. Click the Environment Variables button. The Environment Variables dialog opens.
4. Click the New button under either User variables or System variables, depending on whether
you want to set the variable for the current user or all users. To set environment variables for all
users, you need to have administrator privileges.
5. In the Variable name field, enter the name of the environment variable you want to set. For a list
of the environment variables that Nuke understands, see Nuke Environment Variables.
6. In the Variable value field, enter the value for the variable. The value can be a directory path, for
example.
7. Click OK.
Note: When editing existing system variables, or adding or deleting either user or system
variables, you may need to log off and on again before your changes to environment
variables take effect.
On Mac
On Mac, you can use the launchd.conf file to set environment variables. You may need to create the
launchd.conf file if it doesn’t already exist in the /etc/ directory.
USER GUIDE
1143
Environment Variables | Setting Environment Variables
Note: If you only need to set an environment variable for a single session, or you don't want
to use the Mac launchd.conf file, you can also set variables using the method described in
the On Linux section.
Environment variables set using the launchd.conf file are read both when Nuke is launched from the
Nuke icon and when it’s launched from the Terminal.
1. Open a Terminal window.
2. Create the /etc/launchd.conf file, if it doesn't already exist, and then add the environment
variable(s) and value(s) to the file using the following format:
setenv <VARIABLE> <VALUE>
For example, to set two environment variables, NUKE_PATH and OFX_PLUGIN_PATH, to point to
alternate locations:
setenv NUKE_PATH /SharedDisk/Nuke/
setenv OFX_PLUGIN_PATH /SharedDisk/OFX
For a list of the environment variables that Nuke understands, see Nuke Environment Variables.
Tip: A handy command line tool for creating and editing files in the Terminal is nano. To
start nano with the correct permissions, enter:
sudo nano /etc/launchd.conf
3. To force the OS to read the launchd.conf file at startup, enter:

launchctl < /etc/launchd.conf; sudo launchctl < /etc/launchd.conf
4. Restart your Mac to apply the changes.
On Linux
1. The procedure for setting an environment variable depends on what your default shell is. To get
the name of the shell you are using, launch a shell and enter echo $SHELL.
2. Depending on the output of the previous step, do one of the following:
• If your shell is a csh or tcsh shell, add the following command to the .cshrc or .tcshrc file in your
home directory: setenv VARIABLE value. Replace VARIABLE with the name of the environment
variable and value with the value you want to give it, for example setenv NUKE_PATH
/SharedDisk/Nuke.
• If your shell is a bash or ksh shell, add the following command to the .bashrc or .kshrc file in
your home directory: export VARIABLE=value. Replace VARIABLE with the name of the
environment variable and value with the value you want to give it, for example export NUKE_
PATH=/SharedDisk/Nuke.
USER GUIDE
1144
Environment Variables | To Check if an Environment Variable Exists
For a list of the environment variables that Nuke understands, see Nuke Environment Variables.
To Check if an Environment Variable Exists
From Inside Nuke

• Press X in the Node Graph, check that TCL is enabled, and enter:
getenv <VARIABLE>
OR
• Open the Script Editor and enter:
import os
print os.environ["VARIABLE"]
In both cases, VARIABLE should be replaced by the environment variable you're interested in. For
example, NUKE_TEMP_DIR on Windows returns:
C:/Users/<current_user>/AppData/Local/Temp/nuke
In the Windows Environment

1. Select Start > All Programs > Accessories > Command Prompt.
2. In the command window that opens, enter echo %VARIABLE%. Replace VARIABLE with the name
of the environment variable. For example, to check if NUKE_DISK_CACHE is set, enter echo
%NUKE_DISK_CACHE%.
If the variable is set, its value is displayed in the command window.
In the Mac or Linux Environment

1. Launch Terminal or a shell.
2. Enter echo $VARIABLE. Replace VARIABLE with the name of the environment variable. For
example, to check if NUKE_DISK_CACHE is set, enter echo $NUKE_DISK_CACHE.
If the variable is set, its value is displayed in the Terminal or shell window.
To Display a List of Set Environment Variables
On Windows
1. Select Start > All Programs > Accessories > Command Prompt.
USER GUIDE
1145
Environment Variables | To Display a List of Set Environment Variables
2. In the command window that opens, enter set.
A list of all the environment variables that are set is displayed in the command window.
On Mac or Linux
1. Launch Terminal or a shell.
2. Enter printenv.
A list of all the environment variables that are set is displayed in the Terminal or shell window.
USER GUIDE
1146
Nuke Environment Variables | To Display a List of Set Environment Variables
Nuke Environment Variables

The following table lists the environment variables Nuke recognizes.
Environment Variable Description
FN_CRASH_DUMP_PATH Allows you to specify where Issue Reporter dumps are saved by default.
FN_DISABLE_LICENSE_ By default, if you have installed a temporary license, Nuke displays a

DIALOG dialog at start-up alerting you to the number of days remaining. If you
want to disable this behavior, you can set either of these environment
or variables to 1 to suppress the warning message about imminent license
FN_NUKE_DISABLE_ expiration.
TMPLIC_NOTIFY_DIALOG
Note: You still get a warning if no license is found, for
example if you only have a Nuke license but you try to run
NukeX.
FN_LICENSE_DIALOG_ By default, if you have installed a temporary license, Nuke displays a

DAYS_LEFT_BEFORE_ dialog at start-up alerting you to the number of days remaining. If you
PROMPT want to disable this behavior until a set number of days before expiry,
you can set this environment variables to the required number of days.
Note: You still get a warning if no license is found, for

example if you only have a Nuke license but you try to run
NukeX.
FN_NUKE_DISABLE_GPU_ This variable disables Nuke's CUDA and OpenCL capabilities. When
ACCELERATION enabled, any GPUs installed locally are disabled and cannot be selected
from Preferences > Performance > Hardware > default blink device
dropdown. Any GPU accelerated nodes, such as Kronos and Denoise,
default to processing on the CPU.
FN_SUBSCRIPTION_ On Windows, user names containing non-ASCII characters can cause

LICENSE_DIR subscription licensing to fail. If a licensing error similar to the following
displays:
Unable to create subscription license directory: C:\Users\Zoë
USER GUIDE
1147
Hernández\FoundryLicensing\
Try changing the license directory to an alternate location using this

environment variable.
foundry_LICENSE The location of the Nuke RLM license file, if the following
recommended location is not used:
On Mac and Linux:

/usr/local/foundry/RLM
On Windows:
drive letter:\Program Files\The Foundry\RLM

FOUNDRY_LICENSE_ This variable prints additional licensing information to the command

DEBUG line or Terminal.
FOUNDRY_LICENSE_FILE The location of the Nuke FLEXlm license file, if the following
recommended location is not used:
On Mac and Linux:

/usr/local/foundry/FLEXlm
On Windows:
drive letter:\Program Files\The Foundry\FLEXlm

FOUNDRY_LOG_FILE This variable specifies the location of Nuke Studio’s logfile. If you don’t
specify a logfile, all output is to screen.
FOUNDRY_LOG_LEVEL This variable sets the level of logging Nuke Studio produces during
USER GUIDE
1148
operation. There are four levels of detail, on a sliding scale from

minimal to verbose:
• error
• warning
• message
• verbose
Note: Setting the logging level to verbose can produce large

log files when FOUNDRY_LOG_FILE is specified.
FRAMESERVER_LOG_DIR This variable is used to specify a different location for the Frame Server
to write log files to, if you'd like to keep them separate from the default
NUKE_TEMP_DIR.
See Using the Frame Server on External Machines for more

information.
HIERO_DISABLE_ Set this variable to stop Nuke Studio loading thumbnails.

THUMBNAILS
HIERO_DISABLE_ Set this variable to stop Nuke Studio caching thumbnails for improved
THUMBNAILS_CACHE access once loaded.
Note: This variable does not clear the cache, you must
remove cached files manually.
USER GUIDE
1149
NUKE_AJA_CHANNEL AJA cards take 3G level signal (mostly for 12-bit 444 RGB) and combine it
into a single 3G-B (B denotes B level, hence the 3G-B) stream through
SDI1 by default. Use these environment variables to customize this
output behavior:
• NUKE_AJA_CHANNEL - set this variable to 2, 3, or 4 to output a single
stream through SDI2, SDI3, or SDI4.
• NUKE_AJA_DUALOUTPUT - set this environment variable to 1 to force
the card to separate the single 3G stream into two 1.5G streams
through SDI1 and SDI2.
Combining these two environment variables can force the stream to

NUKE_AJA_DUALOUTPUT split and output through alternate SDI outputs. For example:
• DUALOUTPUT + CHANNEL=1 OR CHANNEL=2 results in two 1.5G
streams coming from SDI1 and SDI2.
• DUALOUTPUT + CHANNEL=3 OR CHANNEL=4 results in two 1.5G
streams coming from SDI3 and SDI4.
Note: Certain modes, such as 12-bit 444, require a 3G stream.

Otherwise, the card uses the single stream on the channel
number specified.
NUKE_ALLOW_GIZMO_ Nuke does not allow you to Overwrite and Save as gizmos by default,
SAVING without copying the gizmo to a Group. Setting this environment
variable to 1 enables this behavior, so artists don't need to copy a
gizmo before editing it.
NUKE_CRASH_HANDLING Breakpad crash reporting allows you to submit crash dumps to

Foundry in the unlikely event of a crash. By default, crash reporting is
enabled in GUI mode and disabled in terminal mode.
When NUKE_CRASH_HANDLING is set to 1, crash reporting is enabled in

both GUI and terminal mode.
When NUKE_CRASH_HANDLING is set to 0, crash reporting is disabled

in both GUI and terminal mode.
NUKE_DEBUG_ When enabled, Comp Viewer image cache data is printed to the
USER GUIDE
1150
IMAGECACHE command line or Terminal. Information on disk space used, the

number of files cached, and the cache location is displayed.
NUKE_DEBUG_MEMORY When working on large images, Nuke may need to free up memory
during rendering. When this happens and NUKE_DEBUG_MEMORY is
set to 1, Nuke prints the following information to the console:
Memory: over maximum usage, trying to reduce usage from 1 GB to

924 MB.
If this variable is not set, you cannot see the debug memory messages.
Note that here, KB, MB, GB, and TB mean units of 1000. For example,
1MB means 1,000,000 bytes.
NUKE_DISK_CACHE The location where Nuke saves all recent images displayed in the
Viewer. Ideally, this should be a local disk with the fastest access time
available.
NUKE_DISK_CACHE_GB The maximum size the disk cache can reach (in gigabytes).
NUKE_EXR_TEMP_DIR On Linux, this is the location Nuke uses for temporary files while
reading PIZ-compressed .exr files. This environment variable is only
relevant on Linux.
If this variable is not set, the location is determined by NUKE_TEMP_

DIR.
NUKE_EXR_TEMP_NAME Changes the naming convention of .exr temporary files during

rendering.
Setting the variable to 1 writes temporary .exr files as

<filename>.exr.tmp, rather than <filehash>.tmp as in previous
releases.
NUKE_FONT_PATH The location that Nuke checks for available font files when the Text
node properties panel is opened.
NUKE_IGNORE_ROTO_ This variable disables the warning dialog displayed when you open
INCOMPATIBILITY scripts containing pre-Nuke 8 RotoPaint nodes.
NUKE_INTERACTIVE The import nuke function checks-out a nuke_r render license by
USER GUIDE
1151
default. If you want to use Nuke interactively, and you have an

interactive license, set this environment variable to 1.
See Nuke as a Python Module for more information.
NUKE_LEGACY_ This variable disables the new channel sorting behavior, where the
CHANNEL_SORTING RGBA layer is sorted first. Enabling this variable causes Nuke to sort
channels alphabetically.
NUKE_LOCALIZATION_ Controls the number of threads available for localization tasks.

NUMWATCHERS Increasing the number of threads can improve localization
performance.
NUKE_MOV64READER_ Set this variable to 0 to disable Nuke's 64-bit mov decoding and fall
ENABLE back to 32-bit decoding.
NUKE_NO_CRASH_ When crash handling is enabled in GUI mode, this allows you to control
PROMPT whether reports are automatically submitted or not:
When NUKE_NO_CRASH_PROMPT is set to 1, crash reports are

submitted automatically without displaying a crash reporter dialog.
When NUKE_NO_CRASH_PROMPT is set to 0, Nuke always displays a

crash reporter dialog before submitting a crash report.
NUKE_NO_VIEWER_GPU Disables Nuke's Comp Viewer OpenGL hardware acceleration.
NUKE_PATH The location where files related to Nuke customizations are stored. For
more information, see Loading Gizmos, NDK Plug-ins, and Python and
Tcl Scripts.
NUKE_TEMP_DIR The location where Nuke saves any temporary files that do not have a
particular place defined for them.
You can find the current location of Nuke's temporary directory from
within Nuke by pressing X on your keyboard, when the focus is on the
Node Graph, and then running the following TCL command:
getenv NUKE_TEMP_DIR
NUKE_ When enabled, data from Nuke's window manager is printed to the
WINDOWMANAGER_ command line or Terminal.
USER GUIDE
1152
DEBUG
OCIO Set this variable to the location of your OCIO configuration file for color
conversion.
Note: If you plan to use the OCIO config file specified in the
Preferences, ensure that the Preferences > Project Defaults
> Color Management > Export > use OCIO nodes when
exportingto a Comp checkbox is enabled.
OFX_PLUGIN_PATH The location where Nuke looks for OFX plug-ins. For more information,
see Loading OFX Plug-ins.
QT_COMPRESS_TABLET_ Due to recent updates to Qt, running Nuke on CentOS 7 Linux

EVENTS distributions with a tablet can cause lag when moving Roto shapes
around the Viewer. Setting this environment variable compresses tablet
events, eliminating the lag.
QT_PLUGIN_PATH The location where Nuke looks for custom Qt libraries if you don't want
to use those shipped with Nuke. Setting this environment variable adds
the custom path to Nuke's Qt library paths.
TIMELINE_DISABLE_PBO_ When enabled, the performance benefit from using pixel buffer objects
UPLOADS (PBO) for texture uploads from RAM to the GPU is disabled.
You can try disabling PBOs if you notice playback degradation.
USER GUIDE
1153
Loading Gizmos, NDK Plug-ins, and Python and Tcl Scripts | To Display a List of Set Environment
Loading Gizmos, NDK Plug-ins,

and Python and Tcl Scripts
On start-up, Nuke scans various directories for files that customize the behavior of Nuke. It looks for
information on favorite directories, menu options, image formats, gizmos, NDK plug-ins, Python
scripts, generic Tcl scripts, and preferences.
Warning: It’s worth saying that you should edit Python files with care as mistakes could stop
Nuke from running.
For more information on Python in Nuke, see The Script Editor and Python or the Python
Developer’s Guide (Help > Documentation).
To make your customizations available to all versions of a particular release, place them in the
following directories:
• Linux: /usr/local/Nuke/<version>/plugins/
• Mac: /Library/Application Support/Nuke/<version>/plugins/
• Windows: C:\Program Files\Common Files\Nuke\<version>\plugins\
If you want Nuke to look for plug-ins somewhere else rather than in these default locations, you can
also define a common plug-in path yourself. Thus, by defining the Nuke plug-in path, you can assign
yourself a common shared directory from which to control Nuke for multiple artists. See Defining the
Nuke Plug-in Path for more information.
Nuke also looks in specific sub-directories of your home directory and the Nuke application directory
in the order shown by platform:
Note: If you place customizations in the application directories, they'll only be available for
that release.
• Linux:
/usr/local/Nuke11.3v5/plugins
/usr/local/Nuke11.3v5/plugins/nukescripts
/home/login name/.nuke
• Mac:
USER GUIDE
1154
Loading Gizmos, NDK Plug-ins, and Python and Tcl Scripts | To Display a List of Set Environment
/Applications/Nuke11.3v5/Nuke11.3v5.app/Contents/MacOS/plugins
/Applications/Nuke11.3v5/plugins/nukescripts
/Users/login name/.nuke
• Windows:
drive letter:\Program Files\Nuke11.3v5\plugins\nukescripts or
drive letter:\Program Files\Nuke11.3v5\plugins
~\.nuke
Note: On Windows, the .nuke folder can be found under the directory pointed to by the
HOME environment variable. If this variable is not set (which is common), the .nuke directory
will be under the folder specified by the USERPROFILE environment variable - which is
generally of the form drive letter:\Documents and Settings\login name\ or drive
letter:\Users\login name\
pointing at, enter %HOME% or %USERPROFILE% into the address bar in Windows Explorer. If
the environment variable is set, the folder it’s pointing at is opened. If it’s not set, you get an
error.
In addition to setting up your Python/plugin environment using the init.py and menu.py files, Nuke
Studio also allows you to run Python code automatically on start-up. The default location for this
Python code is:
<STARTUP_PYTHON_PATH>/Python/Startup
Where by default, <STARTUP_PYTHON_PATH> is $HOME/.nuke as described in the OS breakdown

above.
Any Python .py modules or packages containing __init__.py found within the /Python/Startup directory
is imported when the application launches, and added to the hiero.plugins namespace.
USER GUIDE
1155
Defining the Nuke Plug-in Path | To Display a List of Set Environment Variables
Defining the Nuke Plug-in Path

To define the plug-in path:
1. On each artist’s machine, create an environment variable called NUKE_PATH.
2. Assign the NUKE_PATH environment variable to the path name of the directory where files related
to Nuke customization will reside.
For example, on Mac using a csh or tcsh shell:
setenv NUKE_PATH /SharedDisk/Nuke
Or, if you're using a bash or ksh shell:
export NUKE_PATH=/SharedDisk/Nuke
Loading a plug-in (plugin_find()) searches the NUKE_PATH until the first plug-in is located, ensuring
that only the most local plug-in is loaded. For example if the NUKE_PATH variable contains:
project_dir:studio_dir:company_dir
The path is searched in the following order until the first plug-in is located:
• ~/.nuke
• project_dir
• studio_dir
• company_dir
• nuke_dir
However, the NUKE_PATH environment variable is parsed in reverse order when loading init.py and
menu.py and all discovered copies are used. This allows localized settings to override more global
ones. So, with the directory hierarchy above, init.py scripts would execute as follows:
• nuke_dir/init.py
• company_dir/init.py
• studio_dir/init.py
• project_dir/init.py
• ~/.nuke/init.py
USER GUIDE
1156
Loading OFX Plug-ins | To Display a List of Set Environment Variables
Loading OFX Plug-ins

On start-up, Nuke scans various directories for OFX plug-ins that bring additional functionality to
Nuke. Paths to these directories vary between operating systems, but here are examples of where you
may find them:
• Linux:
/usr/OFX/Nuke
/usr/OFX/Plugins
• Windows:
C:\Program Files\Common Files\OFX\Nuke (or, when using 64-bit Nuke on 64-bit Windows,
\Program Files (x86)\Common Files\OFX\Nuke)
C:\Program Files\Common Files\OFX\Plugins (or, when using 64-bit Nuke on 64-bit Windows,
\Program Files (x86)\Common Files\OFX\Plugins)
• Mac:
/Library/OFX/Nuke
/Library/OFX/Plugins
If you want Nuke to look for OFX plug-ins somewhere else you can. Just define the environment
variable OFX_PLUGIN_PATH to point to the new directory.
For example, on Mac using a csh or tcsh shell:

setenv OFX_PLUGIN_PATH /SharedDisk/OFX
Or, if you're using a bash or ksh shell:

export OFX_PLUGIN_PATH=/SharedDisk/OFX
USER GUIDE
1157
Defining Common Favorite Directories | To Define a Common Set of Favorite Directories
Defining Common Favorite

Directories
Favorite directories can be accessed by artists with a single click from any Nuke file browser. Typically
you would create these favorites for common directories on a project.
To Define a Common Set of Favorite Directories

To define an common set of favorite directories:
1. Create a file called menu.py in your plug-in path directory.
For more information on plug-in path directories, see Loading Gizmos, NDK Plug-ins, and Python
and Tcl Scripts.
2. Add an entry in the following format:
nuke.addFavoriteDir('DisplayName', 'Pathname')
• Replace DisplayName with the string you want as the display name for the directory link, for
example ‘Home’ or ‘Desktop’.
• Replace Pathname with the path name to the directory to which the favorite should point.
3. In the above entry, you can also add the following optional arguments after 'Pathname':
• type. This is an integer and a bit-wise operator, OR a combination of nuke.IMAGE,
nuke.SCRIPT or nuke.FONT.
• nuke.IMAGE restricts the favorite directory to appear only in the image file browser, which
Nuke opens for file Read/Write operations.
• nuke.SCRIPT restricts the favorite directory to appear in the script file browser, which appears
when you select File > Open Comp or a similar menu selection to open or import script files.
• nuke.FONT restricts the favorite directory to appear in the font browser.
• icon='Name'. Replace Name with the name and file extension of the .png (or .xpm) image you
wish to use as the icon for the favorite directory. This image must be stored in your Nuke plug-in
path directory. It should be 24 x 24 pixels in size.
• tooltip='My tooltip'. Replace My tooltip with the string you wish to display as pop-up help.
Example 1
The following entry would create a favorite called DisasterFlickStore which appears on all
File Browsers invoked from Read nodes and which points to the /job/DisasterFlick/img directory.
USER GUIDE
1158
nuke.addFavoriteDir ('DisasterFlickStore', '/job/DisasterFlick/img',

nuke.IMAGE)
The result of example 1.
Example 2
The following entry would create a favorite called Test. It appears on all File Browsers invoked from
Read nodes or by selecting File > Open Comp and points to the /job/Test directory. The entry also
defines Test Images and Scripts as the tool tip for the favorite directory.
nuke.addFavoriteDir ('Test', '/job/Test', nuke.IMAGE | nuke.SCRIPT,
tooltip='Test images and Scripts')
USER GUIDE
1159
The result of example 2.
USER GUIDE
1160
Handling File Paths Cross Platform | To Define a Common Set of Favorite Directories
Handling File Paths Cross

Platform
If your facility uses Nuke on several operating systems, you may want to configure Nuke to replace
the beginnings of file paths so that scripts created on one platform also work on another.
For example, to ensure file paths created on Windows also work on Linux and vice versa, you can do
the following:
1. Create a file called init.py in your plug-in path directory if one doesn’t already exist.
and Tcl Scripts.
2. Open the init.py file in a text editor and add an entry in the following format:
import platform
def filenameFix(filename):
if platform.system() in ("Windows", "Microsoft"):
return filename.replace( "/SharedDisk/", "p:\\" )
return filename.replace( "p:\\", "/SharedDisk/" )
This way, the Windows file paths (beginning with p:\ in the above example) are replaced with the
Linux file paths (beginning with /SharedDisk/) under the hood whenever a Nuke script is used on
Linux. Otherwise, the Windows file paths are used.
Note that the file paths displayed in the graphical user interface (GUI) do not change. If you are
using p:\ in a node control, this is still displayed as p:\. However, on Linux, Nuke interprets p:\ as
/SharedDisk/.
USER GUIDE
1161
Setting Default Values for Controls | To Define a Common Set of Favorite Directories
Setting Default Values for

Controls
You can set default values for node controls by adding a simple line of Python to your init.py file.
Once a default value is set, all controls with matching names default to this value. For example, you
can set default values for file format specific controls in the Read, Write or other file format-
dependent nodes. To set a default value, use the following statement:
nuke.knobDefault()
To specify file format specific defaults, use the class name, followed by the file format extension and
the control name, all separated by periods. For example:
nuke.knobDefault("Read.exr.compression", "2")
Maybe you want the last frame value of the frame range controls in the Project Settings to default to
frame 200. To do this, use the following statement:
nuke.knobDefault("Root.last_frame", "200")
Or you might want to set a different default font style in the Text node:
nuke.knobDefault("Text2.font", "{ Arial : Regular : arial.ttf : 0 }")
USER GUIDE
1162
Defining Custom Menus and Toolbars | To Define a Common Set of Favorite Directories
Defining Custom Menus and

Toolbars
You can freely add custom menus and menu options as well as toolbars and toolbar options to the
Nuke interface. Artists can then use these options to trigger gizmos and plug-ins stored in the plug-in
path directory.
For example, to add a new menu in the default Toolbar with an option to trigger a gizmo called
MyGizmo, you can do the following:
1. In your home directory, create a directory called .nuke (if it doesn't exist already). For more
information on this directory, see Loading Gizmos, NDK Plug-ins, and Python and Tcl Scripts.
2. In the .nuke directory, create a file called menu.py if one does not already exist.
3. In a text editor, modify the file menu.py, adding the lines:
toolbar = nuke.toolbar("Nodes")
toolbar.addCommand( "Test/MyGizmo", "nuke.createNode('MyGizmo')")
This adds a menu labeled "Test" to the default Nodes Toolbar with an item labeled "MyGizmo"
that creates an instance of the node MyGizmo.
Note: Node Class() names occasionally change between major releases, such as Nuke 7 to
Nuke 8. While these changes do not affect legacy scripts, you may not get the results you
were expecting if a node class has been modified. The toolbars.py file, used to create Nuke's
node toolbar, contains all the current node class names and is located in <install_
directory>/plugins/nukescripts/ for reference.
As an example, between Nuke 7 and Nuke 8, the Text node Class() changed from Text to
Text2. In the toolbars.py file for the two releases, the entries for the Text node appear as
follows:
m.addCommand("Text", "nuke.createNode(\"Text\")", icon="Text.png")

m.addCommand("Text", "nuke.createNode(\"Text2\")", icon="Text.png")
It's also possible to add items to other menus in Nuke and even create your own toolbars. The
following sections cover these possibilities in detail.
USER GUIDE
1163
Defining Custom Menus and Toolbars | To Add a Toolbar
To Add a Toolbar
To add a toolbar:
1. Create a file called menu.py in your plug-in path directory if one doesn’t already exist.
and Tcl Scripts.
2. Open the menu.py file in a text editor and add an entry in the following format:
t=nuke.toolbar("ToolbarName")
t.addCommand("NewMenu", "PythonCode", "Shortcut", icon="IconName")
• Replace ToolbarName with the name you want to give to the toolbar. This name appears in the
content menus under Windows > Custom and above the toolbar on the title tab.
• Replace NewMenu with the name of the menu you want to add to the toolbar.
• Replace PythonCode with relevant Python code (usually nuke.createNode), and, if necessary,
use the name of the gizmo, generic Python script, or plug-in file you want the menu option to
invoke. For ease of use, place all such referenced files inside the plug-in path directory.
If you like, you can also replace PythonCode by a Python callable.
• Replace Shortcut with a keyboard shortcut, for example Alt+A, Ctrl/Cmd+A, or Shift+A. The
letter a alone represents lower-case a. F1 represents function key 1. You can combine the Shift,
Ctrl/Cmd, and Alt keys as necessary. If you like, you can also use #A to represent Alt+A, ^A to
represent Ctrl/Cmd+A, and +A to represent Shift+A.
• Replace IconName with the name of the .png (or .xpm) image you wish to use as the menu
icon. This image must be stored in your Nuke plug-in path directory. It should be 24 x 24 pixels
in size.
3. In the above entry, you can also add the following optional arguments in the parenthesis after
"ToolbarName":
• True. This is the default. When True, nuke.toolbar() calls the toolbar with the given name or
creates it if it does not exist. For example, t=nuke.toolbar("Extras", True) would either call an
existing toolbar called Extras or create one if it did not already exist.
• False. When False, the toolbar is not created if it does not already exist and nuke.toolbar()
returns None. You can use this to find out if a toolbar with a given name already exists. For
example, t=nuke.toolbar("Extras", False) would either call an existing toolbar called Extras or
return None if such a toolbar did not exist.
The new toolbar does not appear by default, but is listed under Custom in the content menus. From
there, you can insert it in any pane. Once you are happy with the new toolbar and its position, save
the layout (select Workspace > Save Workspace). Thereafter, the toolbar appears whenever Nuke is
launched in the saved workspace.
USER GUIDE
1164
Defining Custom Menus and Toolbars | To Define a Menu or Toolbar Option
You can build several toolbars for different tasks and save layouts that have one or another present
for easy context switching.
Example 1
The following entry creates a new toolbar called Extras. The toolbar includes an option called Create
VectorBlur that creates a VectorBlur node. The entry also defines v as the keyboard shortcut for the
VectorBlur node.
t=nuke.toolbar("Extras")
t.addCommand("Create VectorBlur", "nuke.createNode ('VectorBlur')", "v")
Example 2
In this example, we add an option called Autoplace to the toolbar created in example 1. This option
places the selected nodes neatly one after another, as illustrated in the following images:
Before using Autoplace After using Autoplace

to tidy up the Node Graph. to tidy up the Node Graph.
The following entry adds the Autoplace option. It also defines Alt+A as the keyboard shortcut for this
option.
def _autoplace():
n = nuke.selectedNodes()
for i in n:
nuke.autoplace(i)
t=nuke.toolbar("Extras")
t.addCommand("Auto&place", "_autoplace()", "Alt+a")
To Define a Menu or Toolbar Option

To define a menu or toolbar option:
USER GUIDE
1165
1. If you haven’t already done so, create a file called menu.py in your plug-in path directory. For
more information on plug-in path directories, see Loading Gizmos, NDK Plug-ins, and Python and
Tcl Scripts.
2. Open the menu.py file in a text editor and add an entry in the following format:
menubar=nuke.menu("MenuType")
m=menubar.addMenu("&NewMenu")
m.addCommand("&NewItem", "PythonCode", "Shortcut", icon="IconName",
index=#)
• Replace MenuType with the type of menu or toolbar you want to add an item to:
Nuke adds an item to the application main menu bar.
Animation adds an item to the menu on the Animation button of all panels, and to the
right-click menu of the Curve editor.
Properties adds an item to the right-click menus of properties panels.
Node Graph adds an item to the right-click menu of the Node Graph.
Nodes adds an item to the default Toolbar.
Viewer adds an item to the right-click menu of the Viewer.
Pane adds an item to the content menus where it appears under Custom.
• Replace NewMenu with the menu name. Using an existing menu name appends any new
options to the existing menu. You can also add options to the default Menu bar and Toolbar.
• Replace NewItem with the underlying item you want to add to the menu. You may precede any
character with an & in order to flag it as keyboard shortcut trigger.
• Replace PythonCode with relevant Python code (usually nuke.createNode) and, if necessary,
use the name of the gizmo, generic Python script, or plug-in file you want the menu option to
invoke. For ease of use, place all such referenced files inside the plug-in path directory.
For more information on plug-in path directories, see Loading Gizmos, NDK Plug-ins, and
If you like, you can also replace PythonCode by a Python callable. This has the advantage that
you get informed about errors in your script on start-up instead of when the menu item is
invoked. For an example of using a lambda function, see Example 3 on page 1168.
• Replace Shortcut with a keyboard shortcut, for example Alt+A, Ctrl/Cmd+A, or Shift+A. The
letter a alone represents lower-case a. F1 represents function key 1. You can combine the Shift,
Ctrl/Cmd, and Alt keys as necessary. If you like, you can also use #A to represent Alt+A, ^A to
represent Ctrl/Cmd+A, and +A to represent Shift+A.
USER GUIDE
1166
Note: By assigning a keyboard shortcut, you can overwrite existing shortcuts. For example,
if you assign the shortcut Ctrl/Cmd+O to a new menu item, it is no longer used for it’s
default purpose, which is opening a file. However, shortcuts are only overwritten in the main
menu bar, the Toolbar, any user-created toolbars, and the menu you are adding the new
menu item to. This means you can add a shortcut into the Node Graph, for example,
without resetting the same shortcut in the Viewer. However, you cannot add a shortcut into
the Node Graph without resetting the same shortcut in the main menu bar or the Toolbar.
• Replace IconName with the name of the .png (or .xpm) image you wish to use as the menu
icon. This image must be stored in your Nuke plug-in path directory. It should be 24 x 24 pixels
in size.
• Replace # with a number that represents the position of the item in the menu or toolbar. If you
do not use an index keyword, the item is added in the end of the menu or toolbar.
Tip: You can also put the menu name in the addCommand call, like this:
nuke.menu("MenuType").addCommand("NewMenu/NewItem", "PythonCode("name")")
Example 1
The following entry creates a new menu and option called Custom > Cue Render in the menu bar. It
inserts a gizmo called “cue_render.” The entry also defines Ctrl+R as the keyboard shortcut for the
gizmo.
menubar=nuke.menu("Nuke")
m=menubar.addMenu("&Custom")
m.addCommand("&Cue Render", "nuke.createNode('cue_render')", "Ctrl+R")
Example 2
For information on how to create a new menu in the default Toolbar with a menu item that triggers a
gizmo, see the example under Defining Custom Menus and Toolbars.
USER GUIDE
1167
Example 3
The following entry creates a menu and options called Custom > Filters > Blur in the menu bar.
Selecting Blur inserts the Blur node.
m.addCommand("Filters/Blur", "nuke.createNode(\"Blur\")" )
You can also do the same with a lambda function:

m.addCommand("Filters/Blur", lambda: nuke.createNode("Blur") )
This way, you don’t have to use the backslashes.
USER GUIDE
1168
Defining Common Image Formats | Defining an Image Format
Defining Common Image Formats

You may wish to define image formats (image resolutions and corresponding pixel aspect ratios) for a
given project. These appear as dropdown menus on Read and Reformat nodes.
Defining an Image Format

To define an image format:
1. If you haven’t already done so, create a file called menu.py in your plug-in path directory.
and Tcl Scripts.
2. Add an entry in the following format:
nuke.addFormat(" ImageWidth ImageHeight LowerLeftCorner LowerRightCorner
UpperRightCorner UpperLeftCorner PixelAspectRatio DisplayName ")
• Replace ImageWidthwith the width (in pixels) of the image format.
• Replace ImageHeightwith the height (in pixels) of the image format.
• If you wish to define an image area that is smaller than the format resolution, replace
LowerLeftCorner, LowerRightCorner, UppperRightCorner, UpperLeftCorner with the
proper x and y coordinates (in pixels) for the lower left, lower right, upper right, and upper left
corners of this smaller image area (optional).
• Replace DisplayNamewith display name of the format. This appears on all relevant dropdown
menus.
Example
The following entry would create a new format called full_aperature_anamorphic with a resolution
of 2048 by 1556 and a pixel aspect ratio of 2.0. (As the image corners are not explicitly defined, the
image covers the entirety of the format area.)
nuke.addFormat (" 2048 1556 2.0 full_aperture_anamorphic ")
USER GUIDE
1169
Creating and Accessing Gizmos | Creating Gizmos
Creating and Accessing Gizmos

Nuke enables artists and technical directors to create gizmos, which are groups of Nuke nodes that
may be reused by other artists. Studios commonly use gizmos to consistently apply certain color
grading techniques, process incoming footage according to a particular conversion formula, and
process outgoing footage in preparation for film printing.
A gizmo is a Group Node that you create and save in a separate .gizmo file in your Nuke plug-in
folder. Nuke scripts can use this gizmo just like any other node type. Saved scripts only contain the
name and control settings for the gizmo; the definition is in the gizmo file and it is read at the same
time the script is loaded into Nuke. Thus, you can alter the implementation of the gizmo and change
all the scripts that are using it.
Note: Unlike other nodes, gizmos cannot be cloned. For more information, see Working
with Nodes.
Tip: You can register gizmos as soft effects in Nuke's Timeline environment. See Soft Effects
Creating Gizmos
A gizmo is a Group Node that you create and save in a separate .gizmo file in your Nuke plug-in
folder. Nuke scripts can use this gizmo just like any other node type.
Using Nuke’s Export gizmo command, you can export nodes inside a Group and explicitly control
which controls may be edited by the artists to ensure the processes within the gizmo are consistently
applied.
To create a gizmo:
1. Select the nodes you want to include in the gizmo.
2. Select Other > Group from the Toolbar (or press Ctrl/Cmd+G) to group the nodes.
3. You may want to rename the group by entering a new name in the Group properties panel title
field. This step is optional, and has no effect on the saved gizmo. However, it is often a good idea
to give the Group the same name you intend to use for the gizmo.
USER GUIDE
1170
Creating and Accessing Gizmos | Creating Gizmos
4. To expose which controls the artists can adjust, see Managing Gizmo Controls.
5. Click the export as gizmo button.
6. In the file browser that appears, click Home. Type .nuke/ after the path displayed at the bottom
of the file browser.
7. Enter a name after the path, and append a .gizmo extension after the name. This is the name of
the command that is written to any saved script that’s using the gizmo. It’s a good idea, and
common practice, to begin the name with a capital letter, because Nuke uses this as an indication
that the command is a node or a gizmo.
8. Click Save.
See Accessing Gizmos in Nuke for information on how to access your gizmo in Nuke.
Note: Nuke does not allow you to Overwrite and Save as gizmos by default, without
copying the gizmo to a Group. If you want to allow this behavior, so artists don't need to
copy a gizmo before editing it, set the NUKE_ALLOW_GIZMO_SAVING environment variable
to 1. See Environment Variables for more information.
USER GUIDE
1171
Managing Gizmo Controls | Adding Knobs Using Drag-and-Drop
Managing Gizmo Controls

You can add controls to your gizmo by picking and editing from the default controls that exist for the
nodes inside your Group node or by adding a control you have created yourself to your gizmo
Properties panel.
In Nuke's Properties panel, you can add controls using the Adding Knobs Using Drag-and-Drop edit
button or by right-clicking in the panel and selecting Adding Knobs Manually. Drag-and-drop
knobs significantly reduce the time spent managing user knobs in a node's Properties panel when
compared to the legacy Manage User Knobs method within Nuke.
Adding Knobs Using Drag-and-Drop

You can drag-and-drop knobs between open node Properties panels or add your own using the knob
icons listed at the top of the panel. You can also order, hide, customize, and delete knobs within the
panel.
Note: The drag-and-drop interface is also useful for managing LiveGroup knobs. See Using
the LiveGroup Node for more information.
Tip: You can float the User Knob Editor by clicking the content menu at the top of the
pane and selecting Windows > User Knob Editor.
To add knobs to a Group:

1. Click the edit button at the top of the Group Properties panel.
USER GUIDE
1172
The User Knob Editor is displayed and the node's properties are now editable.
2. To add a knob, either:

Open the Properties panel of the node containing the knob(s) you want to expose and drag-and-
drop the knob(s) onto the Group node's properties,
USER GUIDE
1173
OR
Drag-and-drop a new knob from the User Knob Editor onto the Group node's properties.
USER GUIDE
1174
Knobs added from nodes in the Group are marked with a gray dot and new knobs are marked
with an orange dot.
See Organizing Knobs for information on how to rearrange knobs in the panel.
3. Click the edit button at the top of the Group Properties panel to exit edit mode.
See Accessing Gizmos in Nuke for information on how to access your gizmo in Nuke.
Editing Knobs
Knobs added to the Properties panel are edited by clicking the gray or orange dot next to the knob
label.
USER GUIDE
1175
The components you can edit depend on the knob type. For example, a translate knob from a node
within the Group allows you to edit the following:
• Name - the scripting name used to call the knob from scripts and expressions. This field is
alphanumeric and cannot contain spaces.
• Label - the display name for the knob in the Nuke interface.
• Tooltip - the information displayed when you hover the pointer over the knob.
Tip: The Nuke Reference Guide lists the labels and knob names for the majority of Nuke
nodes.
A more complex example is the Python Script Button knob from the User Knob Editor. Adding a
Python knob to the Properties panel allows you to edit an additional component, Script. The Script
field adds Python to the knob, which is executed when you click the button. For example:
USER GUIDE
1176
Click the edit button at the top of the Group Properties panel to exit edit mode. See Accessing
Gizmos in Nuke for information on how to access your gizmo in Nuke.
Organizing Knobs
You can re-order, hide, and remove knobs quickly using the drag-and-drop interface. When you're
dragging knobs around the Properties panel, an orange guide line indicates where the knob is placed
when dropped.
You can drag-and-drop multiple knob selections in the same way. Hold Ctrl/Cmd and click to select
multiple knobs in the Properties panel.
USER GUIDE
1177
You can add dividing lines between controls to delineate between groups of knobs, such as transform
and color knobs. To add a divider, drag-and-drop a Divider Line knob between the controls you want
to separate.
Adding Tabs
You can add tabs in the Properties panel to organize knobs into groups, just like Nuke's existing
nodes. To add a tab:
1. Drag a Tab knob from the User Knob Editor and drop it into the Properties.
Dropping a Tab knob above other knobs moves those knobs to the new tab.
USER GUIDE
1178
Tip: If you want to create an empty tab, drop the Tab knob at the bottom of the knob list.
2. Drag-and-drop knobs between tabs to organize them as required.
Hiding and Removing Knobs
You can hide knobs by clicking the icon next to the knob. Hiding a knob doesn't hide it in edit
mode, only when you've finished exposing knobs and clicked the edit button .
To remove a knob you can:

• Select the knob and press backspace or delete,
• Click the button next to the knob, or
• Drag the knob to the bottom-left of the Properties panel and drop it into the trash icon.
USER GUIDE
1179
USER GUIDE
1180
Adding Knobs Manually | Adding Knobs Using Drag-and-Drop
Adding Knobs Manually

You can manually add controls to your Group node before exporting it as a gizmo in two different
ways:
• by picking and editing a control from the default controls that exist for the nodes inside your Group
node. For example, if the Group node contains a Grade node, you can add any of the Grade node
controls to your gizmo Properties panel.
• by adding a control you have created yourself to your gizmo Properties panel.
To Pick Existing Controls

1. Right-click on the dark gray background of the Group properties panel and select Manage User
Knobs. The following dialog opens.
2. To pick a control you want to give the users control of, click on the Pick button. This opens a
dialog that lists all the nodes that the group contains.
USER GUIDE
1181
Expand the list items as necessary to see the controls you can include in the gizmo controls.
Select a control and click OK. You can also select multiple controls by Ctrl/Cmd+clicking on them,
or pick a range of controls by Shift+clicking.
At this point, a new tab called User appears in the Group properties panel. The control you
selected has been added to this tab. In our example, we selected the size control of the Text node.
The control has also been added to the Manage User Knobs dialog, ready for you to edit.
3. To edit the control you added, open the Manage User Knobs dialog and select the control from
the list. Click Edit.
In most cases, you can edit the following:

• Name - Give the new control a unique name here. You need to use this name whenever you
want to reference the control from scripts or via expressions. The name can only contain letters
and digits. Spaces or punctuation are not allowed. This field cannot be left empty.
• Label - Whatever you enter here appears to the left of the control in the gizmo properties panel
(or, in the case of buttons, on the button). If you leave this empty, whatever is in the Name field
is also used as the label.
In the Label fields of check boxes, Tcl script buttons, and Python script buttons, you can also
use HTML. For example, to have your text appear in bold, you can enter <b>text</b>.
USER GUIDE
1182
To add an icon to your check box or Tcl/Python button using HTML, you can enter <img
src="Colorwheel.png"/> in the Label field. This adds the Nuke color wheel icon. You can also
use your own icons in the same way as long as you save them in your plug-in path directory.
Most common image formats work, but we recommend using .png files.
Note that the HTML has been changed to a slightly non-standard form where newlines are
significant. If there is a newline character in your data, a new line is displayed in the label.
• Hide - Check this to hide the control from the users. This can be useful if you want to make a
new control to contain a complex expression that you can then refer to repeatedly by other
controls.
• Start new line - Uncheck this if you want the control to appear on the same line as the previous
control in the gizmo properties panel.
• Tooltip - Enter a short help text here. It appears, along with the Name, in a pop-up tool tip when
the user points the mouse at the control. If you do not provide a tool tip, whatever is in the
Name field is used as the tool tip.
4. If necessary, repeat the previous three steps to add more controls to your Group node (future
gizmo).
5. In the Group node properties panel, the controls are listed in the same order as they are in the
Manage User Knobs dialog. To move a control up or down in the properties panel, select it in the
dialog and use the Up and Down buttons.
9. Click Save.
To Create New Controls:

Knobs. The following dialog opens.
USER GUIDE
1183
2. To add a new control, tab, static text, or divider line to the Group (gizmo) controls, click Add on
the Manage User Knobs dialog and select the option you want to add. This opens a dialog where
you can edit the control, tab or static text you added. In most cases, you can edit the following:
• Name - Give the new control a unique name here. You need to use this name whenever you
want to reference the control from scripts or via expressions. The name can only contain letters
and digits. Spaces or punctuation are not allowed. This field cannot be left empty.
• Label - Whatever you enter here appears to the left of the control in the gizmo properties panel.
If you leave this empty, whatever is in the Name field is also used as the label.
In the Label fields of check boxes, Tcl script buttons, and Python script buttons, you can also
use HTML. For example, to have your text appear in bold, you can enter <b>text</b>.
To add an icon to your check box or Tcl/Python button using HTML, you can enter <img
src="Colorwheel.png"/> in the Label field. This adds the Nuke color wheel icon. You can also
use your own icons in the same way as long as you save them in your plug-in path directory.
Most common image formats can work, but we recommend using .png files.
Note that the HTML has been changed to a slightly non-standard form where newlines are
significant. If there is a newline character in your data, a new line is displayed in the label.
• Hide - Check this to hide the control from the users. This can be useful if you want to make a
new control to contain a complex expression that you can then refer to repeatedly by other
controls.
• Start new line - Uncheck this if you want the control to appear on the same line as the previous
control in the Group properties panel.
• Tooltip - Enter a short help text here. It appears, along with the Name, in a pop-up tool tip when
the user points the mouse at the control. If you do not provide a tool tip, whatever is in the
Name field is used as the tool tip.
3. Use an expression to link the control you just created to a node and its control inside the Group
node. This is important, because for the new control to do anything, you need to refer to it using
USER GUIDE
1184
an expression in some other control on a node inside the Group. For more information, see the
examples below and refer to Expressions.
4. If necessary, repeat the previous four steps to add more controls to your Group node (future
gizmo).
5. In the Group node properties panel, the controls are listed in the same order as they are in the
Manage User Knobs dialog. To move a control up or down in the properties panel, select it in the
dialog and use the Up and Down buttons.
9. Click Save.
To Delete Controls:
Knobs.
2. In the dialog that opens, select the controls that you want to delete from the list and click Delete.
3. To delete an entire tab, select all controls on the tab as well as the tab name and click Delete.
Examples
Below are some examples on how to create new controls for gizmos. To try them out, do the
following preparations:
1. Select Draw > Text and Draw > Rectangle. Create the following setup:
2. Double-click on the Rectangle1 node.
USER GUIDE
1185
3. In the Viewer, resize and reposition the rectangle until it looks like the following:
4. In the Rectangle1 properties panel, go to the Color tab. Click on the 4 button to display multiple
values rather than the slider. Enter 1 as the value for r, and 0 as the value for b, g and a. This
changes the color of the rectangle from white to red.
5. Copy the Rectangle1 node and paste it into the same script. Create the following connections:
6. Double-click on the Rectangle2 node and change the color of the rectangle from red to green (r 0,
g 1, b 0, a 0).
7. Select Merge > Switch to add a Switch node. Create the following connections:
USER GUIDE
1186
8. Select the Text1, Rectangle1, Rectangle2 and Switch1 nodes and press Ctrl/Cmd+G to group them.
This group is the gizmo we add controls to in the following examples.
9. Delete the original four nodes from the Node Graph tab.
10. Select the Group node and append a Viewer to it.
Example 1
In this example, we add a control called Version to the Group node controls. This control is an input
field. Whatever is entered in the field is called by the Text1 node and displayed in the Viewer when
you view the output of the group.
1. Open the Group properties panel and right-click on the dark gray background. Select Manage
User Knobs.
2. In the dialog that opens, select Add > Text input Knob to add a text input field control to your
Group properties panel.
3. Enter version as the Name for the control, Version as the Label, and Enter the version number here
as the Tooltip. Click OK and Done to close the dialogs.
This step created a tab called User in the Group node controls. All the controls you add or pick
are added on this tab by default.
As you can see, the Version control is now there.
4. On the Group1 Node Graph tab, double-click the Text1 node to open its controls. In the message
field, enter the following expression: [value version]. This expression calls the control named
version that you created in the previous step. Therefore, whatever is entered in the Version field
of the Group node (for example, v03), appears as a result of the Text1 node.
USER GUIDE
1187
Example 2
This example teaches you to create a checkbox control that the users can use to specify whether they
want to display or hide the version number added in the previous example.
1. In the Group properties panel, right-click on the dark gray background and select Manage User
Knobs.
2. In the dialog that opens, select Add > Check Box to add a checkbox control to your Group
properties panel.
3. Enter hideversion as the Name for the control, Hide version number as the Label, and Check this to
hide the version number as the Tooltip.
4. To have the new control appear next to the Version control (created in the previous example)
rather than below it on its own line, uncheck Start new line. Click OK and Done to close the
dialogs.
The control you created appears in the Group properties panel now.
5. In the Text1 controls, go to the Node tab. Right-click on the disable control and select Add
expression.
6. In the Expression field, enter hideversion (or, if you want to make it clear that the control is in
the enclosing group, you can also use parent.hideversion). This calls the control you created in
steps 2 and 3. Click OK.
USER GUIDE
1188
From now on, whenever Hide version number is checked in the Group controls, the Text1 node
is disabled and you cannot see the version number it would otherwise create.
Example 3
In this example, we add a control labeled Status to the Group controls. This control is a dropdown
menu with two options: Finished and Unfinished. When Finished is selected, the green rectangle is
displayed. When Unfinished is selected, you’ll see the red rectangle instead.
Knobs.
2. In the dialog that opens, select Add > Pulldown Choice to add a dropdown menu control to your
3. Enter status as the Name for the control and Status as the Label. In the Menu Items field, list the
items you want to appear in the dropdown menu - in this case, Finished and Unfinished.
USER GUIDE
1189
Finally, enter Select the production status here as the Tooltip. Click OK and Done to close the
dialogs.
The Status control should now appear in the Group controls.
4. On the Group1 Node Graph tab, double-click the Switch1 node to open its controls. Right-click
on the which field and select Add expression.
5. In the dialog that opens, enter the following expression: status==0 (or, parent.status==0). This
expression calls the control named status that you created earlier in this example. For the
dropdown menus, the first item is 0, the next 1, the next 2, and so on.
From now on, whenever Finished is selected under Status, the green rectangle is shown. When
Unfinished is selected, the red rectangle is shown.
Example 4
This example teaches you how to visually group and rearrange the controls you created for the Group
properties panel. You can do this by renaming the User tab, and using static text and divider lines to
group the controls on the tab.
First, we’ll rename the User tab in the Group properties panel:
Knobs.
2. In the dialog that opens, select User and click Edit.
3. In the Label field, enter a new name for the tab, for example, Version and status. Click OK and
Done to close the dialogs.
If you now look at the Group controls, you’ll notice that the User tab has been renamed to
Version and status.
Next, we’ll group the two version controls of the Group node under a title called Version controls:
Knobs.
2. In the dialog that opens, select Add > Text to add text to your Group properties panel.
3. Enter versioncont as the Name for the control and Version controls as the Label. Click OK and
Done to close the dialogs.
USER GUIDE
1190
This adds the text Version controls to the Group properties panel. However, the text does not
appear where we want it to appear: on top of the Version and Hide version number controls.
Let’s move it up.
4. Right-click on the Group properties panel again and select Manage User Knobs.
5. Select [Version controls] from the list and click Up three times. Click Done.
The text should now appear on top of the Group properties panel, above the version controls.
Finally, we’ll add a divider line between the version controls and the Status control:
Knobs again.
2. In the dialog that opens, select Add > Divider Line to add a line to divide the controls in your
3. Select the line from the Manage User Knobs dialog, where it is shown as unnamed.
4. Click the Up button once to move the line between the Hide version number and Status
controls. Click Done.
If you now open the Group controls, you’ll notice that a there’s a line between these controls.
5. See Accessing Gizmos in Nuke for information on how to access your gizmo in Nuke.
USER GUIDE
1191
Accessing Gizmos in Nuke | Adding Knobs Using Drag-and-Drop
Accessing Gizmos in Nuke

There are several ways to access your gizmos once you've saved them into your .nuke folder or
NUKE_PATH. See Loading Gizmos, NDK Plug-ins, and Python and Tcl Scripts for more information on
where Nuke looks for your gizmos.
If you're going to use a gizmo often, across multiple sessions, you can add it to a Nuke menu to
provide quick access. See Defining Custom Menus and Toolbars for more information.
Otherwise, you can quickly add it per-session by:

• pressing X on the Node Graph or Properties panel and entering the gizmo name (without the
extension) as a TCL command in the dialog that opens,
• opening the Script Editor and entering nuke.load('gizmo name') where gizmo name is the name of
the gizmo without the extension, or
• selecting Other > All plugins > Update from the node toolbar and then adding the gizmo using the
Tab menu in the Node Graph.
USER GUIDE
1192
Accessing Gizmos in Nuke | Adding Knobs Using Drag-and-Drop
USER GUIDE
1193
Sourcing Custom Plug-ins and Generic Tcl Scripts | Nuke Custom Plug-ins
Sourcing Custom Plug-ins and

Generic Tcl Scripts
Nuke Custom Plug-ins
The Nuke developer’s kit (NDK) allows developers to create and compile their own binary plug-ins.
To Source a Custom Plug-in

1. Place the plug-in file in the plug-in path directory. Its name should include a .dll (on Windows), .so
(on Linux) or .dylib (on Mac) extension.
and Tcl Scripts.
2. Create a menu option referencing the plug-in file (see Defining Custom Menus and Toolbars).
Or instruct artists to invoke the plug-in by opening the Script Editor and entering nuke.load
("plug-in name") where plug-in name stands for the name of the plug-in without the extension.
Sourcing Tcl Procedure

A Nuke script or gizmo is in fact a Tcl procedure (script). Thus, Nuke also allows you to hand code
generic Tcl procedures to automate Nuke in various ways.
To Source a Generic Tcl Procedure

1. Place the Tcl procedure file in the plug-in path directory. Its name should include a .tcl extension.
and Tcl Scripts.
2. Create a menu option referencing the plug-in file (see Defining Custom Menus and Toolbars).
Or instruct artists to invoke the Tcl script by opening the Script Editor and entering nuke.load
("procedure/script file name") where procedure/script file name stands for the name of the
procedure of script file without the extension.
Tip: For some code samples of useful Nuke Tcl procedures, look inside the [Nuke
directory]/plugins directory.
USER GUIDE
1194
Template Scripts | Creating and Using a Template Script
Template Scripts
You can create a template script that is loaded instead of an empty script every time you launch Nuke
or select File > New Comp or File > Close Comp. This allows you to save lookup table (LUT) setups
and favorite arrangements of nodes, for example.
Creating and Using a Template Script

To create and use a template script:
1. Create the script you want to use as a template.
2. Select File > Save Comp As. Navigate to ~/.nuke. The tilde (~) stands for your home directory and
the full stop (.) for a hidden folder.
3. Name your script template.nk and click Save.
The next time you launch Nuke or select File > New Comp or File > Close Comp, Nuke loads the
template from ~/.nuke/template.nk.
Tip: If you’re not sure of the location of your home directory, on Linux and Mac you can
open a terminal window and type echo $HOME. The terminal returns the pathname to your
home directory.
On Windows, you can find the .nuke directory under the directory pointed to by the HOME
environment variable. If this variable is not set (which is common), the .nuke directory will be
under the folder specified by the USERPROFILE environment variable. To find out if the
HOME and USERPROFILE environment variables are set and where they are pointing at, enter
%HOME% or %USERPROFILE% into the address bar in Windows Explorer. If the environment
variable is set, the folder it’s pointing at is opened. If it’s not set, you get an error.
Here are examples of what the path name may be on different platforms:
Linux: /home/login name
Mac: /Users/login name
Windows: drive letter:\Documents and Settings\login name or drive letter:\Users\login name
USER GUIDE
1195
Defining Common Preferences | Defining Default Preferences
Defining Common Preferences

The Nuke Preferences dialog (Edit > Preferences) allows any user to make myriad behavior and
display adjustments to the interface. However, you may wish to assign certain default preferences for
artists.
Defining Default Preferences

To define default preferences:
1. Select Edit > Preferences to display the Preferences dialog.
2. Modify the controls within the dialog as necessary. For descriptions of what the controls do, see
Appendix A: Preferences.
3. Click OK. Nuke writes the modified preferences to a file called preferences11.3.nk, which is
stored inside your [home directory]/.nuke directory.
Tip: If you’re not sure of the location of your home directory, on Linux and Mac you can go
to a terminal window and type echo $HOME. The terminal will return the path name to your
home directory.
On Windows, you can find the .nuke directory under the directory pointed to by the HOME
environment variable. If this variable is not set (which is common), the .nuke directory will be
under the folder specified by the USERPROFILE environment variable. To find out if the
HOME and USERPROFILE environment variables are set and where they are pointing at, enter
%HOME% or %USERPROFILE% into the address bar in Windows Explorer. If the environment
variable is set, the folder it’s pointing at is opened. If it’s not set, you get an error.
Here are examples of what the path name may be on different platforms:
Linux: /home/login name
Mac: /Users/login name
Windows: drive letter:\Documents and Settings\login name or drive letter:\Users\login name
4. Move the resulting preferences11.3.nk file into your Nuke plug-in path directory.
and Tcl Scripts.
Your preferences now act as the defaults for your artists. However, should they make changes using
the Preferences dialog, these changes will override your defaults.
USER GUIDE
1196
Defining Common Preferences | Deleting (and Resetting) the Preferences
Deleting (and Resetting) the Preferences

To delete (and reset) the preferences:
1. Open a terminal (or shell) as described for your operating system at the beginning of this chapter.
2. Using the prompt, go to the .nuke directory, under your home directory.
3. Enter pwd to display and verify the path.
You should see something similar to
• /users/login name/.nuke (on Linux),
• /Users/login name/.nuke (on Mac) or
• drive letter:\Documents and Settings\login name\.nuke or drive letter:\Users\login name\.nuke
This is not always the case, however, because on Windows the .nuke folder can be found under
the directory pointed to by the HOME environment variable or (if HOME is not set) the
USERPROFILE environment variable.
4. Enter rm preferences11.3.nk to delete the preference file.
5. Close the terminal or shell.
The next time you launch Nuke, it rebuilds the file with the default preferences.
USER GUIDE
1197
Altering a Script’s Lookup Tables (LUTs) | Deleting (and Resetting) the Preferences
Altering a Script’s Lookup Tables

(LUTs)
A script’s lookup tables are curves that control the conversion between file or device color spaces and
Nuke’s internal color space. In the Curve Editor, the x axis represents the input pixel values and the y
axis the output pixel values (normalized to the 0-1 range). When applying LUTs, Nuke looks up the
input value along the x axis to determine what the y value is to output.
Nuke provides many nuke-default LUTs, including: linear, sRGB, rec709, Cineon1 , Gamma1.8,
Gamma2.2, Panalog2 , REDLog3 , ViperLog4 , AlexaV3LogC5 , PLogLin6 , SLog7 , and REDSpace8 .
You can also create an unlimited number of additional LUTs and edit or remove existing LUTs in the
script’s settings.
By default, Nuke uses certain LUTs for certain file types or devices. In most cases, you do not need to
touch these defaults. However, there may occasionally be cases when changing the defaults is
necessary: for example, if your material has been shot with a camera that records in a custom color
space, such as Panalog. In those cases, you can change the defaults in the script’s settings so that you
don’t need to change the color space on each Read or Write node.
If you do not want to use the default LUT for reading or writing certain individual images, you can
select the LUT to use in the corresponding Read or Write node’s controls.
1 The Cineon conversion is implemented as defined in Kodak’s Cineon documentation.
2 The Panalog LUT is based on a log2lin conversion with a blackpoint of 64, whitepoint of 681, and a gamma of 0.89.
3 The REDLog LUT is based on a log2lin conversion with a blackpoint of 0, whitepoint of 1023, and a gamma of 1.022.
4 The ViperLog LUT is based on a log2lin conversion with a blackpoint of 0, whitepoint of 1023, and a gamma of 1.0.
5 The Alexa LogC LUT uses the formula provided by ARRI.
6 The PLogLin LUT uses the default values for the formula, mapping log 0.43457 (code value 445 on a 1024 scale) to linear 0.18 (mid-gray)
assuming a negative gamma of 0.6 and a density per code value of 0.002. (This does factor in the fact that value ranges for a log image in
Nuke are still scaled to 0-1 range.)
7 The Sony S-Log LUT comes from the Sony S-Log Whitepaper Version 1.110. For more information, see
http://www.sony.co.uk/res/attachment/file/66/1237476953066.pdf.
8 The REDSpace LUT is implemented as defined in a curve provided by RED.
USER GUIDE
1198
Managing Nuke's Native LUTs | To Display LUT Curves
Managing Nuke's Native LUTs

The Color tab in the Nuke Project Settings allows you to view, add, edit, and delete Nuke's native
LUTs.
Note: You can't adjust LUTs that appear under the OCIO color management scheme. See
OCIO Color Management for more information.
See this Support Knowledge Base article for more detailed information.
To Display LUT Curves

1. Select Edit > Project Settings to open the settings for the script.
2. Go to the Color tab.
3. From the list on the left, select the LUT you want to display in the curve editor. To select several
LUTs, press Ctrl (Mac users press Cmd) while selecting the LUTs. All the selected LUTs are shown
in the curve editor at the same time.
To Create a New LUT

USER GUIDE
1199
Managing Nuke's Native LUTs | To Edit LUTs
3. Click the plus button (+). A dialog opens.

4. Enter a name for the new LUT and click OK.
5. Adjust the lookup curve to suit your needs. Click on the curve to select it. Ctrl/Cmd+Alt+click to
add points on the curve, and drag the points to a new position. To change the shape of the curve,
adjust the tangent handles.
The new LUT is now available in the global LUT settings, and the colorspace dropdown menu of Read
and Write nodes’ properties panels, the Viewer controls.
To Edit LUTs
3. From the list on the left, select the LUT you want to edit.
4. Adjust the lookup curve to suit your needs. Click on the curve to select it. Ctrl/Cmd+Alt+click to
add points on the curve, and drag the points to a new position. To change the shape of the curve,
adjust the tangent handles.
To use the usual editing commands, such as copy and paste, right-click on the curve editor and
select Edit. Then, select the editing command you want to use, just like you would on any curve
editor.
Note: Renaming existing LUTs is currently not possible. If you want to rename a LUT, you
need to add and name a new LUT, copy the information from the old LUT into the new one,
and then remove the old LUT.
To Reset the LUT Curves Back to Their Initial Default

Shapes
3. From the list on the left, select the LUT you want to reset. To select several LUTs, press Ctrl/Cmd
while selecting the LUTs.
4. Click reset.
To Remove LUTs
USER GUIDE
1200
Managing Nuke's Native LUTs | To Remove LUTs

3. From the list on the left, select the LUT you want to remove. Only remove LUTs that you have, for
example, created by accident and are not using in your script. To remove the LUT, click the minus
button (-).
The LUT is removed from the LUT settings, and the colorspace dropdown menu of Read and Write
nodes’ properties panels.
Note: If you remove a LUT that is used in a node, the node continues to refer to the LUT by
name and raises an error.
USER GUIDE
1201
Selecting the LUT to Use | Default LUT Settings
Selecting the LUT to Use

To select the LUT to use when reading or writing an image:
1. Double-click to open the Read or Write node’s properties panel.
2. From the colorspace dropdown menu, select the LUT you want to use. To use the default LUT
defined in Nuke’s settings for the image type in question, select default.
Default LUT Settings

By default, Nuke uses the following LUTs in the following cases:
LUT File Type / Device Default

LUT
working space This determines what colorspace files should be converted to when linear
read and from when written- it's the colorspace used by Nuke under
the hood. In earlier releases of Nuke, this colorspace was hidden
because linear was always chosen as the working space. You may find
that some operations work better in colorspaces other than linear. For
example, some transforms work better in the CLog colorspace.
Note: You can only change the working space if you're using
OCIO color management.
monitor This is used for postage stamps, OpenGL textures, the color chooser sRGB
display, and all other non-Viewer image displays.
8-bit files This is used when reading or writing image files that contain 8-bit data. sRGB
Also used by the Merge node’s sRGB switch, and to convert Primatte
inputs into sRGB and outputs from sRGB.
16-bit files This is used when reading or writing image files that contain 16-bit sRGB
integer data (not half float).
log files This is used when reading or writing .cin or .dpx files. Cineon
float files This is used when reading or writing image files that contain floating- linear
point data.
USER GUIDE
1202
Selecting the LUT to Use | To Change the Default LUT Settings
To Change the Default LUT Settings

3. From the dropdown menus under Default LUT settings, select the LUTs you want to use by
default for each file type or device.
The new defaults are now used for any LUT setting where you have not selected a specific LUT. Any
controls you have set to a specific LUT (that is, not set to default) continues to use the selected LUT,
and only those set to default are affected.
USER GUIDE
1203
Example Cases | Working in Video Colorspace
Example Cases
Below are some examples of situations where you might need to alter the default LUTs.
Working in Video Colorspace

Emulating compositor software that works in video color space is not recommended. However, if you
do need to do so, do the following:
1. Select Edit > Project Settings and go to the Color tab.
2. Under Default LUT settings, change the monitor, 8-bit files, and 16-bit files values to linear.
This prevents Nuke from converting from sRGB into linear. Nuke’s nodes still assume linear data, but
the image processing is applied to your unlinearized video color space images.
Linear Data in 16-Bit Files

Some facilities use linear data in 16-bit files. If this is the case in your facility, do the following:
2. Under Default LUT settings, change the 16-bit files value to linear.
Cineon Displays
Some facilities have adjusted their monitor electronics to correctly display Cineon data. If this is the
case in your facility, do the following:
2. Under Default LUT settings, change the monitor value to Cineon.
Color Management
Although true color management requires using other nodes, it may be useful to approximate it with
a LUT that is used for the monitor setting. This way, texture maps and postage stamps resemble the
final display more accurately.
If your color management is creating a monitor-corrected image, you’ll want to set monitor to sRGB
so you get reasonably monitor-correct output on non-Viewer images.
USER GUIDE
1204
OCIO Color Management | Color Management
OCIO Color Management

Nuke uses OpenColorIO for color management. All of the colorspaces in Nuke, whether those
shipped with the application or custom colorspaces, are defined in OCIO config files.
Depending on the OCIO config file that you are working with, there are a number of colorspace
options and roles (aliases to colorspaces) that you can set in Nuke. There are also default options,
which change depending on what file type you are working with. When the default option is selected,
the colorspace that Nuke has set for it is listed in brackets.
Tip: Use the options in Preferences > Project Defaults > Color Management to apply
them to all new projects.
1. The color management dropdown determines whether Nuke uses the LUTs read from the
configuration specified or the Nuke native LUTs. Selecting OCIO makes the relevant OCIO LUTs
available to the Read and Write nodes in scripts on a per project basis.
All OCIO configurations except nuke-default automatically switch this control to OCIO.
2. Set the OpenColorIO Config you want to use for this project.
Nuke ships with a number of default configurations, but you can use a custom OCIO config file by
selecting custom from the OpenColorIO Config dropdown and then entering the file path.
Changing the configuration updates the Default Color Transforms accordingly. If the selected
configuration is invalid for certain transforms, a warning displays.
3. The working space transform determines what colorspace files should be converted to (Read)
and from (Write) - it's the colorspace used by Nuke under the hood.
USER GUIDE
1205
OCIO Color Management | Color Management
Note: In earlier releases of Nuke, this colorspace was hidden because linear was always
chosen as the working space. You may find that some operations work better in
colorspaces other than linear. For example, some transforms work better in the CLog
colorspace.
4. You can use Default Color Transforms dropdown menus to override how clips in the Viewer,
thumbnails, and so on are converted to and from the working space.
When the Nuke is selected, Reads and Writes work the same as in previous versions of Nuke, with no
integrated OCIO transforms. When OCIO is selected:
• Reads and Writes use OCIO transforms, with no Nuke built-in LUTs applied to the image.
• Read and Write colorspace controls are populated with the list of colorspaces defined in your
currently selected OCIO config.
• The default LUT settings dropdowns are also populated with the list of colorspaces or display
transforms defined in your OCIO config. The default value for each menu match the defaults in a
Nuke Studio project with the same config. These defaults can be overridden using Python callbacks.
See the following path for the default implementation that ships with Nuke:
<install_dir>/plugins/nuke/colorspaces.py
• The working space dropdown allows you to change the colorspace that Nuke uses internally for its
image processing. This automatically sets the in colorspace of Write nodes and Viewer Processes,
and the out colorspace for Read nodes. This defaults to the scene linear role defined in your OCIO
config.
• Nuke Studio-created comps no longer contain automatically injected OCIOColorspace nodes.
Instead, OCIO Color Management is automatically set in the comp’s Project Settings, and the
correct OCIO colorspace is set directly into the Read and Write nodes.
USER GUIDE
1206
Creating Custom Viewer Processes | Color Management
Creating Custom Viewer

Processes
Using look-up tables (LUTs) in Viewer Processes, you can adjust individual Viewer displays to simulate
the way the image looks on output to film or some video display device. Nuke includes some
predefined Viewer Process gizmos, but you can also add your own processes by registering a node or
gizmo as a Viewer Process. You can register as many custom Viewer Processes as you like. If you want
to use one of the 1D LUTs listed in the Project Settings in your Viewer Process, you can use the built-
in gizmo called ViewerProcess_1DLUT.
Tip: There are a couple of commented out examples in the installed init.py file
demonstrating how to use a 3D LUT for a Viewer Process. You can find this file in the
following location:
On Windows:
drive letter:\Program Files\Nuke11.3v5\plugins or
drive letter:\Program Files (x86)\Nuke11.3v5\plugins
On Mac:
On Linux:
All available Viewer Processes (both custom and predefined ones) can be applied from the Viewer
Process dropdown menu in the Viewer controls.
Both predefined and custom

Viewer Processes can be
applied from the Viewer
Process dropdown menu.
USER GUIDE
1207
Creating Custom Viewer Processes | Color Management
Note that Viewer Processes are part of a built-in, fixed pipeline of nodes that are applied to images
before they are displayed in the Viewer. This pipeline is either:
• gain > Input Process > Viewer Process > gamma > dither > channels > cliptest (if viewer input order
has been set to before viewer process in the Viewer settings)
OR
• gain > Viewer Process > Input Process > gamma > dither > channels > cliptest (if viewer input order
has been set to after viewer process in the Viewer settings).
However, depending on what the Viewer Process is doing, this may not be the correct order.
Therefore, if your Viewer Process (or an Input Process) has controls that also exist for the Viewer,
such as controls named gain, gamma, or cliptest, then the Viewer drives them from the
corresponding Viewer controls and does not do that image processing itself. This allows you to
implement these controls in your Viewer Process using whatever nodes and order you want. If your
Viewer Process does not have these controls (and they are not found on any Input Process in use
either), then the Viewer applies the effects in its normal way according to the built-in pipeline.
In the built-in pipeline, dither is applied to diffuse round-off errors in conversion of floating point
data to the actual display bit depth. Although the cliptest is drawn at the end, it is computed on the
image as input to the Viewer.
USER GUIDE
1208
Using a Gizmo as a Custom Viewer Process | To Register a LUT in the Project Settings as a Viewer
Using a Gizmo as a Custom

Viewer Process
To create a custom Viewer Process, you would typically create a gizmo that includes some color
correction like a look-up table (LUT) and register it as a Viewer Process using Python. (For more
information on gizmos, see Sourcing Custom Plug-ins and Generic Tcl Scripts.)
If you want to use one of the 1D LUTs listed in the Project Settings in your Viewer Process, you do not
need to create a custom gizmo. Instead, you can simply register a built-in gizmo called
ViewerProcess_1DLUT. This gizmo takes a parameter for which LUT to use, but does not allow it to be
edited. For more information, see To Register a LUT in the Project Settings as a Viewer Process.
If you want anything more complex than a 1D LUT that can be found on the LUT tab of the Project
Settings, you need to create your own gizmo and register that. For more information, see To Create a
Viewer Process Gizmo and To Register a Custom Viewer Process.
To Register a LUT in the Project Settings as a Viewer

Process
Scripts.
2. To register one of the LUTs in the Project Settings as a Viewer Process, use, for example, the
following function in your init.py:
nuke.ViewerProcess.register("Cineon", nuke.createNode, ("ViewerProcess_
1DLUT", "current Cineon"))
This registers a built-in gizmo called ViewerProcess_1DLUT as a Viewer Process and sets it to use
the Cineon LUT. The registered Viewer Process appears in the Viewer Process dropdown menu as
Cineon.
Note that you can set the built-in gizmo to use any 1D LUT in the Project Settings. For example, to
set it to use the Panalog LUT, use the following function:
nuke.ViewerProcess.register("Panalog", nuke.createNode, ("ViewerProcess_
1DLUT", "current Panalog"))
USER GUIDE
1209
Using a Gizmo as a Custom Viewer Process | To Create a Viewer Process Gizmo
To Create a Viewer Process Gizmo

1. Create the node(s) you want to use as a Viewer Process. For example, you can use a ColorLookup,
Vectorfield (3D LUT), or Colorspace node.
2. Select the node(s) you want to include in the Viewer Process and select Other > Group.
3. To select which controls the users of your Viewer Process can adjust, right-click on the dark gray
background of the Group properties panel and select Manage User Knobs. For more
information on how to add controls to your gizmo, see Creating and Accessing Gizmos.
If you expose controls with the same name as the controls in the Viewer (such as gain or gamma),
then the controls in the Viewer are used to drive these. However, if an Input Process that exposes
the same controls is also in use, the Input Process takes precedence and the Viewer controls drive
it, ignoring the same-named Viewer Process control(s). For more information on Input Processes,
see Using the Viewer Controls > Input Process and Viewer Controls.
4. Once you are happy with the modified Viewer Process group, export it to a gizmo by clicking
export as gizmo on the Node tab of the group controls.
of the file browser. Enter a name after the path, and append a .gizmo extension after the name.
The name should begin with a capital letter. Finally, click Save.
6. Proceed to registering the gizmo as a custom Viewer Process, described below.
Tip: If you like, you can test your Viewer Process gizmo as an Input Process before
registering it. Do the following:
1. In the top right corner of the Viewer, set the Viewer Process dropdown menu to None.
2. Select the gizmo in the Node Graph.
3. To toggle the Input Process on or off, click the IP button in the Viewer controls. If you are
happy with the result, proceed to registering the gizmo as a Viewer Process.
For more information on Input Processes, see Using the Viewer Controls > Input Process
and Viewer Controls.
Tip: If you want to view or modify the internals of an existing Viewer Process, you can do the
following:
1. Select the Viewer Process that you want to modify from the Viewer Process dropdown
menu.
2. Select Edit > Node > Copy Viewer Process to Node Graph. This inserts the Viewer
Process gizmo you selected in the Node Graph.
3. Double-click on the gizmo to open its controls. Go to the Node tab and click copy to
USER GUIDE
1210
Using a Gizmo as a Custom Viewer Process | To Register a Custom Viewer Process
group. This gives you an editable group version of the gizmo.

4. In the Group controls, click the S button to show the internals of the group. They are
shown on a new tab in the Node Graph.
5. Make your changes and export the group to a gizmo by clicking export as gizmo on the
Node tab of the group controls.
Tip: If you use the ViewerLUT node in a Viewer Process gizmo, you can toggle rbg_only in
the ViewerLUT controls to define whether the LUT is applied to all channels or only the red,
green, and blue channels. You can also expose this control in the Viewer Process gizmo's
controls, so that users can set it themselves.
To Register a Custom Viewer Process

Scripts.
2. To register a gizmo or a node as a Viewer Process, use the following function in your init.py:
nuke.ViewerProcess.register()
For example, to register a gizmo called MyProcess.gizmo as a Viewer Process and have it appear
in the Viewer Process dropdown menu as My Custom Process, you would enter the following:
nuke.ViewerProcess.register("My Custom Process", nuke.Node, ("MyProcess",
""))
Your Viewer Process should now appear in the Viewer controls.
If you need to unregister a Viewer Process, you can use nuke.ViewerProcess.unregister(). For
example:
USER GUIDE
1211
Using a Gizmo as a Custom Viewer Process | To Register a Custom Viewer Process
nuke.ViewerProcess.unregister("My Custom Process").

To get help on the use of these statements, you can enter help (nuke.ViewerProcess) in the
Script Editor.
Tip: You can also pass arguments to nuke.ViewerProcess.register(). For example, to

register a Blur node with its size knob set to 10, you would enter the following:
nuke.ViewerProcess.register("Blur", nuke.createNode, ("Blur", "size 10"))
Tip: You can easily register any LUT defined in the project settings as a Viewer Process. For
how to do this, see the installed menu.py file where the built-in Viewer Processes are
registered. You can find menu.py in the following location:
On Windows:
drive letter:\Program Files\Nuke11.3v5\plugins or
drive letter:\Program Files (x86)\Nuke11.3v5\plugins
On Mac:
On Linux:
USER GUIDE
1212
Applying Custom Viewer Processes to Images | To Apply Your Custom Viewer Process to Images
Applying Custom Viewer

Processes to Images
In the Viewer controls, you can apply a custom Viewer process to images displayed in the Viewer and
open the controls for the currently active Viewer process.
To Apply Your Custom Viewer Process to Images

Displayed in a Viewer
Select the process from the Viewer Process dropdown menu in the Viewer controls.
To View the Controls of the Currently Active Viewer

Process
In the Viewer controls, select show panel from the Viewer Process dropdown menu.
This opens the Viewer Process’ properties panel. Any controls with the same name as the controls in
the Viewer (such as gain or gamma) can only be adjusted using the Viewer controls. If these controls
are also exposed on an Input Process and the Input Process has been activated, the Viewer controls
drive the Input Process controls and the Viewer Process controls are disabled.
For more information on Input Processes, see Using the Viewer Controls > Input Process and
Viewer Controls.
USER GUIDE
1213
Expressions
This chapter is intended as a primer on how to apply expressions (programmatic commands) to Nuke
parameters. It explains how to perform some common tasks with expressions (for example, how to
link the values of one parameter to another), and concludes with a table all the functions that you
may include as part of an expression.
Quick Start
1. You enter Nuke expressions in the Expression dialog, which you can open either by pressing the
equals sign (=) on a parameter or by right-clicking on it and selecting Add expression.
2. In the Expression dialog, enter text that either references values from other parameters (creating
a linking expression - see Linking Expressions) or applies mathematical functions of some kind to
the current values (see Adding Mathematical Functions to Expressions). An example of the former
would be parent.Transform1.rotate, which indicates that this control takes its values from the
parent control, Transform node’s rotate control.
3. If necessary, you can also convert expressions between scripting languages (that is, between Nuke
expressions, Tcl, and Python). See Converting Expressions Between Scripting Languages.
USER GUIDE
1214
Linking Expressions |
Linking Expressions
Through expressions, you can link the parameters from one node and control the values of the
parameters in other nodes. When creating a linking expression, type the elements listed in the table
below; remember to separate each element with a period.
Element Description
Node name The node with the source parameter (i.e., Transform1).
Parameter name The name of the parameter with the source value (for example, translate).The
name is defined internally, and may not match the parameter’s label that
appear in the Nuke interface. If necessary, hover over the parameter’s field with
your mouse pointer and its name appears in the pop-up tool tip.
Child parameter Some parameters include child parameters, such as the fields for x and y axes,
name or red, green, and blue color channels. Child parameter names do match the
label that appears before the parameter’s field (for example, x).
(optional)
Time By default, linking expressions pull values from the current frame number, but
(optional) you can read values from other frames, either statically or dynamically (that is,
with a temporal offset).
If you want to read in a static value for a given frame, you just type that frame
number inside a set of parenthesis (for example, (10)).
If you want to read in dynamic values but with an offset in time, type t, the
variable for time, followed by a + (for a forward offset) or - (for a backward
offset), followed by a number representing the number of frames worth of
offset. For example, typing (t-2) would capture values that are two frames back
from the current frame.
Thus, to create a linking expression that pulls the value from a Transform node’s x translation field at
the tenth frame, you would type = on a parameter to open the expression dialog, and then enter
Transform1.translate.x(10) in the dialog’s Expression field.
USER GUIDE
1215
Linking Expressions | Referencing Values from Another Parameter
The steps below recap the process for creating a linking expression.
Referencing Values from Another Parameter

To Reference Values from Another Parameter - Method 1:
1. Click on the destination parameter (the one which receives values from another parameter).
2. To display the expression dialog, right-click on the parameter and select Add expression,
OR type = in the parameter field.
3. In the dialog that opens, type the name of the node containing the source parameter and a
period. (Each node prominently displays its name on its face.)
4. If you want to enter a multi-line expression, you can click the multi-line edit field button .
USER GUIDE
1216
Linking Expressions | Referencing Values from Another Parameter
5. Follow the name of the node by the source parameter’s name and a period. (If you don’t know the
parameter’s name, you can hover over its field in order to see it displayed in a tool tip.)
6. Optionally, type the child parameter’s name and a period.
7. Optionally, type a frame number or offset variable in brackets (for example, (2) or (t-2)) in order
to specify the frame or range of frames from which you pull values.
8. Next to the expression entry field, you can click the Py button to automatically make your
expression a Python callback. You can also toggle the R button to have your expression
interpreted as an expression or as a series of statements. For example, with the multi-line edit
mode and the Python mode on, you could enter the following expression, and get 15 as the
resulting value:
-execlocal
def example():
a = 5
return a
def example2():
b = 10
return b
ret = example()+example2()
9. Click OK. This links the parameters, which turn blue. In the Node Graph, a green arrow appears
between the nodes to indicate that they are linked via an expression.
10. To edit the expression later on, right-click on the parameter and select Edit expression (or press =
on the parameter). You can also click the animation button and select Edit expression to edit
expressions for all the parameters next to the button.
USER GUIDE
1217
Linking Expressions | Linking Channels and Formats Using Expressions
To Reference Values from Another Parameter - Method 2:

1. Ctrl/Cmd+drag the parameter that has the values you want to use on top of the parameter that
receives these values. This links the parameters, which turn blue. In the Node Graph, a green
arrow appears between the nodes to indicate that they are linked via an expression.
To view or edit the expression, right-click on the parameter and select Edit expression.
2. If you want to link several parameters at the same time, Ctrl/Cmd+drag the animation button
next to the source parameters on top of the animation button next to the destination parameters.
To view or edit the expressions used to link the parameters, click the animation button and select
Edit expression.
Linking Channels and Formats Using Expressions

You can also create expression links to connect channel, layer, and format controls with other
controls in various nodes. Since these controls aren’t meant to be animated, you can’t use the full
range of Nuke expressions, nor can you use Python or Tcl languages. You can link controls using the
Link menu next to the control on the properties panel:

1. Click the Link menu and select Set link. An Expression dialog opens.
2. Enter your expression in the Expression field and click OK.
3. You can edit an existing link by clicking the Linkmenu and selecting Editlink.
4. You can also Ctrl/Cmd+drag the Link menu to another control to create a link between the two.
5. To remove a link, click the Linkmenu and select Removelink.
USER GUIDE
1218
Adding Mathematical Functions to Expressions | Linking Channels and Formats Using Expressions
Adding Mathematical Functions

to Expressions
You can incorporate mathematical functions into parameters. For example, you might negate an
expression in order to invert a tracking curve which you wish to use to stabilize an element (such an
expression might resemble the following: -(Transform1.translate.x)).
You can also rely on a function to add more complex mathematical operation to your expressions.
The table below list all the functions which you may incorporate into Nuke expressions.
Function Purpose Operator Related DeepExpression

Usage Functions Compatible
abs (x) Returns the absolute value x See also: fabs.

of the floating-point number
x.
acos (x) Calculates the arc cosine of If x is less than See also: cos,
x; that is the value whose -1 or greater cosh, asin, atan.
cosine is x. than 1, acos
returns nan
(not a
number).
asin (x) Calculates the arc sine of x; If x is less than See also: sin,
that is the value whose sine -1 or greater sinh, acos, atan.
is x. than1, asin
returns nan
(not a
number).
atan (x) Calculates the arc tangent of x See also: tan,

x; that is the value whose tanh, acos, asin,
tangent is x. The return atan2.
value is between -PI/2 and
PI/2.
atan2 (x, y) Calculates the arc tangent of x, y See also: sin,
USER GUIDE
1219

the two variables x and y. cos, tan, asin,

This function is useful to acos, atan,
calculate the angle between hypot.
two vectors.
ceil (x) Round x up to the nearest x See also: floor,

integer. trunc, rint.
clamp (x, Return x clamped to [min ... x, min, max See also: min,
min, max) max]. max.
clamp (x) Return x clamped to [0.0 ... x See also: min,

1.0]. max.
cos (x) Returns the cosine of x. x in radians See also: acos,

sin, tan, cosh.
cosh (x) Returns the hyperbolic x See also: cos,

cosine of x, which is defined acos, sinh, tanh.
mathematically as (exp(x) +
exp(-x)) / 2.
curve (frame) Returns the y value of the optional: See also: value,
animation curve at the given frame, y.
frame. defaults to
current frame.
degrees (x) Convert the angle x from x See also:

radians into degrees. radians.
exp (x) Returns the value of e (the x See also: log,

base of natural logarithms) log10.
raised to the power of x.
exponent (x) Exponent of x. x See also:

mantissa, ldexp.
fBm (x, y, z, Fractional Brownian Motion. x, y, z, octaves, See also: noise,

octaves, This is the sum of octaves lacunarity, random,
lacunarity, gain turbulence.
USER GUIDE
1220

gain) calls to noise(). For each of

them the input point is
multiplied by pow
(lacunarity,i) and the result is
multiplied by pow(gain,i).
For normal use, lacunarity
should be greater than 1 and
gain should be less than 1.
fabs (x) Returns the absolute value x See also: abs.

of the floating-point number
x.
false () Always returns 0 See also: true.
floor (x) Round x down to the nearest x See also: ceil,

integer. trunc, rint.
fmod (x, y) Computes the remainder of x, y See also: ceil,

dividing x by y. The return floor.
value is x - n y, where n is
the quotient of x / y,
rounded towards zero to an
integer.
frame () Return the current frame See also: x.

number.
from_byte Converts an sRGB pixel color_ See also: to_

(color value to a linear value. component sRGB, to_
component) rec709f, from_
rec709f.
from_rec709f Converts a rec709 byte value color_ See also: form_

(color to a linear brightness component sRGB, to_
component) rec709f.
from_sRGB Converts an sRGB pixel color_ See also: to_

(color value to a linear value. component sRGB, to_
USER GUIDE
1221

component) rec709f, from_

rec709f.
hypot (x, y) Returns the sqrt(x*x + y*y). x, y See also: atan2.

This is the length of the
hypotenuse of a right-angle
triangle with sides of length
x and y.
int (x) Round x to the nearest x See also: ceil,

integer not larger in floor, trunc, rint.
absolute value.
ldexp (x, exp) Returns the result of x, exp See also:

multiplying the floating- exponent.
point number x by 2 raised
to the power exp.
lerp (a, b, x) Returns a point on the line f a, b, x See also: step,

(x) where f(0)==a and f(1)==b. smoothstep.
Matches the lerp function in
other shading languages.
log (x) Returns the natural x See also: log10,

logarithm of x. exp.
log10 (x) Returns the base-10 x See also: log,

logarithm of x. exp.
logb (x) Same as exponent(). x See also:

mantissa,
exponent.
mantissa (x) Returns the normalized x See also:

fraction. If the argument x is exponent.
not zero, the normalized
fraction is x times a power
of two, and is always in the
range 1/2 (inclusive) to 1
USER GUIDE
1222

(exclusive). If x is zero, then

the normalized fraction is
zero and exponent() Returns
zero.
max (x, y, ... ) Return the greatest of all x, y, (...) See also: min,
values. clamp.
min (x, y, ... ) Return the smallest of all x, y, (...) See also: max,
values. clamp.
mix (a, b, x) Same as lerp(). a, b, x See also: step,

smoothstep,
lerp.
noise (x, y, z) Creates a 3D Perlin noise x, optional y, See also:

value. This produces a optional z random, fBm,
signed range centerd on turbulence.
zero. The absolute
maximum range is from -1.0
to 1.0. This produces zero at
all integers, so you should
rotate the coordinates
somewhat (add a fraction of
y and z to x, etc.) if you want
to use this for random
number generation.
pi () Returns the value for pi

(3.141592654...).
pow (x, y) Returns the value of x raised x, y See also: log,

to the power of y. exp, pow.
pow2 (x) Returns the value of x raised x, y See also: pow.

to the power of 2.
radians (x) Convert the angle x from x See also:

degrees into radians. degrees.
USER GUIDE
1223

random (x, y, Creates a pseudo random optional x, See also: noise,

z) value between 0 and 1. It optional y, fBm, turbulence.
always generates the same optional z
value for the same x, y and
z. Calling random with no
arguments creates a
different value on every
invocation.
rint (x) Round x to the nearest x See also: ceil,

integer. floor, int, trunc.
sin (x) Returns the sine of x. x in radians See also: asin,

cos, tan, sinh.
sinh (x) Returns the hyperbolic sine x See also: sin,

of x, which is defined asin, cosh, tanh.
mathematically as (exp(x) -
exp(-x)) / 2.
smoothstep Returns 0 if x is less than a, a, b, x See also: step,

(a, b, x) returns 1 if x is greater or lerp.
equal to b, returns a smooth
cubic interpolation
otherwise. Matches the
smoothstep function in
other shading languages.
sqrt (x) Returns the non-negative x See also: pow,

square root of x. pow2.
step (a, x) Returns 0 if x is less than a, a, x See also:

returns 1 otherwise. Matches smoothstep,
the step function other lerp.
shading languages.
tan (x) Returns the tangent of x. x in radians See also: atan,

cos, sin, tanh,
atan2.
USER GUIDE
1224

tanh (x) Returns the hyperbolic x See also: tan,

tangent of x, which is atan, sinh, cosh.
defined mathematically as
sinh(x) / cosh(x).
to_byte (color Converts a floating point color_ See also: form_

component) pixel value to an 8-bit value component sRGB, to_
that represents that number rec709f, from_
in sRGB space. rec709f.
to_rec709f Converts a floating point color_ See also: form_

(color pixel value to an 8-bit value component sRGB, from_
component) that represents that rec709f.
brightness in the rec709
standard when that standard
is mapped to the 0-255
range.
to_sRGB Converts a floating point color_ See also: form_

(color pixel value to an 8-bit value component sRGB, to_
component) that represents that number rec709f, from_
in sRGB space. rec709f.
true () Always Returns 1. See also: false
trunc (x) Round x to the nearest x See also: ceil,

integer not larger in floor, int, rint.
absolute value.
turbulence (x, This is the same as fBm() x, y, z, octaves, See also: fBm,
y, z, octaves, except the absolute value of lucanarity, noise, random.
lucanarity, the noise() function is used. gain
gain)
value (frame) Evaluates the y value for an optional: See also: y,

animation at the given frame, curve.
frame. defaults to
current frame.
USER GUIDE
1225

x () Return the current frame See also: frame.

number.
y (frame) Evaluates the y value for an optional: See also: value,

animation at the given frame, curve.
frame. defaults to
current frame.
USER GUIDE
1226
Converting Expressions Between Scripting Languages | Linking Channels and Formats Using
Converting Expressions Between

Scripting Languages
Depending on where you need to use an expression, you might find that you want to, for example,
convert Nuke expressions to Tcl expressions or embed Python functions in a Nuke expression. The
different languages are used in different parts of Nuke:
• Python can be used in the Script Editor, in the Script Command (File > Comp Script Command) and
in scripts run when Nuke starts (such as init.py and menu.py).For more information, see the Nuke
Python documentation (Help > Documentation).
• Tcl can be used in most string knobs (where text other than just numbers can be entered), in the
Script Command dialog (File > Comp Script Command), to open some compatibility start up scripts
(such as init.tcl and formats.tcl).
• Nuke expressions can be used on the Add Expression dialog with most knobs in Nuke and
expression entry field in the Expression node.
You can use the following functions to use different types of expressions together:
• nuke.expression() to use a Nuke expression in Python code.
• expression to use a Nuke expression in Tcl.
• nuke.tcl() to run Tcl code in Python.
• python to run Python code in Tcl.
• [ ] (square brackets) to embed Tcl in a Nuke expression (or a string knob).
• [python {...}] to embed Python in a Nuke expression.
Tip: Note that putting braces ( { } ) around Python code when embedding it in Tcl may make
the process a bit easier, because this prevents Tcl from performing its own evaluation of the
Python code before passing it through to the Python interpreter. For example: [python
{"hello " + "world"}]
Tip: Note that the "python" Tcl command by default evaluates a single line of code and
returns the result. Use the "-exec" option (for example, "python -exec") if you want to run
multiple lines. Please refer to the Nuke Tcl Scripting documentation (Help > Documentation
> TCL Scripting) for further information.
USER GUIDE
1227
The Script Editor and Python
Nuke ships with a comprehensive Python application programming interface (API), enabling you to
perform user interface actions using Python scripting. This chapter describes how you can use the
Script Editor for executing Python scripts and directs you to sources of more information on Python.
Quick Start
1. Enter Python statements in Nuke’s Script Editor to perform the required actions.
2. Save your script with the extension .py in a directory that is contained in the sys.path variable.
3. Later, when you want to execute the same statement sequence, import the .py file into Nuke’s
Script Editor. Nuke executes the statements in the specified order.
If you need more information on using Python in Nuke, you can always turn to the Nuke Python
Developer’s Guide (Help > Documentation).
Tip: You can import Nuke as a module into a third-party Python 2.7.13 interpreter. See Nuke
as a Python Module for more information.
Tip: You can also run an interactive Python session on the command line with nuke -t.
USER GUIDE
1228
Using the Script Editor | Opening the Script Editor
Using the Script Editor

If you're not using a third-party Python interpreter, you can type Python scripts into Nuke’s Script
Editor.
Opening the Script Editor

To open the Script Editor, click on one of the content menus and select Script Editor from the menu
that opens.
Input and Output Panes

The Script Editor is divided into two parts, as shown in the figure below. You use the lower part (input
pane) to type in and execute your Python statement, and when you have done so, statements and
their outputs appear in the upper part of the editor (output pane). Successfully executed statements
are followed by a hash mark (#).
The two parts of the Script Editor.
To hide the output or input pane, click the Show input only or Show output only button on
top of the Script Editor.
To show both panes again, click the Show both input and output button .
Entering a Statement
To enter a statement in the Script Editor:
1. Click on the input pane of the editor to insert the cursor there.
USER GUIDE
1229
Using the Script Editor | Entering a Statement
2. Type in your statement. To use the usual editing functions, such as copy and paste, right-click on
the editor and select the desired function.
When entering the statement, you’ll notice that any words that are Python’s keywords (such as
print and import) turn green, while strings (basically, anything in quotation marks) become either
red or cyan. Comments are shown in yellow.
If you like, you can change these colors and the font on the Script Editor tab of the Preferences
dialog. To open the preferences, press Shift+S.
Tip: You can also use auto-complete to help you with entering Python statements. Start
writing a command and press the Tab key. If there’s only one way to end your command,
Nuke auto-completes it straight away. If there are several possible completions, Nuke gives
you a pop-up menu listing them. If there’s no known way to complete your command,
nothing happens. Even if your command is automatically completed, it is not executed
automatically, just in case you don’t like surprise side effects.
3. If your statement includes several lines or you want to enter several statements at once, press
Return to move to the next line.
4. To execute the statement, click the Run the current script button on the top of the Editor, or
press Ctrl/Cmd+Return.
Tip: You can also execute statements by pressing Ctrl/Cmd+Enter on the numeric keypad.
By default, successful statements disappear from the input pane, and appear in the output pane.
However, if you want all statements to stay in the input pane after they are executed, you can do the
following:
1. Press Shift+S to open the Preferences dialog.
2. Go to the Script Editor tab.
3. Uncheck clear input window on successful script execution.
4. Click Close to save the preference for the current project only, or Save Prefs to save the
preference for the current and future projects.
If you enter an invalid statement, Nuke produces an error in the output pane of the Script Editor,
leaving the invalid statement in the input pane. Correct the statement and execute it again until you
get it right.
USER GUIDE
1230
Using the Script Editor | Moving Through and Clearing the Script History
Note: Sometimes you may get an error if you copy and paste statements into the Script
Editor from another source, like an e-mail. This may be caused by the mark-up or encoding
of the source you copied the statement from. To fix the problem, re-enter the statement
manually.
If you want to have all executed Python commands appear in the output pane of the Script Editor,
open the Preferences dialog (press Shift+S), go to the Script Editor tab, and check echo all
commands to output window. This applies to both commands executed by yourself and by Nuke.
For example, if you select a node from the Toolbar, the corresponding Python command is displayed
in the output pane. This does not apply to all actions you take in the graphical user interface,
however, but only those that are performed by executing Python script commands.
To only execute part of a script, enter the script in the input pane and select the part you want to
execute. Press Ctrl/Cmd+Return. Nuke runs the selected part of the script, leaving the script in the
input pane.
To repeat a statement, click the Previous Script button on top of the Editor to move back to the
previous statement. You can do this until you reach the statement you want to repeat. To execute the
statement again, press Ctrl/Cmd+Enter.
To increase the indentation in the input window, press Tab.
To decrease the indentation in the input window, press Shift+Tab.
Moving Through and Clearing the Script History

In addition to stepping backwards through the history of your script, you can also step forwards. Click
the Next script button to move forward through your statements.
To clear the history, click the Clear history button.
Clearing the Output Pane

Click the Clear output window button (or press Ctrl/Cmd+Backspace).
USER GUIDE
1231
Automating Procedures | Saving Statements in a Python Module
Automating Procedures
Okay, so you know how to use the Script Editor to type in a sequence of Python statements that take
care of a procedure. But so far, you’ve still sat by your computer typing the statements in. It’s time to
automate the procedure. All you need to do is save your statements, and when you want to use them
again later, import them into the Script Editor.
Saving Statements in a Python Module

To save statements in a Python module:
1. On the top of the Script Editor, click the Save a script button .
2. Save the script with the extension .py (for example firstmodule.py) in a directory that is contained
in the sys.path variable. (To see these directories, enter print sys.path in the Script Editor. To add
a directory to the sys.path variable, enter sys.path.append ("directory") where directory
represents the directory you want to add.)
You have now created your first Python module.
Opening a Python Script in the Script Editor

To open a Python script in the Script Editor:
1. Click the Load a script button . on top of the Script Editor. The Script to open dialog opens.
2. Navigate to the Python module that contains the script you want to open and click Open.
Nuke opens the script in the input pane of the Script Editor, but does not execute it.
Importing and Executing a Python Script

To import and execute a Python script:
1. On top of the Script Editor, click the Source a script button . The Script to open dialog
opens.
2. Navigate to the Python module that contains the script you want to import and click Open.
OR
USER GUIDE
1232
Automating Procedures | Importing and Executing a Python Script
In the input pane, enter:
import module
Where module represents the name of your Python module without the file extension, for example:
import firstmodule
Nuke imports the Python module and performs the procedure defined in the module.
Note: Importing the module is done according to Python’s default rules. During the import,
the module is searched in the following locations and order:
1. In the current directory.
2. In the directories contained in the PYTHONPATH environment variable, if this has been
defined. (To view these directories, enter echo $PYTHONPATH in a command shell.)
3. In an installation-dependent default directory.
During the search, the variable sys.path is initialized from these directories. Modules are
then searched in the directories listed by the sys.path variable. To see these directories,
execute the statement print sys.path in the Script Editor.
USER GUIDE
1233
Nuke as a Python Module | Importing and Executing a Python Script
Nuke as a Python Module

You can import Nuke as a module into a third-party Python 2.7.13 interpreter, granting full access to
the Nuke Python-API, but from within a native Python interpreter instead of Nuke.
Note: Foundry cannot provide customer support for third-party Python interpreters.
To run Nuke as a Python module:

1. Add the file path for Nuke's site-packages directory to the usrlocal.pth file in your Python 2.7.13
install.
For example, if you're running on Windows, add C:\Program Files\Nuke 11.3v5\lib\site-
packages to the usrlocal.pth file.
Tip: You can also use relative paths to the directory containing the usrlocal.pth file.
2. At the Python prompt, use the import nuke declaration to make Nuke’s Script Editor functions
and commands (such as nuke.nodes.Blur() to add a Blur node) available in your chosen Python
interpreter.
The import nuke function checks-out a nuke_r render license by default. If you want to use Nuke
interactively, and you have a nuke_i license, set the NUKE_INTERACTIVE environment variable to 1.
See Environment Variables for more information on setting environment variables.
For more information on using Nuke as a Python module, select Help > Documentation from Nuke's
menu bar and navigate to Python Developers Guide > Nuke as a Python Module.
USER GUIDE
1234
Getting Help | More Documentation
Getting Help
In the scope of this user guide, it’s not possible to go into detail with Python and all the scripts
available. However, there are several sources of more information that you may find useful if you
need help using Python.
More Documentation
In the Python Developer’s Guide, you’ll find everything you need in terms of Python examples and
ways of using Python for various purposes in Nuke. You’ll find the Python Developer’s Guide when
you click Help > Documentation > Python Developers Guide in Nuke.
You may also want to read the Nuke Python API reference documentation under Help >
Documentation > PythonScriptingReference.
Viewing More Examples

We only described a few examples of Python scripts in this chapter, but there are more. You can find
them in the following location:
• On Windows:
drive letter:\Program Files\Nuke11.3v5\plugins\nukescripts or
drive letter:\Program Files (x86)\Nuke11.3v5\plugins\nukescripts

• On Mac:
/Applications/Nuke11.3v5/Nuke11.3v5.app/Contents/MacOS/plugins/nukescripts
• On Linux:
/usr/local/Nuke11.3v5/plugins/nukescripts
To view an example, select one of the .py files and open it in any text editor.
Using the Help Statement

Possibly the quickest way of getting help on the available commands is to enter the following in the
Script Editor:
help (nuke)
USER GUIDE
1235
Getting Help | Python on the Web
This statement lists many of the available Python commands in an alphabetical order together with
their descriptions.
You can also get help on more specific things. For example, the following statement gives you a
description of what the setValue() method does:
help (nuke.Knob.setValue)
Python on the Web

To read more about Python, check out its documentation, or interact with other Python users. Visit
the Python programming language official website at http://www.python.org/.
USER GUIDE
1236
Timeline Editing in Nuke Studio
Nuke Studio's timeline is designed to provide shot management, conform, and playback capabilities
for people creating visual effects and delivers visual effects sequences without resorting to other third
party applications. These are the topics covered:
• Using Tags tells you how to quickly sort or filter source clips and shots for better visibility,
organization, and export.
• Viewing Metadata describes how to examine extra clip data and filter your project to find the
footage you need.
• Conforming Sequences describes the process of matching up the footage from a shoot with the
required edit decisions to create a meaningful timeline.
• Managing Timelines describes the timeline interface and how to add and manage footage on
timelines.
• Soft Effects tells you how to add real-time GPU effects to timeline shots.
• Create Comp explains the difference between comps and regular shots and how to create and
manage them.
• Annotations allow you to add editorial comments to your timeline output, enabling collaborative
work between Nuke Studio and other Nuke workstations.
• Timeline Editing Tools describes how you manipulate your shots directly in the timeline using a
series of modal editorial tools that complement the Multi Tool.
• Versions and Snapshots describes how to record the different states of your workflow as you
progress using versions and snapshots.
• Exporting from Nuke Studio deals primarily with Nuke Studio's shot management and export
functionality when you're farming out shots or sequences to other artists. It also deals with presets,
which dictate how Create Comp passes data between the Timeline and Compositing environments.
USER GUIDE
1237
Using Tags
Tags are used by Nuke Studio to quickly sort or filter clips and shots for better visibility, organization,
and export. Tags are used to mark shots of a particular type or content as you organize your project.
The default tags supplied include Approved, Note, Reference, and other general purpose tags. You
can also create custom tags by right-clicking in the Tags tab or by pressing Ctrl/Cmd+Y. You can apply
tags to clips, shots and Comp shots, individual frames, sequences, and tracks.
Clip and shot tags and notes can be added to exports using the burn-in feature. See Adding Burn-in
Text to Exports for more information.
Tags can also be converted to Final Cut Pro or Premiere markers during export to XML. See Exporting
EDLs and XMLs for more information.
Using Quick Tags

Quick tags allow you to add tags, depending on context, by right-clicking a selection and then
choosing the type of tag to apply. If you’re tagging a large amount of media, you might find it more
convenient to use the drag-and-drop methods described later on.
Quick tags are accessible from bins, spreadsheets, Viewers, and timelines for single or multiple
selections.
1. Select the target clips or sequences.
2. Right-click a highlighted selection, go to Tags, and choose the required action, dependent on
context.
For example, bin items only allow you to Tag Selection, whereas shots allow you to Tag Shot
Selection, Tag Tracks, or Tag this Sequence
Once you’ve selected the tag type, the Add Tag dialog displays.
3. Select the icon to represent the tag using the Icon dropdown.
4. Enter a Name and Note as required.
5. Click Add to mark your selections with the chosen tag.
USER GUIDE
1238
Using Tags |
SeeCreating Custom Tags and Removing Tags for more information.
USER GUIDE
1239
Tagging Using the Viewer |
Tagging Using the Viewer

To apply a tag using the Viewer:
1. Click the Tags tab, or navigate to Window > Tags.
The Tags panel displays.
2. Drag-and-drop the required tag from the Tags panel to the Viewer.
Depending on whether you’re looking at a clip or sequence, drop the tag on Tag this frame, Tag
whole clip, or Tag whole sequence as required.
Tags applied to frames appear above the playback tools in the Viewer frame slider.
Tip: You can use Alt+Shift+left and right arrows to skip to the previous or next tag on the
current clip. You can also reposition tags by dragging them along the Viewer timeline.
Tags applied to clips are displayed above the Viewer.
USER GUIDE
1240
Tagging Shots |
Tagging Shots
To apply a tag to a shot on the timeline:
The Tags panel displays.
2. Drag-and-drop the required tag from the Tags panel to the timeline.
Depending on where the tag is dropped, you’ll mark a shot (or items if you make multiple
selections) or a track.
Tags applied to shots appear on the right of the selected item(s) on the timeline.
Tags applied to tracks appear in the track header on the left of the timeline.
USER GUIDE
1241
Tagging Shots |
USER GUIDE
1242
Adding Notes to Tags |
Adding Notes to Tags

In some cases, a simple tag on a frame or clip may not contain all the information that you wish to
pass on to the next stage of production. Adding notes to a tag can provide that extra detail.
Warning: To delete a note, don’t click the - button because this refers to the tag. Instead,
simply delete the notes in the window and click outside the note dialog.
1. Add notes to tags by clicking on the required tag and entering text or editing the metadata keys
and values.
The example shows a note and metadata key “Artist” added to a clip tag, but you can add notes to
frame and timeline tags in the same way.
2. Click outside the dialog to save the note.
Nuke Studio allows you to “hide” tags using the Python API. Hidden tags are not displayed in the
interface, unless you enable Show Hidden in the Tags popup, but the notes and metadata are still
accessible.
Navigate to Help > Documentation > Hiero Python Developer's Guide for more information on
creating hidden tags.
USER GUIDE
1243
Removing Tags |
Removing Tags
To remove a tag from a frame or shot, click the tag and then click .
You can remove all tags from a source clip or selection of clips by right-clicking your selections in the
bin and choosing Tags > Clear Tags.
To remove a tag from a track or shot, click on a tag icon and select the required tag to remove.
Click to remove your selection.
USER GUIDE
1244
Creating Custom Tags |
Creating Custom Tags

You may find that you require a specific tag or suite of tags that are not provided by default. Creating
custom tags allows you to really control the organization of your media, and you can even create your
own tag icons.
Note: Custom tags can only be created in the Tags panel.
To create a custom tag:

2. Select your project and navigate to Project > New Tag, or press Ctrl/Cmd+Y.
The new tag is placed in the selected project.
3. Double-click the tag to open the Edit Tag dialog box.
4. Click the Icon dropdown menu to select an icon for the custom tag.
Tip: You can import your own image for the tag by selecting Custom to open the browser.
5. Enter a description for the tag in the Name field.

6. Click OK to save your changes.
USER GUIDE
1245
Sharing Custom Tags | Filtering and Flagging Media Using Tags
Sharing Custom Tags

Custom tags can be shared between artists by saving them in a project in a shared network location.
Shared tags are loaded at startup and appear in the Tags tab, below the standard tags that ship with
Nuke Studio.
1. Create the required custom tags as described in Creating Custom Tags.
2. Save the .hrox project in a folder called Templates in a shared network location. For example:
/SharedDisk/NukeStudio/Templates/SharedTags.hrox
Note: If your custom tags use custom icons, save the icon files in the same directory as the
.hrox project.
3. Create a Python file containing the following lines, to direct Nuke Studio to the shared location:
import hiero.core
hiero.core.addPluginPath("/SharedDisk/NukeStudio")
4. Save the .py file in the ~/.nuke/Python/Startup/ directory on all the machines running Nuke
Studio that require access to the tags.
Note: You may need to create the .nuke/Python/Startup/ path manually.
The location of the .nuke directory differs by platform. See Loading Gizmos, NDK Plug-ins, and
Python and Tcl Scripts for more information.
5. Launch Nuke Studio and navigate to Window > Tags or switch to a workspace that contains the
Tags tab.
The custom tags are listed under the project name below the standard tags that ship with Nuke
Studio.
Filtering and Flagging Media Using Tags

You can search for clips containing certain tags, for example, if you wanted to find all clips that you
tagged as Approved.
There are two types of tag search you can perform: Filter and Flag. Select the desired search type by
clicking the magnifier icon in the Project tab.
• Filter - displays all objects that contain the specified tag. This is the default search method.
• Flag - displays all objects and marks the items that don’t match the search tag.
USER GUIDE
1246
Sharing Custom Tags | Filtering and Flagging Media Using Tags
Drag the required tag from the Tags panel into the search box and select the bin or bins you want to
Filter or Flag.
Tip: If you have more than one search criteria, click the icons in the search box to display a
brief description of the icons.
Filters and flags persist until you change the search criteria or click the x icon in the search
box.
The following examples show Filtering a bin to display only clips with the Notes tag applied and
Flagging all clips that don’t have the Notes tag applied.
Filtering ... ... Flagging.
USER GUIDE
1247
Viewing Metadata
Metadata is information that describes media content, separate from the clip itself, in the form of a
table on the Metadata tab. Types of metadata include Duration, File Size, and the Path to the
location of the source media.
Source Clip and Shot Metadata

To view metadata for a source clip, right-click the clip and select Open In > Metadata View, or press
Alt+D.
To view metadata for a shot, select the Metadata tab in the timeline panel and click on the item to
examine.
Tip: You may have to add the Metadata tab manually by clicking the icon and selecting
Window > Metadata.
USER GUIDE
1248
Filtering and Flagging Media Using Metadata |
Filtering and Flagging Media

Using Metadata
If searching your project using tags has not filtered your media effectively, you can search for clips
containing certain metadata. For example, if you wanted to find all clips that had a particular
resolution or frame rate.
To filter or flag using metadata:

1. Right-click the clip that contains the required metadata key and select Open In > Metadata View,
or press Alt+D.
2. Drag-and-drop the required key from the Metadata panel to the bin view search box.
3. Use the metadata key as a filter or flag as described in Filtering and Flagging Media Using Tags.
USER GUIDE
1249
Conforming Sequences
Conforming describes the process of matching up the footage from a shoot with the required edit
decisions to create a meaningful timeline. Nuke Studio accepts sequences either from EDLs (edit
decision lists), AAFs (advanced authoring format), or Final Cut Pro XML files from a specified directory
structure containing the source media files. Nuke Studio attempts to conform the media, warning you
if there are missing media.
Nuke Studio conforms EDLs into single tracks, and AAFs and XMLs into multi-track timelines. You can
either conform into a brand new timeline, or into an existing timeline by adding new tracks. For
example, when conforming multiple EDLs into the same timeline, you would add new tracks for each
EDL sequence conformed.
USER GUIDE
1250
Timeline Environment Project Settings |
Timeline Environment Project

Settings
A good place to start work is by defining default Project Settings before importing sequences,
particularly in the case of EDLs as they may not contain frame rate information. Project Settings only
apply to the current project and override Preferences settings.
Note: You can modify Project Settings later on, for example, when you’re ingesting media.
To define Project Settings:

1. Navigate to Project > Edit Settings.
The Project Settings dialog displays.
2. Click the General sub-menu to set the project Name.
3. Enter a Project Directory if required. This is the location of the .hrox project file and can be used
as the root of the project if you want to use relative paths to source clips. See About Clips and
Shots for more information.
If you want this setting to apply to all new projects, use the Preferences > Project Defaults >
General panel settings.
Tip: Click Hrox Directory to automatically enter an expression that evaluates to the .hrox
location.
4. Set the Poster Frame used by Project bin clips or use the default First frame. See Setting Poster
Frames for more information.
5. Click the Sequence sub-menu to set the default Output Resolution, Frame Rate, and Start
Timecode for new timelines in the current project, and set clip formatting when new clips are
added to the timeline.
6. Click the Views sub-menu to set up multi-view or stereo projects. See Stereoscopic and Multi-
View Projects for more information.
7. Click the Color Management sub-menu to manage the display and file colorspaces for this
project.
See Color Management Settings for more information.
USER GUIDE
1251
Timeline Environment Project Settings | Color Management Settings
8. Click the RED Settings sub-menu to define the Default Video Decode Mode for new R3D files in
the current project. This setting overrides the Preferences > Behaviors > File Handling > default
red clip video decode mode control for existing projects. SeeAppendix A: Preferences for more
information.
Note: Changing this setting does not change the decode setting for R3D media that has
already been imported.
The dropdown contains a sliding resolution scale from FullPremium to SixteenthGood, but bear
in mind that higher resolutions are slower to decode.
9. Lastly, click the Export/Roundtrip sub-menu to select:
• External Media Track Name - sets the default name of the track created when exported media
is brought back into Nuke Studio.
• Export Directory - sets whether the Project Directory, if specified, or a custom directory is
used for exports. If no Project Directory is specified, the project root in the Export dialog is
used.
If you want this setting to apply to all new projects, use the Preferences > Project Defaults >
General panel settings.
• Custom Export Directory - when Export Directory is set to custom, enter the required custom
export directory.
• Shot Preset - sets the default preset to use when you select Create Comp from the timeline.
See Create Comp for more information.
Color Management Settings

Nuke Studio uses OpenColorIO for color management. All of the colorspaces in Nuke Studio, whether
those shipped with the application or custom colorspaces are defined in OCIO config files.
Depending on the OCIO config file that you are working with, there are a number of colorspace
options and roles (aliases to colorspaces) that you can set in Nuke Studio. There are also default
options, which change depending on what file type you are working with. When the default option is
selected, the colorspace that Nuke Studio has set for it is listed in brackets.
USER GUIDE
1252
Timeline Environment Project Settings | Color Management Settings
1. Set the OpenColorIO Config you want to use for this project.
Nuke Studio ships with a number of default configurations, but you can:
• use a custom OCIO config file by selecting custom from the OpenColorIO Config dropdown
and then entering the file path, or
• add your own config to your .nuke file. See Adding OCIO Configurations for more information.
Changing the configuration updates the Default Color Transforms accordingly. If the selected
configuration is invalid for certain transforms, a warning displays. For example, if you choose the
shipped iff configuration, the 8-bit and 16-bit transforms are not compatible.
In this case, the non-compatible transforms are set to the raw colorspace.
2. The Working Space transform determines what colorspace files should be converted to, on
import, and from, during export - it's the colorspace used by the Timeline environment under the
hood.
USER GUIDE
1253
Timeline Environment Project Settings | Adding OCIO Configurations
Note: In earlier releases of Nuke Studio, this colorspace was hidden because linear was
always chosen as the Working Space. You may find that some operations work better in
colorspaces other than linear. For example, some transforms work better in the CLog
colorspace.
3. You can use Default Color Transforms dropdown menus to override how clips in the Viewer,
thumbnails, and so on are converted to and from the Working Space.
4. The Nuke Script Project Settings dropdown determines whether Nuke Studio uses the LUTs
read from the configuration specified or the Nuke native LUTs during export. Selecting OCIO
makes the relevant OCIO LUTs available to the Read and Write nodes in scripts on a per project
basis.
All configurations except nuke-default automatically switch this control to OCIO.
When the Nuke is selected, Reads and Writes work the same as in previous versions of Nuke, with no
integrated OCIO transforms. When OCIO is selected:
• Reads and Writes use OCIO transforms, with no Nuke built-in LUTs applied to the image.
• Read and Write colorspace controls are populated with the list of colorspaces defined in your
currently selected OCIO config.
• The default LUT settings dropdowns are also populated with the list of colorspaces or display
transforms defined in your OCIO config. The default value for each menu match the defaults in a
Nuke Studio project with the same config. These defaults can be overridden using Python callbacks.
See the following path for the default implementation that ships with Nuke:
<install_dir>/plugins/nuke/colorspaces.py
• The working space dropdown allows you to change the colorspace that Nuke uses internally for its
image processing. This automatically sets the in colorspace of Write nodes and Viewer Processes,
and the out colorspace for Read nodes. This defaults to the scene linear role defined in your OCIO
config.
• Nuke Studio-created comps no longer contain automatically injected OCIOColorspace nodes.
Instead, OCIO Color Management is automatically set in the comp’s Project Settings, and the
correct OCIO colorspace is set directly into the Read and Write nodes.
Adding OCIO Configurations

You can add your own OCIO configurations to Nuke Studio as they become available, such as new
versions of ACES. You can also add legacy configs for backward compatibility.
1. Navigate to the location of your .nuke file as shown by platform. You may have to create a .nuke
folder if it doesn't exist.
• Linux: /users/login name/.nuke
USER GUIDE
1254
Timeline Environment Project Settings | Adding OCIO Configurations
• Mac: /Users/login name/.nuke

• Windows: ~\.nuke
Note: On Windows, the .nuke folder can be found under the directory pointed to by the
HOME environment variable. If this variable is not set (which is common), the .nuke
directory is under the folder specified by the USERPROFILE environment variable - which is
generally of the form drive letter:\Documents and Settings\login name\ or drive
letter:\Users\login name\
pointing at, enter %HOME% or %USERPROFILE% into the address bar in Windows Explorer.
If the environment variable is set, the folder it’s pointing at is opened.
2. Recreate the following structure within your .nuke folder:

~/plugins/OCIOConfigs/configs/<config name>
3. Copy the contents of the config into the config name named folder. There should be a luts folder
and .ocio file at the bare minimum.
4. If Nuke Studio is already running, relaunch the application to apply the change.
5. You can now select your configuration from the Project Settings > Color Management >
OpenColorIO Config dropdown.
USER GUIDE
1255
Importing Sequences | Adding OCIO Configurations
Importing Sequences
Nuke Studio allows you to import your EDL, XML, or AAF sequences in one of two ways, depending on
your preferences. Either:
• Navigate to File > Import EDL/XML/AAF, use the browser to locate the EDL, XML, or AAF, and then
select the file and click Open to import the sequence,
OR
• Drag-and-drop the EDL, XML, or AAF files directly from a file browser into the interface.
If you’re importing an EDL, bear in mind that there is no guaranteed frame rate information included
in the file, so an Import Options dialog displays.
1. Select the correct frame rate and use the following check boxes, if required:
• Drop Frame - when enabled, the EDL is assumed to contain drop file information. SeeTimeline
Playback Tools for more information.
• Assume differences in source/destination durations indicate a retime - when enabled, any
disparity between the source clip (Src) and shot (Dst) duration is treated as a retime.
2. Click OK to import.
XMLs and AAFs imported into Nuke Studio support transform, crop, and retime edit decisions
implemented in third-party applications, such as Adobe Premiere, Apple Final Cut Pro, and Avid
Media Composer. The information in the .xml or .aaf is interpreted using soft effects, such as
Transform and Crop. Non-linear retimes are represented by TimeWarp effects. Constant linear
retimes are handled in the same way as in previous versions of Nuke Studio. See Notes on AAF
Sequences for more information.
USER GUIDE
1256
Note: Non-linear animation curves from .xml may not appear as expected when imported.
As a result, you may need to adjust the handles on curves to match footage between
keyframes in the Curve Editor or Dope Sheet.
See Animating Parameters for more information.
Additionally, Premiere Pro .xml exports only support constant, linear retimes. As a result,
retimed shots on the Nuke Studio timeline may not match those on the Premier Pro
timeline, because certain non-linear retime data is not written into the exported .xml file.
After importing the EDL, AAF, or XML the Conforming workspace displays and the spreadsheet and
timeline are populated with offline clips - media with an unknown location.
Note: The Event column represents the clip’s position on the timeline, not its event number
from the edit.
Notice that clicking entries in the spreadsheet highlights the corresponding shots on the timeline?
USER GUIDE
1257
The spreadsheet, timeline, and Viewer are linked together when viewing sequences. If suitable screen
real estate exists within the current workspace, double-clicking a sequence forces the associated
panel to open automatically. If you want to close a single panel in a linked group, hold the Alt
modifier while closing the linked panel, otherwise all panels in the group are closed.
Note: If you imported an XML sequence, you may find that Nuke Studio has automatically
matched media for you.
Any transform, crop, or retime edit decisions from third-party software .xml and .aaf files are
represented using soft effects. These effects are imported along with the shot to which they're
associated.
USER GUIDE
1258
Importing Sequences | Notes on AAF Sequences
Notes on AAF Sequences

Avid Media Composer supports retimes using curves that map frame to frame or frame to speed.
Nuke Studio handles the import differently depending on the retime method.
• frame to frame - describes the retiming in relative terms, such as 'at frame 100 in the output clip,
display frame 50 of the source clip'.
• frame to speed - describes the retiming in terms of overall output duration. For example, half
speed doubles the duration of the clip.
Nuke Studio's TimeWarp effect only supports frame to frame mapping, which means that frame to
speed retimes from .aaf files requires some curve-fitting to describe the required retime. As a result,
the keyframes generated in Nuke Studio don't match those in Avid, but the resulting curve should
match the original very closely.
Note: Nuke Studio currently only supports Fixed Keyframes from Avid Media Composer.
Tip: If you need to adjust the handles on curves, see Animating Parameters for more
information.
Nuke Studio's TimeWarp effect supports the following Spline types when importing .aaf files:
• Shelf
• Linear
• Spline
USER GUIDE
1259
Importing Sequences | Notes on AAF Sequences
• Bezier
USER GUIDE
1260
Conforming Sequences | Notes on AAF Sequences
Conforming Sequences
Once your EDL, AAF, or XML sequences are imported, it’s time to begin the conform process to match
the offline shots in your spreadsheet with the source clips on disk. You can conform sequences by
searching on disk or by pre-ingesting the required clips into Nuke Studio.
ulimit -Sn 2048
USER GUIDE
1261
Conforming Using a Browser | Notes on AAF Sequences
Conforming Using a Browser

To conform a sequence using a browser:
1. After importing a sequence, click Match Media on the spreadsheet and use the browser to locate
the source folder containing the correct media.
Note: Match Media can also be used on selected events in the Spreadsheet view.
2. Click Open to display the Conform Options dialog.
Nuke Studio uses a set of conform Rules and file name Patterns to match candidate media files on
disk to the events, or shots, in a sequence:
• Rules - sets the offline media properties to match to the corresponding spreadsheet entry
during conform.
Rules that rely on information that doesn't exist in the event or candidate clip are ignored, and
some rules compound others to identify a better match.
Rule Description
Umid Match a file’s unique material ID (UMID) – that is written into the file's
metadata on creation – to the candidate media’s UMID. If either, or both, lack
a UMID this rule is ignored.
RedTapeName Match a RED-style camera reel name from the event to the candidate media
USER GUIDE
1262
Rule Description
name.
RedName Look for a RED-style camera file name in the event that matches the
candidate media name.
ReelName Look for the event's reel name in the candidate's media name.
FullPath Match the event's entire filepath to the candidate media’s entire filepath.
The Event is the first field in the Spreadsheet view, the order in which shots
appear on the timeline. Candidate media is the media that Nuke Studio is
testing the conform rules against.
FileName Match only the event's file name (no path) to the candidate media’s file name.
FileHead Match the event's file name head (no path, file extension, or padding) to the
candidate media’s file name.
PartialName Look for the event's name in the candidate media’s name and vice versa.
FolderName Look for the event's name in the filepath of the candidate media.
All rules are enabled by default, but you may occasionally need to disable rules if they cause
incorrect matches between a particular edit and set of source clips.
Tip: Use the Select/Deselect All buttons to quickly enable or disable rules.
• Patterns - sets the inclusion and exclusion parameters during the conform, separated by
spaces. For example, *mov *dpx would only include or exclude .mov and .dpx files.
You could also conform by name, such as BR_Shot*, which would only include or exclude
source clip names starting with that string.
Tip: It’s always a good idea to be as specific as possible with search locations during
conforms, but if the need arises, conform Rules and Patterns can save time.
3. Enable Accept best timecode match... to use the nearest source timecode match to conform the
event, if no rules are matched.
4. When Ignore clips with non-overlapping timecodes is enabled, any potentially matching source
clip whose timecode doesn't overlap the shot in question at all is ignored.
USER GUIDE
1263
Disabling Ignore clips with non-overlapping timecodes causes Nuke Studio to fall back to the
other selected conform rules, even if the timecodes don't overlap.
5. Check Conform shots that already have media if you want to update all timeline shots. By
default, the application doesn’t try to conform events that are not offline.
6. When Split sequences is enabled, any non-contiguous file sequences found by the conform are
split into separate clips, in the same way as when the split seq option is enabled in the file
browser.
7. Click OK to begin the conform process.
Nuke Studio attempts to conform the edits with the selected media.
A dialog box informs you of the success rate once the conform is complete.
Successfully matched media is placed in a new Conform bin in the project.
USER GUIDE
1264
Note: You can display the conform Rules matched for each spreadsheet object by hovering
the cursor over the required entry.
USER GUIDE
1265
Conforming with Pre-ingested Media | Notes on AAF Sequences
Conforming with Pre-ingested

Media
To conform with pre-ingested media:
1. If your source media has been ingested, you can drag-and-drop media from the bin view onto the
Match Media button.
See Ingesting Media for information on getting media into Nuke Studio.
2. Follow the Conform Options instructions described previously to complete the conform process.
If you want to conform a single entry in the spreadsheet, drag-and-drop the media from the bin
view onto the required entry in the spreadsheet.
USER GUIDE
1266
Conforming with Pre-ingested Media | Notes on AAF Sequences
Conforming individual, pre-ingested media doesn’t require Conform Options because Nuke
Studio already knows the exact location of the media and trusts your decision to replace a shot.
USER GUIDE
1267
About the Media Spreadsheet | Sorting and Custom Columns
About the Media Spreadsheet

All events in a sequence are displayed in an easy to read format in the spreadsheet including status,
the track it resides on, length, and the source file location.
After conforming, you can use the spreadsheet to locate source clips or replace shots in the timeline,
as well as massage timecodes if they are invalid.
The media spreadsheet displays each entry’s current media state:

• - the media was successfully conformed and its timecode is correct.
• - the media was successfully conformed, but the timecode is currently incorrect.
• - the media could not be conformed.
Note: Any source or destination field highlighted in yellow indicates that the entry has been
rounded down for display purposes.
See Managing Timelines for more information on importing tracks and reference media.
Sorting and Custom Columns

The spreadsheet can be organized in much the same way as accounting spreadsheets:
• Right-click the column headers to display the list of default columns available. Enable or disable
each column using the checkboxes.
• Click the required column to sort the spreadsheet in ascending or descending order, as indicated by
the arrow in the column header.
• Drag-and-drop column headers to reorder the spreadsheet as required.
• Add custom columns, such as Tags, using the Python API. See Help > Documentation for more
information on the Python API.
Tip: Nuke Studio's Project panel search functionality extends to the spreadsheet, allowing
you to enter strings and apply searches on all or partial matches with the option to include
metadata searches. Nuke Studio searches for items that match any of the input string and
displays only those items by default. See Sorting and Searching Media for more information.
USER GUIDE
1268
About the Media Spreadsheet | Spreadsheet Controls
Spreadsheet Controls
There are also a number of controls, accessed by clicking the cog icon, that determine the
spreadsheet’s appearance and behavior.
• Select Matching Name - when enabled, selecting an item in the spreadsheet highlights all items
with the same name.
• Select Linked - when enabled, selecting an item in the spreadsheet highlights other items linked to
it, such as audio tracks ingested with video tracks.
• Display Speed As - sets the unit used in the Speed column of the spreadsheet. Select either fps
(frames per second) or % (the percentage of the media frame rate).
• Retime method - sets the type of Speed retime applied on the timeline.
• Time Edit Behaviors - sets how source and destination In, Out, and Duration are calculated.
SeeRetiming Shots for more information on retime methods and Timeline Editing Tools for
source/destination calculations.
You can locate, display, reconnect, or rename shots directly from the spreadsheet.
• Hold Alt and click an entry to move the playhead to the shot’s In point on the timeline.
• Hold Alt and double-click an entry to move the playhead to the shot’s In point on the timeline and
zoom to fit the timeline view.
• Right-click a spreadsheet entry and select the required option:
USER GUIDE
1269
About the Media Spreadsheet | Spreadsheet Controls
• Open In - the associated source clip opens in the selected location, such as a Viewer.
• Project View - the associated source clip is highlighted in the bin view.
• Reconnect Media - attempt to reconnect the media from a specified location on disk, such as
when the source was originally on a drive that is no longer connected.
• Replace Clip - replaces the selected entry with a specified source clip. Nuke Studio assumes that
any source clip you choose is acceptable, regardless of timecode.
• Delete - deletes the selected entries from the spreadsheet and timeline.
• Tags - allows you to add tags to your selection directly from the spreadsheet view. See Using
Quick Tags for more information.
• Localize - allows you to control the localization of clips, tracks, and sequences from the
spreadsheet. See Localizing Media for more information.
• Effects - provides access to Create Comp and soft effects directly from the spreadsheet. See
Create Comp andSoft EffectsSoft Effects for more information.
USER GUIDE
1270
Adjusting Timecodes | Spreadsheet Controls
Adjusting Timecodes
You can easily adjust single or multiple event timecodes:
1. Select the invalid entry or entries in the spreadsheet.
2. Double-click in the Src In column.
3. Adjust the timecode as required. You can enter absolute or relative timecode values:
• Absolute - absolute timecodes contain eight digits and specify the new timecode for the event,
regardless of the current timecode.
Example Result
01:05:43:21 Sets the timecode at 1 hour, 05 minutes, 43 seconds, and 21 frames
01054321
• Relative - uses + and - values to alter the timecode relative to its current value. You can also use
h, m, and s to denote hours, minutes, and seconds.
Current Position Example Result
01:05:43:21 +1h 02:05:43:21
-110 01:05:42:11
+10000 01:06:43:21
-6m 00:59:43:21
The media changes state to .
Note: Timelines start at 01:00:00:00 by default, but you can change this to any value using
the Sequence panel.
USER GUIDE
1271
If you’re not sure what the timecode should be, you can:
• Hover the mouse over the target entry in the spreadsheet to view a timecode tooltip.
OR
• Examine the source clip’s metadata and calculate the correct Src In:
1. Right-click the required entry and select Project View.
2. Right-click the clip in the bin and select Open In > Metadata View, or press Alt+D.
The selected clip metadata is displayed in a floating pane.
USER GUIDE
1272
USER GUIDE
1273
Renaming Shots on the Timeline | Spreadsheet Controls
Renaming Shots on the Timeline

Once you’ve conformed your edit, you may want to rename shots on the timeline sequentially for
clarity.
To rename shots:
1. Select the shots to rename using the timeline or spreadsheet view.
2. Right-click on the timeline and select Editorial > Rename Shots.
Tip: You can also navigate to Timeline > Rename Shots or use the Alt+Shift+/ keyboard
shortcut.
The Rename Shots dialog displays.
3. Select the rename type from the dropdown:

• Simple Rename - all shots are replaced by the New Name specified.
• Find and Replace - a simple find and replace shot name. All selected shots containing the
specified Find pattern are substituted with the Replace pattern.
• Sequential Rename - rename shots sequentially using the Pattern, Start #, and Increment
fields.
• Match Sequence - allows you to select a sequence to copy shot names from, providing that
they use the same shots. For example, renaming shots on a 30 second timeline to mirror the
shot names from a 60 second timeline.
Note: You can only use sequences that reside in the same project and shots that have
overlapping frame ranges.
• Clip Name - all shot names are replaced by the name of the source clip they reference. This
option can be used to revert previous rename operations.
USER GUIDE
1274
Renaming Shots on the Timeline | Spreadsheet Controls
• Change Case - the case of all shot names is changed, as specified by the Case dropdown. For
example, selecting Title Case capitalizes the first character of each word.
4. Rename operations also accept token substitutions. The following tokens are recognized:
Token Resolves to
{clip} The name of the source clip referenced by the target shot.
{event} The EDL event number associated with the target shot.
{filename} The file name of the shot's source.
{fps} The frame rate of the sequence containing the rename target.
{sequence} The name of the sequence containing the rename target.
{shot} The name of the shot.
{track} The name of the track containing the rename target.
5. Enable Include Clips From Audio Tracks to rename audio shots as well as video shots.
6. Click Rename.
The selected shots are renamed as specified.
USER GUIDE
1275
Saving and Loading Projects | Spreadsheet Controls
Saving and Loading Projects

You can save your work in a project using the .hrox file extension. Projects can contain .nk scripts in
the form of shots added by using Create Comp. See Create Comp for more information.
If you quit the application without saving, you’ll be prompted to save or discard your changes:
Click the required button or press D for Don’t Save or S to Save.
Note: If you have a .nk script open in the same session, a second prompt is displayed so
you can save your script as well.
To save a project:
1. Navigate to File > Save Project or Save Project As...
OR
Use the Ctrl/Cmd+S or Shift+Ctrl/Cmd+S keyboard shortcuts respectively.
The Save Project dialog box displays.
2. Browse to the save location and enter a name for the project.
3. Click Save.
Your project is saved to the location specified and appends the .hrox file extension automatically.
To load a project:
1. Navigate to File > Open Project.
Tip: If you need to open a project that you worked on recently, you can select Open Recent
Project to quickly locate your work.
OR
Use the Ctrl/Cmd+O keyboard shortcut.
The Open Project dialog box displays.
2. Browse to the file location and click Open.
Your project opens and populates the necessary panel automatically.
USER GUIDE
1276
Autosaved Projects | Spreadsheet Controls
Autosaved Projects
The autosave function creates a temporary project save file at 5 minute intervals, but you can adjust
force project autosave after in the Preferences > General dialog. See Appendix A: Preferences for
more information.
At startup, the application scans for autosaved projects and displays a prompt if autosaves exist.
Click Yes to load the autosave or No to ignore and delete it.
Opening a project also uses the autosave functionality. If the autosave is more recent than the saved
project file, a prompt displays:
Click Yes to load the autosave file or No to load the original project file.
Note: Clicking No does not delete the autosaved project in this case.
USER GUIDE
1277
Managing Timelines
Timelines contain video and audio shots that reference the source clips in your project. Once the
conform process is complete, the timeline displays your clips in context and enables you make finer
edits. Timelines can contain any number of video sequences and audio tracks with each track
containing shots that reference the source clips in your project - making changes to shots in the
timeline does not affect the original source clip.
Nuke Studio also features real-time soft effects on the timeline and the ability to add shots containing
.nk scripts. See Soft Effects and Create Comp for more information.
Note: Conformed EDLs only support one video sequence. If you’ve created multiple EDLs
from the same edit, you can add each one into the timeline using the right-click New Track
> New Track(s) from EDL/XML option or the Import Track button in the spreadsheet tab.
See Adding Tracks to the Timeline.
• Video Toggles - quickly turn off and on video tracks during playback. Hold Alt and click to solo the
selected track. You can also enable and disable track blending and masking. See Blending Tracks on
the Timeline for more information.
• Disk Caching Controls - click and hold to display the disk caching options for the current timeline.
See Caching Frames in the Disk Cache for more information.
USER GUIDE
1278
Managing Timelines |
• Timecode - displays the timecode or frame number depending on the Time Display mode selected.
You can adjust the scale using the Scale Slider or by using the mouse wheel.
• Playhead Position - displays the playhead location synchronized with the contents of the Viewer.
• Video Tracks - contain all video sequences for the current timeline.
• Audio Tracks - contain all the audio clips for the current timeline.
• Audio Toggles - quickly mute audio or set the track output during playback to left, right, or mono.
• Track Lock - secure the selected track to disable all editing tools.
Tip: Selecting tracks while holding Ctrl/Cmd allows tools to affect multiple tracks at once,
such as locking, disabling, and resizing tracks.
Video tracks in multi-track timelines are read from the highest number track downward, for example
Video 3, Video 2, Video 1. As a result, if video is present on track 3, video on track 2 in the same time
slice is obscured.
In this example, although the playhead crosses clips on two video tracks, only the clip in Video 3 is
Audio tracks, on the other hand, play back simultaneously - all the audio tracks crossed by the
playhead in the example play back together, creating a complete audio backing for the video.
USER GUIDE
1279
Managing Timelines |
Tip: Enabling Preferences > Panels > Timeline > show frame end marker draws an extra
line on the timeline to the right of the playhead, indicating the end of the current frame.
USER GUIDE
1280
Adding Tracks to the Timeline |
Adding Tracks to the Timeline

You can add empty tracks to existing timelines or import other EDL, AAF, or XML edits - effectively
another sequence within the timeline.
To import EDL, AAF, or XML edits:

1. Select the required sequence in the project bin, right-click, and select Import > New Track(s)
from EDL/XML,
OR
2. Click Import Track in the spreadsheet tab.
3. Use the browser to locate the EDL, AAF, or XML files, select the file(s) and click Open to import the
sequence.
Note: If you’re importing EDLs, bear in mind that there is no guaranteed frame rate
information included in the file. Select the correct frame rate then click OK in the dialog
supplied.
4. Conform the new track as described in Conforming Sequences.
To add new tracks:

• Drag-and-drop a clip above or below existing tracks as shown,
USER GUIDE
1281
OR
• Right-click in the timeline and select New Track > New Video Track or New Audio Track.
USER GUIDE
1282
Note: You can also collapse and expand existing tracks using the right-click Editorial menu,
and resize the track header to accommodate longer track names.
USER GUIDE
1283
Adding Clips to the Timeline |
Adding Clips to the Timeline

The timeline allows you to add clips by simple drag-and-drop from either the Viewer or bins. Using
the Viewer restricts you to a single clip, the current clip, but you can drag as many clips as you like
from bins.
Tip: You can create a new sequence by dragging a clip to an empty timeline pane.
New timelines pick up their frame rate from the Project > Edit Settings > Sequence sub-menu by
default. Dropping a clip with a different frame rate on a new timeline displays a warning:
However, if the timeline is already populated and the clip you’re adding doesn’t have the same frame
rate as the timeline, you’re prompted to choose whether the clip’s duration or frame count is
retained.
Take care not to overwrite existing shots - the most recent clip overlays any existing shot. To avoid
this, do one of the following:
• Move the playhead to the target area of the timeline in the record Viewer , load the required clip
in a source Viewer , and then use Insert (N) or Overwrite (M) to place the clip into the timeline
at the playhead position on the lowest, unlocked track available.
Note: You can only Insert or Overwrite using clips from the current project.
USER GUIDE
1284
See Insert, Overwrite, and 3-Point Editing for more information on source/record editing.
• Use the Multi or Move/Trim tools to make space for the new clip and then drag-and-drop it in to
the space (see Using the Move/Trim Tool for more information).
• Drag-and-drop the new clip at the end of the sequence, then using the Multi or Move/Trim tools,
drag the new clip to an existing transition, hold down the Alt modifier, and drop the clip to Ripple
all other shots down the timeline.
USER GUIDE
1285
USER GUIDE
1286
Audio and the Timeline |
Audio and the Timeline

Audio tracks on the timeline are handled in much the same way as video tracks. By default, linked
audio and video tracks are edited at the same time, but you can lock either track and move them
independently or hold Alt to select a single track, if required.
Note: Although you can import and edit multi-channel audio, during playback audio is
mixed to 48 KHz stereo output.
Note: Nuke Studio does not currently support any QuickTime audio on Linux. Support for
audio on Linux is scheduled for a later release.
When both Viewer inputs contain clips, the audio output is set by the orange marker in the color
picker information bar, displayed by clicking . In the following example, input A is providing the
audio output:
The volume slider in the upper-right corner of the Viewer controls the output level for that Viewer
only.
Audio output for shots can be toggled between left, right, and mono using the audio toggles in the
track header. Click on the icon to cycle between outputs:
USER GUIDE
1287
You can also control audio on a per track and per shot basis. Audio track headers and shots have
independent volume controls in the timeline and Properties tab.
• Track headers - click and hold the mute icon on the header to display the volume slider.
• Shots - select an audio shot and click on the Properties tab to display the Volume control.
Tip: You can control the volume on multiple shots simultaneously by holding Ctrl/Cmd and
selecting the required items before adjusting the volume slider.
USER GUIDE
1288
The Preferences > Panels > Viewer (Sequence) sub-menu contains audio controls allowing you to
control the volume level for all new Viewers. See Appendix A: Preferences for more information.
Note: If the frame rate drops too low, audio is automatically muted and the speaker
changes to the no audio playback icon.
USER GUIDE
1289
WAV Shots |
WAV Shots
Audio can be recorded at the same time as shooting the video or it can be unrelated to the shoot, for
example sound effects or music. You can add .wav clips to the timeline in two ways:
• Drag-and-drop - drag your .wav clip to a timeline audio track and drop it in to place.
• Navigate to File > Import File(s) or Import Folder(s).
Tip: You can also use the Ctrl/Cmd+I and Ctrl/Cmd+Shift+I keyboard shortcuts.
Use theTimeline Editing Tools to move the clip into place and set its output.
USER GUIDE
1290
Displaying Audio Waveforms |
Displaying Audio Waveforms

Visualizing an audio waveform helps synchronization with video events, and Nuke Studio displays
waveforms in the timeline by default.
Audio shots are manipulated in the same way as video shots, so using waveforms in conjunction with
the Timeline Editing Tools on page 1389 enables you to quickly synchronize audio and video events.
Audio shots also support Fade In, Fade Out, and Dissolve transitions in the same way as video. See
Adding Transitions for more information.
You can toggle the waveform display on and off by right-clicking in the timeline and selecting View >
Audio Waveforms. You can also control how the waveform appears, when enabled. Open the
Preferences and navigate to Panels > Timeline to toggle between full and half waveforms.
Displaying waveforms in audio-heavy projects can cause significant slow down on the timeline, so
Nuke Studio includes a preference to limit how much system memory is available for waveform
display. In the Preferences, navigate to Performance > Caching > Audio Waveforms and set the
waveform memory control to the required amount.
USER GUIDE
1291
Audio Scrubbing |
Audio Scrubbing
Nuke Studio's timeline supports audio scrubbing, allowing you synchronize audio and video more
easily while scrubbing the playhead. Audio scrubbing is enabled by default, but you can disable it by
right-clicking in the timeline tab and clicking Audio > Audio Scrubbing or by pressing
Ctrl/Cmd+Alt+S.
Audio shots cache temporarily to increase responsiveness during scrubbing. If you need to clear the
audio cache, navigate to Cache > Clear Audio Cache.
Note: Audio scrubbing is not currently available through monitor output cards. Audio
scrubbing is only supported through internal audio output devices.
USER GUIDE
1292
Synchronizing Audio and Video |
Synchronizing Audio and Video

Nuke Studio allows you to massage the synchronization between audio and video tracks using audio
latency adjustment during playback in the Viewer, or by a default amount in the Preferences >
Panels > Viewer (Sequence) sub-menu.
Note: Latency adjustments can take a few seconds to affect the audio track.
1. Mark a portion of the timeline containing the target audio and video shots using In and Out
markers.
2. Press or use the L keyboard shortcut to begin playback.

3. Click the Viewer settings icon and increment the latency using the controls in the popup.
4. Adjust the latency until the tracks are in sync.
USER GUIDE
1293
PulseAudio on Linux | Stopping PulseAudio
PulseAudio on Linux
PulseAudio on Linux distributions has been linked with fluctuating frame rates due to the latency
when retrieving audio samples. If detects that your setup is running PulseAudio alongside the
application, a warning message displays.
Stopping PulseAudio
You can disable PulseAudio for the current user or all users on a machine. To stop the daemon, do
the following:
Note: PulseAudio restarts automatically when you restart you machine, but you can prevent
this by navigating to System > Preferences > Startup Applications and disabling the
PulseAudio Sound System.
1. Open the ~/.pulse/client.conf file to disable PulseAudio for the current user,
OR
Open the /etc/pulse/client.conf file to disable PulseAudio for all users.
2. Set the following attribute and ensure the line is not commented out:
autospawn = no
3. Call pulseaudio --kill to end the PulseAudio process.
4. Call ps -e | grep pulse to check the process stopped correctly.
Note: Ending PulseAudio while other applications are running may disable audio output.
Stop and start the application to re-enable audio output. Additionally, the desktop audio
slider may be removed.
USER GUIDE
1294
PulseAudio on Linux | Restarting PulseAudio
Restarting PulseAudio
To start the PulseAudio daemon, do the following:
1. Open the ~/.pulse/client.conf file to enable PulseAudio for the current user,
OR
Open the /etc/pulse/client.conf file to enable PulseAudio for all users.
2. Set the following attribute and ensure the line is not commented out:
autospawn = yes
3. Call pulseaudio --start to start the PulseAudio daemon.
4. Call ps -e | grep pulse to check the process started correctly.
USER GUIDE
1295
Using Reference Media | Restarting PulseAudio
Using Reference Media

Importing a reference version of your timeline enables you to compare your current timeline against
the reference media to avoid issues with continuity, missed frames, and so on.
To import reference media, click Set Reference Media and use the browser to locate the required
file.
The reference media is automatically imported into Reference tracks, pushing existing tracks
outward, and marked with the Reference Media tag.
After importing the reference media, use the show/hide icon or A/B input tools to compare the
current timeline against the reference clip. SeeComparing Media for more information.
USER GUIDE
1296
Comparing Media | Restarting PulseAudio
Comparing Media
The Viewer A/B tools allow you to quickly compare media using the two Viewer input buffers. Select a
clip, sequence, shot or track and press 1 or 2 to place your selection in the Viewer input buffers. You
can also drag-and-drop items into the input buffers using the Viewer hotspots.
the Viewer background being “added” to the image. If you’re working with un-premultiplied
images, set the Viewer background to Black. See Appendix A: Preferences for more
information.
When the Viewer input buffers contain sequences, the A and B dropdowns control what is displayed
in the Viewer using track names and tags. Selecting a track or tag from the dropdown displays the
selected media in the Viewer.
Use the wipe, stack, horizontal, and vertical modes to control how the buffers are displayed in the
Viewer.
Note: If you're working in a multi-view project, using stereo footage for example, you can
set which view is output in the A and B buffers using the Views buttons over the Viewer. See
Displaying Views in the Viewer for more information.
USER GUIDE
1297
Comparing Media | Restarting PulseAudio
The wipe and stack modes also allow you to blend the two buffers together, and in the case of wipe
mode, provides a handle in the Viewer to quickly wipe between the two inputs.
The color picker overlay displays a description of the contents of the A and B inputs, or No Clip
when there is no clip at the playhead, for instance, when there is a gap in a timeline or if a track is
disabled.
The orange triangle in the overlay denotes the clip currently supplying audio and timecode
information in the Viewer.
USER GUIDE
1298
Caching Frames in the Playback Cache | Restarting PulseAudio
Caching Frames in the Playback

Cache
The playback cache places frames in RAM for rapid retrieval during playback, rather than creating files
locally as with Caching Frames in the Disk Cache orLocalizing Media.
The white bar under the Viewer represents the contents of the playback cache, a full bar indicating
that the entire clip or timeline is currently accessible from RAM, optimizing playback. You can:
• Temporarily disable caching using the pause button above the Viewer, or use the P keyboard
shortcut.
Clicking pause again, resumes caching from the playhead position.

• Flush the cache completely by navigating to Cache > Clear Playback Cache. Caching is
automatically paused after flushing, but clicking the pause button resumes caching from the
playhead position.
There are also a number of Preferences that affect how much RAM is available and when caching
should occur. To set the caching behavior:
1. Navigate to Nuke > Preferences (OS X) or Edit > Preferences (Linux and Windows),
OR
Use the Preferences keyboard shortcut Shift+S.
2. Select Performance > Caching and set the total RAM cache available using the playback cache
size field.
Note: You can't set this to a value higher than 80% of the memory available (rounded down
to the nearest half-GB). For example, if you have 6 GB of memory available, the maximum
cache size available 4.5 GB.
3. You can enable free timeline playback RAM to discard any frames cached to RAM (the white bar
in the timeline Viewer) when you switch to the Node Graph within Nuke Studio, freeing the RAM
for use in Nuke.
Note: When you switch back to the timeline, the cached files are re-cached, which can be
time consuming.
USER GUIDE
1299
Caching Frames in the Playback Cache | Restarting PulseAudio
4. Enable pause caching when the application goes to the background to pause playback
caching when the application loses focus.
When you click back into Nuke, caching picks up from where it stopped.
5. Enable clear cache when the application goes to the background to flush the playback cache
when the application loses focus.
When you click back into Nuke, caching starts again from the position of the playhead.
USER GUIDE
1300
Caching Frames in the Disk Cache | Restarting PulseAudio
Caching Frames in the Disk Cache

Timeline Disk Caching provides reliable playback for more complex timelines by rendering frames to
disk using the GPU. The cache provides persistent frames per edit in the timeline that only needs
updating for full changes on the edit, such as adding a soft effect. For editorial changes, only the new
frames need to be cached.
Note: Frames are always cached at the sequence resolution, regardless of source clip
format and Viewer proxy settings.
You can cache whole sequences, selections of clip ranges, and frame ranges specified using In and
Out points. Files in the disk cache are frames identical to what you see rendered in the timeline
Viewer, written into .exr sequences, and saved in NUKE_TEMP_DIR/TimelineCache by default.
Tip: You can find the location of Nuke's general cache directory from within Nuke by hitting
X on your keyboard, when the focus is on the Node Graph, and then running the following
Tcl command:
getenv NUKE_TEMP_DIR
Cached frames are represented in the timeline by the state of the timeline cache icon:
- None of the frames in the current timeline are cached.

- The current timeline is partially cached.
- The current timeline is fully cached.
Cached frames are represented in the Viewer with an orange bar, under the RAM cache bar, which is
white by default.
USER GUIDE
1301
Caching Frames in the Disk Cache | Restarting PulseAudio
See Caching Sequence Ranges, Caching Selected Shot Ranges, and Caching In/Out Ranges for more
information.
You can set the cache directory location, size of the timeline cache, and type of EXR compression used
by:
• Clicking and holding the timeline cache icon and selecting Cache Settings, or
• Opening the Preferences and navigating to Performance > Caching.
USER GUIDE
1302
Caching Sequence Ranges | Restarting PulseAudio
Caching Sequence Ranges

Caching a sequence caches exactly what you see when play through the entire timeline. As the
playhead walks the timeline, all tracks are read from the highest track downwards, so what you see in
the Viewer is what Hiero caches to disk.
Hiero can only cache one sequence at a time, either from the timeline itself, the Cache menu, or the
Project bin:
Tip: You can pause caching at any time by navigating to Cache > Disk Cache > Pause in the
menu bar.
• Select the required sequence in the timeline tab, click and hold the icon, and then select Cache
Sequence Range.
• Select the required sequence in the timeline tab and then select Cache > Disk Cache > Cache
Sequence Range from the menu bar.
USER GUIDE
1303
Caching Sequence Ranges | Restarting PulseAudio
• In the Project bin, right-click the sequence you want to cache, and then select Cache > Cache
Sequence Range.
See Clearing Cached Frames for information on how to clear frames from the disk cache.
USER GUIDE
1304
Caching Selected Shot Ranges | Restarting PulseAudio
Caching Selected Shot Ranges

Caching shot ranges allows you to be more selective than caching an entire sequence, though it works
in the same way. All tracks are read from the highest track downwards, so what you see in the Viewer
for the shot selection is what Hiero caches to disk.
menu bar.
As a result of this, you may not get the frames you expect. For example, caching the shot in the image
produces frames from the three shots on the highest track, including the soft effect, not the selected
reference media.
You can cache clip ranges from the timeline itself or from the Cache menu:
• Select the required clip range in the timeline tab, right-click and then select Disk Cache > Cache
Selected Shot Ranges.
USER GUIDE
1305
Caching Selected Shot Ranges | Restarting PulseAudio
• Select the required clip range in the timeline tab and then select Cache > Disk Cache > Cache
Selected Shot Ranges from the menu bar.
USER GUIDE
1306
Caching In/Out Ranges | Restarting PulseAudio
Caching In/Out Ranges

Caching ranges uses In and Out points set on the timeline to determine what Hiero caches to disk. All
tracks are read from the highest track downwards, so what you see in the Viewer between the In and
Out markers is what Hiero caches to disk. See Using In and Out Markers for more information.
menu bar.
After setting In and Out points, you can cache that range from the timeline itself or from the Cache
menu:
• Right-click in the timeline tab and then select Disk Cache > Cache In/Out Range.
• Select Cache > Disk Cache > Cache In/Out Range from the menu bar.
USER GUIDE
1307
Caching In/Out Ranges | Restarting PulseAudio
USER GUIDE
1308
Clearing Cached Frames | Clearing All Cached Frames
Clearing Cached Frames

Frames in the timeline cache can be cleared in a number of ways, including Clear All frames across
all projects and the more selective Clear Range option.
Clearing All Cached Frames

The Clear All option is only available from the Preferences dialog under Performance > Caching,
because it clears all frames from all projects and should be used with care. Clicking Clear All displays
a warning dialog containing all the projects that are affected.
Click OK to proceed or Cancel to retain all the frames in the cache.
Clearing Sequences
Instead of using the Clear All option, you can clear certain sequences within a project. You can clear a
sequence from the timeline itself, the Cache menu, or the Project bin:
• Select the required sequence in the timeline tab, click and hold the icon, and then select Clear
Sequence Range.
USER GUIDE
1309
Clearing Cached Frames | Clearing Sequences
• Select the required sequence in the timeline tab and then select Cache > Disk Cache > Clear
Sequence Range from the menu bar.
• In the Project bin, right-click the sequence you want to clear, and then select Cache > Clear
Sequence Range.
USER GUIDE
1310
Clearing Cached Frames | Clearing Selected Shot Ranges
Clearing Selected Shot Ranges

You can also clear selected ranges of shots from a sequence, rather than the entire sequence, from
the timeline itself or from the Cache menu:
• Select the required clip range in the timeline tab, right-click and then select Disk Cache > Clear
Selected Shot Ranges.
USER GUIDE
1311
Clearing Cached Frames | Clearing In/Out Ranges
• Select the required clip range in the timeline tab and then select Cache > Disk Cache > Clear
Selected Shot Ranges from the menu bar.
Clearing In/Out Ranges

Clearing frames using In and Out markers is the most fine-grained method for removing frames from
the timeline cache. Only the frames bracketed by the markers are removed, regardless of how the
frames were cached originally. See Using In and Out Markers for more information.
After setting In and Out points, you can clear that range from the timeline itself or from the Cache
menu:
• Right-click in the timeline tab and then select Disk Cache > Clear In/Out Range.
USER GUIDE
1312
Clearing Cached Frames | Clearing In/Out Ranges
• Select Cache > Disk Cache > Clear In/Out Range from the menu bar.
USER GUIDE
1313
Viewing Multi-Format Timelines | Clearing In/Out Ranges
Viewing Multi-Format Timelines

All sequences and shots added to them adopt the output resolution and clip reformat settings
from the Preferences dialog under Project Defaults > General. You can override these settings on
a per Project or per Sequence basis.
The Preferences dialog controls the default settings for

new projects. Existing projects are unaffected by changes
to the preferences.
The Project Settings control the default settings for new

sequences and shots in the current project. Existing
sequences and shots are unaffected.
Project Settings override Preference settings.
The Sequence settings control the current sequence.
Sequence settings override Project Settings and

Preference settings for the current timeline.
The Properties settings control shots in the current

sequence.
Properties settings override Sequence settings, Project

Settings, and Preference settings for shots on the
current timeline.
Reformatting applied to shots on a timeline carry over into any export, including Create Comp. The
reformat options in the Export dialog are applied after the transforms applied here. See Exporting
from Nuke Studio for more information.
USER GUIDE
1314
Viewing Multi-Format Timelines | Setting the Sequence Format
Setting the Sequence Format

The current sequence format is controlled by the Output Resolution dropdown in the timeline
Sequence tab. The selected resolution is applied to all shots on the timeline, but by default, shots
retain their native resolution. For example, setting the Output Resolution to 2k_DCP on a timeline
containing HD_1080 shots results in the output resolution being larger than the shot.
Tip: Enabling the Custom Sequence Format guide at the top of the Viewer makes it easier
to see differences between output resolution and shot format.
A 1920x1080 shot at native The same shot with Output

resolution. Resolution set to 2048x1080.
Similarly, setting the Output Resolution to a format smaller than the shot's native resolution results
in the clip resolution being larger than the sequence resolution.
A 1920x1080 shot at native The same shot with Output

resolution. Resolution set to 720x486.
Shots with a higher resolution than the sequence format enable you to apply soft effects, such as
Transform, more easily. See Soft Effects for more information.
USER GUIDE
1315
Viewing Multi-Format Timelines | Reformatting Shots
Reformatting Shots
Shots on the timeline retain the source clip's resolution by default. You can force shots to the
sequence format using the timeline's Properties tab:
1. Select the target shots on the timeline.
2. Click the Properties tab to display the shot options.
Note: The Volume control is only available when you have an audio shot selected.
3. Click the Clip Reformat dropdown and selected To Sequence Resolution.

4. The shot is reformatted to the Output Resolution specified on the Sequence tab.
5. Depending on the source clip, you may need to change the Resize Type:
• Width - scales the original until its width matches the format’s width. Height is then scaled to
preserve the original aspect ratio.
• Height - scales the original until its height matches the format’s height. Width is then scaled to
preserve the original aspect ratio.
• Fit - scales the original until its smallest side matches the format’s smallest side. The original’s
longer side is then scaled to preserve original aspect ratio.
• Fill - scales the original until its longest side matches the format’s longest side. The input’s
shorter side is then scaled to preserve original aspect ratio.
USER GUIDE
1316
Viewing Multi-Format Timelines | Reformatting Shots
• Distort - scales the original until all its sides match the lengths specified by the format. This
Tip: When Clip Reformat is To Sequence Resolution, disabling center places the clip at
the bottom-left of the Output Resolution, rather than the center.
USER GUIDE
1317
Refreshing and Replacing Shots | Reformatting Shots
Refreshing and Replacing Shots

During the post process, media inevitably changes location or form. Nuke Studio can reload or
replace your media using the refresh, rescan, reconnect, and replace functions.
Though all four options deal with reloading shots, each has a particular use dependent on context:
• Reconnect Media - allows you to redirect the file path when the source file location changes.
• Replace Clip - replaces the selected shot with a specified source clip. Nuke Studio assumes that any
source clip you choose is acceptable, regardless of timecode.
• Refresh Clips (F8) - allows you to reload the shot when the source file location has not changed,
such as when work has been done on the clip offline. Selecting refresh only refreshes the clip’s
current frame range.
• Rescan Clip Range (Alt+F5) - similar to Refresh Clips, above, but rescan also checks for additional
frames that may have been added to the source file and adds them to the shot’s frame range.
• Set Soft Trims - sets the files handles on the selected clip(s). SeeSetting Soft Trims for more
information.
USER GUIDE
1318
Setting Soft Trims | Reformatting Shots
Setting Soft Trims

Soft Trims limit the handles on shots to a pre-defined amount, simulating In and Out points on the
source clips, allowing you to use other timeline tools on the shots such as Slip Clip and Slide Clip.
To set Soft Trims on a shot(s):

1. Select the shot(s) on the timeline.
2. Right-click and select Clip > Set Soft Trims.
The Set Soft Trims dialog displays.
3. Set the number of frames to add to the head and tail of each shot:
• Use full available range - sets the handles to the full extent of the source clip frame range.
• Use Frames - adds the specified number of frames to the head and tail of the shot(s).
4. Click OK to add the specified number of handles.
If the handles requested are not within the available frame range, a warning dialog displays with a
suitable correction for each selected shot.
Click Yes to accept, or No to abort the operation.
USER GUIDE
1319
Setting Soft Trims | Reformatting Shots
Note: With shots used in multiple sequences, click Yes to All to accept the correction in all
instances.
USER GUIDE
1320
Enabling and Disabling Shots | Reformatting Shots
Enabling and Disabling Shots

You can temporarily enable or disable tracks and shots on the timeline to selectively view your media
without removing shot(s), for example if you wanted to view to lower level video tracks within a
timeline.
To enable or disable a track or shot(s):

1. Select the item(s) you want to enable or disable.
2. Right-click on a highlighted item and select Editorial > Disable Track or Disable Items to disable
the selection.
Tip: You can also use the D keyboard shortcut to disable or enable your selection.
Disabled items appear gray, and are effectively removed from the timeline.
3. Right-click the item and select Enable Clip, or press D again, to re-enable the clip.
USER GUIDE
1321
Adding Transitions | Reformatting Shots
Adding Transitions
Nuke Studio supports basic video and audio fade transitions as well as dissolves between shots on
the same track. Transitions come in three flavors:
• Fade in - fades in from black on a single shot.
• Fade out - fades out to black on a single shot.
• Dissolve - fades out from one shot and into the next, by merging frames.
Tip: Once a transition is in place, it can be nudged in the same way as an edit using the ,
(comma) and . (period) keyboard shortcuts, providing the required handles exist.
To add a fade transition:

1. Right-click the target shot and select Editorial > Add Transition > Fade In, Fade Out, Audio Fade
In or Audio Fade Out to add the fade icon.
2. Adjust the fade by dragging the fade icon using the Multi Tool or Move/Trim tool.
To add a dissolve transition:
USER GUIDE
1322
Adding Transitions | Reformatting Shots
Note: You can only add dissolves between shots when they're on the same track and have
sufficient handles available on both sides of the transition.
1. Select the Multi Tool or Roll Edit tool and hover the mouse pointer over an edit between two
shots.
Tip: Clicking and holding the edit point displays available handles as a red overlay.
2. Right-click and select Editorial > Add Transition > Dissolve or Audio Crossfade, or use the
Ctrl/Cmd+T keyboard shortcuts, to add the dissolve icon to the edit.
3. Adjust either side of the dissolve by dragging the icon, in a similar way to using the Multi Tool or
Move/Trim tool.
USER GUIDE
1323
Invalid Transitions | Reformatting Shots
Invalid Transitions
Transitions are controlled in a similar way to shots, in that you can drag-and-drop them, but with the
following restrictions:
• A fade can not be dragged past the ends of the shot it's attached to, and if the item is deleted, the
fade is deleted with it.
• Dissolve ends can not be dragged past the ends of the shots they are attached to, and if both items
are deleted, then the dissolve is also deleted.
If only one of the shots linked by the dissolve is deleted, the transition remains so that another item
can be placed on the other side.
Invalid transitions are colored red on the timeline. In most cases, adjusting the length of the
transition should be enough to correct the error.
USER GUIDE
1324
Retiming Shots | Reformatting Shots
Retiming Shots
In addition to transitions, Nuke Studio supports constant retimes on shots. Decreasing the speed of a
shot causes frames to play more than once, whereas increasing the speed skips frames.
Note: Audio is not currently supported for retimes and is automatically muted to avoid
playback problems.
To retime shots using the Speed column in the spreadsheet:

1. Select the required event(s) in the spreadsheet view.
2. Click the cog icon and select the required Retime method:
• Keep source duration - the shot length is altered on the timeline depending on the retime
applied.
For example, retiming a shot to 50% renders frames 1, 1, 2, 2, 3, 3, 4, 4, and so on in the Viewer,
and as a result, the item’s length is doubled on the timeline.
Retiming a shot to 200% renders frames 1, 3, 5, 7, and so on in the Viewer, but the item’s length
is halved on the timeline.
• Keep timeline duration - the shot length on the timeline is maintained regardless of the retime
applied.
For example, retiming a shot to 50% renders frames 1, 1, 2, 2, 3, 3, 4, 4, and so on in the Viewer,
but the item’s length on the timeline remains the same, effectively removing the second half of
the item.
Retiming a shot to 200% renders frames 1, 3, 5, 7, and so on in the Viewer, but the item’s length
on the timeline remains the same. If no extra frames are available from the source, the item is
filled with black frames.
3. Double-click the Speed column and enter the retime value.
The following example shows a shot and the results of 50% and 200% retimes with the Keep source
duration and Keep timeline duration methods selected.
Notice that the Keep timeline duration method doesn’t change the length of the shot on the
timeline and inserts blank filler frames on the 200% retime?
USER GUIDE
1325
Original clip
50% Source retime 200% Source retime
50% Destination retime 200% Destination retime
You can also retime shots using the Src, Dst, and Duration columns of the spreadsheet, though the
calculation method depends on the Time Edit Behaviors applied.
1. Select the event(s) in the spreadsheet view.
2. Click the cog icon and select the required Time Edit Behaviors:
Modify Using Result
Src In Retime Adjusts the event’s Src In and retimes the remaining frames to maintain
Dst Duration.
Before and after a 2 second Src In increase:
USER GUIDE
1326
Modify Using Result
Src Out Retime Adjusts the event’s Src Out and applies a retime to maintain Dst
Duration.
Before and after a 2 second Src Out increase:
Src Dur Retime Adjusts the event’s Src Dur and Src Out, and applies a retime to maintain
Dst Duration.
Before and after a 50 frame Src Dur increase:
Dst In Retime Adjusts the event’s Dst In and retimes the remaining frames to maintain
the relationship between Dst In and Out.
Before and after a 2 second Dst In increase:
Dst Out Retime Adjusts the event’s Dst Out and retimes the remaining frames to
maintain the relationship between Dst Out and In.
Before and after a 2 second Dst Out increase:
USER GUIDE
1327
Modify Using Result
Dst Dur Retime Adjusts the event’s Dst Dur and Dst Out, and applies a retime to
accommodate the new Dst Duration.
Before and after a 50 frame Dst Dur increase:
3. Adjust the values as required to retime the shot(s) by the specified amount.
Note: Any source or destination field highlighted in yellow indicates that the entry has been
rounded down for display purposes.
To retime a shot using the Timeline menu:

1. Select the required shot(s) on the timeline.
2. Navigate to Timeline > Retime.
The Clip Speed dialog displays.
3. Enter the required retime value as a percentage.
4. Select the required retime method using the dropdown:
• Keep timeline duration - the shot length on the timeline is maintained regardless of the retime
applied. When increasing speed, if no extra frames are available from the source, the shot is
filled with black frames.
• Keep source duration - the shot length is altered on the timeline depending on the retime
applied. For example, a 200% retime halves the length of the item.
• Anchor current frame - the shot length on the timeline and the current frame’s position are
maintained after the retime. When increasing speed, if no extra frames are available from the
source, the shot is filled with black frames.
5. Click OK to retime the shot(s).
USER GUIDE
1328
Using Freeze Frames | Reformatting Shots
Using Freeze Frames

The freeze frame feature enables you to create shots of any length using a single frame. To achieve
this, the application takes the first frame of the shot and applies a 0% retime, which is reversible by
applying a 100% retime.
To freeze frame shots:

1. Select the target item(s) on the timeline.
2. Right-click the item and select Editorial > Make Freeze Frame,
OR
Navigate to Timeline > Make Freeze Frame.
Note: Freeze Frames can also be created using the spreadsheet retime modes to modify
Src Dur to 0, or make Src In and Src Out equal.
3. The selection is retimed to 0% and colored blue on the timeline for easy identification.
4. Use the Multi Tool or Slip Clip to set the freeze frame from the available range.
5. Drag the item’s edit points, using the Multi Tool or Move/Trim as required, to set the length of
the shot. There’s no upper limit for the length of a freeze frame shot.
USER GUIDE
1329
Blending Tracks on the Timeline | Adding New Blend Tracks
Blending Tracks on the Timeline

Nuke Studio allows you to perform merges between tracks in the timeline, for example overlaying a
logo on a shot without heading into the Compositing environment. Tracks that are designated as
blend tracks have a blue header in the timeline for convenience and are blended using a sub-set of
Nuke's Merge node operations.
See the Nuke Online Help for a full description of the available blend modes.
the Viewer background being “added” to the image.
You can add soft effects to blended tracks as normal (see Soft Effects for more information) and
blended tracks are included along with the shot in Export operations. See Exporting from Nuke
Studio and Create Comp for more information.
Adding New Blend Tracks

1. Right-click in the timeline, select New Track > New Video Track Blend, and then choose the
blend mode to apply.
USER GUIDE
1330
Blending Tracks on the Timeline | Converting Tracks to Blend Tracks
A new track is added at the top of the track stack, colored blue to indicate that it's going to be
blended over the track below.
2. Add the required shot to the blend track as you would any other shot. See Adding Clips to the
Timeline for more information.
3. Click and hold the Blend icon to select the blend mode.
4. Click the Blend icon to toggle blending on and off.
Converting Tracks to Blend Tracks

1. Click the Blend icon to toggle blending on,
OR
Right-click in the header of the target track and select Editorial > Blend Mode > Enable Track
Blend.
USER GUIDE
1331
Blending Tracks on the Timeline | Masking Blended Tracks
The selected track is converted into a blend track, colored blue to indicate that it's going to be
blended over the track below.
2. You can add shots to the blend track as you would any other shot. See Adding Clips to the
The Viewer displays the higher track blended with the track below.
3. Click and hold the Blend icon to select the blend mode.
Masking Blended Tracks

Masking limits the effect of the blend track to just those areas covered by the alpha channel in the
blend image. For example, using the Multiply blend mode with masking disabled multiplies the
background plate in non-alpha areas, which may not be the result you require.
USER GUIDE
1332
Blending Tracks on the Timeline | Masking Blended Tracks
Background track B. Blend track A.
A multiplied by B with Masking A multiplied by B with Masking

disabled. enabled.
See the Nuke Online Help for a full description of the available blend modes.
Click the Mask icon to toggle alpha masking on and off. The mask option also carries over into the
Node Graph when you Create Comp for a masked blend operation:
Comp of A over B using the multiply merge The same comp using the multiply merge
operation with masking disabled. operation, but with masking enabled.
USER GUIDE
1333
Stereoscopic and Multi-View
Projects
Nuke Studio provides multi-view support for as many views as you need. The views do not have to be
stereo pairs, but since that is the most obvious application, these pages mainly deal with stereoscopic
projects. See Creating Views in a Project and Importing Source Clips for more information.
Existing views inside a project are managed in the Viewer, timeline, and in the Properties panel of
most soft effects. In the Viewer, all views in the current project are represented by buttons that allow
you to switch between views with a single click. See Displaying Views in the Viewer for more
information.
The timeline employs a views button that allows you to switch between views per track. See
Displaying Views in the Timeline for more information.
USER GUIDE
1334
Stereoscopic and Multi-View Projects |
The Properties panel includes a split button for controls that support multiple views. Split
controls only affect individual views. See Applying Changes to Selected Views for more information.
You can create comps or export multi-view shots and effects in a similar way to regular shots. See
Exporting Multi-View Source Clips for more information.
USER GUIDE
1335
Creating Views in a Project | Creating and Managing Views
Creating Views in a Project

Views in Nuke Studio are managed in the Project Settings. Views can be processed separately or
together and you can see the effect of your changes on each view. If you're working with multi-view
source clips, such as .exr and .sxr, Nuke Studio offers to create views for you on import. See
Importing Multi-View Clips for more information.
Creating and Managing Views

If you're working with single-view clips, it's a good idea to create the project views before you import
your footage so that the views are assigned correctly. If you're working with multi-view clips, the views
are set up automatically on import.
1. Select Project > Edit Settings.
2. Go to the Views sub-menu. The available views are listed in the Views panel.
3. If you want to remove the view called main and add views called left and right, click the Set up
views for stereo button. The two views are assigned colors. To change the colors, double-click on
the color field and select another color from the color picker that opens.
4. Enable Use colors in UI to apply the selected color for each view to the associated button above
the Viewer.
USER GUIDE
1336
Creating Views in a Project | Creating and Managing Views
You can add and remove views using the + and - buttons or move views using the up and down
arrows above the views panel.
Each view has its own button above the Viewer controls.
Tip: If you decide that you only need the main view in the project, click Set up Views for
Mono in the Project Settings.
See Importing Multi-View Clips for information on reading source clips in to Nuke Studio.
USER GUIDE
1337
Importing Source Clips | Importing Single-View Clips
Importing Source Clips

Nuke Studio supports multi-view clips in two formats:
• single-view - formats such as .dpx and .jpg, with file names that Nuke Studio can interpret as multi-
view using specific naming conventions. For example:
myStereo_comp1_left.####.dpx and myStereo_comp1_right.####.dpx
See Importing Single-View Clips for more information.

• multi-view - formats such as .exr, .sxr, and .mov, that support multiple views per file. For example:
myStereo_comp1.####.exr containing two layers called left and right.
Note: Nuke Studio can assign any shot on the timeline to any view in the project, but stereo
is the most common use case.
See Importing Multi-View Clips for more information.
Importing Single-View Clips

After setting up views in the Project Settings, Nuke Studio can interpret single-view clips as multi-
view using specific naming conventions. These conventions can be applied to the file names
themselves or to the directories in which they reside.
File Name Variables

If the images you want to read in contain a view name or the initial letter of one (for example, left,
right, l or r) in their file names, replace this with the variable %V or %v in the file name. Use %V to
replace an entire view name (for example, left or right), and %v to replace an initial letter (for
example, l or r). When a variable is used, Nuke Studio reads in the missing inputs and combines all
inputs into a single output.
Note: You can enable detect views in the file browser to automatically substitute the %V or
%v variable where possible.
To use the %V and %v variables manually:
USER GUIDE
1338
1. Click File > Import File(s) or press Ctrl/Cmd+I to display the file browser.
2. Locate a single-view file, for example pubstitch.left.####.dpx
3. Replace the view name with the %V variable, continuing the example pubstitch.%V.####.dpx
4. Click Open.
Nuke Studio reads in both pubstitch.left.####.dpx and pubstitch.right.####.dpx with the same
Read node, provided that views called left and right exist in the Project Settings.
Note: Mac and Linux operating systems can be case-sensitive or case-insensitive. If your OS
is case-sensitive, you'll need to make sure you use the correct case when naming your left
and right views, as the %v variable can only retrieve the case used in the view name.
Both input images are combined into a single source clip, marked with a icon, which can
display any of the combined views.
USER GUIDE
1339
See Displaying Views in the Viewer and Displaying Views in the Timeline for more information.
Directory Name Variables

You can also use the %V and %v variables at a directory level. For example, if you have set up views
called cam3, left, and right, and you have the following directories and files:
stereo/pubstitchcam3/pubstitch_cam3.####.dpx
stereo/pubstitchleft/pubstitch_left.####.dpx
stereo/pubstitchright/pubstitch_right.####.dpx
If you read in pubstitch_cam3.####.dpx and change the file path to

stereo/pubstitch%V/pubstitch_%V.####.dpx, all three inputs are read in with the same Read node,
providing that the cam3, left, and right views exist in the Project Settings.
USER GUIDE
1340
All three input images are combined into a single source clip, marked with a icon, which can
display any of the combined views.
See Displaying Views in the Viewer and Displaying Views in the Timeline for more information.
USER GUIDE
1341
Importing Source Clips | Importing Multi-View Clips
Importing Multi-View Clips

Nuke offers to automatically create views when you import single files containing multiple views, such
as .exr and .sxr, unless they already exist. For multiple files, views must be created manually. See
Importing Source Clips for more information.
1. Click File > Import File(s) or press Ctrl/Cmd+I to display the file browser.
2. Locate a multi-view file, for example pubstitch_stereo.####.exr
3. Click Open.
The Create missing views? dialog displays.
4. Click Add Views, Replace Views, or No:

• Add Views - add the views in the incoming clip to those that exist in the project.
• Replace Views - replace all existing project views with those in the incoming clip.
• No - import the clip and display only the first view in the file, retaining any existing views in the
project.
You can now access the views in your project. See Displaying Views in the Viewer and Displaying
Views in the Timeline for more information.
Multi-View QuickTime Files

Multi-view .mov files only display one view by default. To enable all views in a multi-view .mov file:
1. Double-click the .mov to display its Properties panel.
2. Disable First track only.
You'll notice that the .mov in the bin is now marked with to denote multiple views.
You can now access the views in your project. See Displaying Views in the Viewer and Displaying
Views in the Timeline for more information.
USER GUIDE
1342
Displaying Views in the Viewer | Displaying a Particular View
Displaying Views in the Viewer

You can only display the views that exist in your Project Settings. To see a list of these views, or to
add or delete views, select Project > Edit Settings and go to the Views tab. For more information,
see Creating Views in a Project.
Displaying a Particular View

1. Double-click the clip or sequence to load it in the timeline Viewer.
2. On top of the Viewer controls, click the view to display. In the example, main, left or right.
Tip: You can also press the ; (semicolon) and ’ (forward single quote) keys to move between
different views in the Viewer.
Note: Nuke Studio lists the views in .exr files in the order they appear in the clip's header,
so a view named 'left' may not always be the first view displayed above the Viewer.
If your views do not appear in the correct order, you can rearrange them in the Project >
Edit Settings > Views tab. See Creating Views in a Project for more information.
Displaying Two Views Next to Each Other

1. Right-click in the Viewer and select the Stereo Modes menu.
2. Select one of the following options:
USER GUIDE
1343
Displaying Views in the Viewer | Displaying a Blend Between Two Views
Displaying a Blend Between Two Views

2. Select one of the following options:
Viewer.
Viewer.
Displaying OpenGL Stereo in Timeline Viewers

The Viewer OpenGL Stereo mode allows you to see both views at once on a 3D monitor for review
purposes. OpenGL Stereo is only supported on NVIDIA Quadro series GPUs and AMD Radeon Pro
series GPUs on Windows and Linux OS.
Note: OpenGL Stereo mode is not supported on Mac due to limitations in macOS and not
supported with AMD GPUs on Linux.
Enabling OpenGL Stereo Output
Windows
1. Right-click on the desktop and select NVIDIA Control Panel.
USER GUIDE
1344
Displaying Views in the Viewer | Displaying OpenGL Stereo in Timeline Viewers
2. Navigate to 3D Settings > Manage 3D Settings > Stereo - Enable.
To enable AMD GPU stereo output:

1. Double-click the AMD taskbar icon and select Advanced Settings.
2. Navigate to the AMD Pro Settings and check Enable Quad Buffer Stereo.
3. Select either Auto-Stereo (Horizontal Interleaved) or Auto-Stereo (Vertical Interleaved) and

click Apply.
USER GUIDE
1345
Linux
1. Open a command prompt and enter:
nvidia-xconfig --stereo=3
Tip: For more information on the nvidia-xconfig utility, please see the man page: man
nvidia-xconfig
Note: OpenGL Stereo mode is not supported with AMD GPUs on Linux.
Switching to OpenGL Stereo Output

2. Navigate to Stereo Modes and select OpenGL Stereo.
If you select OpenGL Stereo mode before enabling your GPU settings, a warning is displayed.
The first time you select OpenGL Stereo, a warning message is displayed.
Tip: You can disable the warning by enabling Don't show again and clicking OK.
USER GUIDE
1346
OpenGL Stereo is displayed in the Timeline Viewer.
Note: Switching to and from OpenGL Stereo mode causes playback to pause. Press play to
resume playback.
USER GUIDE
1347
Displaying Views in the Timeline | Splitting Views to Separate Tracks
Displaying Views in the Timeline

You can only display the views that exist in your Project Settings. To see a list of these views, or to
add or delete views, select Project > Edit Settings and go to the Views tab. For more information,
see Creating Views in a Project.
Adding a multi-view clip to the timeline groups all views into a single track. All the views in the clip are
assigned a Viewer button. Click the assigned view icon to display the views available in the shot. All
Views are visible by default.
Selecting a particular view for a track in the timeline, such as left, means that only the left view is
visible in the Viewer for that track.
Note: You can also import single-view files manually and then assign them views in the
timeline individually, providing that the views exist in the Project Settings.
Splitting Views to Separate Tracks

Working with all views in a single track is not always convenient, such as when you need to apply
different color corrections to stereo cameras. Nuke Studio includes a Split Views to Tracks option
that quickly separates views into individual tracks. You can use Split Views to Tracks with grouped
single- and multi-view shots.
1. Right-click a multi-view shot on the timeline and select Editorial > Split Views to Tracks.
USER GUIDE
1348
A separate track is created for each view in the group or file.
The new track names are suffixed with the view name, for example Video1_left, and the views are
assigned appropriately.
Note: If a view exists in the Project Settings, but there's no corresponding view in the source
files, empty placeholder tracks are added.
2. You can change the view assigned to a track by clicking the icon and selecting from the list of
available views.
USER GUIDE
1349
See Applying Changes to Selected Views for information on adding soft effects to different views.
Displaying Two Views Next to Each Other

2. Select on of the following options:
Displaying a Blend Between Two Views

2. Select on of the following options:
Viewer.
Viewer.
USER GUIDE
1350
USER GUIDE
1351
Applying Changes to Selected Views | Splitting Views to Tracks
Applying Changes to Selected

Views
By default, Nuke Studio applies any changes you make to all views. To apply changes to a particular
view only (for example, the left view but not the right), you can:
• Split the views into separate tracks. Separating views allows you to apply effects, such as a
Transform or Timewarp, to individual views without affecting the other views. See Splitting Views to
Tracks for more information.
• Split the view off in the soft effect’s Properties. Splitting properties allows you to perform the same
operation on multiple views, but with different values for each, such as a Grade or ColorCorrect. See
Splitting Views in the Properties Panel for more information.
Splitting Views to Tracks

1. Right-click a multi-view shot on the timeline and select Editorial > Split Views to Tracks.
2. Right-click the track to which you want to add the soft effect, select Effects and then the required
soft effect. For example, adding a Transform to the left view applies the effect to only the left view
track. See Soft Effects for more details on adding effects to shots.
USER GUIDE
1352
Applying Changes to Selected Views | Splitting Views in the Properties Panel
Splitting Views in the Properties Panel

1. Right-click the track to which you want to add the soft effect, select Effects and then the required
soft effect. For example, adding a ColorCorrect a multi-view shot.
See Soft Effects for more details on adding effects to shots.
2. Select the view you want to make changes to using the buttons above the timeline Viewer.
USER GUIDE
1353
Applying Changes to Selected Views | Showing Separate Values for Each View
3. Open the effect’s Properties and click the view button on the right, next to the control you
want to adjust.
4. Select Split off [view name]. For example, to apply changes to a view called left, select Split off
left. You can also split all the effect’s controls by selecting Split all knobs from the right-click
menu.
An eye appears on the split view button . Any changes you make using the control in question
are only applied to the view you chose to split off. Changes to controls that have not been split off
are still applied to all views.
Showing Separate Values for Each View

Once you have split off a view, you can apply changes to the existing views separately. Click on the
arrow on the left side of a split control to divide the control into values for each view.
In the example, the left view is split for the saturation control and left and cam3 views are split for
the contrast control.
USER GUIDE
1354
Applying Changes to Selected Views | Unsplitting Views
Note: The * (asterisk) denotes there is more than one unsplit view remaining for the
saturation control and ( ) denotes that the right view is the only unsplit view for the
contrast control.
Unsplitting Views
1. In the effect’s controls, click the view button .
2. Select Unsplit [view]. For example, to unsplit a view called left, you’d select Unsplit left.
The view is unsplit, and all changes you made to individual views are lost.
USER GUIDE
1355
Soft Effects
You can add soft effects to your timeline in any of the workspaces. A soft effect is a real-time effect,
processed on GPU instead of CPU.
You can add custom plug-in or gizmo soft effects to the Add Effect menu using Python. Valid custom
soft effects must have a GPUEngine implementation using DD::Image::Iop::gpuEngine type
functions. For more information see Nuke's NDK Reference Guide.
Soft Effects must also be registered after creation. An example of how to register a plug-in or gizmo
as a custom soft effect is located in:
<install_directory>/pythonextensions/site-packages/hiero/examples/custom_
soft_effect.py
Available Soft Effects

Below is a brief summary of the available soft effects. These are similar to the tools in Nuke's Node
Graph. See the Nuke Online Help for more information about them.
Note: Create Comp and Create Comp Special are not soft effects.
Soft Effect Summary
Transform Allows you to translate, rotate, scale, and skew shots from a single control panel.
Mirror Allows you to flip the input image around the center of the format area. A flip on
the x axis mirrors the image vertically. A flop on the y axis mirrors the image
horizontally.
Crop Allows you to cut out the unwanted portions of the image area. You can fill the
cropped portion with black or adjust the image output format to match the
cropped image.
TimeWarp Allows you to slow down, speed up, or even reverse selected frames in a clip
without necessarily altering its overall length. Sequences imported from .xml
and .aff files also support non-linear retimes.
USER GUIDE
1356
Soft Effects |
Soft Effect Summary
Warning: TimeWarp effects are only allowed on tracks with clips as

they are linked to the clips.
Grade Allows you to define white and black points by sampling pixels from the Viewer.
For example, you can use this for matching foreground plates to background
plates.
LUT Allows you to uses the OpenColorIO library to load a colorspace conversion from
a file (usually a 1D or 3D LUT) and apply it. You can also load other file-based
transformations, for example an ASC ColorCorrection XML.
CDL Allows you to apply an ASC CDL (American Society of Cinematographers Color
Decision List) grade based on the OpenColorIO Library. For more information,
see http://opencolorio.org
Colorspace Allows you to convert images from one colorspace to another, for example from
Nuke’s native colorspace to other color spaces more appropriate to a given
process or intended display device. This supports RGB, HSV, YUV, CIE, and CMS
formats (and various sub-formats). It can adjust for different primaries, white
point, and different encodings.
ColorCorrect Allows you to make quick adjustments to saturation, contrast, gamma, gain, and
offset. You can apply these to a clip’s master (entire tonal range), shadows,
midtones, or highlights.
You can control the range of the image that is considered to be in the shadows,
midtones, and highlights using the lookup curves on the Ranges tab. However,
do not adjust the midtone curve - midtones are always equal to 1 minus the
other two curves.
Text Allows you to add text overlays on your images. You can simply type in the text
you want to have displayed or use Tcl expressions (such as [metadata values])
or Tcl variables to create a text overlay. Text overlays can also be animated using
animation layers in the Groups tab, so that their properties (such as position,
size, and color) change over time.
Burn-In Allows you to quickly add standard burn-in elements on the timeline. You can
control the color, opacity, font, scale, and so on, as well as use the dropdowns to
determine what element is added from the file or sequence metadata.
USER GUIDE
1357
Soft Effects |
Soft Effect Summary
You can also reference custom metadata from shots. For example:
hiero/tags/Approved
Extracts the Approved tag from the shot. You can also append note to include
any notes associated with the tag:
hiero/tags/Approved/note
Note: You must precede spaces and slashes in the tag name with \\
(backslashes) to enable Nuke Studio to process the tag name correctly.
For example:
hiero/tags/Blue\\ Screen/note
You can also add burn-in through the Export dialog, see Adding Burn-in Text to
Exports for more information.
ChromaKeyer Allows you to pull a quick chroma key from green or bluescreen areas of your
footage.
Use the screen color selector to choose a color from the Source input to use as
the blue/green screen color. To remove blue/green spill from the foreground
object, use the despill controls to pick skin tones from the source. Use the
matte parameters to improve the matte.
BlinkScript Allows you to run Foundry's Blink framework on the timeline, enabling you to
write code once and run it on any supported device.
Warning: BlinkScript is very flexible, as there are no restrictions on the

code you can write within a kernel. As a result, code complied from the
Kernel Source can cause Nuke to crash, so please use caution!
The BlinkScript soft effect supports a subset of the functionality available in the
full BlinkScript node:
• You can't publish your kernels to Groups or Gizmos.
• Due to the way stacks of soft effects are processed in Nuke Studio, the
BlinkScript soft effect only contains one input source and produces only one
USER GUIDE
1358
Soft Effects | Adding Sequence-Level Soft Effects
Soft Effect Summary
output.
• Currently, the BlinkScript soft effect only supports eAccessPoint data access,
which means that only one point from the input can be accessed at a time, and
only one point from the output can be written, for each iteration position.
• The following functions are not supported by the BlinkScript effect:
• log10
• round
• rsqrt
• abs for integer types
• modf(a, *b)
• sign
• rcp
• max, min, and clamp for integer types
• median, atomicInc and atomicAdd.
• The only data types supported by the BlinkScript effect are int, float, and bool.
Note: The BlinkScript soft effect supports both pixel-wise and

component-wise kernels, but the former is preferred for performance
reasons. For more information about this and a more detailed
description of the language, see
https://learn.foundry.com/nuke/developers/113/BlinkKernels/
Adding Sequence-Level Soft Effects
Warning: Please keep in mind the following:

• Sequence-level soft effects are only permitted on the same track as clips if they're trimmed
to exactly match the in and out points of individual clips. In this case, each effect is linked to
a specific clip.
• Soft effects can be trimmed arbitrarily if they're on tracks with no clips.
As well as adding soft effects using the spreadsheet view (see next section), the timeline provides
some additional ways to add soft effects. You can either:
USER GUIDE
1359
• Right-click a shot on the timeline, select Effects and then select the soft effect you want to apply, or
• Select a shot on the timeline, click the Add Effect button to the left of the timeline, and then select
the soft effect you want to apply.
Note: You can add a soft effect to multiple shots by selecting the required shots first and
then right-clicking on one of them and selecting the soft effect you want to add. A soft effect
is added to each of the selected shots. You can also add a single soft effect for the whole
track by right-clicking on the track header and then selecting the soft effect you want to add.
Note: TimeWarp effects are only allowed on tracks with clips (and therefore linked to clips).
Using the Spreadsheet View

You can add soft effects using the spreadsheet view in any workspace by doing the following:
1. To open the spreadsheet view in any workspace, select Window > New Spreadsheet View.
2. Right-click a sequence in the bin view.
This opens a context menu.
3. If the spreadsheet view is not already populated in the context menu, select Open In >
Spreadsheet View.
USER GUIDE
1360
This loads the sequence in the spreadsheet view that you previously opened.
4. Right-click an event from the list in the spreadsheet view and select Effects to open a list of all
available soft effects.
Note: Create Comp and Create Comp Special are the only items in the Effects list that are
NOT a soft effect.
5. Select the required soft effect from the Effects list.

The sequence-level soft effect is then displayed above the shot and is color coordinated. For
example, if the effect appears in green on the timeline, the corresponding effect properties are
highlighted in the same color in the Properties pane.
When you insert a soft effect, its properties panel opens automatically. If you have it open, the
effect properties panel displays in the Properties pane. If the Properties pane is not open, the
effect's properties panel appears as a floating dialog.
USER GUIDE
1361
Soft Effects | Adding Shot-Level Soft Effects
Adding Shot-Level Soft Effects
Warning: Soft effects added at shot-level, must match the length of the shot on the locked
track. Any soft effect that is trimmed beyond the end of a shot, or a different length from the
shot is marked with red hashing to show that it is invalid.
You can add shot-level soft effects on the timeline, by doing the following:
1. Right-click the shot that you want to add a soft effect to.
2. Select Open In > Timeline View.
3. Click the Effects menu icon and select the soft effect you want to apply. For example, you can
select Grade.
The soft effect is then displayed above the shot as a colored box.
When you close the timeline view of the shot – as it is a shot-level soft effect – the soft effect is
displayed as a colored line within the top of the shot. The color of the line displayed reflects the
highest soft effect added to the shot.
Soft Effect Controls

Adding an effect displays the associated controls in the Properties panel, similar to Nuke nodes.
Adjusting the controls affects the shots underneath the effect in real-time. For example, adding a
Grade effect at sequence level displays the Grade controls in the Properties panel.
USER GUIDE
1362
Soft Effects | Soft Effect Controls
See Properties Panelsfor more information on node controls.
If you intend to animate soft effect controls using keyframes, you can use the Curve Editor and Dope
Sheet to fine-tune the output. To add the Curve Editor or Dope Sheet to the interface, navigate to
Window and select the required panel.
See Animating Parametersfor more information.
USER GUIDE
1363
Soft Effects | Editing Sequence-Level Soft Effects
Editing Sequence-Level Soft Effects

You can copy, move, and cut soft effects just like you can with shots in the timeline. You can perform
these actions by either accessing them from Edit in the right-click menu, or by using the keyboard
shortcuts. You can copy soft effects to different tracks, and different sequences, but not different
projects.
Moving
You can move a sequence-level soft effect by simply clicking and dragging the soft effect to a different
shot, or even onto a different video track.
Copying
Nuke Studio allows you to copy a sequence-level soft effect above the original to create a stack, to a
different track, or to a different sequence. You can also copy a sequence-level soft effect to a shot
open in the timeline view, therefore pasting it as a shot-level soft effect. You can copy a soft effect by
doing the following:
1. Select the soft effect you want to copy by clicking it.
2. Select Edit > Copy (or Ctrl/Cmd+C).
3. Move the playhead to where you want to paste the copy.
4. Select Edit > Paste (or Ctrl/Cmd+V).
Cloning
Nuke Studio allows you to clone a sequence-level soft effect. This copies the soft effect and links it to
the original, which means when one of these is edited, the changes are reflected in the other one. You
can clone a soft effect to a different track or even a different sequence. You cannot clone a soft effect
in different projects.
To clone a soft effect:

1. Select the soft effect you want to clone by clicking on it.
2. Select Edit > Copy (or press Ctrl/Cmd+C).
3. Move the playhead on the timeline to where you want to place the new clone.
4. Select Edit > Clone (or press Alt+K).
The new clone is placed at the current playhead position on the timeline. You can repeat steps 3
and 4 to create multiple clones that are all linked, at different places on the timeline.
USER GUIDE
1364
Soft Effects | Editing Sequence-Level Soft Effects
Clones are indicated by a C highlighted in red in the left of the soft effect.
Note: Cloning animation in soft effects is not supported.
Copying as Clone
You can also copy a sequence-level soft effect as a clone. This means, when you paste a new copy of
the soft effect above a selected shot, it is automatically linked to the original soft effect as a clone.
Therefore, any changes made to either of the cloned soft effects, are reflected in the other.
To copy a soft effect as a clone, do the following:

1. Select the soft effect you want to copy as a clone by clicking on it.
2. Select Edit > Copy as Clones (or press Ctrl/Cmd+K).
3. Click on the shot that you want to clone the soft effect to.
4. Select Edit > Paste (or press Ctrl/Cmd+V).
The soft effect is copied as a clone on your selected shot. You can repeat steps 3 and 4 to create
multiple clones that are all linked above different selected shots.
Clones are indicated by a C highlighted in red in the left of the soft effect.
Note: Cloning animation in soft effects is not supported.
Decloning
To declone a soft effect, simply click on the clone you want to declone and select Edit > Declone (or
press Alt+Shift+K).
Note: For more information about copying, moving, and cutting soft effects, see Timeline
Editing Tools .
Deleting
To delete a soft effect, simply right-click on it and select Edit > Delete (or press Backspace).
USER GUIDE
1365
Soft Effects | Editing Shot-Level Soft Effects
Editing Shot-Level Soft Effects

To edit a shot-level soft effect, you first need to open the shot with the applied soft effect, in the
timeline view. You can do this by right-clicking the shot and selecting Open In > Timeline View. You
can then copy, delete or move the shot-level soft effect in exactly the same way as sequence-level soft
effects.
You can copy and paste the shot-level soft effect on top of the original, creating a stack in the timeline
view. When you close the timeline view, stacked shot-level soft effects are displayed as a single line
within the top of the shot. Nuke Studio allows you to copy a shot-level soft effect, return to the full
sequence, and then paste it as a sequence-level soft effect. You can also paste a shot-level soft effect
to another shot open in the timeline view.
Note: You cannot clone shot-level soft effects.
Enabling and Disabling Soft Effects

You can choose to disable and re-enable soft effects from the output. To enable or disable a
sequence-level soft effect, select the soft effect by clicking on the colored box and then pressing D. To
enable or disable a shot-level soft effect, you first need to open the shot – that includes the soft effect
– in the timeline view. Then you can select the soft effect and press D.
USER GUIDE
1366
Create Comp
In the Timeline environment, you can choose to create a Nuke Comp of a shot to be able to open it in
the Compositing environment. You can then add any necessary compositing work and render out the
script.
You can only create comps from the timeline when the Frame Server is running. You can check the
status of the server in the status bar at the bottom-right of the interface.
Before you create a Nuke Comp, you can choose to change Export settings, and set the required Shot
Preset setting in the Project Settings dialog to get the required result.
It is not necessary to change the Export and Project Settings. If you don't change these settings when
creating a Nuke Comp, the default settings are used.
Note: You can use Nuke Studio's internal Frame Server to take advantage of external slave
machines to process renders faster. See Using the Frame Server on External Machines for
more information.
Edit Export Settings

If you do not change the export settings, the Using Local export preset setting control just uses the
default setting of Basic Nuke Shot With Annotations.
To change this setting, do the following:

1. Select File > Export... or press Ctrl+Shift+E.
USER GUIDE
1367
Create Comp | Edit Project Settings
This opens the Export dialog.

2. In the Using Local export preset: section, you can select the type of shot preset to export when
you use Create Comp.
For more information, see Introduction to the Export Dialog.

3. Select the required setting and then close the Export dialog.
Edit Project Settings

If you do not change the Project Settings, the Shot Preset control just uses the default setting of
Basic Nuke Shot With Annotations.
To change the Shot Preset setting in the Project Settings dialog, do the following:
1. Select Project > Edit Settings.
2. Open the Export / Roundtrip section by clicking on it.
3. Use the Shot Preset dropdown to select your required setting.
Note: You can use custom shot presets, but they must contain a Nuke Project File and
Nuke Write Node Content preset to be valid for selection. See Using Local and Project
Presets, Nuke Project File Settings, and Nuke Write Node Settings for more information.
USER GUIDE
1368
Create Comp | Creating and Editing a Comp
4. After selecting the required setting, close the Project Settings dialog.
Creating and Editing a Comp

To create a Nuke Comp of a shot, do the following:
1. Right-click on the shot that you want use to create a Nuke Comp.
2. Select Effects > Create Comp.
If you have not set a project root folder up in the Project Setting dialog, Nuke Studio asks you to
set a project root folder.
3. Click Choose, select to the required location, and then click OK.
A Nuke Comp is created, and placed directly above the original shot on the next available track.
USER GUIDE
1369
Create Comp | Creating and Editing a Comp
The Nuke Comp is displayed in light red, signifying that the Nuke Comp has not been rendered.
You can choose to render it by either, right-clicking and selecting Render Comp, or by selecting
Render > Render Selected Comp Containers.
Note: You can render all Nuke Comps by selecting Render > Render All Comp Containers.
4. Open the Nuke Comp in the Compositing environment in the same session by double-clicking the
Nuke Comp. You can also open the Nuke Comp in the Compositing environment in the same
session by either right-clicking on the Nuke Comp above the timeline, or in the Project bin and
selecting Open In > Node Graph.
To open the Nuke Comp in a new NukeX session you can either hold Ctrl/Cmd and double-click
on the Nuke Comp, or you can right-click on the Nuke Comp and select Open In > New Nuke
Session.
If you have not set a project root folder up in the Project Setting dialog, Nuke Studio asks you if
you want to save and set a project root folder.
USER GUIDE
1370
Create Comp | Nuke Comp Colors
You can now edit the script and add VFX using any of the tools available in the Compositing
environment. See Using the Compositing Environment.
Nuke Comp Colors

Color of Comp Meaning
Light Red This signifies that the comp has not been
rendered.
Dark Red (with OFF) This signifies that the comp is offline.
USER GUIDE
1371
Create Comp | Creating Multiple Shot Comps
Light Green (with This signifies that the comp is being

progress bar) rendered.
Dark Green This signifies that the comp has been

rendered.
Creating Multiple Shot Comps

You can create Nuke Comps from multiple shots on a single track or across multiple tracks, by doing
the following:
1. Select multiple shots on the timeline by using Shift+click to select a row of adjacent shots and/or
Ctrl/Cmd+click to select multiple non-adjacent shots. These can be across multiple tracks.
2. Right-click on one of the selected shots and select Effects > Create Comp.
This opens a Create Comp dialog.
USER GUIDE
1372
Create Comp | Creating Multiple Shot Comps
3. The Create Comp dialog displays all the selected clips, and allows you to select a master clip from
the list simply by clicking on it.
The master clip is highlighted yellow.
4. To create one Nuke Comp for all the shots, select the Add a single comp across all shots radio
button. Or, if you want to create a Nuke Comp for each of the selected shots, select the Add a
separate comp for each shot radio button.
When you create a Nuke Comp on multiple shots on a single track, the Node trees are
automatically connected and are displayed in reverse chronological order, as shown below.
USER GUIDE
1373
Create Comp | Create Comp Special
When you create a Nuke Comp on multiple shots across different tracks, the Node trees from the
same track are connected, but the Node trees on different tracks are not connected by default
(see below). You can choose to have them connected using the Create Comp Special settings. See
Create Comp Special for more information.
Create Comp Special

You can also choose to create a Nuke Comp using Create Comp Special. When you use the Create
Comp Special option, a dialog containing export options is automatically opened, allowing you to set
the export settings for the selected Nuke Comp.
USER GUIDE
1374
Create Comp | Create Comp Special
To use the Create Comp Special option:

1. Right-click on the shot that you want use to create a Nuke Comp.
2. Select Effects > Create Comp Special.
This opens a Create Comp dialog that contains the available export settings.
You can use this dialog to set the export location, set a version number, define the tracks to
export, filter by tag, and define the range. See Introduction to the Export Dialog for more
information.
3. Select the Nuke script in the Path section of the dialog, and then go to the Content section.
A number of additional controls are displayed. You can use these to specify what your Nuke Comp
includes and how it is created. You can select the Connect Tracks checkbox to connect all the
Node trees – across different tracks – in the Node Graph. See Introduction to the Export Dialog for
more information about these options.
4. After specifying your required settings, click Create Comp.
A dialog appears stating you have changed the export templates. You can choose to keep the
changes by selecting Yes or click No to discard the changes.
5. Double-click on the Nuke Comp to open it in the Compositing environment.
USER GUIDE
1375
Create Comp | Enabling and Disabling a Nuke Comp
You can now add VFX using any of the toolsets available in the Compositing environment. See
Compositing with Nuke.
Enabling and Disabling a Nuke Comp

You can disable and re-enable a Nuke Comp on the timeline by selecting the Nuke Comp you want to
disable and pressing D.
Saving a New Version and Updating the Comp

When you are happy with the changes you've made, you can save a new version of the script by
selecting File > Save New Comp Version (or pressing Alt+Shift+S).
After saving a new version of the Comp, you can update the original Nuke Comp you created. To
version-up the Nuke Comp, do the following:
1. On the timeline, right-click the original Nuke Comp.
2. Select Versions > Version Up.
The Nuke Comp is versioned-up. Depending on your Preferences > Performance >
Threads/Processes > Rendering > background renders setting, the comp may need rendering
manually.
Stereo and Multi-View Comps

Multi-view comps are similar to regular comps, but you can create a single comp with all views or
separate comps per view. The Multi-View Nuke Shot template is selected automatically when you
create a comp from a multi-view project.
1. Right-click on the shot that you want use to create a multi-view Comp. The shot can be a multi-
view track or single views split into separate tracks.
For separate tracks, you can right-click a single track to export all views or select all the per-track
views and right-click.
USER GUIDE
1376
Create Comp | Stereo and Multi-View Comps
2. Select Effects > Create Comp.

3. If you selected multiple per-track views to create the Comp, you can create a single Comp
containing all views or a separate Comp per view and select the master shot.
If you're using independent files per track, that is without importing multi-view files or using %V
functionality, the separate tracks in the Comp script are not connected by default. If you want the
tracks to be connected in the script, click the cog icon to display the Create Comp dialog and
enable Connect Tracks in the Nuke Project File preset.
USER GUIDE
1377
4. Click OK.
A Nuke Comp is created, and placed directly above the original shot on the next available track.
Note: You can render all Nuke Comps by selecting Render > Render All Comp Containers.
5. Open the Nuke Comp in the Compositing environment in the same session by double-clicking the
Nuke Comp. You can also open the Nuke Comp in the Compositing environment in the same
session by either right-clicking on the Nuke Comp above the timeline, or in the Project bin and
selecting Open In > Node Graph.
USER GUIDE
1378
To open the Nuke Comp in a new NukeX session you can either hold Ctrl/Cmd and double-click
on the Nuke Comp, or you can right-click on the Nuke Comp and select Open In > New Nuke
Session.
If you have not set a project root folder up in the Project Setting dialog, Nuke Studio asks you if
you want to save and set a project root folder.
You can now edit the script and add VFX using any of the tools available in the Compositing
environment. See Using the Compositing Environment.
USER GUIDE
1379
Annotations
Annotations can be used as quick instructions for a compositor to implement in Nuke Studio's
Compositing environment. You can add annotations to a clip, a section marked with in and out
points, or a whole sequence.
Annotations can be exported with a Nuke Comp and can then be viewed and/or deleted in the
Compositing environment. When all the suggested changes have been made to the script in the
Compositing environment, this can be saved as a new comp version and then rendered back to the
Timeline environment. If you want to add new annotations to the rendered Nuke Comp, you can
choose to re-export annotations only.
Workflow
The following steps show an example of Nuke Studio workflow for annotations:
1. In the Viewer, open the Annotations menu.
See The Annotations Menu for more information for more information about it.
2. Add an annotation to a shot(s) at sequence or shot-level, by using the Annotation menu tools.
See Adding Annotations for more information.
3. You can choose to edit a sequence or shot-level annotation after it has been created. See Editing
Sequence-Level Annotations or Editing Shot-Level Annotations for more information.
4. Create a Nuke Comp of the shot with the annotations, ensuring annotations are enabled in the
export settings.
5. Open the Nuke Comp by double-clicking it.
See Viewing Annotations in the Compositing Environment for more information.
6. After the suggested changes are made in the Compositing environment, select File > Save New
Comp Version.
7. Return to the timeline, and version up the Nuke Comp by right-clicking it and selecting Versions >
Version Up.
The Nuke Comp is versioned up. Depending on your Preferences > Performance >
Threads/Processes > Rendering > background renders setting, the comp may need rendering
manually.
8. You can add new annotations to the rendered Nuke Comp by ensuring you select the Clip radio
button and then using the Annotations menu tools.
USER GUIDE
1380
Annotations | The Annotations Menu
9. After adding the new annotations, right-click the rendered Nuke Comp and select Export > Re-
Export Annotations.
10. Open the rendered Nuke Comp by double-clicking it.
11. Double-click the Precomp node to open its properties and version it up. For example, if the file
path has v01.nk at the end, change it to v02.nk. See Re-Exporting Annotations from the Timeline
12. Display the Precomp in the Viewer.
Your new annotations are displayed.
The Annotations Menu

You can open the Annotations menu by selecting the paint brush icon at the top-right of the
Viewer.
The Annotations menu displays down the left side of the Viewer.
Note: Annotations on sequences and shots are only visible when you have the Annotations
menu open.
Adding Annotations
To add an annotation, do the following:
1. After you have opened the Annotations menu, move the playhead on the timeline to where you
want to add your annotation.
2. Click the + (addition) icon.
USER GUIDE
1381
Annotations | Adding Annotations
A dialog is displayed containing annotation options.
3. You can choose which level you want to add your annotation to, sequence or shot-level, by
selecting either the Sequence or Clip radio buttons.
When you add an annotation to the clip level, a turquoise line is displayed within the top of the
shot. When you apply an annotation to the sequence level, a turquoise box appears above the
selected section on a separate Annotations track.
4. From the Range dropdown select one of the following:
• Current Item - Applies the annotation to the shot at the current playhead position.
• Current Frame - Applies the annotation to the frame at the current playhead position.
• In/Out Points - Applies the annotation to in and out points that have already been marked on
the timeline.
• All - Applies the annotation to the whole track.
• Custom - When you select this from the Range dropdown, the in and out point fields within the
dialog, become active. You can then use these to set the section that you want your annotation
to appear on.
Note: The timecode displayed in the In and Out fields is derived from the clip's metadata,
not its position in the sequence.
5. When you have set where you want your annotation to appear, click New.
A label is added to the location of your annotations, detailing the clip timecode and name.
You can click and drag this label to place it anywhere in the Viewer.
6. To draw in your annotation, select the paint brush tool from the Annotation menu.
This is highlighted orange when selected.
7. Before drawing in the Viewer, you can set the brush and paint settings:
• Select the brush and/or background color by clicking the paint colors icon .
USER GUIDE
1382
Annotations | Adding Annotations
This opens a color wheel that allows you to select the color and brightness, and an opacity slider
underneath that you can use to set the opacity of the paint.
• Select the paint brush size icon to set the required brush size. You can either use the slider
to drag to your required brush size, or enter it into the brush size field.
Note: You can also edit these settings after drawing by clicking the selection tool in the
Annotations menu, selecting the lines that you've drawn in the Viewer, and then adjusting
the paint brush settings.
8. Click and drag in the Viewer to draw with your selected brush settings.
9. To add text to your annotation, click the text icon in the Annotations menu and then click
anywhere in Viewer to enter your required text.
A text dialog appears allowing you to type your required text, align it horizontally and vertically,
and adjust the text size. You can then click and drag the text box to anywhere in the Viewer.
USER GUIDE
1383
Annotations | Enabling and Disabling Annotations
Enabling and Disabling Annotations

You can choose to disable and re-enable annotations from the output. To enable or disable a
sequence-level annotation, simply select the annotation by clicking on the turquoise box and then
pressing D. To enable or disable a shot-level annotation, you first need to open the shot – that
includes the annotation – in the timeline view. Then you can select the annotation and press D.
Editing Sequence-Level Annotations

You can choose to edit, copy, move, and delete any annotations that were added on the sequence-
level.
To remove an annotation from the sequence-level, simply select the turquoise box representing the
annotation you want to move from the Annotation track, right-click and select Edit > Delete (or press
Backspace or Delete on the keyboard).
To copy an annotation that was added at sequence level:

1. Select the annotation and then click Edit > Copy (or press Ctrl/Cmd+C). You can also access these
tools in the right-click menu.
2. Move the timeline playhead to the position where you want to paste the annotation.
3. Select Edit > Paste (or press Ctrl/Cmd+V).
Note: You can also copy and paste annotations, that were added at sequence-level, between
different sequences.
To move an annotation that was added at sequence level, you can simply click and drag it to the
required location. You can also drag annotations to different track levels.
You can trim an annotation at either end by hovering the cursor over one end of the annotation until
it changes into the trim icon:
Then click and drag to where you want to trim the annotation to.
To edit the actual annotation, you can use the selection tool in the Annotations menu to select the
annotation in the Viewer and move it, delete it, or replace it.
USER GUIDE
1384
Annotations | Editing Shot-Level Annotations
Editing Shot-Level Annotations

You can choose to move, trim, and simply delete shot-level annotations.
To edit a shot-level annotation, you first need to open the annotated shot as a timeline. To do this:
1. Right-click the shot with the annotation you want edit.
2. Select Open In > Timeline View.
The shot opens in the Timeline View, and the annotation now appears as on a separate level from
the video. To view the annotation in the Viewer, ensure the Annotation menu is open.
To move a shot-level annotation from the Timeline view, hover the cursor over the annotation until it
changes into a move icon:
Then simply click and drag it to where you want to move it to.
You can trim an annotation at either end by hovering the cursor over one end of the annotation until
it changes into the trim icon:
Then click and drag to where you want to trim the annotation to.
To delete the annotation, click the turquoise box and select Edit > Delete (or press Backspace or
Delete on the keyboard).
To edit the actual annotation, you can use the selection tool in the Annotations menu to select the
annotation in the Viewer and move it, delete it, or replace it.
USER GUIDE
1385
Annotations | Viewing Annotations in the Compositing Environment
Viewing Annotations in the Compositing Environment

If you want to view annotations in the Compositing environment, you need to ensure annotations are
enabled when creating a Nuke Comp. You can do this by either opening the Export dialog or the
Project Settings dialog and checking the option Basic Nuke Shot With Annotations is selected. See
Create Comp for more information. You can also enable annotations using the Create Comp Special
option, as shown below.
To create a Nuke Comp with annotations, do the following:

1. Create a Nuke Comp by right-clicking on the shot that you want use to create a Nuke Comp, and
selecting Effects > Create Comp Special.
The Create Comp Special dialog opens.
2. In the Using Local export preset setting in the middle-top of the dialog, ensure Basic Nuke Shot
With Annotations is selected. Also, in the Tracks For This Export section in the bottom-left of
the dialog, ensure either all tracks is selected, or that the Annotations checkbox is selected with
certain tracks.
3. Click Create Comp.
A message warns that you've changed the export templates and asks whether you want to keep
them.
4. Click Yes or No.
This creates a Nuke Comp above the selected shot(s). The Nuke Comp is displayed in light red, as
it has not been rendered yet. SeeCreate Comp for more information on Nuke Comp colors.
5. Open the Nuke Comp by double-clicking it.
When the Nuke Comp script opens, an additional Precomp node is created, which is separate
from the Nuke Comp script. The Precomp node contains a copy of the whole script and includes
the annotations.
6. Connect the Viewer node to the Precomp node.
Your annotation is visible in the Viewer over the footage.
Viewing the Annotation Node Group

If you wish, you can view the annotations by doing the following:
1. Double-click the Precomp node to open its properties panel, and the click Open.
This opens the contents of the Precomp node as a node tree in a new instance of Nuke.
2. Ensure you have the Node Graph tab selected.
USER GUIDE
1386
Annotations | Re-Exporting Annotations from the Timeline
The node tree displays in two parts. The top part represents the shot-level settings and the
bottom part represents the sequence-level settings. Depending on where you set your annotation
to be, an Annotations node group is displayed in the node tree.
3. Double-click the Annotations node group to open its properties panel.

4. In the top-right of the node properties, click the node structure icon to display the contents of
the node group.
This displays the contents of the Annotations node group in a new Node Graph tab.
Re-Exporting Annotations from the Timeline

You may want to add new annotations after the Nuke Comp has been edited and rendered. In this
case, you can re-export the annotations only.
To re-export annotations, do the following:

1. Move the playhead to the rendered Nuke Comp.
2. Press the + in the Annotations menu at the side of the Viewer, and select the Clip radio button.
Tip: You cannot add sequence-level annotations to a Nuke Comp.
3. Add an annotation using the brush or text tools in the Annotation menu.
4. Right-click the Nuke Comp, and select Export > Re-Export Annotations.
USER GUIDE
1387
Annotations | Re-Exporting Annotations from the Timeline
5. Return to the Nuke script and double-click the Precomp node to open its properties, and version
it up.
For example, if the file path has v01.nk at the end, change it to v02.nk. You can also version up
the Precomp node by doing one of the following:
• Select the Precomp node and press Alt+Up Arrow.
• Select the Precomp node and click Edit > Node > Filename > Version Up.
6. Ensure the Precomp node is connected to a Viewer node.
Your new annotation is now visible in the Viewer.
USER GUIDE
1388
Timeline Editing Tools
The timeline editing tools allow you to manipulate your shots directly in the timeline, in single- or
multi-view projects, using a series of modal editorial tools that complement the Multi Tool. Select the
tool you need for the job and then select a new tool and continue editing.
The timeline editing tools are grouped for convenience - each tool group contains several tools and
you can cycle between them by clicking the tool or using keyboard shortcuts. The editing tools work
the same way in single- and multi-view timelines.
Icon Tools Description
Multi Tool The Multi Tool’s functionality is equivalent to most of the other tools
combined, but doesn’t require modal tool selection.
Move/Trim The Move/Trim tool allows you to manipulate the position of a shot or
its output by adding or removing handles.
Select The marquee Select tool allows you to make multiple selections quickly
by lassoing shots. Hold Shift to add to the selection and Alt to subtract
from the selection.
Selection by The track selection tools allow you to quickly select multiple items
Track depending on the initial selection. For example, the Select Track to
Right tool selects all shots to the right of the target shot, within a single
track.
Slip Clip The Slip Clip tool allows you to shift a shot’s In and Out points by the
same amount and in the same direction, retaining the original duration
but altering the timeline output.
Slide Clip The Slide Clip tool allows you to move a shot in relation to the item
before and/or after the target item, without changing its length or
timeline output.
Roll Edit The Roll Edit tool enables you to roll a single edit within the available
handles, shortening one shot while lengthening the other, but keeping
the overall duration the same.
USER GUIDE
1389
Timeline Editing Tools |
Icon Tools Description
Ripple Edit The Ripple Edit tool operates similarly to the trim function of the
Move/Trim tool, except that downstream shots are rippled to
automatically close any resulting gaps in the timeline.
Retime Clip The Retime Clip tool allows you to trim a shot’s In or Out point and
automatically retime the clip to fill the new shot duration.
Razor The Razor and Razor All tools allow you to cut shots in to separate
parts so you can remove sections or rearrange items on the timeline.
Join The Join tool can only be used on edit points between two razored
shots, denoted by the yellow arrows at the edit.
The modal editorial tools are mapped to the Q, W, E, R, and T keyboard shortcuts when the timeline
is the active tab.
Note: For a full list of keyboard shortcuts, please see Appendix B: Keyboard Shortcuts.
Pressing a keyboard shortcut multiple times selects the tools within each mode. For example,
pressing E twice, rapidly in succession activates Slide Clip. Pressing them slowly in succession does
USER GUIDE
1390
Timeline Editing Tools |
not achieve the same result, but instead, remains on the first item in the menu. This allows you to
activate a tool without knowing the current state of tool selection.
• mapped to Q, cycles through Multi Tool, Move/Trim, and Select.
• mapped to W, cycles through Track Selection tools.
• mapped to E, cycles through Slip Clip and Slide Clip.
• mapped to R, cycles through Roll Edit, Ripple Edit, and Retime Clip.
• mapped to T, cycles through Razor, Razor All, and Join.
USER GUIDE
1391
Using the Multi Tool |
Using the Multi Tool

Unlike the other editing tools available, the Multi Tool changes function depending on the position of
your pointer in relation to the shot(s) selected.
The Multi Tool’s functionality is equivalent to most of the other tools combined, but doesn’t require
modal tool selection:
• Move - placing the mouse in the center of a shot activates the tool. Drag the selected shot to the
required position on the timeline.
• Trim - placing the mouse at the left or right of the shot activates the tool. Drag the edit point to the
new position and release the mouse to complete the trim.
See Using the Move/Trim Tool for more information.

• Select - click-and-drag to marquee select clips. Hold Shift to add to your selection or Alt to subtract.
See Using the Selection Tools for more information.

• Slip - placing the mouse at the bottom of the shot activates the tool and displays the slip handles.
Drag the shot to the new position and release the mouse to complete the slip.
See Using the Slip Clip Tool for more information.

• Slide - placing the mouse at the top of the shot activates the tool and displays the slide handles.
Drag the shot to the new position and release the mouse to complete the slide.
See Using the Slide Clip Tool for more information.

• Roll - placing the mouse on the edit between shots activates the tool and displays the handles. Drag
the edit to the new position and release the mouse to complete the roll.
See Using the Roll Edit Tool for more information.

• Razor - when using the Multi Tool, Razor cuts are menu driven. Navigate to Timeline > Razor
Selected or Razor All to make cuts at the current playhead position.
USER GUIDE
1392
Using the Multi Tool |
See Using the Razor and Join Tools for more information.
USER GUIDE
1393
Using the Move/Trim Tool | Moving Shots
Using the Move/Trim Tool

The Move/Trim tool allows you to manipulate the position of a shot or its output by adding or
removing handles. Activate the Move/Trim tool by clicking the tool or pressing Q twice.
Moving Shots
Click and drag the selected shot(s) to the required position on the timeline. A time count popup, in
conjunction with the snap to edit function, helps you to reposition the shot(s) accurately.
You can also move shots up and down the track hierarchy using drag-and-drop or the Alt+, (comma)
and Alt+. (period) keyboard shortcuts.
USER GUIDE
1394
The following table describes the Move/Trim tool’s modifiers and actions:
Mode Method Description
Overwrite drag-and-drop The default move mode. The dragged shot overwrites
any items that are present in the move location.
Ripple drag then hold Alt and Drag-and-drop shots on top of other items without
drop overwriting content - items are pushed down the
timeline to accommodate the move.
Duplicate hold Alt and drag then Copy the shot, then drag-and-drop on top of other
release Alt and drop items overwriting existing content - items are not
pushed down the timeline to accommodate the move.
Ripple and hold Alt then drag and Copy the shot, then drag-and-drop items on top of
Duplicate drop while holding Alt others without overwriting content - items are pushed
down the timeline to accommodate the move.
Note: On Linux, hold Ctrl+Alt for Duplicate and RippleDuplicate modifiers.
USER GUIDE
1395
Action Keyboard Shortcut Description
Delete Backspace Delete the selected shot(s) or gap(s).
Ripple Shift + Backspace Remove the selected shot(s) and ripple items down stream to
Delete close gaps in the timeline.
Note: The ripple effect may not close gaps entirely,

because Nuke Studio does not allow linked tracks to
become desynchronized during rippling.
If you need to nudge shots horizontally by just a frame or two, you can select the items on the
timeline and press , (comma) to nudge it left or . (period) to nudge it right. Press Shift+, (comma) or .
(period) to nudge the shot horizontally by the FrameIncrement set under the Viewer.
Note: You cannot overwrite other shots on the timeline horizontally using the nudge keys.
However, vertical nudging (Alt+, and Alt+.) overwrites other tracks.
To Move Shots Using the Spreadsheet View:

1. Select the required events in the spreadsheet view.
2. Click the cog icon and select the required Time Edit Behaviors:
Modify Using Result
Dst In Move Adjusts the event’s Dst In and Out by the same amount, moving the
Destination shot’s position on the timeline by the specified amount, while
maintaining Speed.
Dst Out Move Adjusts the event’s Dst Out and In by the same amount, moving the
USER GUIDE
1396
Using the Move/Trim Tool | Trimming Shots
Modify Using Result
Destination shot’s position on the timeline by the specified amount, while

maintaining Speed.
3. Adjust the Dst In or DstOut to move the shot(s) by the specified amount.
Trimming Shots
Click-and-drag the edit point to the new position and release the mouse to complete the trim.
Tip: Use the Ripple Edit tool, activated by pressing R twice, to ripple downstream shots
automatically.
USER GUIDE
1397
The Viewer displays the new In or Out point (depending on whether you’re adjusting the beginning or
end of the shot), allowing you to accurately gauge the new output.
Note: Trimming multiple shots simultaneously trims each item by the same amount and in
the same direction.
Alternatively, click an edit point and nudge the edit using the Ctrl/Cmd+ßà (numeric pad) keys or
hold Shift to nudge by the Frame Increment set under the Viewer.
By holding Ctrl/Cmd and dragging an edit, you can add blank frames past the end of the shot’s
handles. Blank frames are colored red on the timeline for clarity:
USER GUIDE
1398
To Trim Shots Using the Spreadsheet View:

1. Select the required events in the spreadsheet view.
2. Click the cog icon and select the required Time Edit Behaviors depending on whether you’re
using the In or Out points or duration:
Modify Using Result
Src In Trim In Trims the event’s Src In, Dst In, and durations while maintaining speed.
Src Out Trim Out Trims the event’s Src Out, Dst Out, and durations while maintaining
speed.
USER GUIDE
1399
Modify Using Result
Src Dur Trim Out Trims the event’s Src Dur, Dst Dur, and Src/Dst Out while maintaining
speed.
Before and after a 50 frame Src Dur increase:
Dst In Trim In Trims the event’s Dst In, Src In, and durations while maintaining speed.
Dst Out Trim Out Trims the event’s Dst Out, Src Out, and durations while maintaining
speed.
Dst Dur Trim Out Trims the event’s Dst Dur, Src Dur, and Dst/Src Out while maintaining
speed.
Before and after a 50 frame Dst Dur increase:
USER GUIDE
1400
3. Adjust the values as required to trim the shot(s) by the specified amount.
USER GUIDE
1401
Using the Selection Tools | Trimming Shots
Using the Selection Tools

The timeline editing tools include a marquee selection tool and several context dependent track
selection tools.
The marquee Select tool, activated by clicking the tool or pressing Q three times, allows you to make
multiple selections quickly by lassoing shots.
Hold Shift to add to the selection:
The track selection tools, activated by clicking the tool or by pressing W, selects multiple items
depending on the initial selection:
• Select Track to Right or Left - all shots right or left of the target shot are selected, within a single
track.
USER GUIDE
1402
• Select All in Track - all shots on the target track are selected, regardless of the item selected.
• Select All Tracks Right or Left - all shots right or left of the target item are selected, regardless of
which track they occupy.
USER GUIDE
1403
USER GUIDE
1404
Using the Slip Clip Tool | Trimming Shots
Using the Slip Clip Tool

The Slip Clip tool allows you to shift a shot’s In and Out points by the same amount and in the same
direction, retaining the original duration but altering the timeline output. Activate the Slip Clip tool by
clicking the tool or pressing E.
Note: The target shot must have handles to use the Slip tool.
The Slip Clip tool displays different Viewer previews depending on whether the playhead is on the
target shot or not, but the basic principles are the same.
Click the target clip to display the available handles and then drag the shot to the new position.
Release the mouse to complete the slip.
Note: Using the Slip Clip tool does not move the shot on the timeline, only the output is
changed.
Alternatively, nudge the slip using the , (comma) or . (period) keys or hold Shift to nudge by the
Frame Increment set under the Viewer.
Tip: If you’re using the Multi Tool, you can nudge using the “slip bar” by clicking at the
bottom the shot.
The following Viewer previews are displayed, depending on the playhead position:
Note: The Viewer background always displays the playhead’s current position.
USER GUIDE
1405
Using the Slip Clip Tool | Slipping Using the Spreadsheet View
• When the playhead is not currently on the target shot, the Viewer displays the In frame (1) and Out
frame (2), allowing you to accurately gauge the new output.
• When the playhead is on the target shot, the Viewer displays the In frame (1), the current frame (2),
and Out frame (3), allowing you to accurately gauge the output of the shot against the current
frame.
• When the playhead is on the target shot and A/B compare is active, the Viewer displays the target
shot (1) and the reference shot (2), allowing you to synchronize your working track against the
reference track.
Slipping Using the Spreadsheet View

You can slip shots using the Src In and Src Out columns of the spreadsheet:
1. Select the required event in the spreadsheet view.
USER GUIDE
1406
Using the Slip Clip Tool | Slipping Using the Spreadsheet View
2. Click the cog icon and select the required Time Edit Behaviors depending on whether you’re
using the In or Out point:
Modify Using Result
Src In Slip Adjusts the Src In and Src Out by the same amount, slipping the event
Source while maintaining speed.
Src Out Slip Adjusts the Src Out and Src In by the same amount, slipping the event
Source while maintaining speed.
3. Adjust the Src In or Src Out to slip the shot(s) by the specified amount.
USER GUIDE
1407
Using the Slide Clip Tool | Slipping Using the Spreadsheet View
Using the Slide Clip Tool

The Slide Clip tool allows you to move a shot in relation to the item before and/or after the target
item, without changing its length or timeline output. Activate the Slide Clip tool by clicking the tool or
pressing E twice.
The shot either side of the target are shortened or lengthened within the limits of their handles to
accommodate the slide.
Note: The surrounding shots must have handles to use the Slide tool.
Click the target shot and then drag it to the new position and release the mouse to complete the
slide.
For example, if you slide the target shot (2) five frames to the right, the preceding item (1) ends five
frames later and the next item (3) starts five frames later.
The first image shows a timeline containing three shot, and the second shows the same shots with
the target (2) sliding to the right.
USER GUIDE
1408
Using the Slide Clip Tool | Slipping Using the Spreadsheet View
The Viewer displays the new end point of the previous shot on the left and the new start point of the
next shot on the right, allowing you to accurately gauge the slide.
The two center images (2) represent the start and end frames of the target shot, which don’t change.
USER GUIDE
1409
Using the Roll Edit Tool | Slipping Using the Spreadsheet View
Using the Roll Edit Tool

The Roll Edit tool enables you to roll a single edit within the available handles, shortening one shot
while lengthening the other, but keeping the overall duration the same. Activate the Roll Edit tool by
clicking the tool or pressing R.
Note: At least one of the target items must have handles to use the Roll tool.
1. Click an edit point between two shots to display the available handles as a red overlay.
2. Drag the edit to the new position and release the mouse to complete the roll.
For example, if you roll a number of frames at the end of one shot (1), the next item (2) starts that
number of frames later. The first image shows a timeline containing two shots, and the second
shows the same items with the edit point “rolled” to the right.
USER GUIDE
1410
Using the Roll Edit Tool | Slipping Using the Spreadsheet View
The Viewer displays the pre-edit shot on the left and the post-edit item on the right, allowing you to
accurately gauge the new position of the edit.
Alternatively, click the edit point between the shot and nudge the edit using the , (comma) or . (period)
keys or hold Shift to nudge by the Frame Increment set under the Viewer.
USER GUIDE
1411
Using the Retime Clip Tool | Slipping Using the Spreadsheet View
Using the Retime Clip Tool

The Retime Clip tool allows you to trim a shot’s In or Out point and automatically retime the clip to
fill the new shot duration. Activate the Retime Clip tool by clicking the tool or pressing R three times.
Click-and-drag the edit point to the new position and release to complete the trim and retime. For
example, trimming a 50 frame shot to 25 frames retimes the clip to 200%.
Alternatively, click an edit point and nudge the edit using the , (comma) or . (period) keys or hold Shift
to nudge by the Frame Increment set under the Viewer.
USER GUIDE
1412
Using the Retime Clip Tool | Slipping Using the Spreadsheet View
Tip: By holding Ctrl/Cmd and dragging an edit, you can retime past the end of the shot’s
handles.
USER GUIDE
1413
Using the Razor and Join Tools | Slipping Using the Spreadsheet View
Using the Razor and Join Tools

The Razor tools allow you to cut shots in to separate parts so you can remove sections or rearrange
items on the timeline. Activate Razor and Razor All by clicking the tool or pressing R.
Place the cursor on the target shot, and if the cut is permissible, click to razor the shot or all shots
depending on which tool you have selected.
Tip: The Razor cursor indicates whether a cut is permissible or not, such as on existing
edits.
You can also apply cuts at the playhead position from the menu bar using Timeline > Razor
Selected, or all tracks using Timeline > Razor All.
Tip: Use the C (with the shot under the playhead selected) and Shift+C keyboard shortcut, or
the right-click context menu, to perform the respective cuts.
The Join tool can only be used on edit points between razored shots, denoted by the yellow arrows at
the edit.
USER GUIDE
1414
Using the Razor and Join Tools | Copying Cuts Between Tracks
Copying Cuts Between Tracks

The Copy Cuts function allows you to quickly apply cuts from one track to other tracks on the
timeline. For example, in the timeline shown, you could copy the cuts from the second video track to
the Reference audio tracks.
To copy cuts:
1. Select the shots containing the cuts to copy, or if you intend to copy all the cuts from a track you
don’t need to make a selection.
2. Right-click in the timeline and select Editorial > Copy Cuts.
The Copy Cuts dialog displays.
USER GUIDE
1415
Using the Razor and Join Tools | Copying Cuts Between Tracks
3. If you made a selection on the timeline, use the dropdown to select Copy All Cuts or Copy
Selected Cuts as required.
This dropdown is not displayed if no shots were selected.
4. Click the From dropdown to select the source track.
5. Check all the destination tracks in the To field to which you want to copy the cuts.
6. Choose whether or not the resulting shots are named identically to the source track.
Selecting None retains the destination clip name.
7. Click OK to copy the cuts to the destination track(s).
USER GUIDE
1416
Insert, Overwrite, and 3-Point Editing | Inserting Clips
Insert, Overwrite, and 3-Point

Editing
Insert and Overwrite edits are applied at the current playhead position by default, but the use of In
and Out points in the clip Viewer and/or sequence Viewer can give you greater control over the result.
3-point editing, makes use of In and Out points in the clip Viewer and an In or Out in the sequence
Viewer to control where the clip is placed on the timeline.
Inserting Clips
By default, Insert places the entire contents of the clip Viewer into the timeline at the current
playhead position, on the lowest available track. All shots downstream of the playhead are rippled to
make room for the clip. No items are overwritten or removed.
Pre-insert timeline Post-insert timeline
Note: If the playhead is not positioned at an edit point, or there are shots on other tracks,
the Insert action cuts the shot(s) at the playhead and ripples the cut downstream.
For example, the Post-insert image shows the audio shot being cut and rippled, even though
it doesn’t reside on the same track.
You can select a track before inserting if you don’t want to target the lowest available track. Even if the
target track is empty, shots on all other unlocked tracks are rippled by the same amount.
You can also use In and Out points to control where the clip is inserted and how many frames are
included. See 3-Point Editing for more information.
To insert a clip at the playhead:

1. Navigate to Workspace > Editing to display the 2-up Viewer layout.
USER GUIDE
1417
Insert, Overwrite, and 3-Point Editing | Overwrite Edits
2. Double-click your sequence in the bin view to load it into the right-hand sequence Viewer .
3. Double-click the source clip to load it into the left-hand clip Viewer .
4. Place the playhead at the required edit point or timecode and select the target track, if necessary.
5. Navigate to Clip > Insert, or press N, to insert the clip into the timeline.
All shots downstream of the clip are rippled to make room for the duration of the edit.
To insert a clip at an In or Out point:

4. Place an In or Out point on the timeline to determine the clip’s position:
• In point - the source clip is inserted so that the first frame is at the In point specified.
• Out point - the source clip is inserted so that the last frame is at the Out point specified.
5. Navigate to Clip > Insert, or press N, to insert the clip into the timeline.
All shots downstream of the In or Out point are rippled to make room for the duration of the edit.
Overwrite Edits
Unlike inserting, Overwrite does not incorporate downstream ripple and doesn’t alter the length of
your sequence. Any shots you overwrite are destroyed, though they can easily be recovered from the
source clips in the bin view.
Pre-overwrite timeline Post-overwrite timeline
You can select a track before overwriting if you don’t want to target the lowest available track.
You can also use In and Out points to control what the clip overwrites and how many frames are
included. See 3-Point Editing for more information.
To overwrite at the playhead:

USER GUIDE
1418
Insert, Overwrite, and 3-Point Editing | 3-Point Editing
4. Place the playhead at the required edit point or timecode and select the target track, if necessary.
5. Navigate to Clip > Overwrite, or press M, to overwrite from the playhead for the duration of the
source clip.
All shots under the source clip are overwritten.
To overwrite from an In or Out point:

4. Place an In or Out point on the timeline to determine the clip’s behavior:
• In point - the source clip begins overwriting from its first frame at the In point specified
downstream for the duration of the clip.
• Out point - the source clip begins overwriting from its last frame at the Out point specified
upstream for the duration of the clip.
5. Navigate to Clip > Overwrite, or press M, to overwrite from the In or Out point for the duration of
the source clip.
All shots under the source clip are overwritten.
3-Point Editing
Setting the output of a source clip and then editing the clip into a timeline at a specific point is
sometimes referred to as 3-point editing. Using this method, you can insert and overwrite edits in an
existing timeline or quickly construct scratch timelines from your source clips.
Firstly, set the output of your source clip using In and Out points in a clip Viewer, then set the
reference In or Out point on your timeline to determine the clip’s position. Finally, add the clip to the
timeline using Insert or Overwrite.
Tip: You can set both In and Out points on the timeline, but bear in mind that there may be
insufficient source frames for the range specified. If this is the case, blank frames are added
and highlighted in red.
You can select a track before editing if you don’t want to target the lowest available track. When
inserting, even if the target track is empty, shots on all other unlocked tracks are rippled by the same
amount.
USER GUIDE
1419
2. Double-click the required source clip to load it into the left-hand clip Viewer .
3. Set the required frame range using In and Out points.
5. Set In and/or Out points on the timeline to specify where the clip should be added and use Insert
(N) or Overwrite (M) as required.
As an example, assuming your clip Viewer and timeline are represented by the following image,
and the Overwrite function is used:
• No In or Out points - insert or overwrite at the current playhead position, for the range
currently set in the clip Viewer.
USER GUIDE
1420
• In point but no Out point - insert or overwrite from the In point position downstream, for the
range currently set in the clip Viewer.
• Out point but no In point - insert or overwrite from the Out point position upstream, for the
range currently set in the clip Viewer.
USER GUIDE
1421
• In and Out points - insert or overwrite at the current In point position, for the duration set by
the timeline's In and Out points. If there are insufficient source frames for the range specified,
blank frames are added highlighted in red.
USER GUIDE
1422
Versions and Snapshots
In addition to the regular project save and restore options, Nuke Studio can record the different
states of your workflow as you progress using versions and snapshots.
Versions are children of clips. You can have any number of versions per clip as long as they follow
the correct naming conventions, as shown in Using Versions. Versions can only be applied to source
clips and shots and can be swapped in and out without overwriting existing work.
Snapshots are time-stamped copies of a sequence, allowing you to save its current state without the
inconvenience of saving the entire project. When you restore a snapshot, a warning displays prior to
the restore reminding you that edits since the snapshot was taken are lost. See Using Snapshots for
more information.
USER GUIDE
1423
Using Versions |
Using Versions
Versions can be added to source clips and shots to allow greater flexibility in your workflow. You can
have as many versions as required and cycle through them quickly using keyboard shortcuts.
Note: You cannot use versions when a clip is opened as a timeline, that is, by using the
right-click Open In > Timeline View option.
The application relies on specific file naming or folder structure conventions to discover versions:
Convention Description Example
File name constants
Clip name The file name base must remain the same. myClip_v1.0001.dpx
myClip_v2.0001.dpx
myClip_v3.0001.dpx
Version prefix The delineation between the file name and myClip_v1.0001.dpx
version information must be either _
(underscore) or . (period) and remain the myClip_v2.0001.dpx
same for all versions. myClip_v3.0001.dpx
File name variables
Version The version number padding in the clip myClip_v1.0001.dpx

padding name can be increased or decreased.
myClip_v002.0001.dpx
myClip_v03.0001.dpx
Frame padding The frame padding in the clip name can be myClip_v1.01.dpx
increased or decreased.
myClip_v1.1.dpx
myClip_v1.0001.dpx
Extension The file format is interchangeable. See myClip_v1.01.png

Appendix C: Supported File Formats for
myClip_v1.0001.dpx
USER GUIDE
1424
Using Versions |
Convention Description Example
more information. myClip_v1.mov
Note: If the file extension is a movie format, such as .r3d or .mov, the Frame padding
can be omitted.
Folder name constants
Root folder The root folder name must remain the ~/version/v1/myClip_v1.0001.dpx
same for all folders containing versions.
~/version/v2/myClip_v2.0001.dpx
~/version/v3/myClip_v3.0001.dpx
Folder name variables
Version The version number padding in the folder 09_WF_Shot004_v1

padding name can be increased or decreased.
09_WF_Shot004_v002
09_WF_Shot004_v03
USER GUIDE
1425
Versions in Bins |
Versions in Bins
Versions behave similarly in both bins and sequences, and in both cases, you first have to ingest an
existing version.
Ingest and locate the versioned clip, then:

1. Right-click and select Version > Scan for Versions to search for available versions.
A dialog box lets you know how many versions were discovered.
2. Use the right-click Version menu to:
• Go to the next Version Up or Version Down.
• Go to the Minimum or Maximum Versions.
Tip: You can also use the Alt+Up/Down Arrow keyboard shortcuts to increment versions or
Alt+Shift+Up/Down Arrow to go to the maximum or minimum.
When you reach the end of the discovered versions, incrementing the version automatically scans
for new versions that may have become available.
3. For source clips only, you can right-click the clip and select Open In > Versions Bin to display all
discovered versions of the target clip.
The versioning conventions may allows clips into the Version Bin that you weren’t expecting. You
can disable versions by selecting them and pressing D or by selecting the Set Active Version of a
clip using the right-click Version menu.
The Active Version is the version displayed when you drag the source clip to the timeline,
denoted by the orange V in the top left-hand corner of the thumbnail.
USER GUIDE
1426
Versions in Bins |
4. Once you’ve sorted all the available versions, select a clip in the bin view and press V to display all
versions for that clip in a convenient window. Disabled versions are not displayed.
5. Select the required clip to set the Active Version and apply it to the clip.
USER GUIDE
1427
Versions in Sequences |
Versions in Sequences
As mentioned previously, versions behave similarly in both bins and sequences, but swapping
versions in sequences allows you to compare results more easily.
Note: You cannot use versions when a clip is opened as a sequence, that is, using the right-
click Open In > Timeline View option.
Locate the ingested version clip and drag it to the timeline, right-click and select the Version menu:
• Scan For Versions to locate new versions of the clip.
• Version Up or Version Down to increment the version by one.
• Go to the Minimum or Maximum Version.
Tip: You can also use the Alt +Up/Down Arrow keyboard shortcuts to increment versions
or Alt+Shift+Up/Down Arrow to go to the maximum or minimum.
Once you’ve scanned for versions, select a shot on the timeline and press V to display all available
versions for that item in a convenient window.
Select the required shot version to set the Active Version.
USER GUIDE
1428
Using Snapshots |
Using Snapshots
Within a project you can save the current state of a sequence as a snapshot, including a comment or
tag to describe that snapshot. You can see what snapshots exist for a sequence in the bin view and
flip it back to any previously saved state.
An example workflow might appear as follows:
1. Two snapshots of the sequence (SQ) are recorded after edits. See Creating Snapshots.
2. Snapshot 1 is then restored. See Restoring Snapshots.
3. Further edits are made, then the sequence is recorded as snapshot 3.
USER GUIDE
1429
Creating Snapshots |
Creating Snapshots
To create a snapshot for a sequence:
1. Locate the sequence in the bin view.
2. Right-click the sequence and select Snapshot > Add Snapshot.
The Add new snapshot dialog box displays.
3. Enter a comment, or use the default date and time supplied.

4. Click Add to create the snapshot.
Snapshots are indicated in the bin view with a camera icon containing the number of snapshots
available.
USER GUIDE
1430
Creating Snapshots |
USER GUIDE
1431
Restoring Snapshots |
Restoring Snapshots
To restore a snapshot:
1. Locate the sequence in the bin view.
2. Right-click the sequence and select Snapshot > Restore Snapshot.
3. Select the required snapshot from the list.
A warning displays reminding you that edits since the snapshot was taken are lost.
4. Click OK to restore the sequence to the point at which the snap was recorded.
USER GUIDE
1432
Exporting from Nuke Studio
This section deals primarily with shot management and export functionality when you're farming out
shots or sequences to other artists. It also deals with the presets, which dictate how Create Comp
passes data between the Timeline environment and Compositing environment.
The export suite can transcode, export clip selections from a timeline or bin, write out EDLs and
XMLs, or bake out an entire timeline as a single clip in your required delivery format. The Export
presets are also used to manage how Create Comp sends clips back and forth between Nuke Studio's
Timeline environment and Compositing environment using Local and Project Presets.
Nuke Studio ships with several context-sensitive and ad hoc export options:
• Exporting Sequences and Shots - the process of preparing a sequence or individual shots for export
and paving the way for VFX work to come back into Nuke Studio.
• Transcoding - converts one file format to another. You can transcode sequences, timeline
selections, and clips from the bin view.
• Ad Hoc Exports - an umbrella covering exports that you might not perform on a per project basis,
such as EDL or XML exports.
• Create Comp - edit or create the presets used during Create Comp, passing source clips or shots to
the Compositing environment and sending rendered Write nodes from the Node Graph back to the
Timeline environment.
With the addition of Python bindings to perform the same functions, this infrastructure provides a
massive amount of flexibility, whether your pipeline is GUI or command line orientated.
Note: Nuke Studio is non-destructive and can slot into your pipeline if you setup your shot
template to mirror the existing file structure.
USER GUIDE
1433
Introduction to the Export Dialog |
Introduction to the Export Dialog

Nuke Studio uses presets and shot templates to perform export operations, including round-tripping
and EDL/XML creation. The Export dialog controls what is exported and where, and whether or not to
expect versioned clips as part of a round-trip from Nuke.
The Export dialog is accessed from the File menu, from the right-click bin and timeline menus, or by
using the keyboard shortcut Ctrl/Cmd+Shift+E.
The Shot Template is also used to create the presets used during Create Comp, passing shots from
the Compositing environment and sending rendered Write nodes from the Node Graph back to the
Timeline environment.
Nuke Studio uses Content Presets in all shot templates, enabling you to create commonly used
export conditions, which are then available across all projects. Some presets are only available with
certain exporters, for example, the EDL Exporter preset cannot be used with Process as Shots
exports.
USER GUIDE
1434
Introduction to the Export Dialog |
You can filter your exports using the Tracks for this Export and Filter by Tag lists, exporting only
certain tracks or shots marked with a particular tag. See Using Tags for more information.
For your convenience, Nuke Studio ships with a number of ready-made Content Presets, but you can
edit these as required:
• Transcode Images - defines transcode parameters allowing you to save your most-used file type
conversions.
• Nuke Project File - defines the script name and paths used by Nuke Read and Write nodes during a
round-trip or Create Comp.
• Nuke Write Node - defines the render format for Nuke Write nodes. Add multiple Nuke Write Node
presets to create multiple Write nodes in the resulting Nuke script.
• Nuke Annotations File - defines the script name and paths used by Nuke Write nodes and
Precomp group during a round-trip or Create Comp.
• Render with - selects how Nuke Studio renders your export: Single Render Process or Frame
Server. This dropdown defaults to Frame Server using the number of slave processes specified in
the Preferences > Performance > Threads/Processes. If you set this preference to 0, Nuke Studio
relies on external machines set up as render slaves. See Using the Frame Server on External
Machines for more information.
• EDL Exporter - used to export a sequence to the EDL format.
• SymLink Exporter - creates symlinks to the location of the source files, rather than making copies.
• XML Exporter - used to export a sequence to XML format.
• Copy Exporter - creates copies of the source files to a specified location, rather than symlinking.
• Audio Export - copies any audio tracks to .wav files in a specified location.
USER GUIDE
1435
Using Local and Project Presets | Using the Shot Template
Using Local and Project Presets

Presets are containers for export preferences, such as file structure and format, and filters for tracks,
tags, and frame range. Two types of Presets are available to construct commonly used export tasks:
• Local Presets - these presets are used to set up round-trips between artists on different platforms
and also to manage passing files between the Timeline and Compositing environment. See Create
Comp for more details. Local Presets are saved in a Task Presets folder using the XML file format.
• Project Presets - you can drag-and-drop Local Presets into this panel to save the preset within a
project .hrox file. This option is designed for collaborative work, allowing you to quickly share your
export presets.
Using the Shot Template

The shot template sets up the folder hierarchy and naming conventions for export presets such as
Basic Nuke Shot with Annotations and Transcode Clip DPX, and how Create Comp sends clips
back and forth between the Timeline environment and Compositing environment. Any folders added
to the template are created during export unless they already exist, in which case the export writes to
the existing structure.
Nuke Studio ships with default templates for your convenience, but you can quickly create custom
templates using folders and “tokens”, which are replaced with the relevant information during export.
Tip: Clicking an entry in the shot template displays a preview file path with the tokens
resolved under the Version token number field.
Exports can resolve the following tokens:
Token Resolves to
_nameindex The index of the shot name in the sequence, preceded by _

(underscore), to avoid clashes with shots of the same name.
{ampm} The local equivalent of either AM or PM.
{binpath} The bin structure to preserve. Including this token recreates your bin
structure up to the nearest parent bin.
USER GUIDE
1436
Token Resolves to
{clip} The name of the clip used in the shot processed.
{day} The local weekday name, abbreviated to Mon, Tue, and so on.
{DD} The day of the month as a decimal, 01, 02, and so on.
{event} The timeline event number associated with the shot to process.
{ext} The extension of the file to output, such as .dpx or .mov
{filebase} The base of the clip name to process. For example, the filebase of
Shot01_####.dpx is Shot01_####.
{fileext} The format of the clip to process, such as .dpx or .mov
{filehead} The source clip filename not including frame padding or extension.
For example, the filehead of Shot01_####.dpx is Shot01.
{filename} The source clip name of the media to process.
{filepadding} The source filename padding, which you might use for formatting
frame indices.
{filepath} The full file path to the source media referenced in the export.
{fullbinpath} The full bin structure to preserve. Including this token recreates the
bin structure up to the project level.
{fullday} The local full weekday name.
{fullmonth} The local full month name.
{hierotemp} The temp directory as specified in the Preferences.
{hour12} The export start time hour component (12-hour clock).
{hour24} The export start time hour component (24-hour clock).
{MM} The month of the year as a decimal, 01, 02, and so on.
{minute} The export start time minute component.
{month} The local month name, abbreviated to Jan, Feb, and so on.
USER GUIDE
1437
Token Resolves to
{project} The name of the parent project of the export item.
{projectroot} The root export file path as specified in the Timeline Environment
Project Settings.
{second} The export start time second component.
{sequence} The sequence name to process.
{shot} The name of the shot to process.
{timestamp} The export start time in the 24-hour clock format (HHMM).
{track} The name of the track to process. Exporting EDLs using this token
generates a separate EDL for each track.
{user} The current username.
{version} The string v#, defined by the number (#) set in the Version section of
the export dialog
{YY} The year of the century as a decimal, 01, 02, and so on.
{YYYY} The year, including century.
Tip: Double-click the path column, right-click, and then choose Select Keyword to display a
list of available export tokens, though only valid tokens for the current selection are listed.
USER GUIDE
1438
Using Local and Project Presets | Custom Shot Templates
Token substrings are valid if you need to extract a certain part of an evaluated string. For example, if
{shot} resolves to JB100, then:
• {shot [0:2] } - resolves to JB
• {shot [-3:] } - resolves to 100
Similarly, anything within the outer brackets is evaluated as a Python string. For example, if {shot}
resolves to JB_10_20, then:
• {shot.split(’_’) [0] } - resolves to JB
• {shot.split(’_’) [2] } - resolves to 20
Custom Shot Templates

The shot template enables you to create as many Nuke Read and Write nodes as required for a
project. A typical use case might be creating .jpg clips for review and .dpx resolution clips for
finishing.
Multi-format Exports
The following example describes how to build a shot template to export a sequence of .mov clips,
create .dpx and .jpg Write nodes in Nuke, and bring the .dpx clips back into the timeline.
1. In the Export dialog, select Basic Nuke Shot in the Local Presets panel to auto-complete the
shot template with the preset values.
2. Click Duplicate selected preset and give the new preset a name.
3. Rename the renders folder renders_dpx.
4. Select the nuke folder and click the folder icon to add a new folder. Name the new folder
renders_jpg.
5. Select the renders_jpg folder and click to add a new entry.

6. Replace the {filename} token with {shot}_comp{_nameindex}_{version}.####.{ext}, the same as
the existing entry under renders_dpx.
Note: The #### marks represent frame numbers for image sequences. If you were creating
.mov clips, they’d be omitted.
The shot template should look something like this:
USER GUIDE
1439
7. Click the Content column and select Nuke Write Node.
Note: When using a third party application to produce the VFX work, select
ExternalRender instead of Nuke Write Node.
8. In the Content settings tab, use the File Type dropdown to select jpeg.
USER GUIDE
1440
Notice that the settings available change depending on the File Type selected.
9. Click Nuke Project File in the shot template and check that both Write nodes are enabled.
10. Set up the rest of the export as described in Exporting Sequences and Shots and click Export.
USER GUIDE
1441
Adding Burn-in Text to Exports | Custom Shot Templates
Adding Burn-in Text to Exports

Nuke Studio can burn-in text during the export process using a simple Nuke gizmo. The gizmo is
accessed from the Nuke Write Node preset's Content panel under Burn-in.
The gizmo contains controls for the font style and fields denoting the position of the text. You can
also add burn-in directly on timeline using the Burn-In soft effect. See Soft Effects for more
information.
Note: The Font field only accepts the full file path and name of the font file to use. For
example, on Mac ~/Library/Fonts/Verdana.ttf
Click Edit to display the available controls.
You can mix-and-match the following methods to create burn-in text:

• Enter text manually, what you see is what you get in the burn-in.
• Use any of the tokens valid in the shot template as burn-in tokens. For example:
{shot}_comp
Extracts the shot nam from the timeline and appends _comp.
See Using Local and Project Presets for more information.

• Use metadata from tags applied to clips and shots. For example:
USER GUIDE
1442
Adding Burn-in Text to Exports | Custom Shot Templates
[metadata hiero/tags/Approved]
Extracts the Approved tag from the clip or shot. You can also append note to include any notes
associated with the tag:
[metadata hiero/tags/Approved/note]
Note: You must precede spaces and slashes in the tag name with \\ (backslashes) to enable
Nuke Studio to process the tag name correctly. For example: [metadata hiero/tags/Blue\\
Screen/note]
Tip: If you're not sure what metadata keys and values are available on a shot, you can add a
Text soft effect containing the Tcl expression [metadata values] to display all metadata in
the Viewer. See Soft Effects for more information.
USER GUIDE
1443
Adding Additional Nodes During Export | Custom Shot Templates
Adding Additional Nodes During

Export
Nuke Studio can include additional nodes, in any Nuke Project File or Transcode export in the Shot
Template, by simply copying and pasting scripts from the Node Graph.
You can add nodes to shots, tracks, or sequences, or include them as unconnected ad hoc nodes in
the script, filtered by tags if necessary.
1. In the Content tab, scroll down to the Additional Nodes control and click Edit.
The Additional Nodes Setup dialog displays.
2. Click to add an entry.
3. Click the Apply To field and select what the current entry applies to:
• Shot - the additional nodes are added to the script for each shot in the export.
• Track - the additional nodes are added to the script for each track in the export.
• Sequence - the additional nodes are added to the script for the entire sequence.
• Unconnected - the additional nodes are added to the script, but are not connected to the
export node tree.
• None - temporarily disables the current entry.
4. Select the Tags that you intend to use to filter which items receive the additional nodes.
If you want to affect only the Reference track, for example, select the Reference tag. All items
without that tag are ignored.
5. Copy and paste a node from the Node Graph into the Script panel.
USER GUIDE
1444
Adding Additional Nodes During Export | Custom Shot Templates
Note: If you need more than one node, you might consider creating a Group in the Node
Graph and pasting that into the Script panel.
6. Click OK to accept the additional nodes.

7. Select the Additional Nodes checkbox and complete the export process as described Exporting
Sequences and Shots.
USER GUIDE
1445
Using the Frame Server on External Machines | Configuring the Frame Server on External Machines
Using the Frame Server on

External Machines
Although Nuke Studio is capable of rendering frames internally, running the Frame Server on an
external machine can accelerate the process considerably by sharing work across a network of
machines.
Note: The Frame Server requires a Nuke Studio license (nukestudio_i) on the main
workstation, but only a Nuke render license (nuke_r) on the slave machines. Local Frame
Server processes use ports 5558-5662.
If you want to use an interactive license (nuke_i) on the slave machines, add the --
useInteractiveLicense argument to the runframeserver.py command described below.
Configuring the Frame Server on External Machines

Nuke Studio's Frame Server can be set up on an external machine (or a number of machines) to
render from your Nuke Studio session. To do this, you need to run the runframeserver.py script on
the external machines, found inside the Python site-packages, with specific command line
arguments.
Warning: In order for everything to work smoothly, you need to ensure that both your
external slave machines and main Nuke Studio session can read and write files to a shared
location, such as an NFS share.
Depending on platform this can be done by manipulating your default umask setting, but
be aware that this alters the permissions of the created files.
Additionally, Mac and certain Linux distributions, such as RHEL, can not function as the main
workstation if the firewall is blocking the communication port 5560. You can configure the
firewall to allow certain ports through the firewall using the iptables command, but use
caution when doing so. For example:
sudo iptables -I INPUT -p tcp --dport 5560 --syn -j ACCEPT
USER GUIDE
1446
Using the Frame Server on External Machines | Configuring the Frame Server on External Machines
Please refer to the documentation on firewalls for your particular platform for more
information.
The Frame Server uses a number of worker processes on the external machine, each of which
requires allocated resources, such as threads, memory, and so on. There are a number of arguments
that you must pass to runframeserver.py for the server to work correctly:
• --numworkers - this is the number of concurrent Nuke processes that are launched when you run
this server render node.
• --nukeworkerthreads - the number of threads that each worker is allocated. This is similar to
setting the -m argument when running Nuke from the command line.
• --nukeworkermemory - the amount of memory, in MB, allocated to each frame server worker.
• --workerconnecturl - the TCP port address of the main workstation you want to serve. For
example:
tcp://bob:5560
where bob is the resolved hostname of a machine you wish to serve. You can also use an IP address.
Tip: To ensure that you're entering a valid URL, try using the ping command to see if you get
a response.
• --nukepath - the path to the Nuke application on the slave workstation.
Tip: On Windows, if there are spaces in the file path, remember to place the path in quotes.
For example, --nukepath="C:\Program Files\Nuke11.3v5\Nuke11.3.exe"
On a Linux slave machine, an example command prompt entry running from the install directory
./python ./pythonextensions/site-
packages/foundry/frameserver/nuke/runframeserver.py --numworkers=2 --
--nukepath=./Nuke11.3
On a Windows slave machine, an example command prompt entry running from the install directory
python.exe pythonextensions\site-
packages\foundry\frameserver\nuke\runframeserver.py --numworkers=2 --
--nukepath=Nuke11.3.exe
USER GUIDE
1447
Using the Frame Server on External Machines | Frame Server Logs
In the examples, we specify that the slave uses two Nuke workers, with four threads and 8 GB RAM
each, and are slaved to the main Nuke Studio workstation running on bob.
Tip: If your slave machines run a different OS than your main Nuke Studio machine, you can
use the --remap command line argument to convert file paths between them. The host file
path is read first followed by the slave file path. Nuke Studio expects all file paths to use /
(forward slash) between directories. For example:
--remap "P:/,/mnt/renders/"
converts host paths beginning with P:/ (Windows style) to slave paths beginning with
/mnt/renders/ (Linux style).
You can check that the Frame Server and workers are connected by running the following lines in the
Script Editor on the main workstation:
from hiero.ui.nuke_bridge.FnNsFrameServer import frameServer
print [worker.address for worker in frameServer.getStatus(1).workerStatus]
Successful connections should report something similar to the following in the output panel:
['Worker 0 - henry.local - 192.168.1.11', 'Worker 0 - bob.local -
192.168.1.111', 'Worker 1 - henry.local - 192.168.1.11']
Where henry.local is the name of the remote slave, and bob.local is the name of the main Nuke
Studio session.
Note: If the workers cannot contact the Frame Server, an exception is printed in the Script
Editor's output panel.
Frame Server Logs

Broker and Worker logging can to help diagnose Frame Server issues. The logs are written to NUKE_
TEMP_DIR/logs by default, and take the form:
broker.log
worker-0.log
worker-1.log
worker-2.log
USER GUIDE
1448
Using the Frame Server on External Machines | Frame Server Logs
Note: Running the Frame Server using Python, as described above, always writes log files to
the specific OS temporary directory. For example, on Windows C:\temp is used.
Tip: You can use the FRAMESERVER_LOG_DIR environment variable to force Frame Server
logs into a different location. See Environment Variables in the Nuke Online Help for more
information.
USER GUIDE
1449
Exporting Sequences and Shots | Frame Server Logs
Exporting Sequences and Shots

1. Select an entire sequence in the bin view, or shots in the timeline, and navigate to File > Export...
The Export dialog displays.
2. Select Process as Shots from the Export dropdown.
3. Enter the Export To directory or click Choose... and browse to the location.
The Export To directory is the starting point from which the shot template builds your shot
hierarchy.
4. Select the Basic Nuke Shot preset under Local Presets to auto-complete the shot template or
build a custom shot template by copying an existing template and editing as required using Path
tokens, the Contents field, and the folder and +/- buttons.
Basic Nuke Shot creates a folder for each clip, or shot, containing nuke, script, and renders
folders.
The tokens in the Basic Nuke Shot template break down as follows:
• {shot} simply extracts the shot names as they appear in the timeline.
• {shot}_comp{_nameindex}_{version}.nk extracts the shot name for each clip and the version
selected in the Tracks and Handles controls. For example, Shot01_comp_v03.nk
• {shot}_comp{_nameindex}_{version}.####.{ext} appends padding and the specified file
extension. For example, Shot01_comp_v03.0001.dpx
Note: The {_nameindex} token is included to avoid conflicts with non-unique shot names.
Tip: Select a file entry in the shot template to display a preview of the file path with all the
tokens resolved.
5. Proceed to Nuke Project File Settings to determine the Nuke script's behavior.
USER GUIDE
1450
Nuke Project File Settings | Frame Server Logs
Nuke Project File Settings

Note: Custom shot presets can only be selected from the Project Settings if they contain a
Nuke Project File and Nuke Write Node Content preset.
1. Click the Nuke Project File Content preset to display the script settings.
2. Select Write Nodes from the dropdown and check which path from the shot template should be
used for the Nuke Write node. For example:
{shot}/nuke/renders/{shot}_comp{_nameindex}_{version}.####.{ext} to resolve the render
path where Nuke Studio expects to find the files when they're rendered.
Note: If you included a Nuke Annotations File Content preset, enable the Annotations
Precomp creator. See Annotations for more information.
3. If you're exporting retimed media, set how you want the Nuke script to handle the retime:
• None - no retime is applied.
• Motion - vector interpolation is used to calculate the in between frames. This is the most
accurate retime method, but takes longer to render.
• Frame - the nearest original frame is displayed.
• Blend - a mix between two frames is used for the in between frames. This is quick to render and
is useful when tweaking the timing in the Curve Editor before setting the method to Motion.
4. Soft Effects added to shots in your export are included in the resulting Nuke script by default. If
you don't need the soft effects, disable Include Effects to omit them from the script. See Soft
Effects for more information.
5. Select the required Reformatting options:
• Plate Resolution - exports at the clip's original resolution, regardless of what is set in the
timeline.
• To Sequence Resolution - exports at the resolution set in the timeline Sequence panel Output
Resolution dropdown.
• Custom - activates the Reformat controls allowing you to customize the export resolution.
6. Enable Collate Shot Timings or Collate Shot Name to create additional Nuke Read nodes in the
same script for clips that would normally be hidden by clips higher up the track hierarchy or clips
on the same track with the same shot name.
USER GUIDE
1451
Nuke Project File Settings | Frame Server Logs
Note: If you have a Read node selected, you can’t enable the Collate functions.
For example:
• Collate Shot Timings - Items on track 1 that would otherwise be hidden by track 2.
Timeline environment Compositing environment
Note: Shots on different tracks are not connected by default. If you want all the exported
clips to be connected to the Nuke script Write node, enable Connect Tracks.
This applies to stereo and multi-view sequences as well when you use separate files for the
tracks per view. See Stereoscopic and Multi-View Projects for more information.
• Collate Shot Name - Two items on the same track with the same shot name.
Timeline environment Compositing environment

7. If you want to add additional nodes to the script on export, enable Additional Nodes and click
Edit. See Adding Additional Nodes During Export for more information.
8. Proceed to Nuke Write Node Settings to determine the Write node's behavior.
USER GUIDE
1452
Nuke Write Node Settings | Frame Server Logs
Nuke Write Node Settings

Note: Custom shot presets can only be selected from the Project Settings if they contain a
Nuke Project File and Nuke Write Node Content preset.
1. Click the Nuke Write Node Content preset to display the write settings.
2. Set the following controls common to all file types:
• Channels - set the channels to export using the dropdown. If you want to export a non-standard
channel, type the name of the channel into the field manually.
• Write Node Name - if you intend to create more than one Nuke Write node, define the name
here. The default, Write_{ext}, appends the individual Write nodes with the file extension being
written. You can, however, use any of the tokens Nuke Studio recognizes.
• Colorspace - use the dropdown to set the colorspace to render, such as linear, REDLog, or raw.
3. Select the file type to render using the dropdown and complete the relevant fields, dependent on
the file type selected.
Note: Selecting mov from the dropdown provides additional QuickTime specific controls,
allowing you to choose a codec, encoder, and in some cases, YCbCrMatrix. The matrix
control enables you to use the new Rec 601 and Rec 709 or the Legacy encoding methods,
which are the methods used previously in Nuke. There's also an Advanced dropdown
containing mov32 and mov64 encoder specific controls.
Similarly, selecting exr provides an additional metadata dropdown allowing you to export or
round-trip selected metadata along with your .exr output.
4. Create Directories is enabled by default, which enables the corresponding control in the .nk
script's Write node. This control allows Nuke to create the required directories when you render
out a new version from the Write node.
Disabling this control causes versioned renders to fail because the target directories don't exist.
You can manually create the correct directories or enable Create Directories in the script's Write
node if this happens.
5. Use the Reformat controls to determine how the Write node is set up in the Nuke script:
• None - the clip or sequence resolution is used, no additional formatting is applied during
export.
• To Sequence Resolution - exports at the resolution set in the timeline Sequence panel Output
Resolution dropdown. This option also allows you to set the Filter used to resize the output.
USER GUIDE
1453
Note: The filters available are generally listed in order of quality and processing time. Cubic
can be faster to render, but Lanczos4 may be produce better results. See Choosing a
Filtering Algorithm for more information on filters in Nuke.
• To Scale - activates all the Reformat controls, except Format, allowing you to customize the
export resolution.
• Custom - activates all the Reformat controls, except Scale, allowing you to customize the
export resolution.
Reformat Control To To Custom

Sequence Scale
Resolution
Format - sets the format to render out in Nuke, such as 1920x1080

HD 1080.
Scale - sets the proportion by which to scale the output format.
Resize - sets the method by which you want to preserve or override

the original aspect ratio:
• width - scales the original until its width matches the format’s
width. Height is then scaled in such a manner as to preserve the
original aspect ratio.
• height - scales the original until its height matches the format’s
height. Width is then scaled in such a manner as to preserve the
original aspect ratio.
• fit - scales the original until its smallest side matches the format’s
smallest side. The original’s longer side is then scaled in such a
manner as to preserve original aspect ratio.
• fill - scales the original until its longest side matches the format’s
longest side. The input’s shorter side is then scaled in such a
manner as to preserve original aspect ratio.
• distort - scales the original until all its sides match the lengths
specified by the format. This option does not preserve the original
aspect ratio, so distortion may occur.
Center - when enabled, transform image to the center of the output.

When disabled, the image is aligned with the bottom-left corner of
the output.
USER GUIDE
1454
Reformat Control To To Custom

Sequence Scale
Resolution
Filter - sets the filtering algorithm used to transform pixels. See

Choosing a Filtering Algorithm for more information on filters in
Nuke.
6. You can apply text burn-in to the media using a Nuke gizmo. Enable Burn-in Gizmo and click Edit
to define the information applied. See Adding Burn-in Text to Exports for more information.
7. Proceed to Tracks, Range, and Handles Settings to select which items are processed during
export.
USER GUIDE
1455
Tracks, Range, and Handles Settings | Frame Server Logs
Tracks, Range, and Handles

Settings
The tracks, tags, and handles controls in the Export dialog allow you to select the frame range or
shots to export.
1. Set the Version number for the export, if applicable. Use the arrows to increment the version
number and the +/- buttons to increase or decrease the padding. You can also type directly into
the numeric field.
Note: See Using Versions for more information on how versioning works in Nuke Studio.
2. Select Tracks for this export by enabling or disabling the tracks in the list. Nuke Studio exports
all tracks by default.
3. Enable or disable tags using the Filter by tag panel. Click the checkbox to cycle through the
available tag states.
4. If you're exporting a sequence, set the Range controls as required:
• Select Whole Sequence or In/Out Points to export only the selected frames.
• Set how clip Start Frames are derived using the dropdown menu:
• Sequence - use the sequence’s start frame.
• Custom - specify a start frame for all clips using the field to the right.
5. If you're exporting shots, set the Handles controls as required:
• Clip Length - exports the full clip length available, as if the clip was opened as a Viewer.
• Cut Length - exports only the cuts included on the timeline.
Note: Selecting Cut Length allows you to add handles to each clip, up to the maximum
available source clip length.
• Check Apply Retimes to export any retimes present on the timeline.
Note: When Apply Retimes is disabled, which is the default state for Create Comp, any
TimeWarp soft effects are not included in the resulting Nuke script. When the new shot is
USER GUIDE
1456
Tracks, Range, and Handles Settings | Frame Server Logs
created through Create Comp or Build Track from Export Tag, TimeWarp soft effects are
copied from the original shot to the new one.
• Source - use the source clip’s start frame.
6. Set how Nuke Studio should render your export using the Render with dropdown. Nuke Studio
provides the following options:
• Frame Server - uses multiple Nuke processes to speed up render times and shares a render
queue with any Nuke Comp renders in the timeline, improving resource management.
See Using the Frame Server on External Machines for more information.
• Single Render Process - uses a single Nuke process to render your export. Rendering
QuickTimes falls back to this setting, but it's also used when a problem is detected with the
Frame Server.
• Custom Render Process - uses a custom render process. Nuke Studio requires a Python script
to pass exports to your render farm of choice. Scripts must be located in specific directories,
dependent on platform, as listed in Loading Gizmos, NDK Plug-ins, and Python and Tcl Scripts.
If no scripts exist, the dropdown only contains the default render processes.
7. Click Export.
The Export Queue window displays an estimate of how long each component of the export is
expected to take.
Once the export is complete, the file structure specified in the shot template is created. You can
then import the Nuke clips on a separate track when they’re ready.
Tip: Click the magnifying glass icon to reveal the file structure in a browser window.
When clips are exported from Nuke Studio, they are marked with a Nuke tag flagging which clips
have an export history. Clips tagged in this way can be used to build VFX tracks quickly as
described in Building VFX Tracks and Comp Clips.
USER GUIDE
1457
Building VFX Tracks and Comp Clips | Building Tracks From Export Structure
Building VFX Tracks and Comp

Clips
When the compositing work is complete, the clips are ready to re-ingest. The shot template defines
where the Nuke files reside, so all you need to do is instruct Nuke Studio to build tracks from
previous exports. See Building Tracks From Export Structure for more information.
Alternatively, if you have a history of rendered VFX clips, different versions and so on, you can also
build tracks from export tags to select from a list of available clips. This method allows you to add
Comp Clips to the timeline, which act as containers for Nuke .nk scripts, or placeholders for Nuke
renders. See Building Comp Clips From Export Tags or Building Render Placeholders From Export
Tags for more information.
Building Tracks From Export Structure

1. Select the required clips on the timeline and right-click to display the context sensitive menu.
Tip: You may find it easier to select clips in the spreadsheet and then right-click on the
timeline.
2. Click Build Track > From Export Structure.

The Build Track From Export Structure dialog displays.
3. Enter a Track Name or use the default VFX.
USER GUIDE
1458
Building VFX Tracks and Comp Clips | Building Tracks From Export Structure
4. Select an Export Preset using the dropdown menu. In this case, select the same preset used
during the export.
5. Enter the file path of the Export Root directory or click Choose and browse to the location.
Note: The root directory is the location entered in Export To when exporting the project.
6. Select the Content preset you intend to ingest from the shot template. In this case, the Nuke
Write Node.
7. Click Build to create the VFX track.
Note: Nuke Studio warns you if no selection is made in the Content column.
Nuke Studio automatically creates a new track containing the VFX clips, if they exist, or offline
place holders if the clips are a work in progress.
If a shot already exists in any of the target tracks, a new track is created to hold the new shots.
The clips are automatically updated when work is complete as long as they are saved with the
expected name and location, as specified in the shot template.
USER GUIDE
1459
Building VFX Tracks and Comp Clips | Building Comp Clips From Export Tags
Building Comp Clips From Export Tags

When you build a track from an export tag, Nuke Studio imports Comp Clips containing the Nuke
script, by default. Comp Clips are shots that reference Nuke scripts, rather than being placeholders
for offline clips as in the Building Render Placeholders From Export Tags workflow.
timeline.
2. Click Build Track > From Export Tag.

The Build Track From Export Tag dialog displays.
4. Select the required export tag in the left-hand panel to display tag information in the right-hand
panel.
Nuke Studio imports the .nk Comp Clip by default.
If you just want to import the offline renders when they're finished, disable the Create Comp
Clips checkbox. See Building Render Placeholders From Export Tags for more information.
Nuke Studio automatically creates a new track containing the Comp Clips. If a shot already exists
in any of the target tracks, a new track is created to hold the new shots.
USER GUIDE
1460
Building VFX Tracks and Comp Clips | Building Render Placeholders From Export Tags
6. You can double-click Comp Clips to open them in Nuke Studio's Compositing environment to
make edits as required.
Building Render Placeholders From Export Tags

When you build a track from an export tag, you can choose to import the renders from the .nk script,
rather than Comp Clips which contain the Nuke script.
timeline.
2. Click Build Track > From Export Tag.

The Build Track From Export Tag dialog displays.
4. Select the required export tag in the left-hand panel to display tag information in the right-hand
panel.
5. Disable the Create Comp Clips checkbox to import the offline renders when they're finished.
USER GUIDE
1461
Building VFX Tracks and Comp Clips | Building Render Placeholders From Export Tags
If you want to import the .nk Comp Clips, enable the Create Comp Clips checkbox. See Building
Comp Clips From Export Tags for more information.
Nuke Studio automatically creates a new track containing the VFX clips, if they exist, or offline
place holders if the clips are a work in progress.
If a shot already exists in any of the target tracks, a new track is created to hold the new shots.
USER GUIDE
1462
Exporting Multi-View Source Clips | Building Render Placeholders From Export Tags
Exporting Multi-View Source Clips

Multi-view exports are similar to regular exports, but the DPX Multi-View or Multi-View Nuke Shot
templates are used to created the required export tree using %v functionality, just like Nuke's Node
Graph. The DPX Multi-View example preset is designed for exporting sequences and the Multi-View
Nuke Shot preset is designed for shots.
See Stereoscopic and Multi-View Projects for more information on working with multi-view footage.
1. Right-click on the sequence or shot that you want to export.
Sequences are exported from the bin view.
Shots can be a multi-view or single views split into separate tracks. See Displaying Views in the
For separate tracks, you can right-click a single track to export all views or select all the per-track
views and right-click.
2. Select Export.
USER GUIDE
1463
Exporting Multi-View Source Clips | Building Render Placeholders From Export Tags
The Export dialog is displayed.

3. Select Process as Sequence or Process as Shots from the Export dropdown.
If you're exporting a sequence, select the Log10 Cineon DPX Multi-View example preset. If you're
exporting shots, select the Multi-View Nuke Shot (%v) example preset.
4. Set the Nuke Project File and Nuke Write Node presets as described in Nuke Project File
Settings and Nuke Write Node Settings.
If you're using independent files per track, that is without importing multi-view files or using %V
functionality, the separate tracks in the export script are not connected by default. If you want the
tracks to be connected in the script, enable Connect Tracks in the Nuke Project File preset.
5. Set the track and handle preferences as described in Tracks, Range, and Handles Settings and
then click Export.
Tip: Click the magnifying glass icon in the Export Queue to reveal the file structure in a
browser window.
When clips are exported from Nuke Studio, they are marked with a Nuke tag flagging which clips
have an export history. Clips tagged in this way can be used to build VFX tracks quickly as
described in Building VFX Tracks and Comp Clips.
USER GUIDE
1464
Transcoding | Transcoding a Sequence
Transcoding
Transcoding in Nuke Studio uses a background render process to convert one file format to another.
You can transcode sequences, timeline selections, and clips from the bin view.
Transcoding a Sequence
1. Select a sequence in the bin view and navigate to File > Export...
2. Select Process as Sequence and the preset you intend to use, or use the default dpx preset.
4. Click the Content column in the shot template to display the transcode options.
5. Set the following controls common to all file types:
• Channels - set the channels to export using the dropdown. If you want to export a non-standard
channel, type the name of the channel into the field manually.
• Colorspace - use the dropdown to set the colorspace to render, such as linear, REDLog, or raw.
6. Select the File Type to render using the dropdown and complete the relevant fields, dependent on
the File Type selected.
Note: Selecting mov from the dropdown provides additional QuickTime specific controls,
allowing you to choose a codec, encoder, and in some cases, YCbCrMatrix. The matrix
control enables you to use the new Rec 601 and Rec 709 or the Legacy encoding methods,
which are the methods used previously in Nuke. There's also an Advanced dropdown
containing mov32 and mov64 encoder specific controls.
Similarly, selecting exr provides an additional metadata dropdown allowing you to export or
round-trip selected metadata along with your .exr output.
7. Use the Reformat controls to determine how the Write node is set up in the Nuke script:
• None - the clip or sequence resolution is used, no additional formatting is applied during
export.
• To Scale - activates all the Reformat controls, except Format, allowing you to customize the
export resolution.
• Custom - activates all the Reformat controls, except Scale, allowing you to customize the
export resolution.
USER GUIDE
1465
Transcoding | Transcoding a Sequence
Reformat Control To Custom

Scale
Format - sets the format to render out in Nuke, such as 1920x1080 HD 1080.
Scale - sets the proportion by which to scale the output format.
Resize - sets the method by which you want to preserve or override the original
aspect ratio:
• width - scales the original until its width matches the format’s width. Height is
then scaled in such a manner as to preserve the original aspect ratio.
• height - scales the original until its height matches the format’s height. Width is
then scaled in such a manner as to preserve the original aspect ratio.
• fit - scales the original until its smallest side matches the format’s smallest side.
The original’s longer side is then scaled in such a manner as to preserve original
aspect ratio.
• fill - scales the original until its longest side matches the format’s longest side.
The input’s shorter side is then scaled in such a manner as to preserve original
aspect ratio.
• distort - scales the original until all its sides match the lengths specified by the
format. This option does not preserve the original aspect ratio, so distortion may
occur.
Center - when enabled, transform image to the center of the output. When
disabled, the image is aligned with the bottom-left corner of the output.
Filter - sets the filtering algorithm used to transform pixels. See Choosing a
Filtering Algorithm for more information on filters in Nuke.
8. Complete the general controls common to all file types:

• Select the Retime Method to apply, if applicable.
• Soft Effects added to shots in your export are included in the resulting Nuke script by default. If
you don't need the soft effects, disable Include Effects to omit them from the script. See Soft
Effects for more information.
• Include Audio - when enabled, any audio tracks are exported alongside the video.
• Delete .WAV File - when disabled, delete the .wav file used to create the export. This control
only applies to formats that support audio, such as .mov files.
9. Enable the Burn-in Gizmo to burn-in text using a Nuke gizmo. Click Edit to define the information
applied during burn-in. See Adding Burn-in Text to Exports for more information.
USER GUIDE
1466
Transcoding | Tracks and Range Settings
10. Specify any Additional Nodes required during export by clicking Edit. See Adding Additional
Nodes During Export for more information.
11. Check Keep Nuke Script if you require the .nk files after the transcode operation.
Note: The following controls may improve render times for certain exports:
• For .dpx exports, you can enable Read All Lines, which can speed up transcode times for
I/O heavy scripts.
• For machines with multiple CPU sockets using a Single Render Process, you may find that
limiting the render with Use Single Socket may improve render times.
Tracks and Range Settings

the numeric field.
3. If you set in and out point on the sequence, enable In/Out Points to export only the selected
frames.
4. Set how clip Start Frames are derived using the dropdown menu:
5. Set how Nuke Studio should render your export using the Render with dropdown. The following
options are available:
Frame Server.
USER GUIDE
1467
Transcoding | Tracks and Range Settings
6. Click Export.
expected to take.
Once the export is complete, the file structure specified in the shot template is created containing
the transcoded files.
Tip: Click the magnifying glass icon to reveal the exported file in a browser window.
USER GUIDE
1468
Transcoding a Sequence as Shots | Tracks and Range Settings
Transcoding a Sequence as Shots

1. Select the required sequence in the bin view and navigate to File > Export...
2. Select Process as Shots and use the default, Transcode Shots DPX, or build a shot template
using the Path and Contents fields and the folder and +/- buttons.
The default:
Creates a folder for each shot, containing a clip with the {shot} name and the required file
padding (####) and extension {ext}.
hierarchy.
4. In the Content tab, complete the File Type specific and general controls common to all file types
as described in Transcoding a Sequence.
5. Click the Tracks and Handles tab, select the Tracks For This Export by enabling or disabling the
tracks in the list. Nuke Studio exports all tracks by default.
6. Enable or disable tags using the Filter by Tag panel. Click the checkbox to cycle through the
7. Set the Range and Handles, as required:
8. Check Apply Retimes to export any retimes present on the timeline.
Note: When Apply Retimes is disabled, which is the default state for Create Comp, any
TimeWarp soft effects are not included in the resulting Nuke script. When the new shot is
USER GUIDE
1469
Transcoding a Sequence as Shots | Tracks and Range Settings
created through Create Comp or Build Track from Export Tag, TimeWarp soft effects are
copied from the original shot to the new one.
the numeric field.
Note: SeeUsing Versions for more information on how versioning works in Nuke Studio.
Frame Server.
12. Click Export.
expected to take.
Tip: Click the magnifying glass icon to reveal the file structure in a browser window.
USER GUIDE
1470
Transcoding Timeline Selections | Tracks and Range Settings
Transcoding Timeline Selections

Transcoding an entire timeline can be time consuming, or even unnecessary, if all you’re looking for
is a new version of a selection of shots.
To transcode a selection of clips from a timeline:

1. Select the required shots on the timeline.
2. Right-click a highlighted item and select Export...

3. Refer to Transcoding a Sequence to complete the export.
USER GUIDE
1471
Transcoding from the Bin View | Tracks and Range Settings
Transcoding from the Bin View

To transcode directly from the bin view:
1. Select the bin(s) to export from the bin view.
2. Right-click a highlighted bin and select Export...

3. Select Process as Clips and modify the shot template, if required.
4. Follow the steps under Transcoding a Sequence to complete the export.
USER GUIDE
1472
Ad Hoc Exports | Tracks and Range Settings
Ad Hoc Exports
This section covers exports that you might not perform on a per project basis, such the EDL or XML
Exporters and Copy Exporter. Exporters are available for sequences, shots, and clips as described in
the following table.
Exporter Source
Sequences Shots Source Clips
EDL Exporter
XML Exporter
Audio Exporter
Copy Exporter
SymLink Exporter
USER GUIDE
1473
Exporting EDLs and XMLs | Tracks and Range Settings
Exporting EDLs and XMLs

Nuke Studio supports export to EDL and XML using very similar methods, the main difference being
that EDL doesn’t support multiple video tracks in a single file whereas XML does.
Note: Nuke Studio can read AAF files, but not write them out.
To export to EDL or XML:

2. Select Process as Sequence from the Export dropdown.
3. Select the CMX 3600 EDL or Final Cut Pro XML preset, or duplicate one and create your own
preset.
Note: EDLs only support one video track per file. If you have more than one track, include
the {track} token in the shot template to write out an EDL for each track preset.
For example, {filename}_{track}.{ext} might produce a separate EDL for each track on your
timeline called myTimeline_Video1.edl, myTimeline_Video2.edl, and so on.
hierarchy.
5. If you're exporting to EDL, set the additional EDL Exporter controls in the Content tab, if required:
• Reel Name - define the reel name written into the EDL, independent of the clip's reel name.
Enter text or standard shot-level tokens in this field. See Using the Shot Template for more
information.
If the field is left blank, the reel name from the clip is used or the name of the shot, if no reel
name exists.
• Truncate Reel Name - restricts the Reel name to eight characters.
• Use Absolute Path - adds the full file path for each clip to the EDL comments field.
• From Clip Name - define the text appended to “from” comment fields in EDLs, such as *FROM
CLIP NAME. Text and/or standard shot-level tokens are valid in this field: {shot}, {clip}, {track},
{sequence}, {event}, {fps}, and the default {filename}.
OR
USER GUIDE
1474
Exporting EDLs and XMLs | Tracks and Range Settings
If you're exporting to XML, you can enable Include Markers to convert any frame tags present
in the sequence to markers in Final Cut Pro or Premiere. See Tagging Using the Viewer for more
information on adding tags to frames.
6. Click the Tracks and Range tab and select the Tracks For This Export by enabling or disabling
the tracks in the list. Nuke Studio exports all tracks by default.
7. If you set in and out point on the sequence, enable In/Out Points to export only the selected
frames.
the numeric field.
Frame Server.
dependent on platform, as listed in Loading Gizmos, NDK Plug-ins, and Python and Tcl Scripts. If
no scripts exist, the dropdown only contains the default render processes.
11. Click Export.
expected to take.
Once the export is complete, the file structure specified in the shot template is created.
USER GUIDE
1475
Using the Audio Exporter | Exporting Audio from Sequences
Using the Audio Exporter

The Audio Exporter allows you write audio to separate .wav files. You can extract audio from whole
sequences, shots, and source clips.
Exporting Audio from Sequences

2. Select Process as Sequence from the Export dropdown.
3. Select the Log10 Cineon DPX preset, duplicate it, and give it a name.
hierarchy.
5. Build a custom shot template using Path tokens, the Contents field, and the folder and +/-
buttons.
An example shot template is shown below:
the numeric field.
8. Set the Range controls as required:
USER GUIDE
1476
Using the Audio Exporter | Exporting Audio from Shots
Frame Server.
10. Click Export.
expected to take.
Exporting Audio from Shots

1. Select the required shots in the timeline and navigate to File > Export...
2. Select Process as Shots from the Export list.
3. Select the Transcode Shots DPX preset, duplicate it, and give it a name.
hierarchy.
buttons.
USER GUIDE
1477
Using the Audio Exporter | Exporting Audio from Shots
the numeric field.
Note: SeeUsing Versions for more information on how versioning works in Nuke Studio.
8. Enable or disable tags using the Filter by tag panel. Click the checkbox to cycle through the
9. If you're exporting a sequence, set the Range controls as required:
10. If you're exporting shots, set the Handles controls as required:
USER GUIDE
1478
Using the Audio Exporter | Exporting Audio from Source Clips
Frame Server.
13. Click Export.
expected to take.
Exporting Audio from Source Clips

1. Select the required source clips and navigate to File > Export...
2. Process as Clips is selected automatically from the Export list.
3. Select the Transcode Clips DPX preset, duplicate it, and give it a name.
hierarchy.
buttons.
USER GUIDE
1479
Using the Audio Exporter | Exporting Audio from Source Clips
the numeric field.
Frame Server.
9. Click Export.
expected to take.
USER GUIDE
1480
Using the Copy Exporter | Exporting Audio from Source Clips
Using the Copy Exporter

Copying media from various locations is very time consuming and can waste disk space. The Copy
Exporter allows you to consolidate sequences containing only your project media in a named file
structure using the shot template.
To copy media to a named location:

hierarchy.
buttons.
the numeric field.
• Source - use the source start frame.
USER GUIDE
1481
Using the Copy Exporter | Exporting Audio from Source Clips
Frame Server.
9. Click Export.
expected to take.
USER GUIDE
1482
Using the SymLink Generator | Exporting Audio from Source Clips
Using the SymLink Generator

The SymLink Generator allows you to create symbolic links to your project media in a named file
structure using the shot template.
Note: Windows only: Symbolic links are only supported by Windows Vista, or later. if you're
linking across file systems, the remote file servers must also be running Windows Vista, or
later. Additionally, you may need administrator privileges and a local NTFS drive to create
symbolic links.
To create symbolic links to a named location:

hierarchy.
buttons.
the numeric field.
• Source - use the source start frame.
USER GUIDE
1483
Using the SymLink Generator | Exporting Audio from Source Clips
Frame Server.
9. Click Export.
expected to take.
USER GUIDE
1484
Advanced Compositing with
NukeX and Nuke Studio
These pages explain in detail key feature of NukeX and Nuke Studio. For more information on the
differences between the various applications, see Nuke Products. These are the topics covered:
• VectorGenerator, Kronos, and MotionBlur - VectorGenerator, Kronos, and MotionBlur use

Foundry’s advanced motion estimation technology to produce images containing motion vector
fields, slow down or speed up footage, and add motion blur. For more information, see Retiming
and Motion Blur.
• LensDistortion - The LensDistortion node gives you multiple ways to analyze image sequences and
lens grids, resulting in a lens model and the ability to warp and un-warp in order to compensate for
lens distortion. For more information, see Working with Lens Distortion.
• PlanarTracker - The PlanarTracker is a powerful tool for tracking surfaces that lie on a plane in
your source footage. You can use your tracking results to replace the tracked plane with another
image for instance. For more information, see Tracking with PlanarTracker.
• CameraTracker - With the fully integrated 3D CameraTracker node, you can do you own camera
solves and create reference geometry and cards positioned at tracked points in the 3D scene. For
more information, see Camera Tracking.
• MatchGrade - The MatchGrade node allows you to automatically calculate a grade to match the
colors in the Source input to the colors in the Target input. For more information, see Using
MatchGrade.
• Smart Vector Toolset - The Smart Vector Toolset allows you to work on one frame in a sequence
and then use motion vector information to accurately propagate work throughout the rest of the
sequence. For more information, see Using the Smart Vector Toolset.
• DepthGenerator - The DepthGenerator node provides a method to produce a per-frame Z-depth
map from the input 2D footage. It additionally requires a camera solve which can be obtained using
the CameraTracker node. For more information, see Generating Depth Maps.
• PointCloudGenerator - You can create dense point clouds from your footage using the
PointCloudGenerator and CameraTracker. For more information, see Creating Dense Point Clouds.
• PoissonMesh - With the PoissonMesh node, you can use a dense point cloud to create a 3D mesh
from your 2D footage. For more information, see Using the PoissonMesh Node.
• ModelBuilder - ModelBuilder provides an easy way to create 3D models for a 2D shot, given a
tracked camera. You can build a model by creating shapes and then editing them, and align models
over your 2D footage by dragging vertices to their corresponding 2D location. For more information,
see Using ModelBuilder.
USER GUIDE
1485
Advanced Compositing with NukeX and Nuke Studio |
• Particles - The Particle node set is a solution for creating particles in a 3D environment. You can use
the Particle nodes for emitting, manipulating and displaying limitless types of particles in your 3D
scene. For more information, see Creating 3D Particles.
• PrmanRender - PrmanRender is a render node that works together with Pixar’s PhotoRealistic
RenderMan® Pro Server software to give you an even better quality render result. PrmanRender is
an alternative to the ScanlineRender node for rendering 3D scenes, and it gives you control over
features such as shadows, reflections, refractions and depth-of-field. For more information, see
PrmanRender .
• FurnaceCore - This plug-in bundle consists of Foundry’s best Furnace tools, including regraining,
rig-removal, and more. For more information, see Using F_DeFlicker2, Using F_ReGrain, Using F_
WireRemoval Using F_Align, Using F_RigRemoval, and Using F_Steadiness.
Note: The following FurnaceCore nodes have been replaced by other nodes and can no
longer be found in the FurnaceCore menu:
F_DeGrain and F_Denoise were replaced by Denoise (Filter > Denoise) in Nuke 6.3.
F_Kronos was replaced by Kronos (Time > Kronos) in Nuke 7.0.
F_MotionBlur was replaced by MotionBlur (Filter > MotionBlur) in Nuke 7.0.
F_VectorGenerator was replaced by VectorGenerator (Time > VectorGenerator) in Nuke 7.0.
• BlinkScript - The BlinkScript node runs Foundry's Blink framework enabling you to write your code
once and run it on any supported device. For more information, see Using the BlinkScript Node on
page 1817.
USER GUIDE
1486
Retiming and Motion Blur
This chapter looks at creating motion vectors using VectorGenerator, retiming sequences using
Kronos, and adding motion blur using MotionBlur.
Before MotionBlur. After MotionBlur.
USER GUIDE
1487
VectorGenerator |
VectorGenerator
VectorGenerator allows you to produce images containing motion vector fields. A vector field for an
image in a sequence has the same dimensions as the image, but contains an (x,y) offset per pixel.
These offsets show how to warp a neighboring image onto the current image. Clearly, as most of the
images in a sequence have two neighbors, each can have two vector fields:
1. The backward vector field: the x and y offsets per pixel that, when applied to the previous frame
in the sequence, allow you to reconstruct an approximation to the current frame.
2. The forward vector field: the x and y offsets needed to transform the next frame into an
approximation to the current one.
The output from VectorGenerator is stored in the vector channels. The images below show the
different vector images for a clip.
The source sequence. Forward and backward motion vectors.
Forward motion vectors. Backward motion vectors.
When viewing 'forward' or 'backward' motion vectors, the Viewer represents the x values as amounts
of red and y values as amounts of green. Motion vectors can be positive or negative, where zero
represents no motion.
In general, once you have generated a sequence of motion vector fields that describe the motion in a
particular clip well, they are suitable for use in any nodes which can take vector inputs. These include
Kronos and MotionBlur. If you are going to be using more than one of these effects in your project, it
might be worth generating the vector fields beforehand with VectorGenerator, so that they can be
reused.
USER GUIDE
1488
Quick Start |
Quick Start
1. Add VectorGenerator to your node tree. See Connecting VectorGenerator below.
2. View and refine the results. See Viewing and Refining the Results.
take effect.
USER GUIDE
1489
Connecting VectorGenerator |
Connecting VectorGenerator
To connect VectorGenerator:
1. Select Time > VectorGenerator to insert a VectorGenerator node after the sequence from which
you want to generate motion vectors.
2. Attach a Viewer to the output of the VectorGenerator node.
3. If your sequence is composed of a foreground object moving over a background, the motion
estimation is likely to get confused at the edge between the two. To fix this, add a matte of the
foreground region to the Matte input. Then, use Matte Channel in the VectorGenerator
properties to select which component of the matte to use.
This helps the motion estimation algorithm inside VectorGenerator understand what is
foreground and background in the image, so that the dragging of pixels between overlapping
objects can be reduced.
White areas of the matte are considered to be foreground, and black areas background. Gray
areas are used to attenuate between foreground and background
4. If you supplied a foreground matte in the previous step, you can set Output in the
VectorGenerator properties to:
• Foreground - to output the vectors for the foreground regions.
• Background - to output the vectors in the background regions.
5. Proceed to Viewing and Refining the Results.
USER GUIDE
1490
Viewing and Refining the Results |
Viewing and Refining the Results

To view and refine the results:
1. Select the required vector calculation type from the Method dropdown:
• Local - uses local block matching to estimate motion vectors. This method is faster to process,
but can lead to artifacts in the output.
• Regularized - uses semi-global motion estimation to produce more consistent vectors between
regions.
Note: Scripts loaded from previous versions of Nuke default to Local motion estimation for
backward compatibility. Adding a new VectorGenerator node to the Node Graph defaults the
Method to Regularized motion estimation.
2. To view the generated motion vector fields, click the channels dropdown menu on top of the
Viewer and select:
• motion - to view both forward and backward motion vectors.
• forward - to view the forward motion vectors.
• backward - to view the backward motion vectors.
3. If the calculated motion vector fields do not produce the results you’re after when used with other
nodes (such as Kronos and MotionBlur), try adjusting Vector Detail in the VectorGenerator
properties.
This determines the resolution of the vector field. The larger vector detail is, the greater the
processing time, but the more detailed the vectors should be. A value of 1.0 generates a vector at
each pixel. A value of 0.5 generates a vector at every other pixel. For some sequences, a high
vector detail near 1.0 generates too much unwanted local motion detail, and often a low value is
more appropriate.
4. If you're using the Regularized vector calculation method, adjust the Strength control to
determine the strength of pixel matching between frames. Higher values allow you to accurately
match similar pixels in one image to another, concentrating on detail matching even if the
resulting motion field is jagged. Lower values may miss local detail, but are less likely to provide
you with the odd spurious vector, producing smoother results.
Note: The default value should work well for most sequences.
5. If you're using the Local vector calculation method, adjust the Smoothness control to improve
your results. A high smoothness can miss lots of local detail, but is less likely to provide you with
USER GUIDE
1491
Viewing and Refining the Results |
the odd spurious vector, whereas a low smoothness concentrates on detail matching, even if the
resulting field is jagged.
Note: The default value should work well for most sequences.
6. If there are variations in luminance and overall flickering in your Source sequence, enable Flicker
Compensation to avoid problems with your output.
bodies of water within a layer that reflects light in unpredictable ways.
Note: Enabling Flicker Compensation increases rendering time.
7. By default, VectorGenerator analyzes motion based on the brightness of the image. Specifically
this is the mean of the red, green, and blue channels. However, you can bias this using the
Weight Red, Weight Green, and Weight Blue controls under Tolerances. For example, if you set
the Weight Red and Weight Green parameters to zero, VectorGenerator only looks for motion in
the blue channel.
8. Once you’re happy with the results, we recommend that you insert a Write node after
VectorGenerator to render the original images and the vector channels as an .exr file. This format
allows for the storage of an image with multiple layers embedded in it. Later, whenever you use
the same image sequence, the motion vector fields are loaded into Nuke together with the
sequence.
USER GUIDE
1492
Kronos |
Kronos
Kronos is NukeX’s retimer, designed to slow down or speed up footage. It works by calculating the
motion in the sequence in order to generate motion vectors. These motion vectors describe how each
pixel moves from frame to frame. With accurate motion vectors, it is possible to generate an output
image at any point in time throughout the sequence by interpolating along the direction of the
motion.
Simple mix of two frames to Kronos vector interpolation

achieve an in-between frame. of the same two frames.
By default, Kronos is set to perform a half-speed slow down. This is achieved by generating a new
frame at position 0.25 and 0.75 between the original frames at 0 and 1. Frames are created at a
quarter and three quarters instead of zero (an original frame) and a half so as not to include any
original frames in the re-timed sequence. This avoids the pulsing that would otherwise be seen on
every other frame on a half-speed slow down, and can introduce motion blur.
Kronos only interpolates between input frames, as it cannot extrapolate images before the first frame
or after the last frame. A retime with a constant speed s "stretches" the output time by a factor 1/s,
and generates the required images to fill in all intervals between the input frames.
In the following table, | denotes where we need to have images to fill the video sequence. In this
example, we assume that the input sequence has 5 frames, denoted by X, and we want to retime it
with constant speed 0.5 (the default setting). This operation corresponds to stretching the time
between the input frames by a factor of 2, which leaves a number of gaps that Kronos fills by
generating the images denoted by O.
Samples | | | | | | | | |
Input X X X X X
Input Stretched X X X X X
Output X O X O X O X O X
USER GUIDE
1493
Kronos |
The output sequence is number_of_intervals / speed + 1 frames long. For example, if speed is 0.5
and the input frame is in the range [1,10], the output sequence length will be (10 - 1) / 0.5 + 1 = 19
frames long.
Example Input Speed Calculation

Let's say that the input sequence is [t_start, t_end]. Retiming the input sequence in the Timing >
Input Speed mode varies the speed at which time flows in the input sequence, so that a point in time
tau, in the input time frame, maps to the output time t_o as follows:
The start frame t_start is the same for both the input and output sequence. If the speed parameter s
is not animated, and therefore s is a constant, we get:
So, if we want to know where the last frame of the sequence maps to, we use:
Let's look at an example. If the input sequence is in the range [1, 10] and speed is 0.5, the output
sequence has a range [1, 19], as follows:
t_o_start = 1
t_o_end = 1 + 2 * (10 - 1) = 19
USER GUIDE
1494
Quick Start |
Quick Start
1. Create a node tree with Kronos. See Connecting Kronos.
2. Retime your footage. See Retiming a Sequence.
3. If you’re not happy with the results, adjust the vector generation parameters. See Refining the
Results.
4. If necessary, add motion blur to the retimed footage. See Adding Motion Blur.
take effect.
USER GUIDE
1495
Connecting Kronos |
Connecting Kronos
To connect Kronos:
1. Select Time > Kronos to insert a Kronos node after the sequence you want to retime.
The Input Range is set automatically by the source clip when you first create the node. After that,
it is only updated if you click Reset.
2. If you intend to use Output Speed or Frame timing, attach a Viewer to the output of the Kronos
node,
OR
If you intend to use Input Speed timing, attach a Viewer before the Kronos node.
3. Enable Use GPU if available to render output on the Local GPU specified, if available, rather than
the CPU.
For more information on the minimum requirements, please see Windows, Mac OS X and macOS,
or Linux or refer to the Nuke Release Notes available in Help > Release Notes.
4. Select the channels you want to apply the retime to using the channels dropdown. Kronos
retimes all channels by default.
foreground region to the Matte input. Then, use Matte Channel in the Kronos properties to
select which component of the matte to use.
This forces the motion of the foreground to be calculated separately to the motion of the
background and so should produce fewer artifacts in the retimed sequence.
6. If the motion in your input sequence has been estimated before (for example, using
VectorGenerator or third-party software), you can supply the motion vectors to the Background
Vectors (BgVecs) and Foreground Vectors (FgVecs) inputs to save processing time.
Note: The BgVecs input appears as an arrowhead on the side of the node.
If you have separate vectors for the background and foreground, you should connect them to the
appropriate inputs and supply the matte that was used to generate them to the Matte input. If
you have a single set of vectors, you should connect it to the FgVecs input.
7. You can set Output in the Kronos controls to:
• Result - displays both the foreground and background retimed footage.
• Matte - displays the retimed Matte input.
• Foreground - displays the retimed foreground. The background regions outside the matte input
may show garbage.
USER GUIDE
1496
Connecting Kronos |
• Background - displays the retimed background. The foreground regions inside the matte input
may show garbage.
8. If your Source sequence is very noisy and interfering with the motion estimation, you can supply
a smoothed version of the sequence in the MoSrc input. When a MoSrc input is connected,
motion vectors are calculated from that sequence and applied to the Source sequence.
9. Proceed to Retiming a Sequence.
USER GUIDE
1497
Retiming a Sequence | Retiming a Sequence Using Output Speed
Retiming a Sequence
Kronos allows you to retime a sequence by speed or by frame. Speed retimes describe the retime by
a percentage of full speed, either against the output or input frames. For example, the default setting
Output Speed = 0.5 is equal to a 50% retime on the output frames. Frame retimes describe the
retiming in terms of ‘at frame 100 in the output clip, I want to see frame 50 of the source clip‘.
Note: Input Range in the Kronos properties defines which frames are used for the retime.
When you first create the node, the range is automatically set to the frame range of the
Source clip. If you change the Source clip later, you need to click the Reset button to make
sure Input Range matches the current input.
Retiming a Sequence Using Output Speed

To apply a constant retime:
1. Ensure that the Viewer is connected to the Kronos node and Timing is set to Output Speed.
2. Select the required interpolation Method using the dropdown:
• Blend - a mix between two frames is used for the in-between frame. This is quick to render and
is useful when tweaking the timing on a curve before setting the Method to Motion.
• Motion - vector interpolation is used to calculate the in-between frame.
3. Enter a value for the Output Speed control. Values below 1 slow down the clip and values above 1
speed up the clip. The default value is 0.5, creating a half-speed slow down. Quarter-speed would
be 0.25.
4. Play through the sequence to view the retime.
5. If you're not happy with the results of the retime, try Refining the Results.
To use varying retimes:
Tip: If you want to retime at a certain event in the clip, you may find that using Input Speed
is easier to predict where keyframes should be placed.
1. Ensure that the Viewer is connected to the Kronos node and Timing is set to Output Speed.
USER GUIDE
1498
Retiming a Sequence | Retiming a Sequence Using Input Speed

3. Move the playhead to the frame you want to start retiming and enter the required Output Speed.
4. Click the Animation button and select Set Key.
5. Move the playhead to the frame you want to change the retime value and enter the required
Output Speed.
A new keyframe is added automatically.
6. Once you've added all the required keyframes, click the Curve Editor tab to view your keyframes
in context of the sequence.
Retiming a Sequence Using Input Speed

1. Ensure that the Viewer is connected to the tree before the Kronos node and Timing is set to Input
Speed.
3. Enter a value for the Input Speed control. Values below 1 slow down the clip and values above 1
speed up the clip. The default value is 0.5, creating a half-speed slow down. Quarter-speed would
be 0.25.
4. Connect the Viewer to the Kronos node, and then play through the sequence to view the retime.

1. Ensure that the Viewer is connected to the tree before the Kronos node and Timing is set to Input
Speed.
USER GUIDE
1499
Retiming a Sequence | Retiming a Sequence Using Frame
3. Move the playhead to the frame you want to start retiming and enter the required Input Speed.
4. Click the Animation button and select Set Key.
5. Move the playhead to the frame you want to change the retime value and enter the required
Input Speed.
A new keyframe is added automatically.
6. Once you've added all the required keyframes, connect the Viewer to the Kronos node, and then
click the Curve Editor tab to view your keyframes in context of the output sequence.
Retiming a Sequence Using Frame

1. Ensure that the Viewer is connected to the Kronos node and Timing is set to Frame.
3. Go to a frame in the timeline, and set Frame to the input frame you want to appear at that output
position. For a constant retime, you’ll need two key frames to retime the clip.
For example, to slow down a 50 frame clip by half, you can set Frame to 1 at frame 1, and to 50 at
frame 100. To do a four times slow down, set Frame to 1 at frame 1, and to 25 at frame 100.
4. Play through the sequence to view the retime.

1. Ensure that the Viewer is connected to the Kronos node and Timing is set to Frame.
USER GUIDE
1500
Retiming a Sequence | Retiming a Sequence Using Frame
3. Go to a frame in the timeline, and set Frame to the input frame you want to appear at that output
position. You’ll need atleast two key frames to retime the clip.
For example, to speed up a 50 frame clip and then return to normal speed at the end:
• Set Frame to 1 at frame 1,
• Set Frame to 50 at frame 50.
These keyframes produce the following curve in the Curve Editor.
USER GUIDE
1501
Refining the Results | Retiming a Sequence Using Frame

1. To have the motion vectors displayed in the Viewer, expand the Advanced control and enable
Overlay Vectors. Forward motion vectors are drawn in red, and backward motion vectors in
blue.
Overlay Vectors enabled.
Note: Motion vectors displayed in the Viewer are added to your output if you don’t turn off
the overlay before rendering.
2. To set the spacing between motion vectors displayed on the Viewer, adjust Vector Spacing. The
default value of 20 means every 20th vector is drawn. Note that Vector Spacing only affects the
Viewer overlay rather than the retimed result.
A Vector Spacing value of 4. A Vector Spacing value of 40.

3. Adjust Vector Detail to vary the density of the vector field. The larger vector detail is, the greater
the processing time, but the more detailed the vectors should be. A value of 1 generates a vector
at each pixel. A value of 0.5 generates a vector at every other pixel. For some sequences, a high
vector detail near 1 generates too much unwanted local motion detail, and often a low value is
more appropriate.
USER GUIDE
1502
Areas of unwanted local motion detail. Lower Vector Detail is more appropriate
in this case.
4. Vector fields usually have two important qualities: they should accurately match similar pixels in
one image to another and they should be smooth rather than noisy. Often, it is necessary to trade
one of these qualities off against the other. A high Strength misses lots of local detail, but is less
likely to provide you with the odd spurious vector. A low Strength concentrates on detail
matching, even if the resulting field is jagged. The default value of 1.5 should work well for most
sequences.
Jagging as a result of a low Strength Minimal jagging as a result of a high

value. Strength value.
5. Set the type of resampling applied when retiming:
levels. You can use Bilinear to preview a retime before using one of the other resampling types
to produce your output.
6. If the overall brightness in your Source footage changes between frames, enable Flicker
Compensation. This allows Kronos to take into account variations in luminance and overall
flickering, which could otherwise cause problems with your output.
bodies of water that reflect light in unpredictable ways.
USER GUIDE
1503
Note that using Flicker Compensation increases rendering time.

7. In order to reduce processing time, much of the motion estimation is done on luminance only -
that is, using monochrome images. In most cases this is perfectly acceptable, but the parameters
in the Tolerances group allow you to concentrate on a particular feature in an image by adding
bias to individual colours. You may, for example, wish to increase Weight Red to allow the
algorithm to concentrate on getting the motion of a primarily red object correct, at the cost of the
rest of the objects in a shot.
USER GUIDE
1504
Adding Motion Blur | Retiming a Sequence Using Frame
Adding Motion Blur

To add motion blur:
1. In the Shutter controls, select a suitable Shutter Time, depending on the amount of blur you
wish to add. This sets the equivalent shutter time of the retimed sequence. For example, a shutter
time of 0.5 is equivalent to a 180 degree mechanical shutter, so at 24 frames per second the
exposure time is 1/48th of a second.
The larger the value, the more motion blur is produced.
A low Shutter Time value. A high Shutter Time.

Alternatively, you can enable Automatic Shutter Time to have Kronos automatically calculate the
shutter time throughout the sequence. Note that this only produces motion blur when the
retimed speed is greater than the original speed.
2. Render the sequence to see the motion blurred result.
3. If you can see that the motion blur has been created from a few discrete images, try increasing
Shutter Samples. This results in more in-between images being used to generate the motion blur
and so produces a smoother blur but takes longer to render.
USER GUIDE
1505
MotionBlur | Retiming a Sequence Using Frame
MotionBlur
MotionBlur adds realistic motion blur to a sequence.
Before MotionBlur. After MotionBlur.
It uses the same techniques and technology as the motion blur found in Kronos, but presents the
controls in a less complex, more user friendly way. However, if you need precise control over the
motion vectors used for adding blur, or a large temporal range (that is, a very high shutter time), you
should use Kronos.
USER GUIDE
1506
Quick Start | Retiming a Sequence Using Frame
Quick Start
1. Create a node tree with MotionBlur. See Connecting MotionBlur.
2. Adjust the amount and quality of the motion blur produced. See Adjusting MotionBlur Controls.
take effect.
USER GUIDE
1507
Connecting MotionBlur | Retiming a Sequence Using Frame
Connecting MotionBlur
To connect MotionBlur:
1. Select Filter > MotionBlur to insert a MotionBlur node after the clip you want to add motion blur
to.
foreground region to the Matte input. Then, use Matte Channel in the MotionBlur controls to
select which component of the matte to use.
This forces the motion of the foreground to be calculated separately to the motion of the
background and so should produce fewer artifacts in the motion blur.
3. If the motion in your input sequence has been estimated before (for example, using
VectorGenerator or third-party software), you can supply the motion vectors to the Background
Vectors (BgVecs) and Foreground Vectors (FgVecs) inputs to save processing time.
If you have separate vectors for the background and foreground, you should connect them to the
appropriate inputs and supply the matte that was used to generate them to the Matte input. If
you have a single set of vectors, you should connect it to the FgVecs input.
4. Attach a Viewer to the output of the MotionBlur node.
5. Proceed to Adjusting MotionBlur Controls below.
USER GUIDE
1508
Adjusting MotionBlur Controls | Retiming a Sequence Using Frame
Adjusting MotionBlur Controls

To adjust MotionBlur controls:
1. Select a suitable Shutter Time, depending on the amount of blur you wish to add. This sets the
equivalent shutter time of the retimed sequence. For example, a shutter time of 0.5 is equivalent
to a 180 degree mechanical shutter, so at 24 frames per second the exposure time is 1/48th of a
second.
The larger the value, the more motion blur is produced.
A low Shutter Time. A high Shutter Time.

2. Set the method of calculating motion estimation vectors:
• Local - uses local block matching to estimate motion vectors. This method is faster to process,
but can lead to artifacts in the output.
• Regularized - uses semi-global motion estimation to produce more consistent vectors between
regions.
Note: Scripts loaded from previous versions of Nuke default to Local motion estimation for
backward compatibility. Adding a new MotionBlur node to the Node Graph defaults the
Method to Regularized motion estimation.
3. Use Vector Detail to alter the density of the calculated motion vector field. A value of 1 generates
a vector at each pixel, whereas a value of 0.5 generates a vector at every other pixel.
Bear in mind that higher Vector Detail values require longer to process, but can pick up finer
movement and enhance your results in some cases.
4. Set the type of re-sampling applied when retiming:
levels. You can use Bilinear to preview a motion blur before using one of the other re-sampling
types to produce your output.
5. Render the sequence to see the motion blurred result.
USER GUIDE
1509
Adjusting MotionBlur Controls | Retiming a Sequence Using Frame
6. If you can see that the motion blur has been created from a few discrete images, try increasing
Shutter Samples. This results in more in-between images being used to generate the motion blur
and so results in a smoother blur.
Shutter Samples = 2 Shutter Samples = 4
Shutter Samples = 8 Shutter Samples = 30
USER GUIDE
1510
Working with Lens Distortion
Lens distortion can make composting work more difficult because the characteristics of the lens can
cause areas of the footage to warp, making correct placement of assets problematic. One way around
this problem is to undistort your plate before starting work on a comp and then redistort and merge
only the VFX work to produce the final comp.
The original plate The undistorted plate and comp
The comp after redistortion
It's a good idea to shoot grids with the same lenses you use to create your footage as this make
estimating distortion easier. If you have the lens properties, such as Focal Length and Sensor Size,
and a grid shot with the same lens you'll find this process much less painful.
Nuke's LensDistortion node allows you to undistort or distort an image according to several radial
distortion models that ship with Nuke or a custom model you define yourself. You can calculate the
warp for use on the input image or output the warp to an STMap for use elsewhere in the script.
Quick Start
USER GUIDE
1511
Working with Lens Distortion |
1. Read in an input sequence, connect it to a LensDistortion node (Transform > LensDistortion),

and connect the output to a Viewer.
2. If you have a grid shot using the same lens as your sequence, use Grid Detection to calculate the
distortion, and then apply the warp to your sequence. See Estimating Lens Distortion Using a Grid
and Removing Lens Distortion from an Image for more information.
3. If you do not have a grid, or Grid Detection did not work, draw features and lines manually to
estimate distortion and undistort the sequence. See Estimating and Removing Lens Distortion
Using Lines for more information.
Note: If you want, you can calculate the lens distortion on one image and apply that
distortion to another image with the help of an STMap node. For more information, see
Working with STMaps.
4. Add your VFX work to the undistorted sequence.

5. Redistort the VFX work and merge it over the original plate. See Applying Lens Distortion to an
Image for more information.
USER GUIDE
1512
Estimating Lens Distortion Using a Grid |
Estimating Lens Distortion Using

a Grid
Grid analysis estimates the distortion from a checkerboard or thin line grid. As a general rule, if you
have a grid you can use to calculate your lens distortion, you should use grid analysis. Grids are
constructed from Features and Links, which can be calculated separately to allow for manual
corrections or all at once automatically.
The original distorted shot The distorted grid associated with the shot
To estimate distortion using a grid:

1. Read in your grid and connect a LensDistortion node, followed by a Viewer.
2. Set the lens Type and Projection using the dropdowns in the LensDistortion Properties panel. In
this example, the lens is Spherical and Rectilinear, but there are a number of other presets
included.
3. If you're using a fisheye lens, you need to 'defish' the lens. You can do this by selecting the correct
Projection setting, such as Fisheye Equisolid, entering the Focal Length and Sensor Size.
Tip: If you don't know the Focal Length, switch the Output Mode to Undistort and adjust
the Focal Length until the curved lines in the image appear approximately straight. Don't
forget to switch back to Mode > STMap before continuing.
4. Set the Distortion model preset to use for the estimation. Nuke ships with several models for use
with CaraVR and 3DEqualizer, as well as a NukeX Classic model.
The model you choose sets appropriate Distortion Model controls and populates the read only
Distortion Equation fields showing the math involved in the estimation.
USER GUIDE
1513
5. The LensDistortion node adds a keyframe on the current frame automatically, but you can add
more using the or Key All buttons.
Adding keyframes can produce a better result for long sequences, but it takes longer to calculate
the result.
6. Click Detect to start the grid calibration. By default, Nuke looks for Features on the grid and then
creates Links between those features to create the distortion grid.
Tip: For difficult shots, you can make adjustments between Features and Links detection to
improve the results. See Adjusting Grid Detection Parameters for more information.
A grid overlay is displayed when the detection step is complete.
Tip: If the grid does not cover most of the image, try increasing the Number of Features
value on the Grid Detection tab and clicking Detect again.
7. Click Solve to estimate the distortion using the grid.

Nuke estimates the distortion and displays a new overlay and overall Solve Error.
USER GUIDE
1514
Green lines represent links that fall within the Threshold value and red lines those that fall
outside the Threshold. You can hover over links to display their solve error.
8. Clicking Solve again can improve the result in some cases. You can also refine the detection and
linking before solving again using the controls on the Grid Detection tab. See Adjusting Grid
Detection Parameters for more information.
9. When you're happy with the solve, switch the Output Mode to Undistort.
Note: You can also output the distortion as an STMap for use in other images. STMaps
contain the pre-computed warp data in the motion channel, allowing you apply the warp
quickly and easily. See Working with STMaps for more information.
Nuke takes the estimated distortion and uses the result to 'straighten' the feature links.
USER GUIDE
1515
You may notice that the grid expands beyond the bounding box of the footage, which can mean
you're losing image data.
10. Set the Format control to Reformat and select a larger format to include the grid. For example, if
your original grid is UHD_4K, you might reformat to 8K_LatLong.
Reformatting the grid deforms it into a characteristic bow tie shape. Any areas of the image where
there is no data are overscanned, meaning the last pixel available is duplicated to the edge of the
format. The areas at the top and bottom of the image in the center show overscan.
11. If you don't want to reformat the image, or only a small amount of the image is missing, you can
bring more image data into the format bounds using the Adjust Format controls.
12. Enable Override and then use the Add Pixels and Adjust Aspect controls to enlarge the format
as required.
USER GUIDE
1516
The new format is called lens_distortion_adjusted by default, but you can enter any name you
choose in the adjustedFormatName control. The new name is then added to the Project
Settings > Format dropdown after adjustment to avoid changing Nuke's default formats.
Tip: You can also preserve overscan data by increasing the bounding box. The Add Pixels
control allows you to add an equal number of pixels on all sides of the image or you can edit
the bounding box manually by enabling Override and using the outputBBox controls.
13. Proceed to Removing Lens Distortion from an Image or Adjusting Grid Detection Parameters to
fine tune your grid.
USER GUIDE
1517
Adjusting Grid Detection Parameters | Feature Detection
Adjusting Grid Detection

Parameters
The default settings on the LensDistortion node's Grid Detection tab are suitable for a wide range of
input images, but in some cases, you'll need to make some adjustments in order to achieve the
required coverage of Features and viable Links. You'll need to click Detect again after making
changes to the Grid Detection controls to update the results.
Tip: For particularly hard solves, you can split the Features and Links detection into two
steps, rather than using Grids, which performs both steps automatically.
Feature Detection
The Features in a solve form the building blocks for accurate Links, so it's important to get the
maximum coverage possible, while still maintaining quality.
Increasing the maximum Number of Features can improve coverage relatively easily. You can use
the Detection Threshold control to reject bad features automatically. If you enter a low detection
threshold value, features are detected evenly on all parts of the image, even if they would otherwise
have been rejected.
Low Number of Features High Number of Features
The Patch Size control determines how much bias is placed on detecting features on saddle points.
High values force features towards saddle points, which might not be desirable.
USER GUIDE
1518
Adjusting Grid Detection Parameters | Link Detection
Low Patch Size High Patch Size
The Feature Separation control sets the distribution of features in relation to each other. High
values spread features at even distances over the image. It is important that the features do not
cluster together, if this is the case, try increasing this value.
Increasing the separation too far can reduce the number of features detected dramatically, so use
caution.
Low Feature Separation High Feature Separation
Link Detection
The Links in a solve depend on solid Features detection, so it's important to make sure your features
coverage is good before linking. Links provide the curvature data from the lens so that the solve can
be as accurate as possible.
The Angle Threshold control sets how much offset tolerance is allowed between potential features
before they can be linked. Try increasing the Angle Threshold if there are missing links between
features, but higher values can introduce links between features that may be incorrect.
The image on the right shows some potentially poor links on the left-hand side.
USER GUIDE
1519
Adjusting Grid Detection Parameters | Link Detection
Low Angle Threshold High Angle Threshold
The Distance Threshold control determines how far apart Links can be before they are merged into
a single, averaged link. Try increasing the Distance Threshold if features are removed incorrectly
after Links Detection.
Low Distance Threshold High Distance Threshold
The Peak Threshold control determines how much directional tolerance is allowed between
potential features before they can be linked. Reducing the Peak Threshold can increase viable links
with low contrast lens reference grids.
Low Peak Threshold High Peak Threshold
Proceed to Removing Lens Distortion from an Image when you're satisfied with your grid.
USER GUIDE
1520
Removing Lens Distortion from an Image | Link Detection
Removing Lens Distortion from

an Image
After estimating the lens distortion from a reference grid, you can use the warp to remove the
distortion from your sequence before performing comp work. The easiest way to do this is to copy
and paste the LensDistortion node that was used to generate the warp and connect it to your
sequence.
Note: You can also use STMaps to remove and apply distortion. See Working with STMaps
1. Read in the sequence from which the reference grid was taken.
2. Copy and paste the LensDistortion node containing the grid analysis (described in Estimating Lens
Distortion Using a Grid), then connect it to the sequence and a Viewer.
The distortion is removed from the sequence in the Viewer.
USER GUIDE
1521
The original plate The undistorted plate

You may notice that the lines expand beyond the bounding box of the footage, which can mean
Reformatting the grid deforms it into a characteristic bow tie shape. Any areas of the image where
there is no data are overscanned, meaning the last pixel available is duplicated to the edge of the
format. The areas at the top and bottom of the image in the center show overscan.
You can now apply your VFX work before redistorting the plate back into its original state.
as required.
USER GUIDE
1522
Note: You can use expressions to control the number of pixels added as a percentage of
the format by right-clicking the Add Pixel control and selecting Add Expression. For
example, input.width*0.05 adds pixels equal to 5% of the input format's width.
The undistorted format The adjusted format

6. You can now track, matchmove, composite, and so on in the undistorted space before
redistorting the image. See Applying Lens Distortion to an Image for more information.
USER GUIDE
1523
Estimating and Removing Lens Distortion Using Lines | Link Detection
Estimating and Removing Lens

Distortion Using Lines
Line analysis estimates the distortion from lines drawn manually along features in the input that are
known to be straight. This can be useful if there is no grid available or if you have a grid for your
sequence but the grid analysis failed, for instance due to bad lighting.
The LensDistortion node has five tools for selecting and drawing features and lines in the Viewer:
Select - a multi-purpose tool used to select, move, and delete features and lines.
Select Feature - used to select, move, and delete features. You cannot affect lines using this
tool.
Select Line - used to select, move, and delete lines. You cannot affect features using this tool.
Add Feature - used to add, select, move, and delete features. You cannot affect lines using this
tool.
Add Line - used to add and delete lines. You can connect existing features or create new lines
using this tool.
To estimate lens distortion using lines, do the following:

1. Examine the source footage for distorted lines that are known to be straight. If you have a grid,
this is simple, but real life examples include walls, railings, roads, and so on.
2. Select the Add Feature tool and start to add points along the distorted lines by clicking in the
Viewer.
Tip: You can also draw lines immediately using the Add Line tool, but adding features first
can be more accurate.
USER GUIDE
1524
3. When the features are complete for a particular line in the image, switch to the Add Line tool.
4. Select the first feature in the Viewer and then click the next feature to add a line between the two.
Tip: When you've finished drawing a line, either select a different tool or press Enter to
finish the current line before starting another.
Tip: You can adjust the position of features and lines using the selection tools in the Viewer.
5. The solve requires at least as many lines as there are Distortion Parameters to calculate the
distortion. In the case of NukeX Classic, two vertical and two horizontal to cover the
Denominator and Centre parameters. Good line drawing practices include:
USER GUIDE
1525
• Drawing lines with three or more features - lines with only two points are ignored.
• Drawing longer lines - they contain more useful information on the curvature resulting from
lens distortion.
• Distributing lines evenly - avoid biasing the solve by covering as much of the image as
possible.
The following example shows a typical set of lines.
The original distorted shot Example lines in the alpha channel

6. Click Solve to calculate the distortion, there's no need to run the Detect step with manual
features and lines.
Green lines represent links that fall within the Threshold value and red lines those that fall
outside the Threshold. You can hover over links to display their solve error.
Tip: Clicking Solve again can improve the result in some cases.
7. When you're happy with the solve, switch the Output Mode to Undistort.
USER GUIDE
1526
Note: You can also output the distortion as an STMap for use in other images. STMaps
contain the pre-computed warp data in the motion channel, allowing you apply the warp
quickly and easily. See Working with STMaps for more information.
Nuke takes the estimated distortion and uses the result to 'straighten' the feature links.
You may notice that the lines expand beyond the bounding box of the footage, which can mean
Reformatting the feature lines deforms it into a characteristic bow tie shape. Any areas of the
image where there is no data are overscanned, meaning the last pixel available is duplicated to
the edge of the format. The area at the bottom of the image in the center shows overscan.
You can now apply your VFX work before redistorting the plate back into its original state.
USER GUIDE
1527
as required.
Note: You can use expressions to control the number of pixels added as a percentage of
the format by right-clicking the Add Pixel control and selecting Add Expression. For
example, input.width*0.05 adds pixels equal to 5% of the input format's width.
The undistorted format The adjusted format

11. You can now track, matchmove, composite, and so on in the undistorted space before
redistorting the image. See Applying Lens Distortion to an Image for more information.
USER GUIDE
1528
Applying Lens Distortion to an Image | Link Detection
Applying Lens Distortion to an

Image
The LensDistortion node's Redistort mode allows you to warp your VFX work to match the original
plate, which avoids the increased processing time the filter hit on the entire image would incur.
Note: You can also use STMaps to remove and apply distortion. See Working with STMaps
1. Copy and paste the LensDistortion node containing the grid analysis (described in Estimating Lens
Distortion Using a Grid), then connect it downstream of your VFX work.
2. Change the Mode control to Redistort to reverse the warp that was applied during the
undistortion of the plate.
3. Merge the redistorted VFX work back over the plate to complete the comp.
The undistorted plate and comp The comp work after redistortion
USER GUIDE
1529
Working with STMaps | Link Detection
Working with STMaps

STMaps allow you to warp an image or sequence according to the pre-calculated distortion from a
LensDistortion node, stored in the motion layer. The motion channels represent the absolute pixel
positions of an image normalized between 0 and 1, which can then be used on another image to
remove or add distortion without the warp estimation step. See Estimating Lens Distortion Using a
Grid and Estimating and Removing Lens Distortion Using Lines for more information on warp
estimation.
To remove distortion using the STMap node, do the following:

1. Read in your sequence and add an STMap node downstream. Make sure the src input is
connected to the sequence.
2. Copy and paste the LensDistortion node containing the grid analysis with the Mode control set to
STMap,
OR
Read in the file containing the motion layer, if you wrote it out separately.
3. Connect the stmap input to the source of the motion layer.
4. Set the STMap UV channels control to the motion layer to remove the distortion.
You can now track, matchmove, composite, and so on in the undistorted space before
redistorting the image.
5. After the comp work is complete, you can warp the VFX work by applying the backward layer in
the UV channels control using another STMap node.
USER GUIDE
1530
Working with STMaps | Link Detection
6. The final step is merging the warped VFX work back over the original image to create the comp.
The output from nodes placed before the second STMap node are distorted by the same warp as
the sequence.
The undistorted plate and comp The comp after redistortion
USER GUIDE
1531
Tracking with PlanarTracker
PlanarTracker is a powerful tool for tracking surfaces that lie on a plane in your source footage.
Planar tracking is often better than tracking individual points (with the Tracker node for instance) as it
takes a great deal more information into account and gives you a more accurate result. A rigid
surface, like a wall or a side of an object, are good planes to track. You can use your tracking results
to replace the tracked plane with another image for instance. You can define the region to track by
creating and animating roto shapes.
USER GUIDE
1532
Tracking a Plane | Drawing a Plane to Track
Tracking a Plane
Before you can track a plane, you need to draw one using the Roto node.
Drawing a Plane to Track

You can use the PlanarTracker to track rigid objects and objects that deform slightly throughout the
track. As the PlanarTracker tries to fit a plane to the object to be tracked, rigid objects obtain better
track results than objects that deform.
For instance, a wall or a flat side of an object are good planes, but you can also get good results
tracking faces or people. It’s also important that the plane you’re tracking has some texture and that
the plane isn’t completely obscured at any point of the tracking. Tracking surfaces without texture and
few features to track is not likely to produce good results.
1. You can do one of the following:
• Insert a PlanarTracker node by either selecting Transform > PlanarTracker, or by pressing tab,
typing PlanarTracker, and pressing Return. This inserts a Roto node that is already in
PlanarTracker mode. You can use this to draw a Bezier shape, which is automatically added as a
track object. The shape's boundary appears in purple, denoting this, and the shape is
automatically added to a layer called PlanarTrackLayer1 in the stroke/shape list.
OR
• Create a Roto or RotoPaint node and use it to draw a Bezier shape around the plane you want to
track. Your new shape’s boundaries appear in red in the Viewer, and a Bezier shape item
appears in the stroke/shape list. The shape remains as a normal roto shape until it is converted
into a track object.
USER GUIDE
1533
Tracking a Plane | Tracking the Footage
2. If you’re drawing more than one shape, you can arrange them in the stroke/shape list to tell
PlanarTracker that they are different layers. Order your shapes from closest to camera (top of the
list) to furthest away (bottom of the list), and PlanarTracker automatically holds out any track
layers above the current one.
3. Make sure you’re still on the same frame as you used to draw the Bezier shape.
Note: When you select a frame to draw a roto shape on, that frame becomes your
reference frame. When you proceed with tracking your plane it’s important that you’re
always starting on the same reference frame. Move to your reference frame using the Go to
Reference Frame button . In the course of the process you may decide to change your
reference frame. You can do this by clicking Set Reference Frame in the Viewer.
Tracking the Footage

You can now proceed to track the plane you’ve drawn:
Depending on whether you have chosen to insert a PlanarTracker node, or a Roto (or RotoPaint)
node, do one of the following:
If you have chosen to insert a PlanarTracker node:
USER GUIDE
1534
Tracking a Plane | Tracking the Footage
1. Use the tracking tools above the Viewer to track the Roto shape.
2. See Tracker Menu Options in the Viewer for more information.
If you have chosen to insert a Roto or a RotoPaint node:

1. Right-click the Roto shape and select one of the following options:
• planar-track this shape - Select this to convert the roto shape into a track object that doesn't
contain any tracking data (if it's not already).
• planar-track this shape (fwd) - Select this to track your shape while the track is played
forwards.
• planar-track this shape (bkwd) - Select this to track your shape while the track is played
backwards.
2. If you selected one of the latter two options, your shape is now tracked throughout the timeline. A
progress dialog is displayed while the shape is tracked.
If you selected the first option to convert your roto shape into a track object without any tracking
data, you can use the tracking tools above the Viewer to track forwards or backwards. See Tracker
Menu Options in the Viewer for more information. When you select any of the options above the
tracker menu tools are displayed above the Viewer.
The shape is now automatically added to a layer called PlanarTrackLayer1 in the stroke/shape
list and is displayed in purple in the Viewer, signifying that it has been converted into a track
object.
3. In the stroke/shape list, a purple rectangle appears in the PT column next to your shape(s). You
can toggle the purple rectangle on and off to revert the shape to a normal roto shape instead of a
track object.
Note: You can only toggle a shape between being a roto shape or a track object, after
converting the Roto shape into a track object.
4. Select the Tracking tab in the Roto node properties to display the tracking data, as shown below.
USER GUIDE
1535
Tracking a Plane | Tracker Menu Options in the Viewer
Tracker Menu Options in the Viewer

You can use the tracker menu options to perform more advanced tracking. These options are only
visible in the Viewer after you have converted a shape into a track object.
Track Motion Controls

Before tracking an object, you can use the track motion controls to specify the type of movement that
the PlanarTracker can expect. The following options are available:
• Translation - Select this so the tracker expects translation.

• Rotate - Select this so the tracker expects rotation.
• Scale - Select this so the tracker expects scaling.
• Shear - Select this so the tracker expects shearing.
USER GUIDE
1536
Tracking a Plane | Tracking Controls
• Perspective - Select this so the tracker expects changing perspective.
Tracking Controls
Use the Tracking buttons to track backwards or forwards throughout the whole footage, or if you
want, select a specified frame range.
You can also track on a frame by frame basis. For example, if the plane you’re tracking is visible
throughout the footage, forward tracking might be all you need. If the plane is only partially visible
over a part of the footage though, it might be a good idea to first track the part with the whole plane
visible, and then track the rest of the footage separately.
With the Tracking controls you can track and re-track your footage. For more information see
Tracking and Stabilizing .
Clear Tracking Data Controls

You can clear tracking information that you’ve already created with the clear buttons.
• clear all - clear all tracking information created by PlanarTracker.

• clear backwards - clear all tracking information backwards from the current frame.
• clear forwards - clear all tracking information forwards from the current frame.
USER GUIDE
1537
Tracking a Plane | Planar Tracker Surface Controls
Planar Tracker Surface Controls

After you have tracked your shape or shapes, you can review your results by scrubbing back and forth
on the timeline. You can adjust the planar surface shape if, for example, your roto shape drifts in the
course of the tracking, or you simply want to change it. To do this, you can use the planar surface
controls.
• center planar surface - Select this to center the planar surface in the Viewer during playback.
• display planar surface - Select this to display the boundary of the planar surface.
• enable planar surface editing - Select this so that you can edit the planar surface by dragging the
corner points in the Viewer.
• display grid - Select this to display a grid over the planar surface.
• set planar surface to image bounding box - Select this to change the planar surface to be the
same as the image bounding box.
Reference Frame Controls

You can use the reference controls above the Viewer to either go to the current reference frame, or
change the reference frame to the current frame.
Keyframe Controls
You can set and remove keyframes, and go to previous or next keyframes using the keyframe
USER GUIDE
1538
Tracking a Plane | Layer Dropdown
Layer Dropdown
You can use the layer dropdown to quickly select various PlanarTracker layers, or to add a new layer.
CornerPin Dropdown
You can use the CornerPin dropdown to insert a CornerPin node. The CornerPin2D tool is designed
to map the four corners of an image sequence to positions derived from tracking data. In practice,
this allows you to replace any four-cornered feature with another image sequence.
The CornerPin dropdown provides you with several different kinds of CornerPin you can choose
from:
• relative - to warp the image according to the relative transform between the current frame and the
reference frame. The image remains unchanged in the reference frame. You can also pick the baked
version of the relative CornerPin node. A baked node has the keyframe values copied from
PlanarTracker, rather than having them expression linked.
USER GUIDE
1539
Tracking a Plane | Tracking Tab Menu Options
• absolute - to use the to and from controls to place the image exactly within the selected plane. This
may skew the image throughout the footage. This attaches and automatically sets its format to the
dimensions of any currently selected node. You can also pick the baked version of the absolute
CornerPin node. A baked node has the keyframe values copied from PlanarTracker, rather than
having them expression linked.
• stabilize - to apply an inverse transform on the image, which effectively locks the image in its place
and stabilizes it in the footage. This type of corner pinning is most useful for drift corrections and
making sure your tracking results are reliable. You can also select Tracker to create a baked Tracker
node for further reducing jitter. You can also pick the baked version of the stabilize CornerPin node.
A baked node has the keyframe values copied from PlanarTracker, rather than having them
expression linked.
• Tracker - creates a Tracker node with four tracks set to each of the tracked corner points. These
tracked points give you access to the Tracker's comprehensive functions, such as track averaging
and enhanced smoothing controls. See Tracking and Stabilizing for more information.
Tracking Tab Menu Options

Select the Tracking tab in the Roto node's properties to access more tracking options.
Settings
track channels Use this dropdown to select the channel set on which you want to
track.
image channels Use this to select an individual channel from the channel set. You do
not need to select an individual channel, and instead can leave it as
none, the default.
pre-track filter Before image patches are compared, the selected filter is applied. You
can select one of the following options:
• none - This disables all pre-filtering, which allows you to have full
control of tuning the input image for tracking.
• adjust contrast - This stretches the image contrast to better suit the
tracking algorithm. This option is recommended.
• grayscale - This converts any input RGB channels to a grayscale
image for faster processing.
adjust for luminance If your footage changes brightness over time, either gradually or
USER GUIDE
1540
changes suddenly, enabling this option performs extra pre-filtering to help

compensate for the changes. This may make the track slower and less
accurate and therefore should only be selected when you need to
handle changes in luminance.
clamp super-white, sub- Select this to clamp the tracking patch pixel values to lie between 0 and
zero footage 1.
Note: If you want to track using the full dynamic range of

your super-white or sub-zero footage, disable this option.
hide progress bar As it is possible to stop a track using the tools above the Viewer, you
can choose to hide the progress dialog that appears when you track a
shape.
Export
You can use the CornerPin dropdown to select the type of corner pin node you want to insert. See
CornerPin Dropdown for more information about the different kinds of CornerPin. After you have
selected the required type of CornerPin, you can press Create to insert it.
Select the link output checkbox to link to the PlanarTracker output so that exported nodes are
updated with the track.
Correction
The CornerPin points are populated automatically when you track an object. When you draw a roto
shape and convert it into a track object, Nuke automatically places 4 corner pins around the shape.
These are the points that are tracked.
You can correct the four automatically placed points by offsetting any or all of the four points. To
offset a point, simply click and drag it in the Viewer to the correct position.
Reference Frame
This is the frame used as a reference to compare all other frames containing tracking data. It defaults
to the first frame used to track. You can change this by entering a new frame number in the
reference frame input field.
USER GUIDE
1541
USER GUIDE
1542
Reusing a Track Result | Reusing a Tracked Plane
Reusing a Track Result

You can reuse an entire tracked plane or a single track.
Reusing a Tracked Plane

You can use a plane you’ve already tracked to analyze a larger plane situated on the same plane.
1. Make sure you’re in the same frame as you used to draw the previous shape.
2. Draw a shape on the same plane as your previously tracked shape. Make sure the new shape is in
the same PlanarTrackLayer as the first one in the stroke/shape list.
3. PlanarTracker has now placed your new roto shape on the same plane as the old one. If you now
scrub in the timeline, you should find that your new shape sticks with the camera movement.
Reusing a Single Track

You can also reuse your tracked points to verify that the overall results are sticking to your footage.
1. Track a plane as described under Tracking a Plane, and adjust the results if necessary.
2. When you’re happy with the tracking results for that plane, you can drag a corner of your tracked
shape and place it on a specific detail that may not be getting tracked correctly. Note that you're
moving the roto shape locally, and not affecting the actual track results.
3. Scrub through the frames to see if the point stays on the detail as you expect.
4. If drifting occurs, drag the point to correct it.
USER GUIDE
1543
Placing an Image on the Planar Surface | Reusing a Single Track
Placing an Image on the Planar

Surface
When you’ve tracked a planar surface on your footage, you might want to place an image on it. To
place an image on the tracked planar surface, do the following:
1. Ensure you’re on the reference frame you’ve drawn the roto shape on, and select the
PlanarTrackLayer you want to use in the Create New Track dropdown on top of the Viewer.
2. Check show plane and correct plane on the properties panel to make your planar
surface visible and to enable modifying it. You can also click the corresponding buttons above the
Viewer.
3. A rectangle indicates your planar surface in the Viewer. If the rectangle is very large, you can click
the Resize Planar Surface to Image button in the Viewer.

4. Drag the corners of the rectangle to cover the area over which you want to place an image. Click
the Show grid lines button in the Viewer to use a guide grid in positioning your rectangle.
USER GUIDE
1544
This shows grid lines corresponding with the current plane and it helps with realigning your plane.
5. Scrub in the timeline to make sure the planar surface sticks to the area you want.
If you need to adjust it, you can do one of the following:

• Adjust the points in the reference frame to change the planar surface over the whole footage.
This way you’re adjusting the actual dimensions of the planar surface rectangle. The rectangle
appears in yellow.
• Adjust the points in other frames to change the planar surface in the current frame and its
adjacent frames. This way you’re correcting small drifts in the planar surface rectangle without
changing its real dimensions. The rectangle appears in blue.
USER GUIDE
1545
6. The values in the CornerPin points show how your plane has warped from the reference frame to
the current frame. Adjust them in the Curve Editor by the selecting Curve Editor tab above the
Node Graph.
Note: While you can drag the CornerPin points data to another node or control by simply
dragging selecting all the points and dragging on any of the animation buttons. If you press
Ctrl/Cmd+drag, this creates an expression link to the CornerPin points data instead of
placing them in the new location.
7. Read in the image you want to place on the tracker planar surface.
8. To convert the new image to the same format, insert a Reformat node and select the correct
image format from the output format dropdown.
9. You can now add a CornerPin node to help you place an image on the plane. Click the Create
CornerPin2D Node dropdown in the Viewer (or use the Export dropdown in the properties panel)
and select the type of corner pinning you want to use.
See the 'Tracker Menu Options in the Viewer' section in Tracking a Plane for more information about
the different types of CornerPin.
10. Insert a Merge node and connect the CornerPin node to the A input, and the Roto node to the B
input. Your node tree should now look similar the one shown below:
USER GUIDE
1546
11. You can now make any other required changes to your image. For example, you can adjust the
mix control in the Merge node's properties, to change the opacity of the image.
12. When you're happy, close all the node property panels to see the result clearly in the Viewer. You
can scrub through the timeline and make sure the position of the image in sorrect throughout.
USER GUIDE
1547
Camera Tracking
Nuke’s CameraTracker node is designed to provide an integrated camera tracking or match-moving
tool, which allows you to create a virtual camera whose movement matches that of your original
camera. Tracking camera movement in a 2D footage enables you to add virtual 3D objects to your 2D
footage.
Introduction
With the CameraTracker node, you can track the camera motion in 2D sequences or stills to create an
animated 3D camera or a point cloud and scene linked to the solve. You can automatically track
features, add User Tracks or tracks from a Tracker node, mask out moving objects using a Bezier or
B-spline shape, and edit your tracks manually. CameraTracker can solve the position of several types
of cameras as well as solve stereo sequences.
Quick Start
The tracking process is outlined below, whether you intend to track a sequence or a set of stills:
1. Connect the CameraTracker node to the sequence you want to track. See Connecting the
CameraTracker Node.
2. Mask out any areas of the image that may cause CameraTracker problems, such as movement
within the scene or burn-in. See Masking Out Regions of the Image.
3. If you're tracking stereoscopic or multi-view images, set the Principal View on the
CameraTracker or Settings tabs. See Working with Multi-View Scripts for more information.
4. Set the camera parameters, such as Focal Length and Film Back Size, if they are known. These
are described under Setting Camera Parameters.
5. Set the Source dropdown to Sequence or Stills, and then:
• If you intend to track a continuous Sequence of frames, set the tracking preferences using the
Settings tab Features and Tracking controls. See Tracking in Sequence Mode for more
information.
• If you're using Stills, you can track all frames in the same way as sequence tracking, or a subset
of Reference Frames using the +/- keyframe buttons above the Viewer or in the properties
panel. See Tracking in Stills Mode for more information.
6. You can place User Tracks to improve difficult solves, use an entirely manual tracking approach,
or set 3D survey points. You can use 3D survey points to tie your sequence to a known 3D world,
such as those created using stills. See Working with User Tracks for more information.
USER GUIDE
1548
Camera Tracking |
Tip: 3D survey points have replaced the ProjectionSolver workflow, but you can still add
ProjectionSolver nodes by pressing X in the Node Graph and entering ProjectionSolver as a
Tcl command.
7. Click Track to begin tracking the sequence.

8. Solve the Camera position by clicking Solve and refine it, if necessary. For more information, see
Solving the Camera Position
9. Set the ground plane, if required, and adjust your scene. See Adjusting the Scene.
10. Select what to export from the solve using the Export dropdown and click Create.
You can export an animated camera, a stereoscopic or multi-view rig, a 3D scene and point cloud,
lens distortion, or cards. See Using Solve Data.
11. If you have multiple footage sources of the same scene or content available, you can also use
survey points to solve each of your sources and then register them all in the same world. See
Combining Solves.
12. Add your 3D virtual objects to the footage. See Placing Objects in the Scene.
13. By default, any 3D objects you added to your footage do not have lens distortion applied to them.
As a result, they can look like they weren't shot with the same camera. To fix this, see Accounting
for Lens Distortion.
USER GUIDE
1549
Connecting the CameraTracker Node |
Connecting the CameraTracker

Node
To connect the CameraTracker node:
1. Read in and select the clip you want to track.
2. Click 3D > CameraTracker.
3. If you want to omit a part of the scene from being tracked, connect a matte to the Mask input.
Note that, unlike the Source input, this input is hidden and appears as a small triangle on the left
hand side of the node. For more information about masking, see Masking Out Regions of the
Image.
4. Click Image > Viewer to insert a Viewer node and connect it to the CameraTracker node.
USER GUIDE
1550
Masking Out Regions of the Image |
Masking Out Regions of the

Image
Tracking works best on fixed, rigid parts of the scene so that each track can create a single, fixed 3D
point. The solver uses these 3D points to work out the camera path. Moving elements and burn-ins
do not have a fixed 3D point in the world and should be masked out before tracking.
To mask regions of your sequence, attach a matte to the Mask input to define image regions that
should not be tracked. You can also use the source input’s alpha channel as a matte.
1. If you want to use a separate matte for masking, connect a Roto node to the CameraTracker Mask
input.
2. Scrub through the sequence and keyframe the roto shapes to cover the areas you don't want to
track.
You don't have to be too accurate with the mask, it's only intended to cover areas that are likely to
cause CameraTracker problems. For example, in the image shown, the actors and the copyright
burn-in are masked.
3. In the Properties panel, set Mask to the component you want to use as a mask:
• None - Track features in the whole footage.
• Source Alpha - use the alpha channel of the source clip to define which areas to ignore.
• Source Inverted Alpha - use the inverted alpha channel of the source clip to define which areas
to ignore.
• Mask Luminance - use the luminance of the mask input to define which areas to ignore.
USER GUIDE
1551
• Mask Inverted Luminance - use the inverted luminance of the mask input to define which
areas to ignore.
• Mask Alpha - use the mask input alpha channel to define which areas to ignore.
• Mask Inverted Alpha - use the inverted mask input alpha channel to define which areas to
ignore.
4. Track as normal using the automated Analysis Track button. See Tracking in Sequence Mode.
Note: There is no need to mask areas of the image when tracking manually - you specify
where User Tracks are placed.
USER GUIDE
1552
Working with Multi-View Scripts |
Working with Multi-View Scripts

CameraTracker can track and solve stereoscopic or multi-view projects in much the same way as
single view projects.
1. Connect the CameraTracker node as described in Connecting the CameraTracker Node.
2. Use the CameraTracker or Settings tab Principal View dropdown to select the view used to
create the tracks. Tracks in any other views present in the script are calculated from the Principal
View.
3. Follow the workflow described under Camera Tracking to track and solve the camera.
4. If you create any User Tracks, note that they currently only support at most two views. To select
which views are to be used as your left and right views for user tracks, use the User Track Views
control on the UserTracks tab of the CameraTracker properties.
5. To visualize both cameras for your views, create a multi-view rig with a camera for each of your
views by selecting Camera rig from the Export section of the CameraTracker tab. See Creating
Camera Nodes for more information.
USER GUIDE
1553
Setting Camera Parameters |
Setting Camera Parameters

Camera settings relate to the physical aspects of the camera used on set. Accurate physical camera
data produces a better camera track and solution.
1. Select the motion type of the on set camera from the CameraTracker tab Camera Motion
dropdown menu. This is linked to a control of the same name on the Settings tab.
• Rotation Only - select this option if the camera is static and rotating, for example, if you’re
using a tripod mounted camera for nodal pans.
• Free Camera - select this option if the camera is both translating and rotating.
• Linear Motion - select this if the camera has a straight, linear path.
• Planar Motion - select this if the camera has a flat path, moving in a two-dimensional plane
only.
2. If you have already used a separate LensDistortion node (see Working with Lens Distortion on
page 1) to remove lens distortion from your footage, you can leave the Lens Distortion control
set to No Lens Distortion.
Otherwise, set Lens Distortion to Unknown Lens before you solve the camera position to force
CameraTracker to calculate the distortion.
3. Select the Focal Length type for the camera from the dropdown menu:
• Known - select this option if the focal length is available and enter a value in the Length
control.
• Approximate Varying - select this option if an approximate focal length is available and enter
keyframed focal length values in the Length control.
• Approximate Constant - select this option if an approximate focal length is available and there
is no zoom, and enter a focal length value in the Length control.
Note: CameraTracker attempts to refine the focal length during the solve if you select an
Approximate option.
• Unknown Varying - select this option if the focal length is unknown and changing.
• Unknown Constant - this is the default option. Use this option if the focal length is unknown
and there is no zoom.
4. Either choose a Film Back Preset from the dropdown to populate the Film Back Size controls, or
if your camera isn't in the list, enter the Film Back Size manually.
USER GUIDE
1554
Setting Camera Parameters |
Tip: You can add your own defaults by editing the ../NukeScripts/camerapresets.py file in
the Nuke installation package.
USER GUIDE
1555
Tracking in Sequence Mode |
Tracking in Sequence Mode

In Sequence mode, CameraTracker tracks the footage attached to the Source input and defines a set
of 2D feature tracks that correspond to fixed points in the scene.
Note: If you intend to remove lens distortion manually using a separate LensDistortion
node, you should do that before you track the sequence. See Working with Lens Distortion.
Otherwise, set Lens Distortion to Unknown Lens before you solve the camera position to
force CameraTracker to calculate the distortion.
Before tracking, use CameraTracker's properties panel to control Viewer output and determine
tracking behavior:
1. On the CameraTracker tab, ensure that Source is set to Sequence.
2. If you intend to mask out parts of your image, set the Mask control to the matte source.
For example, if you're using the alpha channel from a Roto node attached to the Mask input,
select Mask Alpha. For more information, see Masking Out Regions of the Image.
Note: For stereoscopic or multi-view footage, set the Principal View on the
CameraTracker or Settings tab to the main tracking view. Any masking should be applied
to this view, which is then used to calculate the secondary camera.
3. Use the Range dropdown to determine which frames are analyzed:

• Input - the default value, sets the frame range to the length of the sequence attached to the
Source input.
• Global - sets the tracking frame range to the range set in the Project Settings frame range
controls.
If no frame range is defined, the frame range of the first image you read in is used as the Global
frame range.
• Custom - sets a tracking range of frames described by the from and to fields.
4. On the Settings tab, define the starting points for feature tracking:
• Number of Features - define the number of features you want to track in each frame.
Ideally, you should use more than 100 tracks per frame. In most cases, the default 150 should be
sufficient, but in difficult sequences you may consider using a higher number.
• Detection Threshold - set the distribution of features over the input image.
USER GUIDE
1556
If you enter a low detection threshold value, features are tracked evenly on all parts of the image
and vice versa.
Low threshold distribution High threshold distribution

• Feature Separation - set the distribution of features in relation to each other.
To force feature separation and spread features evenly over the image at even distances, enter a
high feature separation value.
Low separation value High separation value

5. Check Refine Feature Locations to lock detected features to local corners. If you activate this,
CameraTracker finds the closest corner point in your footage and locks the feature to it.
6. Check Preview Features to view the current distribution of tracking features.
Preview comes in handy when you want to tweak the tracking parameters further before tracking -
it updates dynamically when controls are adjusted.
It's important to make sure that the previewed features are distributed evenly and not clustered
together too densely before tracking using the distribution and separation controls.
USER GUIDE
1557
Previewing features in a sequence

7. Once you're happy with the feature distribution in the preview, click Track to begin analyzing the
sequence.
8. CameraTracker begins reading frames sequentially and tracking the features present. Tracks that
don't meet the quality thresholds set in the Settings tab Tracking controls are reseeded so that
the number of tracks remains constant.
See Troubleshooting Sequence Tracks for more details on adjusting these controls.
When the playhead reaches the end of the sequence, it begins a verifying pass by reading the
frames sequentially backwards. Any tracks that were reseeded due to error are tracked back past
the point where they were created, if they remain viable.
USER GUIDE
1558
Viewing Track Data | Reviewing AutoTracks Curves
Viewing Track Data

Once tracking is complete, scrub through the timeline to examine the tracked features. The points
represent the features and the vectors the track length calculated for the associated feature. Hover
over a point to display its length, in frames.
Consider masking out areas where tracks are consistently seeded and rejected, such as highly
reflective surfaces, and adjusting the Settings tab Tracking controls to increase track reliability.
See Masking Out Regions of the Image and Troubleshooting Sequence Tracks for more information.
Reviewing AutoTracks Curves

Detailed tracking information is displayed in curves in the properties panel on the AutoTracks tab.
You can select all track curves at once or get a more detailed view of a single curve, such as num
tracks.
Tip: Pressing F with focus on a curve maximizes the selected curve in the space available.
USER GUIDE
1559
Viewing Track Data | Reviewing AutoTracks Curves
Selecting all track curves. Framing a single curve.
Curves can indicate areas of the sequence where tracking encountered problems. For example, the
number of tracks curve on the right shows a significant dip on the current frame, as indicated by the
playhead position.
In this instance, you could try adjusting the Number of Features or Minimum Length and
retracking. See Troubleshooting Sequence Tracks
USER GUIDE
1560
Troubleshooting Sequence Tracks | Pre-Tracking Checks
Troubleshooting Sequence Tracks

Some sequences are inevitably going to cause problems. There are a number of pre-tracking checks
and post-tracking refinement controls to assist CameraTracker.
Pre-Tracking Checks
• Play through the sequence before tracking and mask out any problem areas in the scene. Large
moving objects can confuse the tracking and solving process as can appear to be fixed areas of the
scene.
See Masking Out Regions of the Image for more information.
Tip: If you're compositing on moving elements, such as faces, try tracking and solving on
just the moving element using masks. You'll then get a camera that moves around the
element as if it was fixed in a single position.
• Textureless areas of the sequence, such as greenscreens, can cause features to cluster together in
other areas, affecting the track and solve.
Turn on Preview Features and use the distribution and separation controls to even out the
features. See Tracking in Sequence Mode
• Regular edges, such as striped patterns, in the scene can confuse CameraTracker.
Again, turn on Preview Features and use the Detection Threshold to force CameraTracker to use
corner-like images textures. See Tracking in Sequence Mode
• You can improve tracking data by adding User Tracks manually, see Working with User Tracksfor
more information.
Reviewing and Refining Tracking Data

You can use the threshold controls on the AutoTracks tab to dynamically reject tracks and remove
them, before solving, to improve accuracy. Using the num tracks curve as an example:
1. Click the AutoTracks tab in the properties and select the num tracks curve.
Tip: Press F in the curve display to fit the selected track(s) to the available screen space.
USER GUIDE
1561
Troubleshooting Sequence Tracks | Reviewing and Refining Tracking Data
In the example, you can see that the number of tracks has dropped significantly around frames
180-200.
2. Move the Viewer playhead to the affected area and examine the tracked features in the Viewer.
3. If you increase the Min Length control in the properties panel, you'll start to see tracks turn red
as they fall below the threshold.
Min Length set at the default value, 3 frames. Min Length set to 50 frames.
4. Click Delete Rejected to remove all the tracks that fall below the specified threshold.
Tip: You can also remove tracks manually, by selecting them in the Viewer, right-clicking,
and choosing tracks > delete selected.
5. You can then retrack the affected frame range by clicking Update Track. See Retracking Partial
Frame Ranges for more information.
USER GUIDE
1562
Troubleshooting Sequence Tracks | Re-tracking Using Tracking Settings
Re-tracking Using Tracking Settings

Once you have refined the tracking data, review the sequence by scrubbing the playhead to make
sure you have tracks across the image on the rigid fixed background. If you see lots of tracks that drift
or jump, select them manually, right-click, and go to tracks > delete selected.
You can also add User Tracks to improve tracking data before solving, see Working with User Tracks.
You can improve a set of feature tracks using the following controls and then retrack to improve your
chances of getting a good solve:
• Minimum Length - sets a threshold value for the minimum acceptable track length. Tracks that fail
to stick to the associated feature for this amount of frames are rejected.
In long slow camera moves, it is best to solve from long tracks generated in the shot. If there are
lots of short tracks, this can lead to noise on the calculated camera. Try removing the short tracks
before solving.
• Track Threshold - CameraTracker’s tolerance to change along the track is determined by this
control.
Reducing this threshold makes tracking more tolerant to image changes, potentially producing
longer tracks.
• Track Smoothness - sets the threshold for smooth track generation. Adjusting this value can be
useful in preventing poor tracks in complex sequences.
Increasing the smoothness value removes tracks that fail over time.
USER GUIDE
1563
Troubleshooting Sequence Tracks | Re-tracking Using Tracking Settings
• Track Consistency - sets the threshold for how inconsistent a feature track can be before
CameraTracker discards it and reseeds in a different location.
Higher values allow for less inconsistency and vice versa.
Click Delete Rejected to remove all the tracks that fall below the specified thresholds.
USER GUIDE
1564
Extending Existing Camera Tracks | Re-tracking Using Tracking Settings
Extending Existing Camera Tracks

CameraTracker allows you to extend an existing set of tracking data by adding new frames, such as
when more frames become available from a shoot.
Note: You can update tracking data as often as you like before solving, but once you've
solved the camera position, updating tracking data should only be used to add a relatively
small number of frames. See Updating Solves with Extended Tracking Data for more
information.
To add new frames to your tracking data:

1. Read the new frames into Nuke using the Read node's frame range controls.
2. Open the CameraTracker properties panel and click Update Track.
A dialog displays allowing you to set the frame range to update.
3. Set the required frame range and click OK.
CameraTracker tracks the specified range and combines the tracking data with the existing tracks.
For example, if your original AutoTracks > num tracks curve appeared as shown on the left, the
image on the right represents the new tracking data after the update is complete.
Original frame range track data. Updated track including the extended frame range.
Tip: If the transition between the ranges seems abrupt, you can retrack the frames around
the join using the method described in Retracking Partial Frame Ranges.
4. Proceed to Solving the Camera Position.
USER GUIDE
1565
Retracking Partial Frame Ranges | Re-tracking Using Tracking Settings
Retracking Partial Frame Ranges

After refining your feature points, you may not need to retrack the entire sequence. You can use
Update Track to analyze a specific frame range.
1. Refine your features using the controls described previously.
2. Click Update Track.
CameraTracker retracks the selected range and combines the tracking data with any existing
tracks.
Note: You may notice that CameraTracker performs some analysis outside your selected
frame range. This can be necessary to extend existing tracks into the update range, avoid
creating duplicate tracks, and to combine the new tracks with existing tracking data.
USER GUIDE
1566
Tracking in Stills Mode | Re-tracking Using Tracking Settings
Tracking in Stills Mode

In Stills mode, CameraTracker tracks the reference frames in the Source input and analyzes the input
in two stages. Tracking defines the set of 2D feature tracks that correspond to fixed rigid points in the
scene, then the solver calculates the camera path and projection that creates a 3D point for each
reference frame.
Note: If you intend to remove lens distortion manually using a separate LensDistortion
node, you should do that before you track the stills. See Working with Lens Distortion.
Otherwise, set Lens Distortion to Unknown Lens before you solve the camera position to
force CameraTracker to calculate the distortion.
USER GUIDE
1567
Still Photography Guidelines | General Guidelines
Still Photography Guidelines

Still tracking relies on good input to produce good tracking output, so it's vital that you capture good
still photographs that CameraTracker can interpret correctly.
CameraTracker has different requirements depending on the subject of the stills. For example, stills
for a near-flat scenes are captured differently to those required for an interior set or model.
General Guidelines
• Don't crop your stills or transform them in anyway, such as rotation.
• Avoid dramatic changes in scale and angle between stills.
• To ensure you capture every part of the scene in 3 to 4 images, aim for a maximum change of 20-
25% in content between frames.
• Avoid unnecessary redundancy where still frames contain large portions of the previous frame.
• Avoid tracking stills containing multiple occlusions - features with complex overlap can result in
drastic changes in content between frames.
Shooting Near-flat Scenes

The best stills for near-flat scenes are captured head-on, facing the subject. In Nuke terms, that would
mean taking your stills from the camera positions shown on the right.
Incorrect camera positioning. Correct camera positioning.
Shooting 3D Objects
When moving around an object, a photo every 15-25 degrees should be adequate, so a minimum of
16 shots for a full 360. Of course, taking more can improve the result.
USER GUIDE
1568
Still Photography Guidelines | Shooting Interiors
Too few stills. Minimum number of stills.
Shooting Interiors
For enclosed spaces, such as interiors, capture stills from the center facing outward and then around
the perimeter facing inward, rather than from different points within the space.
Incorrect camera positioning from different points.
Capturing stills facing outward. Capturing stills facing inward.
USER GUIDE
1569
Tracking Still Frames | Selecting the Frames to Track
Tracking Still Frames

Before you do anything else, you need to tell CameraTracker which still frames you want it to track
and how it should distribute the tracking features on each frame. Then, you can track your still
frames.
Selecting the Frames to Track

1. On the CameraTracker tab of the CameraTracker properties, ensure that Source is set to Stills.
2. Use the Range dropdown to determine which frames are analyzed:
• Input - the default value, sets the frame range to the length of the sequence attached to the
Source input.
• Global - sets the tracking frame range to the range set in the Project Settings frame range
controls.
If no frame range is defined, the frame range of the first image you read in is used as the Global
frame range.
• Custom - sets a tracking range of frames described by the from and to fields.
• Reference Frames - allows you to manually set keyframes on the frames you want to analyze.
3. If you set Range to Reference Frames, do one of the following:
• To set all frames in the input as reference frames, set the Add dropdown menu to Add All . Any
missing frames on the input are skipped.
• To set a specific frame range as reference frames, set the Add dropdown menu to Add Range.
In the dialog that opens, enter the frame range in the Frame range field and click OK.
• To set individual frames as reference frames, set the Thumbnails dropdown menu to All at the
top of the Viewer. Then, scrub to a frame you want to set as a reference frame and click the add
reference frame button . The current frame is set as a keyframe and displayed in the
thumbnail gallery at the bottom of the Viewer.
USER GUIDE
1570
Tracking Still Frames | Adjusting the Features to Track
Scrub to the next frame you want to add as a reference frame and repeat the process until
you're happy with your set of reference frames.
Tip: If necessary, you can also use the delete reference frame button or the Delete
dropdown menu in the CameraTracker properties to remove frames from the set of frames
to analyze.
4. Proceed to Adjusting the Features to Track below.
Adjusting the Features to Track

1. If you intend to mask out parts of your image (such as burn-ins), set the Mask control to the
matte source.
For example, if you're using the alpha channel from a Roto node attached to the Mask input,
select Mask Alpha. For more information, see Masking Out Regions of the Image.
Note: For stereoscopic or multi-view footage, set the Principal View on the
CameraTracker or Settings tab to the main tracking view. Any masking should be applied
to this view, which is then used to calculate the secondary camera.
2. On the Settings tab, define the starting points for feature tracking:
• Number of Features - define the number of features you want to track in each frame.
USER GUIDE
1571
Tracking Still Frames | Adjusting the Features to Track
Ideally, you should use more than 200 tracks per frame. In most cases, the default 250 should be
sufficient, but in difficult solves you may consider using twice that number.
• Detection Threshold - set the distribution of features over the input image.
If you enter a low detection threshold value, features are tracked evenly on all parts of the image
and vice versa.
Low threshold distribution High threshold distribution

• Feature Separation - set the distribution of features in relation to each other.
To force feature separation and spread features evenly over the image at even distances, enter a
high feature separation value.
Low separation value High separation value

3. Check Refine Feature Locations to lock detected features to local corners. If you activate this,
CameraTracker finds the closest corner point in your footage and locks the feature to it.
4. Check Preview Features to view the current distribution of tracking features.
Preview comes in handy when you want to tweak the tracking parameters further before tracking -
it updates dynamically when controls are adjusted.
It's important to make sure that the previewed features are distributed evenly and not clustered
together too densely before tracking using the distribution and separation controls.
USER GUIDE
1572
Tracking Still Frames | Tracking the Selected Frames
Previewing features in a sequence

5. Once you're happy with the feature distribution in the preview, proceed to Tracking the Selected
Frames below.
Tracking the Selected Frames

1. In the CameraTracker properties, click Track to begin analyzing the still frames.
CameraTracker begins by detecting viable features, aligning the reference frames, and then
tracking forwards. Tracks that don't meet the quality thresholds set in the Settings tab Tracking
controls are reseeded so that the number of tracks remains constant.
When the playhead reaches the end of the sequence, it begins a verifying pass by reading the
frames sequentially backwards. Any tracks that were reseeded due to error are tracked back past
the point where they were created, if they remain viable.
2. Proceed to Viewing Reference Frames and Track Data.
USER GUIDE
1573
Viewing Reference Frames and Track Data | Tracking the Selected Frames
Viewing Reference Frames and

Track Data
Once tracking is complete, you can review how CameraTracker connected the analyzed frames:
• Switch the Thumbnails control above the Viewer to Tracked. CameraTracker displays the
frames that are connected to the current frame in the thumbnail gallery at the bottom of the
Viewer. This allows you to simply step through the sequence frame-by-frame to evaluate the
connections.
• Switch the Thumbnails control above the Viewer to All to have CameraTracker display all frames
in the tracked range in the thumbnail gallery. This allows you to hover the cursor over a
thumbnail to view its connections to adjacent reference frames highlighted in orange.
Disconnected frames are highlighted in red. See Disconnected Frame Sets for more information.
Tip: You can click-and-drag to scroll the gallery left and right, or click a thumbnail to move
the playhead to the corresponding reference frame. To adjust the size of the thumbnail
gallery, drag the top of the gallery up and down.
2. To display track information, hover over a feature or marquee select several features.
Note: If you make multiple selections, you won't see the track lengths in frames, only the
tracks themselves.
USER GUIDE
1574
Viewing Reference Frames and Track Data | Reviewing AutoTracks Curves

Detailed tracking information is displayed in curves in the properties panel on the AutoTracks tab.
You can select all track curves at once or get a more detailed view of a single curve, such as num
tracks.
Selecting all track curves. Framing a single curve.
Curves can indicate areas of the sequence where tracking encountered problems. For example, the
number of tracks curve on the right shows a significant dip on the current frame, as indicated by the
playhead position.
In this instance, you could try adjusting the Number of Features or Minimum Length and
retracking. See Troubleshooting Still Tracks
USER GUIDE
1575
Disconnected Frame Sets | Reviewing AutoTracks Curves
Disconnected Frame Sets

Auto-tracking stills is not perfect and may throw out some disconnected frames, most likely due to a
large change in viewpoint. After tracking, set the Thumbnails control above the Viewer to All and
hover over images in the thumbnail strip. Thumbnails highlighted in red were not matched with the
reference frames around them and are flagged with an error message in the Viewer.
CameraTracker can deal with disconnected frames in two ways:
• work with the connected set by deleting the unconnected reference frames using the button or
Delete dropdown and tracking again, or
• if all the frames in the sequence are needed, create User Tracks that span the sets to connect them.
Make sure to define the tracks in as many frames as possible, and then track again based on the
User Tracks. See Linking Still Reference Frames.
Tip: When a camera track starts and ends in the same place, CameraTracker may not
automatically connect the end frame with the starting frame. You can set the Thumbnail
control to Tracked to review the connections created between stills, and then add User
Tracks between the two frames and click Update Track. See Working with User Tracks for
more information.
USER GUIDE
1576
Troubleshooting Still Tracks | Pre-tracking Checks
Troubleshooting Still Tracks

Some stills sequences are inevitably going to cause problems. There are a number of pre-tracking
checks and post-tracking refinement controls to assist CameraTracker.
You can improve tracking data by adding User Tracks. See Working with User Tracks for more
information.
Pre-tracking Checks
• For standard sequences, avoid too much or too little redundancy between still frames, that is,
frames that contain large portions of the previous frame. As a guide, try to capture every part of the
scene in 3 to 4 images, aiming for a maximum change of 20-25% in content between frames.
• Try to capture the same object in at least 3, but preferably 4 photos.
For example, when moving around an object, a photo every 15-25 degrees would be adequate
(approximately 16 shots for a full 360).
Post-Tracking Refinements
You can refine a set of feature tracks using the following controls and then retrack to improve your
chances of getting a good solve:
• Minimum Length - sets a threshold value for the minimum acceptable track length. Tracks that fail
to stick to the associated feature for this amount of frames are rejected.
In long slow camera moves, it is best to solve from long tracks generated in the shot. If there are
lots of short tracks, this can lead to noise on the calculated camera. Try removing the short tracks
before solving.
• Track Threshold - CameraTracker’s tolerance to change along the track is determined by this
control.
Reducing this threshold makes tracking more tolerant to image changes, potentially producing
longer tracks.
USER GUIDE
1577
Troubleshooting Still Tracks | Retracking Partial Frame Ranges
Retracking Partial Frame Ranges

After refining your feature points, you may not need to retrack the entire sequence. You can use
Update Track to analyze a specific frame range.
1. Refine your features using the controls described above.
2. Click Update Track.
CameraTracker retracks the selected range and combines the tracking data with any existing
tracks.
Note: You may notice that CameraTracker re-detects features on existing frames before re-
tracking. This can happen when the cached feature data has been purged and needs to be
rebuilt.
USER GUIDE
1578
Working with User Tracks | Retracking Partial Frame Ranges
Working with User Tracks

User Tracks are placed manually, rather then being automatically seeded by CameraTracker, and can
be used to improve or even replace auto tracking data. They can also be used to link unmatched
reference frames together when tracking stills.
You can create User Tracks before tracking and solving to lock the camera to a particular part of the
shot or afterwards to help improve the results.
These pages cover the following topics:

• Adding and positioning User Tracks in the Viewer. See Adding and Positioning User Tracks.
• Track User Tracks in several different ways. See User Tracking Methods.
• Using User Tracks to improve auto-tracking data. See Tracking Assists.
• Tracking a scene manually using only User Tracks. See Tracking a Scene Manually.
• Using User Tracks to link reference frames when tracking stills. See Linking Still Reference Frames.
• Assigning User Tracks as 3D survey points in order to tie your sequence to a known 3D world, such
as those created using stills. See Assigning 3D Survey Points.
USER GUIDE
1579
Adding and Positioning User Tracks | Retracking Partial Frame Ranges
Adding and Positioning User

Tracks
You can add as many User Tracks as required, depending on what you intend to accomplish. For
example, auto-track assist may only require one or two User Tracks, whereas manual tracking would
require at least eight User Tracks.
1. Enable the fast-add button at the top of the Viewer and click in the Viewer to add User Tracks
or click Add Track on the UserTracks tab in the properties panel.
Note: Clicking Add Track places the new track in the center of the current Viewer.
Tip: You can quickly add User Tracks by holding Ctrl/Cmd+Alt and clicking in the Viewer.
Track zoom windows display in the top left of the Viewer.
There are three zoom windows initially:

• Reference frame (green) - the first frame in the track, which remains constant, allowing you to
locate the feature more accurately in subsequent frames.
• Current frame (orange) - the playhead position. Use this window to adjust the tracking anchor
on the current frame by comparing it to the reference frame.
USER GUIDE
1580
Adding and Positioning User Tracks | Retracking Partial Frame Ranges
• Keyframe (blue) - the first keyframe in the sequence. Adding more keyframes adds a zoom
window per keyframe.
Tip: You can click on a keyframe zoom window to instantly jump to that frame.
Tip: You can hide the zoom windows by setting the Zoom Window Viewer control to Off.
2. Drag the crosshair in the center of the anchor over the feature to be tracked or manually adjust
the track x and y fields in the UserTracks list.
3. Refine the track position by clicking and dragging in the current frame zoom window.
Tip: You can change the magnification of zoom windows by holding Shift and dragging the
magnifying glass cursor away from the center of the window.
To resize a zoom window, drag its lower right corner to a new location.
4. Once you're happy with the track's position, proceed to User Tracking Methods.
USER GUIDE
1581
User Tracking Methods | Auto User Tracks
User Tracking Methods

User Tracks can be tracked automatically or manually, extracted from auto-tracking data, and
imported from a Tracker node.
Auto User Tracks

The easiest way to get started with User Tracks is using the Autotrack feature, but you can create
tracks totally independent of auto-tracking. After placing your tracks on the required features in the
sequence:
1. Click on the UserTracks tab in the properties panel to display the track table.
2. Select all the tracks you want to calculate in the table and then click Autotrack.
CameraTracker attempts to track the selected features, in a similar way to Auto Tracks, except that
when a track fails it is not reseeded.
3. You can examine features during and after tracking to determine if they failed:
• During tracking - the anchors are colored yellow until they fail.
• After tracking - any failed tracks don't have keyframes all the way through the sequence.
The image shows a good track, user 1, highlighted in yellow and a failed track, user 3, which has
started to drift on frame 116.
4. If a track fails or drifts, adjust the position of the tracking anchor by dragging the anchor in the
zoom window.
USER GUIDE
1582
User Tracking Methods | Manual User Tracks
5. Scrub through the sequence a few frames and repeat. Continue on through to the end of the
sequence.
6. At each frame, a new keyframe window is added to the right of the zoom window.
Manual User Tracks

Manual tracking can produce the best results, but is time consuming because all keyframes are
tracked manually by you, rather than CameraTracker.
Tip: Generally, you would only manually create User Tracks in stills. Nuke's Tracker node
can create reliable tracks for continuous sequences which you can then import as User
Tracks.
After placing your tracks on the required features in the sequence:

2. Select the required track from the tracks list.
3. Scrub the playhead forward a few frames and move the tracking anchor to the correct position in
the new frame using the current zoom window.
A new keyframe zoom window is added to the Viewer.
4. At the top of the Viewer, set Thumbnails to User Tracks.

All keyframes for the currently selected user track are displayed in the thumbnail gallery.
5. To see a User Track's reference frame, hover over the track in the Viewer. This can help you
quickly decide if the object you're tracking is in the current frame or not.
USER GUIDE
1583
User Tracking Methods | Extracting User Tracks
Tip: The keyframe windows at the top of the Viewer show you all of the keyframes for the
currently selected User Track. You can click on a keyframe window to jump directly to that
frame.
6. When you've finally keyframed your User Track in all of the desired frames, the tick marks on the
Viewer's timeline are set on all of those keyframes. By using Alt+left or right arrow, you can
quickly jump to the previous or next keyframe. Doing this while looking at the zoom window
helps show you what keyframes your User Track is slightly off in, making this a very useful trick
for getting perfectly placed User Tracks.
Extracting User Tracks

After auto-tracking, you can use the data to extract User Tracks. CameraTracker considers User Tracks
to be important and tries to lock down on them over auto-tracks. As a result, extracting User Tracks
from an auto-track can be useful if you have an auto-track that appears to be reliable.
You might also extract auto-tracks to User Tracks if you want to export some auto-tracks to be used
in a Tracker node for 2D work.
1. Right-click on an auto-track in the Viewer and select tracks > extract user track.
The selected track is prefixed with extr. and placed in the User Tracks table.
2. If a track fails or drifts, adjust the position of the tracking anchor by dragging the anchor using the
zoom window.
3. Scrub through the sequence a few frames and repeat. Continue on through to the end of the
sequence.
4. At each frame, a new keyframe window is added to the right of the zoom window.
Importing a Tracker
You can quickly create User Tracks by importing keyframes and tracking data from a Tracker node
within the same script.
2. Click Import Tracker.
3. The Select Tracker to Import From dialog displays.
4. Select the required Tracker node from the tracker node dropdown.
A new User Track is added to the tracks list.
USER GUIDE
1584
Tracking Assists | Importing a Tracker
Tracking Assists
Auto-tracking data can be improved by seeding User Tracks in a sequence, as CameraTracker
assumes manually placed tracks are superior to auto-tracking. Re-tracking after adding User Tracks
forces CameraTracker to recalculate auto-tracks using more accurate data.
1. Track your sequence using the steps described in Tracking in Sequence Mode or Tracking in Stills
Mode
2. Track one or two User Tracks using the methods described in User Tracking Methods.
Tip: Adding tracks is time consuming, but the more User Tracks you add, the better the
result is likely to be.
3. Either re-track the entire sequence by clicking Track,

OR
If you only added a few keyframes to a particularly troublesome area of the sequence, click
Update Track to re-track just that range of frames.
4. If the tracking data requires more work, try adjusting some of the controls described in
Troubleshooting Sequence Tracks orTroubleshooting Still Tracks.
USER GUIDE
1585
Tracking a Scene Manually | Importing a Tracker
Tracking a Scene Manually

Sometimes, a scene can be difficult to track automatically. For example, you might be tracking a set
of still photos of a scene with a very wide baseline between images. Alternatively, you might already
have 2D track data from an earlier process that you want to get a solve from. In cases like this, it can
be easier to only set up and work with User Tracks as your track data.
The key thing to remember is that you typically need at least 8 to 16 User Tracks per frame to get a
reliable solve.
1. Use the steps described in Manual User Tracks to create the required number of manual tracks.
USER GUIDE
1586
Linking Still Reference Frames | Importing a Tracker
Linking Still Reference Frames

During stills tracking, the links between reference frames can fail. This can occur for a number of
reasons, such as a lack of overlap between stills. You can help CameraTracker by adding User Tracks
on features common to the frames before auto-tracking.
1. Use the steps described in Working with User Tracks to create at least four User Tracks in as many
frames as possible.
For example, in the stills shown, the tops of the pillars and the stone work on the corner of the
building.
Reference frame A Reference frame B

2. Move to the next frame and reposition the User Tracks to the corresponding features in the still.
Continue throughout the sequence of stills.
3. Once all your frames have been linked in this way, click Track.
CameraTracker uses the reliable information from the User Tracks to create a set of Auto Tracks.
Note: The above workflow describes how to use User Tracks to help connect frames that
weren't automatically tracked. Sometimes, there are so many poorly tracked frames that
instead of hoping that linking frames and clicking Track works, it might be easier to start
from scratch and perform an entirely manual track. See Tracking a Scene Manually for more
details.
USER GUIDE
1587
Assigning 3D Survey Points | Importing a Tracker
Assigning 3D Survey Points

If you create a User Track, you can assign it as a known 3D survey point. This tells CameraTracker
which points in your 2D footage go with their counterparts on your 3D model, allowing it to solve the
camera to match the known 3D points and achieve the best results.
Note: 3D survey points have replaced the ProjectionSolver workflow, but you can still add
ProjectionSolver nodes by pressing X in the Node Graph and entering ProjectionSolver as a
Tcl command.
Tip: You can create a 3D model for your footage using ModelBuilder or an external
application.
To create 3D survey points in your scene:

1. To define the full rotation, translation, and scale between the scene and your survey points, create
and track at least three User Tracks. If you track more than three, you'll get the best fit result. See
Working with User Tracks for more information.
2. Switch the Viewer to 3D mode, either by pressing Tab in the Viewer or using the View selection
dropdown.
3. Set the Thumbnail control above the Viewer to All to display all reference frame thumbnails as an
overlay in the Viewer.
4. Use the selection mode button above the Viewer to swap to Vertex selection mode.
5. Select a vertex on the 3D model and the User Track you want to be a survey point in the
thumbnail gallery.
6. Right-click on the User Track and select user tracks > snap to 3D vertex selection.
USER GUIDE
1588
Assigning 3D Survey Points | Importing a Tracker
7. In the CameraTracker properties panel, click the UserTracks tab to display a list of tracks.
The user track table displays the pixel error when matching the 3D vertex to the User Track. You
can use this to review how well the camera solve fits the 3D survey points.
If the solver could not match a point, there will be a high error value. If this happens, try looking
at the 2D feature positions to check they correspond to the same point and also double check the
3D vertex on the model.
8. Repeat the process for the required User Tracks in the scene, and then enable the s checkbox for
each track in the User Tracks table.
The s designates the points as known 3D survey positions during the solve, allowing
CameraTracker to create a camera positioned correctly for the geometry.
Tip: If you have multiple footage sources of the same scene or content available, you can
also use survey points to solve each of your sources and then register them all in the same
world. See Combining Solves.
USER GUIDE
1589
Solving the Camera Position | Importing a Tracker
Solving the Camera Position

When you’re happy with the features that you’ve tracked, you can solve the camera position.
CameraTracker uses the tracking information to calculate the camera position and add position
information to the tracked feature points in the Viewer.
In the CameraTracker properties panel, click Solve. The solver selects a subset of keyframes
positioned relatively far apart so that there is sufficient parallax in the images to define the 3D points
for the tracks.
When the solve is complete, the Viewer displays the tracked features with solve data applied.
USER GUIDE
1590
Viewing Solve Data | Importing a Tracker
Viewing Solve Data

Once the solve data is placed in the Viewer, you can zoom in to display the points more clearly. You
can control what appears in the Viewer using the Display controls on the Settings tab:
• Show tracks - show or hide the 2D tracking information.
• Show projected 3D points - show or hide the 2D position of the 3D points.
• Show key tracks only - only show the longest tracks used to calculate the solve.
• Show 3D marker - show or hide 3D markers at each point in the 3D Viewer.
A traffic light scheme applies to the 2D tracks to help find good thresholds for track rejection in the
AutoTracks tab - amber by default, green for good tracks, and red for poor tracks.
The circles and crosses are reprojected solved auto-tracks. The closer the 3D point circles are to the
feature points, the better the solve.
Hover over a point to display its solve information.
CameraTracker also creates a point cloud in the 3D Viewer allowing you to cross-check the relative
positions of the 2D and solved 3D points. Select some points in 2D and then press Tab in the Viewer
to switch between 2D and 3D space to check the point positions.
USER GUIDE
1591
Viewing Solve Data | Reviewing AutoTracks Curves
Several points selected in the 2D sequence. The same points shown in the 3D Viewer.

Detailed solve information is displayed in curves in the properties panel on the AutoTracks tab. You
can select all solve curves at once or get a more detailed view of a single curve, such as error - rms
(root mean square).
Selecting all solve curves. Framing a single curve.
Curves can indicate areas of the sequence where solving encountered problems. For example, the
root mean square error curve on the right shows a significant blip on the current frame, as indicated
by the playhead position. In this instance, you could try adjusting the Keyframe Spacing or
Smoothness and updating the solve. See Troubleshooting Solves
The curves in the graph show the following solve information:
USER GUIDE
1592
Viewing Solve Data | Reviewing Solved User Tracks
• Solve Error - displays the constant SolveError parameter.

• error min - the minimum reprojection error at each frame (in pixels).
• error rms - the root mean reprojection error at each frame (in pixels).
• error track - the maximum root mean reprojection error calculated over the track lifetime at each
frame (in pixels).
• error max - the maximum reprojection error at each frame (in pixels).
• Max Track Error - displays the constant Max RMS Error parameter.
• Max Error - displays the Max Error threshold parameter.
Reviewing Solved User Tracks

When a User Track is solved, you can review its error in the User Track table on the UserTracks tab. If
the error is high, try reviewing the 2D x,y feature track. Alternatively, you can uncheck the User Track's
e (enable) column to remove it from the scene before clicking Solve again.
The User Track's 3D position (X,Y,Z) is also shown in the table, as well as in the 3D Viewer.
Tip: You can create a User Track after solving the camera in order to extract a specific 3D
position in the shot. To produce an accurate 3D point, the User Track must be defined in 3
or more frames with good parallax. When you've created the User Track, select it in the User
Tracks table and click the Update XYZ button to triangulate the 3D position from the current
solve.
Tip: After auto-tracking, you might want to select an auto-track and extract it as a User
Track. This hints to the solver that this is an important track and should be locked down on.
This is particularly useful if a particular part of the scene is not solving properly, despite the
auto-tracks appearing to have tracked well in that region.
For more information, see Extracting User Tracks.
Previewing Matchmove Quality

After calculating the solve, you may want to review its quality by previewing how well objects added to
the 3D scene are going to stick to your footage. If you set Lens Distortion to Unknown Lens before
creating the solve, there are two ways to do this:
USER GUIDE
1593
Viewing Solve Data | Previewing Matchmove Quality
Method 1
1. In the 2D Viewer, right-click on a track point (or several selected points, ideally on the same plane)
and select create > cube, for example.
CameraTracker creates a Cube node and places it in the average position of the selected points.
You are going to use the cube to test how well it sticks to the input footage.
Tip: If necessary, you can adjust the size of the cube using the uniform scale control in the
Cube properties.
2. In the CameraTracker properties, set the Export menu to Scene+ and click Create.
CameraTracker creates a 3D Scene with a Camera, PointCloud, ScanlineRender, and a
LensDistortion node set to undistort the input.
3. Connect your Cube node to one of the Scene node's inputs.
4. View the output from ScanlineRender.

The cube you created in step 1 and the point cloud CameraTracker generated are displayed on top
of the undistorted footage in the 2D Viewer.
USER GUIDE
1594
5. Scrub through the timeline to see whether the cube and the point cloud stick to the footage. If
they do, delete any nodes you no longer need (such as the Cube node) and proceed to Adjusting
the Scene. If they don't, proceed to Troubleshooting Solves.
Method 2
1. In the CameraTracker properties, set the Export menu to Camera and click Create.
CameraTracker creates a Camera node that emulates the camera used on set.
2. In the 2D Viewer, right-click on a track point (or several selected points, ideally on the same plane)
and select create > cube, for example.
CameraTracker creates a Cube node and places it in the average position of the selected points.
You are going to use the cube to test how well it sticks to the input footage.
USER GUIDE
1595
Tip: If necessary, you can adjust the size of the cube using the uniform scale control in the
Cube properties.
3. Enable Undistort Input.

CameraTracker removes lens distortion from the input footage.
4. Make sure that both the CameraTracker and Camera properties panels are open.
5. Press Tab on the Viewer to switch to the 3D view.
6. Set the camera look through menu in the top right corner of the Viewer to the camera you created
in step 1 and click the button to lock the 3D view to that camera.
7. At the top of the Viewer, set the Viewer composite dropdown menu to over, so that you are
viewing your CameraTracker node over the same CameraTracker node.
This overlays the input image currently undistorted by CameraTracker over the point cloud in the
3D Viewer.
8. Scrub through the timeline to see whether the cube you created in step 2 and the point cloud
CameraTracker generated stick to the footage. If they do, delete any nodes you no longer need
(such as the Cube node) and proceed to Adjusting the Scene. If they don't, proceed to
Troubleshooting Solves.
USER GUIDE
1596
Troubleshooting Solves | Using Curve Thresholds to Delete Tracks
Troubleshooting Solves
CameraTracker has several troubleshooting workflows available to improve solve accuracy, but
ultimately, good solves rely on good tracking data.
Using Curve Thresholds to Delete Tracks

You can use the threshold controls on the AutoTracks tab to dynamically reject tracks and remove
them to improve accuracy. Using the error - rms curve as an example:
1. Click the AutoTracks tab in the properties and select the error - rms curve.
In the example, you can see that the error curve has increased significantly around frame 195.
2. Move the Viewer playhead to the affected area and examine the tracked features in the Viewer.
3. If you decrease the Max Track Error control in the properties panel, you'll start to see tracks turn
red as they fall below the threshold.
USER GUIDE
1597
Troubleshooting Solves | Using Curve Thresholds to Delete Tracks
Max Track Error set to 6. Max Track Error set to 2.

4. Click Delete Rejected to remove all the tracks that fall below the specified threshold.
Tip: You can also remove tracks manually, by selecting them in the Viewer, right-clicking,
and choosing tracks > delete selected.
5. You can then resolve the affected frame range by clicking Update Solve.
6. Improve the solve further using the Settings tab Solving controls:
• Camera motion - set the camera movement that CameraTracker should account for during the
solve calculation.
• Keyframe Separation - adjusts the separation value between keyframes.
High separation values generate fewer keyframes with a greater spread and are typically used
for slower camera motion.
Low values generate more keyframes with a tighter spread and are typically used for rapid
camera motion.
Note: When enabled, the Reference Frame control determines where the first keyframe is
placed.
• Smoothness - higher values can help smooth a camera path by adding 'weight' to iron out the
path.
• Reference Frame - allows you to specify the start frame for updating camera solves, as well as
determining where the first keyframe is placed.
Note: Check Set reference frame to enable this control.
USER GUIDE
1598
Troubleshooting Solves | Refining a Solve
Refining a Solve
On the AutoTracks tab, in the Refinement section, there are three refinement controls to help
improve your solve. If the Error and per frame controls on the CameraTracker tab show a relatively
high value, try refining the solve using the inlier tracks. The inliers are defined by the curve thresholds
and can refine the focal length, camera position or camera rotation (or a combination of these). You
can manually edit the camera solve first on the Output tab, then select:
• Focal Length - Check to refine the camera’s focal length.
• Position - Check to refine the camera position.
• Rotation - Check to refine the camera rotation.
Finally, click Refine Solve.
USER GUIDE
1599
Updating Solves with Extended Tracking Data | Refining a Solve
Updating Solves with Extended

Tracking Data
CameraTracker allows you to add more frames to an existing camera solve using updated tracking
data. It is usually not possible to extend the range very far (about 10-15% of the original range),
because the existing solve locks the 3D points in place, so matching new 2D tracking data to 3D
points can produce high tracking errors, which CameraTracker rejects.
For example, if you've calculated a solve from frames 1-10, the 3D point for a given track is based on
the data from only those frames. Updating that track by extending it to frame 20 attempts to fit the
2D point from frames 11-20 to the calculated 3D point. This generally doesn't produce a good fit, and
the solve error for the extended range is increased.
To update a solve with updated tracking data:

1. Track the additional frames as described in Extending Existing Camera Tracks.
If there are a lot of rejected tracks, highlighted in red, try using fewer frames in the tracking
update.
2. Click Update Solve to calculate the new tracking data.
If there are a lot of rejected tracks, you may find it easier to retrack the entire range and solve the
camera position from the new, complete data. Re-solving from scratch uses all of the parallax
information over all frames (rather than the initially solved frames), and so you often get a lower
overall solve error.
USER GUIDE
1600
Adjusting the Scene | Refining a Solve
Adjusting the Scene

When the tracking and solving processes are complete, you can use the solve data to align and scale
the scene before adding cameras, point clouds, objects, and so on.
You can:
• Set the ground plane, axes, or origin for the scene. See Setting the Ground Plane and Axes.
• Manually adjust the scene using the Scene tab's Scene Transform controls. See Transforming the
Scene Manually.
• Use known distances, measured on set, to scale a scene more accurately. See Using Scene
Constraints.
USER GUIDE
1601
Setting the Ground Plane and Axes | Refining a Solve
Setting the Ground Plane and

Axes
A solved camera has no notion of where the ground plane is in the scene, which can produce an
unexpected offset and angle relative to where you'd expect the ground to be in 3D space.
Setting the ground plane or set of axes is designed to provide CameraTracker with a sensible frame of
reference. While doing so is not strictly necessary, it can simplify working in the 3D environment.
Before placing the ground plane, scan through your sequence to find a frame with good track and
solve data in an area that you know to be at ground level.
1. Hover over likely points to display the track and solve data, typically points with RMS error less
than 1.
2. Hold Shift and select your chosen points, from multiple frames if necessary.
3. Switch between the 2D and 3D Viewers to confirm the position of the selected points.
Ground plane points selected in the 2D Viewer. The same points viewed in the 3D Viewer.
4. Switch back to the 2D Viewer and right-click a selected point to display the track options menu.
5. Select ground plane > set to selected to orient the camera in relation to the selected points.
Points designated as the ground plane are highlighted in magenta and the camera is reoriented in
relation to the new 'ground'.
USER GUIDE
1602
Setting the Ground Plane and Axes | Setting Axes Individually
Camera orientation before the ground Camera orientation after the ground
plane shift. plane shift.
Setting Axes Individually

In some situations, you may not have a clearly defined ground plane to reference. Using
CameraTracker, you can set the X, Y, and Z axes separately instead of grouping them on the ground
plane.
1. Select two or more points that you know to be on the required axis using the 3D point cloud to
assist you.
2. Right-click a point and select ground plane > set X, Y, or Z as required.
Points designated as an axis are highlighted in yellow and the camera is reoriented in relation to
the new axis.
Setting the Origin Within a Scene

Defining the origin within a layer can assist you when placing objects in your composition - if you
know where 0, 0, 0 is located, you can easily place objects precisely where you want them on any axis.
1. Select a single track as an origin point from your solve data.
2. Right-click the point and select ground plane > set origin.
The point designated as the origin is highlighted in yellow and the point cloud is translated in
relation to the new origin.
USER GUIDE
1603
Setting the Ground Plane and Axes | Setting the Origin Within a Scene
Point before the origin shift. Point after the origin shift.
USER GUIDE
1604
Transforming the Scene Manually | Transforming Using Channel Files
Transforming the Scene Manually

In addition to defining the ground plane, axes, and origin automatically, you can manually adjust the
scene using the Scene tab's Scene Transform controls. These controls operate in exactly the same
way as other transform tools in Nuke.
To transform a scene, use the following controls:

• rotation order - set the operation order for rotations from the dropdown menu which displays all
the possible axial combinations.
• translate - set translate values for the scene. You can adjust translate values by clicking and
dragging the axis in the 3D viewer.
• rotate - set values to rotate your camera or use Ctrl/Cmd to rotate the camera in the Viewer.
• scale - set scale values on individual axes, rather than uniformly or use Ctrl/Cmd+Shift to scale the
camera in the Viewer.
Note: This control is only displayed if you've scaled the scene on individual axes.
• uniform scale - set a scale for your scene. You can also scale the scene using actual measurements
taken on set, see Using Scene Constraints for more information.
Alternatively, you can also use the Local matrix to perform transforms on your scene.
Note: The World matrix is read-only and cannot be edited.
Transforming Using Channel Files

Channel files contain a set of cartesian coordinates for every frame of animation in a given shot. You
can create and export them using Nuke or 3D tracking software, such as 3D-Equalizer, Maya, or
Boujou.
You can transform the scene using channel files by clicking the button:
• Import chan file - import a channel file and transform the input object according to the
transformation data in the channel file.
• Export chan file - export the translation parameters that you’ve applied to the input object as a
channel file. This is a useful method of sharing setups between artists.
USER GUIDE
1605
Transforming the Scene Manually | Transforming Using Snap To Position
Transforming Using Snap To Position

The Snap To menu allows you to match the position, orientation, and scale of existing points.
Select the required points and then click the button:

• Match selection position - the scene is snapped to a new position depending on the points
selected.
• Match selection position, orientation - the scene is snapped to a new position and orientation
depending on the points selected.
• Match selection position, orientation, size - the scene is snapped to a new position, orientation,
and size depending on the points selected.
USER GUIDE
1606
Using Scene Constraints | Transforming Using Snap To Position
Using Scene Constraints

Scene constraints enable you to use known distances, measured on set, to scale a scene more
accurately. They are applicable to both Auto Tracks and User Tracks.
The unit of measure used is up to you, because Nuke equates any value as being equal to one unit in
3D space. The image shows a point cloud and camera with two User Tracks marked out. The blue
square represents one unit of measure, so the smaller squares represent one tenth of that unit. So, if
you measured on set in meters, one small square could be equal to a meter, centimeter, or
millimeter.
To create scene constraints:

1. Select two solved points in the Viewer.
2. Right-click a highlighted point and select scene > add scale distance.
The image shows User Tracks, but the same workflow applies to Auto Tracks. For example, the
Auto Tracks located on the window panes could easily be measured.
USER GUIDE
1607
Using Scene Constraints | Transforming Using Snap To Position
A connecting line, labled dist. 1, is drawn in the Viewer.

3. In the properties panel, click the Scene tab to display the Scale Constraints table.
4. Click the distance field for dist. 1 and enter the measured distance.
The measured distance is now taken into account when you adjust the Scene Transform
controls.
USER GUIDE
1608
Using Solve Data | Transforming Using Snap To Position
Using Solve Data

Once your camera solve is correctly aligned, you can create animated or baked cameras, multi-view
rigs, 3D scenes and point cloud, lens distortion, or cards from the camera data. See:
• Creating Camera Nodes.
• Creating Scenes.
• Creating Point Clouds.
• Creating Cards.
USER GUIDE
1609
Creating Camera Nodes | Creating a Camera Rig
Creating Camera Nodes

Using the tracking and solve data, CameraTracker can create linked or baked cameras that emulate
the original camera track on set. You can create a single camera or a camera rig for multi-view
projects, depending on your script.
1. Select Camera from the Export dropdown menu.
2. Enable or disable the Link output control to determine whether the Camera node is expression
linked or baked:
• When enabled, CameraTracker creates an expression linked camera so that any adjustments
made in the properties panel Output > Camera controls update the camera.
• When disabled, any adjustments made are ignored by the camera.
3. Click Create.
A Camera node is added to the Node Graph. Expression linked cameras are linked to the
CameraTracker node with a green expression arrow.
Camera node data can be used in various Nuke workflows, such as Creating Dense Point Clouds
and Generating Depth Maps.
Creating a Camera Rig

CameraTracker creates as many cameras as you have views in you script, but the most common use
of rigs is for stereoscopic scripts with left and right views.
1. Select Camera rig from the Export dropdown menu.
2. Enable or disable the Link output control to determine whether the Camera nodes are
expression linked or baked:
• When enabled, CameraTracker creates expression linked cameras so that any adjustments
made in the properties panel Output > Camera controls update the cameras.
• When disabled, any adjustments made are ignored by the cameras.
USER GUIDE
1610
Creating Camera Nodes | Creating a Camera Set
3. Click Create.
Note: If you try to create a rig with only one view, an error is displayed.
A Camera node is added to the Node Graph for each view present in the script. Expression linked
cameras are linked to the CameraTracker node with a green expression arrow.
Camera node data can be used in various Nuke workflows, such as Creating Dense Point Clouds
and Generating Depth Maps.
Creating a Camera Set

CameraTracker can automatically create a camera for each solved frame.
1. Select Camera set from the Export dropdown menu.
2. Enable or disable the Link output control to determine whether the cameras are expression
linked or baked:
3. Click Create.
If you're processing a large amount of frames, a confirmation dialog displays.
4. Click Yes to continue or No to cancel the export.
CameraTracker adds a Group node to the Node Graph containing Camera nodes for every frame
specified connected to a Scene node.
5. Double click the Group to open the properties panel, then click the S above the panel to open the
group in the Node Graph.
USER GUIDE
1611
Creating Camera Nodes | Creating a Camera Set
You could then write the camera data to an .fbx file ready for import into Modo or Maya for
projection painting work.
USER GUIDE
1612
Creating Scenes | Creating a 3D Scene with ScanlineRender and LensDistortion
Creating Scenes
CameraTracker can create a ready-to-use 3D scene containing a point cloud, camera, and Scene node
from the track and solve data. The Scene+ option adds LensDistortion and ScanlineRender nodes in
addition to the standard scene nodes.
1. Select Scene from the Export dropdown menu.
2. Enable or disable the Link output control to determine whether the scene's Camera node is
• When enabled, CameraTracker creates an expression-linked camera so that any adjustments
made in the properties panel's Output > Camera controls update the camera.
3. Click Create.
CameraTracker adds Camera, CameraTrackerPointCloud, and Scene nodes to the Node Graph.
Expression linked cameras are linked to the CameraTracker node with a green expression arrow.
Creating a 3D Scene with ScanlineRender and

LensDistortion
1. Select Scene+ from the Export dropdown menu.
2. Enable or disable the Link output control to determine whether the scene's Camera node is
• When enabled, CameraTracker creates an expression-linked camera so that any adjustments
made in the properties panel's Output > Camera controls update the camera.
3. Click Create.
USER GUIDE
1613
Creating Scenes | Creating a 3D Scene with ScanlineRender and LensDistortion
CameraTracker adds Camera, CameraTrackerPointCloud, Scene, ScanlineRender, and

LensDistortion nodes to the Node Graph. The LensDistortion node is set to undistort the 2D
footage in CameraTracker's Source input and use the result as the background for the 3D scene.
Expression-linked cameras are linked to the CameraTracker node with a green expression arrow.
USER GUIDE
1614
Creating Point Clouds | Creating a 3D Scene with ScanlineRender and LensDistortion
Creating Point Clouds

CameraTracker can create a baked point cloud, that is, points that don't update when you make
changes to the CameraTracker properties panel controls.
1. Select Point cloud from the Export dropdown menu.
2. Click Create.
CameraTracker adds a CameraTrackerPointCloud node to the Node Graph.
3. Double-click the CameraTrackerPointCloud node to open the properties panel.
4. Set how the points should appear, both in the Viewer and when rendered out, using the display
and render dropdowns:
• off - hides the 3D points.
• wireframe - displays only the points.
• solid - displays all points using the color of the pixel they represent.
• solid+wireframe - displays all points using the color of the pixel they represent and the points
themselves.
• textured - displays only the surface texture.
• textured+wireframe - displays the the surface texture and the points.
5. Enable or disable shadow casting and receiving when the point cloud is rendered to 2D using the
checkboxes.
See Casting Shadows for more information on lighting and casting shadows.
6. Set the size of the rendered points using the Point Size control.
USER GUIDE
1615
Creating Cards | Creating a 3D Scene with ScanlineRender and LensDistortion
Creating Cards
CameraTracker can automatically create a 3D card for each solved frame using the camera to project
the image at that frame onto the card.
1. Select Cards from the Export dropdown menu.
2. Enable or disable the Link output control to determine whether the cameras are expression
linked or baked:
3. Click Create.
If you're processing a large amount of frames, a confirmation dialog displays.
4. Click Yes to continue or No to cancel the export.
CameraTracker adds a Group node to the Node Graph containing Card, FrameHold, and Camera
nodes for every frame specified connected to a Scene node.
5. Double click the Group to open the properties panel, then click the S above the panel to open the
group in the Node Graph.
6. Use the Group node's z slider to control the card distance from the camera. You can use this
setting to create a pan and tile dome at this distance from the camera.
The Cards are scaled automatically using the camera settings.
USER GUIDE
1616
Creating Cards | Creating a 3D Scene with ScanlineRender and LensDistortion
The z depth set to 1, the default. The z depth set to 3.
USER GUIDE
1617
Combining Solves | Creating a 3D Scene with ScanlineRender and LensDistortion
Combining Solves
Sometimes, you might have multiple footage sources of the same scene or content available. For
example, you might have footage from witness cameras, or someone has taken detailed still photos
of the scene. CameraTracker provides a way to solve each of your sources and then register them all
in the same world. This allows you to leverage a high-quality camera track from a secondary source
that is good to solve from, and use it in another source that's difficult to solve. You can also use this
technique to tie detailed close-up stills to wide-angle shots.
The key idea is that you pick one of your sources as the 'master' solve. This should be footage that
has good parallax, for example a set of wide-angle stills taken on set. You also need some geometry
created from the solve data. The geometry is to establish 3D survey points, which are then used to tie
together the 'master' and the satellite camera solves.
The workflow is something like this:

1. Pick one of your sources as the 'master' solve and use CameraTracker to track and solve the
camera.
2. Create a linked camera using CameraTracker's Export menu. See Creating Camera Nodes.
3. Use the camera data to create some reference geometry. Either:
• connect the Camera straight into the ModelBuilder node to create the geometry, or
• export the camera data, create the geometry in an external application, then import the
geometry back into Nuke.
See Using ModelBuilder or Exporting Geometry, Cameras, Lights, Axes, or Point Clouds .
4. Set some 3D survey points on the geometry in the 'master' solve. See Assigning 3D Survey Points.
5. Create a separate CameraTracker for each of your satellite cameras, and track the footage as you
see fit.
6. Create User Tracks in the satellite cameras, corresponding to the 3D Survey points in the 'master'
solve, making sure to also mark them as survey points. See Assigning 3D Survey Points.
7. Solve the satellite cameras to align them consistently with the 'master' solve.
Tip: You can copy the X, Y, Z position from a track in one CameraTracker node to another by
right-clicking on a track and selecting copy > translate.
USER GUIDE
1618
Placing Objects in the Scene | Creating a 3D Scene with ScanlineRender and LensDistortion
Placing Objects in the Scene

You can use the camera and point cloud to add geometry to the scene. You can add objects manually,
but placing them inside the camera’s field of vision at the desired position can be time consuming.
CameraTracker provides an automatic creation function to help you achieve the results you need.
1. In the 2D view, select points on the plane to hold the geometry. For example, you might place a
card on a vertical or horizontal plane.
Tip: You might find that swapping between the 2D plate and 3D point cloud helps to locate
potential points.
2. Drag a marquee over the required points or hold down the Shift key and click individual points.
3. Right-click a selected point in the 2D Viewer and choose the create menu to display the available
geometry.
4. Select the required shape to place it in the scene using the average position of all selected points.
The following example shows two cards placed in the scene using points from the vertical and
horizontal planes.
USER GUIDE
1619
Accounting for Lens Distortion | Distorting Your CG Elements to Match Your 2D Footage
Accounting for Lens Distortion

By default, any CG elements you add to your 3D scene do not have lens distortion applied to them. As
a result, when you combine them with your 2D footage, they can look like they weren't shot with the
same camera. To fix this, you can:
• Use CameraTracker to calculate the lens distortion on your 2D footage and generate a
LensDistortion node that applies the same distortion to your CG elements. See Distorting Your CG
Elements to Match Your 2D Footage.
• Use CameraTracker to calculate the lens distortion on your 2D footage and generate a
LensDistortion node that undistorts the 2D footage to match your CG elements. See Undistorting
Your 2D Footage to Match Your CG Elements (Using LensDistortion).
• Use CameraTracker to both calculate the lens distortion on your 2D footage and undistort the
footage to match your CG elements. See Undistorting Your 2D Footage to Match Your CG Elements
(Using CameraTracker).
Note: The above assumes you set LensDistortion to Unknown Lens before creating the
solve. If you set Lens Distortion to No Lens Distortion and used a separate LensDistortion
node to remove lens distortion from your footage before camera tracking, you can use the
same LensDistortion node to either apply the distortion to your CG elements or to undistort
your 2D footage. See Working with Lens Distortion for more information on using the
LensDistortion node in Nuke.
Distorting Your CG Elements to Match Your 2D Footage

1. In the CameraTracker properties, set the Export menu to Distortion and click Create.
CameraTracker creates a LensDistortion node that is set to apply the lens distortion that exists in
your 2D footage.
2. Connect the LensDistortion node to the output of your ScanlineRender node.
The LensDistortion node distorts the CG elements in your 3D scene to match your 2D footage.
3. Composite your CG elements and 2D footage together as necessary.

(Using LensDistortion)
1. In the CameraTracker properties, set the Export menu to Undistortion and click Create.
USER GUIDE
1620
Accounting for Lens Distortion | Undistorting Your 2D Footage to Match Your CG Elements
CameraTracker creates a LensDistortion node that is set to remove the lens distortion that exists
in your 2D footage.
2. Connect the LensDistortion node to the output of your 2D footage.
The LensDistortion node removes the lens distortion in your 2D footage to match your CG
elements.
3. Connect the LensDistortion node to the bg input of ScanlineRender.
The undistorted 2D footage is used as a background image for the 3D scene.

(Using CameraTracker)
1. In the CameraTracker properties, enable Undistort Input.
CameraTracker removes the lens distortion that exists in its input footage.
2. Connect the CameraTracker node to the bg input of ScanlineRender.
The undistorted 2D footage is used as a background image for the 3D scene.
USER GUIDE
1621
About the Lens Model | Undistorting Your 2D Footage to Match Your CG Elements
About the Lens Model

Normally, you do not need to set the Lens parameters on the Output tab, as CameraTracker does
this for you, but it may be helpful to know a little about the equations CameraTracker uses to account
for lens distortion. There are two modes, depending on the lens type detected:
• Spherical - compensates for different spherical lenses. This is the simpler of the two lens
corrections and uses the following equation:
In normalized coordinates from the distortion center, Xu and Xd are equal to the same point in an
undistorted plate and a distorted plate respectively.
d1 is equal to Radial Distortion 1.
r2 is equal to the distance of the point from the Distortion Center.
r4 is equal to the square of r2.

• Anamorphic - compensates for anamorphic lenses. Anamorphic correction uses three additional
parameters (Ax, Ay, and Asq) and requires two equations because, unlike with spherical lenses, the
amount of distortion parallel to the x-axis is not the same as that parallel to the y-axis.
In normalized coordinates from the distortion center, (xu, yu) and (xd, yd) are equal to the same
point in an undistorted plate and a distorted plate respectively.
r2 is equal to the distance of the point from the Distortion Center.
r4 is equal to the square of r2.
Ax is equal to Asymmetric Distortion X.
USER GUIDE
1622
About the Lens Model | Undistorting Your 2D Footage to Match Your CG Elements
Ay is equal to Asymmetric Distortion Y.
Asq is equal to Anamorphic Squeeze.
USER GUIDE
1623
Using MatchGrade
This chapter teaches you how to use MatchGrade to automatically calculate a grade to match the
colors in the Source input to the colors in the Target input. You can use MatchGrade to:
• extract a baked-in grade if the Target clip that you want to match is simply a graded version of the
Source clip. See Extracting a Baked-In Grade.
• match the grade between two different clips to create the same look. See Matching a Grade Between
Two Different Clips.
In both cases, you can also mask the grade to only match certain elements between the Source and
Target clips, and export LUT or CDL files to re-use the calculated grade elsewhere.
The Source image. The Target image.
The color-matched result.
Special thanks to The Mill for use of the above footage, used throughout this chapter.
USER GUIDE
1624
Extracting a Baked-In Grade |
Extracting a Baked-In Grade

1. Select Color > MatchGrade to insert a MatchGrade node into your script.
2. Do the following:
• Connect MatchGrade's Source input to the clip to which you want to apply the color transform.
• Connect the Target input to the clip you want to match. This should be a graded version of the
Source.
• Connect a Viewer to the output of MatchGrade.
3. In the MatchGrade properties, make sure Task is set to Match Graded Source.
4. To extract a baked-in grade, MatchGrade requires pixel-to-pixel correspondence between the
Source and Target clips. If the clips are not aligned in time and space, such as when the target
clip has been reformatted, click Align Target to Source to add a Transform node and a Reformat
node upstream of the MatchGrade node.
USER GUIDE
1625
Extracting a Baked-In Grade | Analyzing Reference Frames
Tip: If the target contains a region that isn't present in the source, for example a black
border, you can enable Crop Target. Clicking Align Target to Source when Crop Target is
enabled generates a rectangular crop for the Target input.
5. If you want MatchGrade to calculate a single, global grade from selected reference frames that
cover the characteristic colors in the sequence and apply that grade to every frame of the Source
sequence, proceed to Analyzing Reference Frames on page 1626. In this mode, you can export the
grade to an OCIOFileTransform or OCIOCDLTransform node to re-use it elsewhere in the script.
If you'd rather have MatchGrade calculate a local grade frame-by-frame, so that the color
transform updates according to the current frame, proceed to Auto-Analyzing Per Frame on page
1629. In this mode, you cannot export the grade to re-use it elsewhere.
Analyzing Reference Frames

1. In the MatchGrade properties, make sure Analysis is set to Analyze Reference Frames.
2. If you only want to color match certain areas of the image, make sure there is a mask in the
Source input's alpha channel. Then, set Mask to either:
• Alpha - to use the alpha channel as a mask. The grade is limited to the non-black areas of the
alpha channel.
• Inverted Alpha - to invert the alpha channel and use that as a mask. The grade is limited to the
non-white areas of the alpha channel.
Note: When the Mask control is set to anything other than None, an additional Apply
Grade to Masked Region control is displayed.
This control is enabled by default, but you can disable it to apply the grade to the whole
image instead, allowing you to compute the grade from a selected region and apply it to the
whole image without having to export the LUT.
3. Set Output to Target.

MatchGrade displays the Target footage in the Viewer.
4. To select reference frames for MatchGrade to use in the analysis, play through the sequence.
When you find frames that cover the characteristic colors in the sequence, click on the button
to set the current frame as a reference frame. If necessary, you can also use the button to
remove a frame from the set of reference frames.
Typically, the more reference frames you set, the better MatchGrade is able to match the colors.
USER GUIDE
1626
Tip: You can also set Target to Matched and Analysis to Auto-analyze Per Frame to
preview where to set reference frames.
5. When you're done setting reference frames, use the Transform dropdown menu to choose how
to calculate the grade:
• 3D LUT - Calculate the grade as a 3D look-up table (LUT). This allows you to export the grade to a
.csp format, which you can use with the OCIOFileTransform node.
• CDL - Calculate the grade as a color decision list (CDL). This allows you to export the grade to an
OCIOCDLTransform node.
Note: The CDL transform is limited and cannot model all types of color transformation. In
most cases, selecting 3D LUT gives the best results.
6. Set Output back to Matched and click Analyze Reference Frames.

MatchGrade calculates the transform needed to match the Source image to the Target image.
7. To evaluate the results, select the MatchGrade node and press D repeatedly to toggle between the
original Source image and the graded output.
USER GUIDE
1627

8. If you're not happy with the results, try one or more of the following:
• Set more reference frames.
• Use the Pre LUT dropdown menu to specify a 1D shaper LUT to use for the analysis: Linear or
Logarithmic.
• Increase the LUT Resolution value. This is the resolution of the look-up table (LUT) in which
MatchGrade stores the color transformation. Higher values can improve the results, but also
increase processing time. The maximum value is 64.
To update the results, click Analyze Reference Frames again.
9. If you set Transform to CDL in step 5, you can fine-tune the grade manually by adjusting slope,
offset, power, and saturation. The lock controls on the right prevent the values from being
recalculated when you click Recalculate CDL, allowing you to set some values manually and
estimate others automatically. For example, you can adjust and lock slope and offset and then
click Recalculate CDL to automatically recalculate power and saturation.
10. If necessary, you can export the calculated grade and apply it to other nodes in your script. How
you do this depends on whether you chose to calculate the grade as a 3D LUT or a CDL:
• If you set Transform to 3D LUT, you can write the LUT into a .csp file and create an
OCIOFileTransform node that you can use elsewhere in the script to apply the same grade. To do
so, enter a file path and name for the .csp file in the LUT output file field and click Write to
export the LUT. Then, click Create OCIOFileTransform to create an OCIOFileTransform node
that applies the color transform from the .csp file.
USER GUIDE
1628
Extracting a Baked-In Grade | Auto-Analyzing Per Frame
• If you set Transform to CDL, you can click Create OCIOCDLTransform to create an
OCIOCDLTransform node that uses the values in the CDL Output section of the MatchGrade
properties.
Auto-Analyzing Per Frame

1. In the MatchGrade properties, set Analysis to Auto-analyze Per Frame.
2. If you only want to color match certain areas of the image, make sure there is a mask in the
Source input's alpha channel. Then, set Mask to either:
• Alpha - to use the alpha channel as a mask. The grade is limited to the non-black areas of the
alpha channel.
• Inverted Alpha - to invert the alpha channel and use that as a mask. The grade is limited to the
non-white areas of the alpha channel.
Alternatively, you can use the Output dropdown menu in the MatchGrade properties to choose
what to display in the Viewer:
• Matched - View the color matched result.
• Source - View the Source input.
USER GUIDE
1629
Extracting a Baked-In Grade | Auto-Analyzing Per Frame
• Target - View the Target input.

4. If you're not happy with the results, try increasing the LUT Resolution value. This is the resolution
of the look-up table (LUT) in which MatchGrade stores the color transformation.
Higher values can improve the results, but also increase processing time. The maximum value is
64.
5. If necessary, you can also use the Pre LUT dropdown menu to specify a 1D shaper LUT to use for
the analysis:
• Auto Detect - Automatically detect the best pre-LUT to use.
• Linear - Use a linear pre-LUT.
• Logarithmic - Use a logarithmic pre-LUT.
USER GUIDE
1630
Matching a Grade Between Two Different Clips | Auto-Analyzing Per Frame
Matching a Grade Between Two

Different Clips
1. Select Color > MatchGrade to insert a MatchGrade node into your script.
2. Do the following:
• Connect MatchGrade's Source input to the clip to which you want to apply the color transform.
• Connect the Target input to the clip you want to match. This should be a completely different
clip, a stereo pair for the Source clip, or the Source clip in a different format.
• Connect a Viewer to the output of MatchGrade.
3. In the MatchGrade properties, set Task to Match Different Clip.

4. Set Output to Source and then Target to compare the Source and Target images.
5. When you find frames that cover similar content in the Source and Target clips, click on the
buttons next to Source and Target to set the current frame as a reference frame. Only frames set
as reference frames are used in the MatchGrade analysis.
For example, if you want grass in the Source clip to have the same color as grass in the Target
clip, you need to choose frames where the relative amount of grass pixels is roughly the same.
The more similar the content, the better MatchGrade is able to match the colors.
Note: You do not need to use the same number of Source and Target reference frames or
set the same frames as reference frames for both.
USER GUIDE
1631
If necessary, you can also use the button to remove a frame from the set of reference frames.
6. If you only want to match certain elements between the Source and Target clips, supply a mask
in both the Source and Target inputs' alpha channels and set Mask to either:
• Alpha - to use the alpha channel from both inputs as a mask. Only the non-black areas of the
Source and Target inputs' alpha channels are used in the MatchGrade analysis, and the grade is
limited to the non-black areas of the Source alpha.
• Inverted Alpha - to use the inverted alpha channel from both inputs as a mask. Only the non-
white areas of the Source and Target inputs' alpha channels are used in the MatchGrade
analysis, and the grade is limited to the non-white areas of the Source alpha.
Masks can also be useful if you cannot find frames that cover approximately the same amount of
similar content. For example, if the Source and Target inputs have a different amount of grass
and sky, you can use a Roto node in both inputs to create masks that cover the same amount of
grass and sky. Bear in mind that if you do so, the grade is only applied to the areas indicated by
the mask in the Source input. To apply it to the entire frame, you need to export the grade first.
For more information, see step 12.
The Source image. In this case, you can ...and draw a mask like this on the Target
set this image to have a fully image to ensure both inputs cover
opaque alpha... similar content.
7. When you're done setting reference frames, use the Transform dropdown menu to choose how
to calculate the grade:
• 3D LUT - Calculate the grade as a 3D look-up table (LUT). This allows you to export the grade to a
.csp format, which you can use with the OCIOFileTransform node.
• CDL - Calculate the grade as a color decision list (CDL). This allows you to export the grade to an
OCIOCDLTransform node.
USER GUIDE
1632
Note: The CDL transform is limited and cannot model all types of color transformation. In
most cases, selecting 3D LUT gives the best results.
8. Set Output back to Matched and click Analyze Reference Frames.


10. If you're not happy with the results, try one or more of the following:
• Look for better reference frames or use a mask in the Target input to make sure both inputs
cover approximately the same amount of similar content.
• Use the Pre LUT dropdown menu to specify a 1D shaper LUT to use for the analysis: Linear or
Logarithmic.
• Increase the LUT Resolution value. This is the resolution of the look-up table (LUT) in which
MatchGrade stores the color transformation. Higher values can improve the results, but also
increase processing time. The maximum value is 64.
USER GUIDE
1633
• Set Colorspace to a different value. The correct setting to use depends on the nature of the
transformation. Try each option to see which works best with your footage.
• Increase the Iterations value. This determines the number of refinement passes. Higher values
can produce a better color match, but also take longer to process.
To update the results, click Analyze Reference Frames again.
11. If you set Transform to CDL in step 8, you can fine-tune the grade manually by adjusting slope,
offset, power, and saturation. The lock controls on the right prevent the values from being
recalculated when you click Recalculate CDL, allowing you to set some values manually and
estimate others automatically. For example, you can adjust and lock slope and offset and then
click Recalculate CDL to automatically recalculate power and saturation.
12. If necessary, you can export the calculated grade and apply it to other nodes in your script. How
you do this depends on whether you chose to calculate the grade as a 3D LUT or a CDL:
• If you set Transform to 3D LUT, you can write the LUT into a .csp file and create an
OCIOFileTransform node that you can use elsewhere in the script to apply the same grade. To do
so, enter a file path and name for the .csp file in the LUT output file field and click Write to
export the LUT. Then, click Create OCIOFileTransform to create an OCIOFileTransform node
that applies the color transform from the .csp file.
• If you set Transform to CDL, you can click Create OCIOCDLTransform to create an
OCIOCDLTransform node that uses the values in the CDL Output section of the MatchGrade
properties.
USER GUIDE
1634
Tip: You can also apply the grade to a different source by simply connecting a new node to
MatchGrade's Source input.
USER GUIDE
1635
Using the Smart Vector Toolset
The Smart Vector toolset allows you to work on one frame in a sequence and then use motion vector
information to accurately propagate work throughout the rest of the sequence. The vectors are
generated in the SmartVector node and then piped into the VectorDistort node to warp the work.
That way, you only need to render the motion vectors once.
Note: All images in this chapter are courtesy of Hollywood Camera Work. See
www.hollywoodcamerawork.com for more information.
Quick Start
The toolset workflow is outlined below:
1. Connect the SmartVector node to the source sequence to generate the required motion vectors.
You can also write them to file using the .exr format. See Generating Motion Vectors.
2. Find a suitable reference frame and add the required paint corrections or an image to the frame.
See Adding Paint to the Source or Adding an Image to the Source.
3. Connect the VectorDistort node to the SmartVector node (or a Read node referencing the .exr
files rendered from the SmartVector node) and the source sequence.
4. Premultiply the VectorDistort output and merge it over the source sequence to complete the
comp. See Applying Motion Vectors to the Source.
USER GUIDE
1636
Generating Motion Vectors |
Generating Motion Vectors
The SmartVector node generates motion vectors for use in the VectorDistort node. You can connect a
SmartVector node directly to the VectorDistort or VectorCornerPin nodes or write motion vectors to
the .exr format, which are then used to drive the warp nodes to reduce overheads later on.
To generate motion vectors:

1. Read in your source sequence and then connect a SmartVector node to the Read node.
2. Double-click the SmartVector node to open its Properties panel, if it's not already open.
3. Set the Vector Detail control to the required detail level. The default value of 0.3 is sufficient for
sequences with low detail and movement, but you may want to increase the detail to improve the
vector quality in some cases.
For example, if the area you're working on is relatively small, the default value of 0.3 might not
capture movement properly. Try increasing the control to 1.0 to capture more detail.
Note: High detail vectors take longer to render, but can improve the results you get from
the VectorDistort node.
4. Set the Strength control to force pixel matching between frames. Higher values allow you to
accurately match similar pixels in one image to another, concentrating on detail matching even if
the resulting motion field is jagged. Lower values may miss local detail, but are less likely to
provide you with the odd spurious vector, producing smoother results.
Note: The default value works well for most sequences.
5. If there is a lot of movement in the foreground, you might want to add a mask. See Masking
Foreground and Background Areas for more information.
Tip: You can examine the vectors produced by connecting a Viewer to the SmartVector node
and switching the channels control above the Viewer to smartvector_<value>.
6. If you want to write the vectors to disk, click Export Write to automatically add a Write node to
the script. The Write node's controls are set to channels > all and .exr output automatically.
You can only write motion vectors to the .exr format. Don't forget to add frame padding in the
form of hashes or printf notation, depending on your Preferences > Behaviors > File Handling
settings.
USER GUIDE
1637
Generating Motion Vectors | Masking Foreground and Background Areas
7. Enter a file path in the Write node's controls and then click Render.
Note: SmartVector does not currently output motion, forward, and backward channels by
default. If you require these channels, add a VectorToMotion node after the SmartVector
node. VectorToMotion converts the vectors to motion that can be used with VectorBlur to
create motion blur, without using a VectorGenerator.
8. Enable Advanced > Flicker Compensation to reduce variations in luminance and overall

flickering, such as light reflecting from shiny surfaces, which can cause problems with your vector
output.
9. Tolerances in the Advanced section allow you to tune the weight of each color channel when
calculating the image luminance.
These parameters rarely need tuning, but you may wish to increase the weighting on a particular
color to allow the algorithm to concentrate on getting the motion for a certain object correct, at
the cost of the rest of the items in a shot.
10. Proceed to Adding Paint to the Source or Adding an Image to the Source.
Masking Foreground and Background Areas

If your sequence is composed of a foreground object moving over a background, the motion
estimation is likely to get confused at the edge between the two. To reduce artifacts, you can add an
alpha mask to the Matte input.
To apply a mask to the sequence:

1. Add a Roto node downstream of the source.
2. Draw a matte around the area you want to mark as foreground or background.
USER GUIDE
1638
Generating Motion Vectors | Masking Foreground and Background Areas
3. Set the Matte Channel to Matte Alpha and the Output control to either Foreground or
Background.
Foreground vectors Background vectors

4. If you're working with Background vectors, you can enable Inpaint Matte Region to
automatically fill in the missing foreground vectors using the nearest available background
vectors. The Matte Dilation slider controls amount of dilation applied to the matte before
inpainting is applied.
Background vectors Inpainted foreground

You can use inpainting to effectively ignore the foreground, allowing you to work with background
vectors in the masked area.
5. Generate the motion vectors as described in Generating Motion Vectors.
USER GUIDE
1639
Adding Paint to the Source | Masking Foreground and Background Areas
Adding Paint to the Source

The Smart Vector toolset minimizes the paint work needed to create your final comp by propagating
paint through a sequence from a single source frame.
To add paint to your sequence:

1. Scrub through your sequence to find a good reference frame. Good reference frames should:
• be at the point of least motion, also known as the motion apex,
• contain minimal motion blur, and
• avoid areas of occlusion.
2. Add a RotoPaint node and apply your paint corrections to the selected reference frame. Ensure
that all your paint is sourced from the background image by navigating to Stroke > source and
selecting background once a stroke has been added.
Original image Premultiplied background paint
Merged result
USER GUIDE
1640
Adding Paint to the Source | Masking Foreground and Background Areas
3. Proceed to Applying Motion Vectors to the Source.
USER GUIDE
1641
Adding an Image to the Source | Masking Foreground and Background Areas
Adding an Image to the Source

The Smart Vector toolset can also propagate an image through a sequence from a single source
frame.
To add an image to your sequence:

1. Scrub through your sequence to find a good reference frame. Good reference frames should:
• be at the point of least motion, also known as the motion apex,
• contain minimal motion blur, and
• avoid areas of occlusion.
2. Merge your image into the node tree at the selected reference frame.
Original image Premultiplied image
Merged result
3. Proceed to Applying Motion Vectors to the Source.
USER GUIDE
1642
Applying Motion Vectors to the Source | Warping the Vectors
Applying Motion Vectors to the

Source
Once you've generated your motion vectors and added your paint to the source sequence, the
VectorDistort node takes the paint from the reference frame and propagates it through the rest of the
sequence using the motion vectors from the SmartVector node.
Warping the Vectors

After generating your vectors, applying your corrections, and masking out areas of the sequence you
don't want to process, use VectorDistort to push your work through the rest of the sequence.
1. Add a VectorDistort node to the Node Graph and connect the Source input to your sequence,
downstream of the paint or image you intend to propagate.
2. Connect the SmartVector input to either the SmartVector node or a Read node referencing the
.exr files containing the motion vector data.
3. Double-click the VectorDistort node to open its Properties panel, if it's not already open.
4. Scrub to the frame containing your paint or image and click set to current frame to set the
VectorDistort reference frame.
Tip: If you know what frame your paint or image is applied to, you can type the frame
number into the Reference Frame control manually.
Scrubbing in the Viewer at this stage introduces undesirable warping. As you move farther from
the reference frame, the warping increases.
USER GUIDE
1643
Warp close to the reference frame. Warp far from the reference frame.
5. Set the maximum Frame Distance over which vectors are calculated. Sequences with rapid
motion typically require values closer to 1 frame, whereas sequences with slower motion typically
require values closer to 64 frames.
See Improving Warps for more information on the Frame Distance control.
6. To apply the warped paint or image to the original sequence, premultiply the output from the
VectorDistort nodes and then merge the result over the source.
For short sequences, or sequences with minimal movement and detail, your corrections are
propagated nicely as shown.
USER GUIDE
1644
If you see that your corrections warp incorrectly over time, you can use VectorDistort's properties to
massage the results (see Improving Warps) or use the VectorCornerPin node or multiple
VectorDistort nodes to apply warps for several reference frames (see Warping Images Using
VectorCornerPin and Warping Multiple Reference Frames).
USER GUIDE
1645
Improving Warps | Warping the Vectors
Improving Warps
For short sequences, or sequences with minimal movement and detail, your corrections are
propagated nicely. If you see that your corrections warp incorrectly over time, you can use
VectorDistort's Frame Distance control to massage the results.
Tip: Warping relies on good vectors from SmartVector and a good reference frame in
VectorDistort. Before doing anything else, try increasing the Vector Detail control in the
SmartVector node and re-rendering the vectors. You can also try selecting a different
Reference Frame in the VectorDistort node's controls.
See Generating Motion Vectors or Applying Motion Vectors to the Source for more

information.
At Frame Distance > 1 frame, VectorDistort calculates warp for every frame in the sequence, that is,
frames 1-2, 2-3, 3-4, 4-5, and so on. For frames close to the Reference Frame with lots of movement
between frames, this is a good thing as the warp needs to change significantly from frame to frame.
However, for frames farther away from the Reference Frame with little movement between frames,
this isn't required as the warp doesn't change significantly from frame to frame.
Increasing the Frame Distance essentially reduces the number of warps calculated between the
Reference Frame and the current frame. In the example sequence, the Reference Frame is set to
57, and the warp at frame 230 has started to slip on the subject's forehead with Frame Distance set
to 1 frame. Increasing the Frame Distance works well up to around 2 frames, but increasing the
distance distorts the warp too much.
The warp with frame distance 1 frame. The warp with frame distance 2 frames.
USER GUIDE
1646
Improving Warps | Warping the Vectors
The warp with frame distance 4 frames.
For longer sequences with local distortion, you can try increasing the Blur Size control to blur the
internally calculated STMap. Increasing the Blur Size can remove local distortions in the warped
result, particularly in longer sequences.
In some cases, no amount of adjustment is going to improve the warp over a large number of frames
- particularly if there's a lot of movement or detail in the sequence. See Warping Images Using
VectorCornerPin and Warping Multiple Reference Frames for more information on how to use
VectorCornerPin and multiple VectorDistort nodes to minimize manual paint work.
USER GUIDE
1647
Warping Images Using VectorCornerPin | Warping the Vectors
Warping Images Using

VectorCornerPin
Some sequences involve movement and detail that can't be propagated from a single frame
correction. In these cases, you can use the VectorCornerPin node to minimize the amount of
correction work needed. The to1-4 Viewer widgets allow you to add keyframes, which then use vector
information from SmartVector to drive the warp.
The final result should be an image with minimal slip and distortion.
1. Set up your image as described in Adding an Image to the Source.

2. Connect the VectorCornerPin's SmartVector input to either a SmartVector node or a Read node
referencing the .exr files containing the motion vector data.
USER GUIDE
1648
Tip: If the image you're warping displays at the wrong resolution, switch to the
VectorCornerPin's From tab, click Set To Input, and then switch back to the
VectorCornerPin tab and click Copy 'from' to resize it automatically.
3. Scrub to the reference frame in your sequence and place the image as required using the to1-4
Viewer widgets.
USER GUIDE
1649
4. In the VectorCornerPin Properties panel, click add keyframe to set a keyframe at the current
frame.
Note: You can remove the keyframe at the current frame using or remove all keyframes
from the sequence using .
5. Play through the sequence until the image begins to slip or warp.
Tip: After adding a keyframe, you can click Bake Corners to calculate the positions of the
pins at each frame in a range using the vectors from the SmartVector input.
USER GUIDE
1650
6. Adjust the image using the to1-4 Viewer widgets. A new keyframe is added at the current frame.
Frames on which no keyframe is present are interpolated automatically, so you may not need to
adjust each frame individually.
7. You can smooth distortion on interpolated frames using the Frame Distance and Blur Size
controls on the SmartVector tab.
Frame Distance 1 frame and Blur Size 0.
USER GUIDE
1651
Frame Distance 4 frames and Blur Size 10.
USER GUIDE
1652
Warping Multiple Reference Frames | Warping the Vectors
Warping Multiple Reference

Frames
Some sequences involve movement and detail that can't be propagated from a single frame
correction. In these cases, you can use multiple VectorDistort nodes with different reference frames
to minimize the amount of correction work needed.
1. Set up your first paint correction as described in Adding Paint to the Source and Applying
Motion Vectors to the Source.
2. In the VectorDistort node's properties, disable the Hold Frame control.
3. Add a FrameHold node before the RotoPaint node containing your corrections.
4. Set the first frame control to the reference frame specified in the VectorDistort node.
5. Set the lifetime of your paint strokes in the RotoPaint node Lifetime tab to the same frame range
as the VectorDistort node.
6. Repeat the process for the required number of reference frames to complete your corrections. An
example node tree might appear as follows:
USER GUIDE
1653
Warping Multiple Reference Frames | Warping the Vectors
USER GUIDE
1654
Generating Depth Maps
You can use the DepthGenerator node in NukeX to generate a depth map from your footage. The
node uses information from a tracked camera to create a channel that displays variations in depth.A
depth map is an image that uses the brightness of each pixel to specify the distance between the 3D
scene point and the virtual camera used to capture the scene.
A source image. The generated depth map.
DepthGenerator also allows you to output depth as normals and position passes, and create a Card
node positioned in 3D space and displaced according to the depth channel.
Note that DepthGenerator can only create a depth map from a sequence, so you can’t use a still
image.
Quick Start
1. Add DepthGenerator to your NukeX script. See Connecting DepthGenerator.
2. Select the type of depth map you want to output, and whether to also output normals and
position passes. See Selecting What to Output.
3. Analyze the depth values for your footage. See Analyzing Depth.
4. Review and refine the results. See Refining the Results.
5. Use the results with other Nuke and NukeX nodes. See Using the Results.
USER GUIDE
1655
Connecting DepthGenerator |
Connecting DepthGenerator
To connect DepthGenerator:
1. To use the DepthGenerator node, you need a tracked camera that matches your footage. If you
don’t already have one, you can create one with the CameraTracker node (see Camera Tracking).
2. Create a DepthGenerator node by clicking 3D > DepthGenerator.
3. Attach a Viewer to the output of DepthGenerator.
4. Connect the Source input to your footage and the Camera input to your Camera node.
A simple DepthGenerator node tree.

5. Depth can only be calculated where the real world 3D position of objects doesn’t change. If there
are moving objects in your Source footage, DepthGenerator is likely to struggle to create an
accurate depth map for those regions. To prevent this, you can exclude moving foreground
regions from the depth calculation by connecting a matte to the Mask input and setting Ignore
Mask in the DepthGenerator properties to the channel that contains the matte. Note that
DepthGenerator expects values of either 0 (for regions to use) or 1 (for regions to ignore).
A Source image. An ignore mask.

The Mask input only appears once you’ve connected the other two inputs.
USER GUIDE
1656
Connecting DepthGenerator |
Using the Mask input.

6. Proceed to Selecting What to Output.
USER GUIDE
1657
Selecting What to Output |
Selecting What to Output

To select what to output:
1. In the DepthGenerator properties, use Depth Output to select the type of depth map you want to
generate:
• Depth (1/Z) - output 1/Z where Z is the distance along the Z axis for the camera. This matches
the depth output of the ScanlineRender node.
This mode also allows you to later create a Group node with a Card positioned in 3D space and
displaced according to the depth channel.
• Distance - output the distance along the ray from the camera center to the 3D surface point.
In the figure below, each pixel forms a ray, and AB measures the physical distance from the
camera to the 3D point, whereas AC measures the distance along the Z axis for the camera.
Calculating the depth.

2. If you also want to output depth as a position pass, set Surface Point to the channels where you
want to store this (for example, you could create a new layer called ppass, which contains the
channels X, Y, and Z).
The position pass includes the X, Y, and Z coordinates for each pixel in the image. You can use it
with the Relight node, or with the PositionToPoints node to visualize depth as a point cloud.
A position pass.
USER GUIDE
1658
Selecting What to Output |
3. Similarly, to output depth as a normals pass, set Surface Normal to the channels where you want
to store the normals pass (for example, create a new layer called npass, which contains the
channels X, Y, and Z).
The normals pass contains three vectors of information for each pixel in the image: X direction, Y
direction, and Z direction. In other words, it stores the direction in which each point in the image
is facing. You can use a normals pass with the Relight node.
A normals pass.
4. Proceed to Analyzing Depth below.
USER GUIDE
1659
Analyzing Depth |
Analyzing Depth
To analyze the depth:
• Use the Frame Separation control to select the offset between the current frame and the frame
against which to calculate depth for your footage. For example, if your current frame is 100, and
your frame separation is 2, DepthGenerator uses frame 98 and frame 102 to generate the depth
map.
Ideally, you want frames close together so that the images are similar and include the same bits
of the world. However, you get more accurate depth when the frames are further apart and the
baseline between the cameras is bigger. So, for fast camera moves, you need a small frame
separation, and for slow camera moves, you can use a larger frame separation. To change the
separation for fast and slow movements, you can animate the Frame Separation control.
Note that this control does NOT affect the number of frames used in the calculation, as that is
always 2.
• To have DepthGenerator attempt to calculate the best Frame Separation automatically, click
Analyze Sequence. For this to work, the camera in the Camera input must be defined for all
frames within the frame range.
DepthGenerator goes through the entire sequence, attempts to work out the correct frame
separation, and animates the Frame Separation value.
• Alternatively, you can also click Analyze Frame to have DepthGenerator automatically calculate
the frame separation for the current frame. This gives you more control than Analyze
Sequence, as you can work through the timeline, analyze particular frames, and if necessary
tweak the Frame Separation value manually.
2. Scrub through the timeline and keep an eye on the Calculated Accuracy value. This displays the
depth accuracy calculated when analyzing frame separation. Values closer to 1 are considered
accurate, and values closer to 0 inaccurate.
You can use the frames that produce accurate values later when placing elements in 3D (for
example, by clicking Create Card or using a PositionToPoints node).
If you’re not getting accurate depth values, try adjusting Frame Separation or using an Ignore
Mask.
3. View the depth channel in 2D by selecting it from the list of channels in the Viewer. To see more
details in the depth map, it can be a good idea to temporarily adjust the gain and gamma controls
in the Viewer.
USER GUIDE
1660
Analyzing Depth |
A depth map. The same depth map after

adjusting Viewer gain.
4. If you have set Surface Point or Surface Normal to generate position and normals passes, you
can also view those by selecting them from the list of channels in the Viewer.
A position pass. A normals pass.

5. Proceed to Refining the Results.
USER GUIDE
1661
Refining the Results |

1. Adjust Depth Detail to vary the resolution of the images used to calculate the depth map. The
default value of 0.5 equals half the image resolution. Lower values speed up processing and
deliver a smoother result. Higher values pick up finer details, but also increase processing time.
Depth Detail = 0.15 Depth Detail = 0.85

2. If you find that the depth map is noisy in flat image regions, try increasing the Noise value in the
Depth Generation parameter group. This sets the amount of noise DepthGenerator should
ignore in the input footage when calculating the depth map. The higher the value, the smoother
the depth map.
Noise = 0 Noise = 0.15

3. If you think the depth map should still be smoother, you can also increase Smoothness. Rather
than affecting the depth calculation, this applies an intelligent blur on the result.
A high smoothness can miss lots of local detail, but is less likely to produce noisy depth values. A
low smoothness concentrates on detail matching, even if the resulting depth map is jagged.
4. The Strength parameter defines the strength of pixel matching between frames. You can usually
leave it at the default value. However, if the depth map does not appear to fit some edges of the
image, you can increase Strength to force matches where fine details are missed. You can also
reduce Strength to smooth the depth map, but generally it’s a good idea to leave it at a relatively
high value to make sure the algorithm is matching the images and calculating the depth correctly.
USER GUIDE
1662
Strength = 0.5 Strength = 1.5

5. If you aren’t happy with the object boundaries in your depth map, adjust Sharpness. Use a high
value to produce distinct boundaries between the objects, or a low value to smooth the
boundaries.
Sharpness = 0.05 Sharpness = 0.95

6. If necessary, use Near Clip Plane and Far Clip Plane to set the minimum and maximum values
allowed in the depth map. All depth values that are outside this range are clipped to these values.
7. To see if there are any regions where the depth calculation is ambiguous, check Mark Bad
Regions.
Any ambiguous regions are marked as very large values in the depth map. Such regions might
occur if CameraTracker hasn’t been able to calculate the depth of all pixels using the camera data,
due to certain camera movements for instance.
Mark Bad Regions enabled.

8. Once you’ve set the parameters so that you’re happy with the depth map, view the normals pass (if
you decided to generate one earlier). Normals are calculated from depth, and the normal
direction is defined by the change in depth between pixels. The normals pass can be noisy, as
USER GUIDE
1663
small changes in depth lead to large changes in the normals. To smooth out the normal map by
calculating normals using a lower resolution depth map, reduce Normal Detail. This integrates
the change in depth across a wider range of pixels, reducing noise in the normals pass.
The default value of 0.25 causes the normals pass to be calculated at a quarter resolution.
Normal Detail = 0.2 Normal Detail = 0.8

9. Proceed to Using the Results.
USER GUIDE
1664
Using the Results |
Using the Results

Once you are happy with the results from DepthGenerator, there are several ways you can use them
in Nuke and NukeX:
• If you set Depth Output to Depth 1/Z, you can click Create Card in the DepthGenerator controls to
create a Group node with a Card positioned in 3D space and displaced according to the depth
channel. This allows you to visualize depth as a surface in 3D space, making it easier to place
elements (such as Cards) at the correct depth in a live-action shot.
A Card node displaced according to the

depth channel.
• If you set DepthGenerator to create a position pass, you can supply the position pass to the
PositionToPoints node (3D > Geometry > PositionToPoints) to create a 3D point cloud. This is
another good way to visualize depth in 3D space.
For more information, see Creating a Dense Point Cloud Using the PositionToPoints Node.
• If you used DepthGenerator to create both a position pass and a normals pass, you can supply
those to the Relight node (3D > Lights > Relight). This node applies a shader to its input image
based on the position and normals passes, creating realistic lighting.
For more information, see Relighting a 2D Image Using 3D Lights.

• You may also be able to use the depth map if you want to introduce fog and depth-of-field effects
into a shot. In Nuke, the ZDefocus node requires a depth map in its input.
For more information, see Simulating Depth-of-Field Blurring.
USER GUIDE
1665
Creating Dense Point Clouds
Dense point clouds are a useful starting point for 3D modeling and can be helpful in positioning 3D
objects into a scene. Using the PointCloudGenerator node, you can create a dense point cloud based
on the information generated by CameraTracker and use the points to create a 3D mesh of your 2D
footage.
Quick Start
1. Make sure your PointCloudGenerator node is connected to the appropriate footage and Camera
node. For more information, see Connecting the PointCloudGenerator Node.
2. Set keyframes in the footage to determine which frames are tracked to create the point cloud. For
more information, see Setting Keyframes in a Sequence.
3. Track the footage to create a dense point cloud. For more information, see Tracking a Dense
Point Cloud.
A 2D source image. The resulting 3D point cloud.

4. Filter points from the result to adjust it and eliminate bad points. For more information, see
Filtering Your Point Cloud.
5. Create groups within the cloud to aid visualization or bake out grouped sections. For more
information, see Grouping, Labeling, and Baking Points.
6. Then if you need to, you can move on to creating meshes. For more information, see Creating a
Mesh from a Point Cloud.
7. Alternatively, use the PoissonMesh node to create a mesh. For more information, see Using the
PoissonMesh Node.
USER GUIDE
1666
Connecting the PointCloudGenerator Node |
Connecting the
PointCloudGenerator Node
To create a dense point cloud, PointCloudGenerator needs a Camera node with a solved camera path,
either from CameraTracker or from a third party 3D application, and the tracked 2D footage.
To connect the node, do the following:

1. Click 3D > Geometry > PointCloudGenerator.
2. Connect the necessary nodes to the inputs:
• Source - the 2D footage from which the camera track was solved.
• Camera - the solved Camera node.
3. If you want to mask part of the scene to avoid tracking moving objects or reflections, connect a
matte to the Mask input. For more information about masking, see Masking Out Regions of the
Image.
4. Click Image > Viewer to insert a Viewer node and connect it to the PointCloudGenerator node.
5. Proceed to Setting Keyframes in a Sequence.
USER GUIDE
1667
Masking Out Regions of the

Image
If you don’t want to track all regions of your image, for example, if there are moving objects or
reflections in the image, you can attach a matte to the Mask input to define image regions that
should not be tracked. You can also use the source input’s alpha channel as a matte.
1. If you want to use a separate matte for masking, connect it to the Mask input in the Node Graph.
2. In the properties panel, set Ignore Mask to the component you want to use as a mask:
• None - Track features in the whole sequence.
• Source Alpha - Use the alpha channel of the source clip to define which areas to ignore.
• Source Inverted Alpha - Use the inverted alpha channel of the source clip to define which
areas to ignore.
• Mask Luminance - Use the luminance of the mask input to define which areas to ignore.
• Mask Inverted Luminance - Use the inverted luminance of the mask input to define which
areas to ignore.
• Mask Alpha - Use the mask input alpha channel to define which areas to ignore.
• Mask Inverted Alpha - Use the inverted mask input alpha channel to define which areas to
ignore.
Note: PointCloudGenerator can track multiple ranges, adding the new points to the existing
cloud, so feel free to mask and re-track as much as necessary.
USER GUIDE
1668
Setting Keyframes in a Sequence | Setting Keyframes Automatically
Setting Keyframes in a Sequence

Keyframes represent the best frames in a sequence to perform a particular function, in this case
tracking. PointCloudGenerator can analyze the sequence for you, automatically picking frames with
good parallax, or you can set keyframes manually by marking the frames yourself.
Setting Keyframes Automatically

1. Click Analyze Sequence.
PointCloudGenerator walks the sequence, adding keyframes in suitable areas.
2. Once the process is complete, scrub the playhead and use the Calculated Accuracy field to
evaluate the keyframes - values closer to 1 indicate a high accuracy.
Note: You can also add keyframes to the sequence manually. See below for more
information.
3. Proceed to Tracking a Dense Point Cloud.
Setting Keyframes Manually

1. Scrub the playhead to the frame you want to mark as a keyframe and click Add,
OR
Set the Frame Spacing field to the number of frames between keyframes throughout the
sequence and click Add All.
For example, if you set this value to 2, keyframes are set at every other frame throughout your
sequence.
2. If necessary, click Delete to remove the keyframe at the current playhead position or click Delete
All to remove all keyframes.
Note: The keyframes generated automatically using Analyze Sequence are not deleted.
3. Proceed to Tracking a Dense Point Cloud.
USER GUIDE
1669
Tracking a Dense Point Cloud | Setting Keyframes Manually
Tracking a Dense Point Cloud

The next step towards creating a dense point cloud is to track your footage for more 3D feature
points using the information from keyframes in the sequence and the solved Camera.
1. Click Track Points to track the sequence and create a dense point cloud.
An example 3D point cloud.

The default settings work well on most sequences, but you can adjust the Dense Tracking
controls to alter the appearance of the cloud:
2. Adjust the Point Separation control to set the separation value, in pixels, between points in the
point cloud.
As you can see, lower values produce denser clouds, but at the expense of processing time.
Point Separation 0. Point Separation 10.

3. Set the Track Threshold to compare the similarity of features over a number of frames and reject
them if they exceed the threshold. You can adjust this value to test whether a track is reliable.
Low threshold. High threshold.
USER GUIDE
1670
Tracking a Dense Point Cloud | Setting Keyframes Manually
Note: If the threshold is set too high, you may find that you don’t get any reliable tracks at
all. The red pixels in the above image are rejected tracks.
4. Proceed to Filtering Your Point Cloud.
USER GUIDE
1671
Filtering Your Point Cloud | Setting Keyframes Manually
Filtering Your Point Cloud

You can filter your point cloud by adjusting the number and quality of points it includes. NukeX filters
the tracked point cloud directly, without recalculating, streamlining your workflow.
1. Ensure that Display rejected points is enabled and adjust the threshold controls to update the
3D Viewer dynamically. Rejected tracks are highlighted in red.
2. Adjust the Angle Threshold to set the minimum acceptable angle to triangulate 3D points (in
degrees). Points with a large triangulation angle tend to be more accurate.
Set a threshold of 0 to triangulate all points or increase the threshold to highlight the least accurate
points.
Tip: As a rule of thumb, anything below 5 degrees is likely to be incorrect.
3. Adjust the Density Threshold to set the minimum acceptable point density. Isolated points tend
to be less accurate.
Set a threshold of 0 to output all points or increase the threshold to highlight the most isolated,
less accurate points.
4. The Viewer wipe tool can help you locate isolated points by comparing the 2D source footage to
the point cloud.
Set Input A to the PointCloudGenerator node and Input B to the Read node for the source image.
See Using the Viewer Controls for more information.

5. Proceed to Removing Rejected Points.
USER GUIDE
1672
Removing Rejected Points | Setting Keyframes Manually
Removing Rejected Points

Once you’ve identified and highlighted the less accurate points:
1. Click Delete Rejected Points to automatically remove all highlighted points.
2. You can then be a little more selective by switching the Viewer to Vertex selection mode.
3. Use the Output controls to assist you when visualizing your point cloud:
• Point Size - set the pixel size of points in your cloud.
• Output points per frame - by default, all tracked points in your sequence are displayed in the
Viewer. Enabling this control allows you to view only the points generated at the current frame.
Note: If you’re having trouble viewing points individually, try reducing the 3D handle size in
the Preferences > Panels > Viewer Handles tab.
4. Manually select points in the cloud using the 3D Viewer by:

• Dragging a marquee over the required points in the Viewer, or
• Holding Shift, and dragging a marquee over several selection areas in the Viewer to select all
points at once.
USER GUIDE
1673
Removing Rejected Points | Setting Keyframes Manually
Note: You can remove points from a selection by holding Shift+Alt and re-selecting
highlighted points.
5. Press Delete/Backspace or right-click in the Viewer and select delete selected to remove the
points.
6. Proceed to Grouping, Labeling, and Baking Points.
USER GUIDE
1674
Grouping, Labeling, and Baking Points | Setting Keyframes Manually
Grouping, Labeling, and Baking

Points
Point clouds can be awkward to interpret, especially if the Point Separation control is set to a
relatively high value. Grouping points in the cloud can help to visualize the scene, particularly when
sensibly labeled and colored in the Viewer.
Groups can also be baked out as separate point clouds or converted into meshes. See Creating a
Mesh from a Point Cloud for more information.
To create a group from a point cloud:

1. Manually select group points in the 3D Viewer by:
• Dragging a marquee over the required points in the Viewer, or
• Holding Shift, and dragging a marquee over several selection areas in the Viewer to select all
points at once.
Note: You can remove points from a selection by holding Alt + Shift and re-selecting
highlighted points.
2. Right-click on a highlighted point and select create group.
The new group is added to the Groups tab.
Note: If groups already exist, the option to add the selected points to a group is enabled.
3. Add more groups as required to build up the visualization of the scene.

4. In the properties panel, select the Groups tab to display the list of existing groups. Use the
Groups controls to determine the appearance of your groups:
USER GUIDE
1675
Grouping, Labeling, and Baking Points | Setting Keyframes Manually
• Display groups in overlay - when enabled, groups are highlighted in the Viewer using their
associated RGB color.
• Create Group - click to add the selected points in the Viewer to a new group.
• Delete Selected Groups - click to delete all the selected groups in the list.
5. Double-click the table columns to edit group names or colors, and toggle visibility.
6. Enable Output visible groups only to display only selected groups in the Viewer.
7. After creating the required groups, you can split them off into individual point clouds, using Bake
Selected Groups, or create group meshes.
8. Proceed to Creating a Mesh from a Point Cloud
USER GUIDE
1676
Creating a Mesh from a Point Cloud | Setting Keyframes Manually
Creating a Mesh from a Point

Cloud
PointCloudGenerator can create meshes from grouped points in the point cloud that you can use as
stand-alone 3D objects, for example, in 3D modeling. You can also use these meshes to quickly
project the 2D sequence onto the mesh using the Project3D node.
PointCloudGenerator meshes are based on the Poisson Surface Reconstruction calculation method.
The original source code and paper were created by Michael Kazhdan, Matthew Bolitho, and Hugues
Hoppe (see http://www.cs.jhu.edu/~misha/Code/PoissonRecon/ for more information).
To create a mesh from a group:

1. On the Groups tab, select the required group(s) from the Groups list.
You can mesh a single group or all the groups available in the list.
2. Click Bake Selected Groups to Mesh.
A mesh is created automatically using the information from the PointCloudGenerator node.
Note: You can use the resulting group mesh node in Nuke as well, though you can’t edit the
geometry.
3. Proceed with Adding Texture to a Mesh.
Many factors influence the quality of any mesh, chief among which are the contents of the scene and
the camera track obtained from the sequence. If you consider a simple scene, such as that shown in
the example images, the resulting mesh can be quite accurate.
A mesh from a simple sequence.
USER GUIDE
1677
Adding Texture to a Mesh | Setting Keyframes Manually
Adding Texture to a Mesh

You can quickly add texture to your point cloud mesh using the Project3D node. Do the following:
1. Click 3D > Shader > Project3D.
2. Connect the solved Camera node to the Project3D cam input.
3. Connect the other input of the Project3D node into the sequence.
4. Connect the Group1_Mesh node img input to the Project 3D node.
The mesh is now textured with your original 2D sequence.
Tip: You can export your camera, point cloud, and meshes as Alembic (.abc) objects to
other applications. For more information about exporting .abc files, see Exporting
Geometry, Cameras, Lights, Axes, or Point Clouds .
USER GUIDE
1678
Using the PoissonMesh Node | Setting Keyframes Manually
Using the PoissonMesh Node

The PoissonMesh node uses information from a dense point cloud to generate a mesh that you can
further use as a 3D object, in 3D modeling for instance. The PoissonMesh node is based on the
Poisson Surface Reconstruction calculation method. The original source code and paper were created
by Michael Kazhdan, Matthew Bolitho and Hugues Hoppe (see
http://www.cs.jhu.edu/~misha/Code/PoissonRecon/ for more information).
Before using the PoissonMesh node, you need a dense point cloud with normals such as the one
created by the PointCloudGenerator node. For more information on creating a dense point cloud, see
Tracking a Dense Point Cloud on page 1670. When you’ve got one, you’re ready to create a mesh:
1. Connect the PoissonMesh node to a PointCloudGenerator node.
2. A mesh is created automatically using the information from the PointCloudGenerator node and
you can view it in the 3D view.
3. If you want to, you can tell PoissonMesh to use the selected points only in the point cloud to
create a mesh. To do this, check the Use Selection box in the properties panel. You can also use
the Filtering box to use the PointCloudGenerator filtering to refine your mesh.
4. You can also adjust other properties for your mesh on the PoissonMesh tab:
• Depth - set the maximum depth (in integers) for the calculation tree used to create your mesh.
You can enter values up to 10, but increasing this value increases the solver’s memory usage, so
it’s a good idea to keep this as low as possible.
• Scale - set the ratio between the 3D cube used to create the mesh and the bounding box of the
footage it’s created from.
• Solver Divide - set the depth at which the solver equation is used in the creation of your mesh.
This can help you reduce your memory usage when generating complex meshes.
• Iso Divide - set the depth at which the iso-surface extractor is used in extraction. This can help
you reduce your memory usage when generating complex meshes.
• Samples per Node - set the minimum number of sample points used to create your mesh. Use
larger values if your original footage is very noisy.
• Confidence - check to use the size of the normals as confidence information when generating
the mesh. This may take longer, but gives you better results since your point cloud point
creation gets double-checked.
Note: You can only use the PoissonMesh node in NukeX as the information used to create
the mesh is not stored in the node and therefore cannot be recreated for viewing in Nuke.
USER GUIDE
1679
Adding Texture to the PoissonMesh | Setting Keyframes Manually
Adding Texture to the

PoissonMesh
You can add texture to your point cloud mesh using the Project3D and ApplyMaterial nodes. Do the
following:
1. Connect your texture node to the Project3D node.
2. Then connect the Project3D node to the mat input of the ApplyMaterial.
3. Connect the other input of the ApplyMaterial node into the PoissonMesh node. The mesh is now
textured with your texture.
Tip: You can export both your point cloud and the mesh as FBX objects to other
applications, but don’t forget to export the Camera node too. For more information about
exporting FBX files, see Exporting Geometry, Cameras, Lights, Axes, or Point Clouds .
USER GUIDE
1680
Using ModelBuilder
The ModelBuilder node provides an easy way to create 3D models for 2D shots. You can build a
model by creating shapes and then editing them, and align models over your 2D footage by dragging
vertices to their corresponding 2D location.
Prerequisites
To be able to align models, ModelBuilder needs a tracked camera and an input image for visual
reference. You can also use other 3D geometry and point clouds as a reference if you already have
these for your scene.
Note: You can use ModelBuilder without a camera, image sequence, or reference geometry.
In this case, all the Edit mode features work just fine; it just means you aren’t able to do
anything in Align mode.
Creating and editing models using ModelBuilder requires a NukeX license, but the resulting geometry
can also be used in Nuke.
A 2D source image with shapes created using ModelBuilder.
Special thanks to for use of the above footage, used throughout this chapter.
USER GUIDE
1681
Using ModelBuilder | Quick Start
Quick Start
1. Connect the ModelBuilder node to your footage and a changing Camera. For more information,
see Connecting the ModelBuilder Node.
2. As a starting point for your model, create one or more of the built-in 3D shapes using the toolbar
on the left hand side of the Viewer.
See Creating Shapes.
3. If necessary, adjust the display characteristics of your shapes so you can better see them in the
Viewer. You can do this at any point during the process.
See Editing Shapes’ Display Characteristics.
4. Position the shape using your 2D footage or existing 3D geometry as a reference.
See Positioning Shapes.
5. Edit the shape until you’re happy with your model.
See Editing Shapes.
6. Once you're happy with your model, you're ready to texture it.
See Applying Textures.
7. If necessary, you can also create a separate geometry node for the selected shapes and use that
as you would use any other geometry node in Nuke.
See Exporting Shapes to Separate Geometry Nodes.
USER GUIDE
1682
Connecting the ModelBuilder Node | Quick Start
Connecting the ModelBuilder

Node
To connect the ModelBuilder node:
1. If you don’t already have a tracked camera that matches your footage, create one using the
CameraTracker node (see Camera Tracking).
ModelBuilder needs a changing Camera (with a change of more than 5 degrees) in order to
calculate 3D points from your 2D footage.
2. Select 3D > Geometry > ModelBuilder to create a ModelBuilder node.
3. Connect your Camera node to ModelBuilder’s cam input, and your 2D footage to its src input.
Note that the objects in your src footage that you want to model should not be moving, as
ModelBuilder works best with static objects.
4. If you have 3D geometry (such as a shape or a point cloud) that you want to use as a reference
when positioning new shapes, connect it to ModelBuilder’s geo input.
For example, you can use the PointCloudGenerator node (see Creating Dense Point Clouds) to
create a 3D point cloud for your shot, and connect that to the geo input. When creating new
shapes, you can then select vertices on the point cloud to automatically align the shapes with the
point cloud. This gives you an approximate initial position for your shapes.
5. If you have a 2D texture image that you want to display as a background in the UV preview
window and on the model in the 3D Viewer, connect it to ModelBuilder's tex input.
6. Attach a Viewer to the output of the ModelBuilder node.
7. Proceed to Creating Shapes below.
USER GUIDE
1683
Creating Shapes | Quick Start
Creating Shapes
To create shapes:
1. Double-click on the ModelBuilder node to open its properties.
2. Use the shape creation menu in the ModelBuilder toolbar on the left hand side of the Viewer
to select the basic 3D shape that best matches the object you are trying to model. You can select
between Point, Card, Cube, Sphere, Cone, Cylinder, and Polygon (when creating a polygon, click
on the Viewer to create vertices where you need them, and when you’re done, press Return).
A colored border appears around the Viewer window to indicate that the action of creating a
shape is in progress.
Tip: If you want to model a simple wall, a Card might be a good starting point. For an entire
building, you may want to use a Cube instead.
You can adjust the shapes (for example, the number of rows or columns in them) on the Shape
Defaults tab. These controls only apply when creating a shape - you cannot adjust them after the
fact.
3. Hover over the Viewer. You should see a yellow circle appear under the cursor. You can now:
USER GUIDE
1684
• click in the Viewer to set where you want the new shape created. If you click over empty space,
the shape is created at a distance specified by the Distance control. There are no specific units
for this value; low values position the shape close to the camera, and higher values mean
further away.
• hold down Shift and click to snap to the surface of any geometry under the cursor.
• drag the mouse left or right to scale the shape.
The shape you selected appears in the Viewer and in the Scene list in the ModelBuilder properties.
4. If the shape was created too close to the camera or too far away, Activate Edit Mode in the
ModelBuilder toolbar and make sure the selection mode menu is set to Select objects .
Then, select the shape in the Viewer, and drag the transform handles to reposition it.
5. If you have more than one object that you want to model in your src footage, you can repeat the
above steps to create several shapes using a single ModelBuilder node.
6. Proceed to Editing Shapes’ Display Characteristics below.
Tip: You can also use the Create buttons on the Shape Defaults tab to create shapes. This
is exactly the same as using the shape creation menu in the Viewer.
Tip: If you want to use 3D geometry as a visual reference and connected it to the geo input,
before creating a shape, you need to:
1. Activate Edit Mode in the ModelBuilder toolbar on the left hand side of the Viewer
and set the selection mode menu to Select vertices .

2. Drag in the Viewer to select vertices on the object that you want to model, bearing in mind
that when you create a new shape, it’s automatically aligned with the selected vertices.
If you cannot see the object in the geo input, make sure Pass Through Geo is enabled in the
ModelBuilder properties.
3. Use the shape creation menu in the ModelBuilder toolbar to select the basic 3D
USER GUIDE
1685
shape that best matches the object you are trying to model.
ModelBuilder creates the shape to align with the selected vertices.
USER GUIDE
1686
Editing Shapes’ Display Characteristics | Quick Start
Editing Shapes’ Display

Characteristics
To edit shapes' display characteristics:
1. If you created a lot of shapes, you may want to organize them into groups. Click the + button in
the ModelBuilder properties to create a Group item and drag shapes inside it to move them into
the group.
2. To rename shapes or groups, click on them in the Scene list and enter a new name.
3. If the Viewer seems too cluttered, you can toggle the visibility of both shapes or groups by clicking
in the Scene list.

Note that when an item is hidden, it doesn’t appear in the Viewer or renders.
4. If you have connected geometry to the geo input but don’t currently need it as a reference, you
can also uncheck Pass Through Geo. This tells ModelBuilder to not output that geometry.
Note that when Pass Through Geo is enabled, the geometry appears both in the Viewer and
renders.
5. By default, the src image is displayed in the 3D Viewer whenever the Viewer is locked to the input
camera. If you don’t want this, uncheck Show Source Image.
Show Source Image enabled. Show Source Image disabled.

6. Use the display and render controls to select how you want your shapes to appear in the Viewer
and renders. These controls affect all the existing shapes.
For example, you may want to set display to wireframe initially in order to better see what you
are doing when positioning your shapes, but change it to textured to review the final result.
For more information on the available options, see Object Display Properties.
7. If you set display to textured or textured+wireframe, you can also use the Texture control to
select the frame used to texture the shapes:
• Current frame - Project the current frame onto the shape.
USER GUIDE
1687
Editing Shapes’ Display Characteristics | Quick Start
• Locked frame - Project the frame specified in the field on the right onto the shape. This can
help you line up the shape against your source footage.
• Frame difference - Subtract the frame specified in the field on the right from the current frame
and project the resulting frame onto the shape. This can help you line up the shape against your
source footage.
8. If you want to disable shape selection in the Viewer, uncheck selectable. This only has an effect
when the selection mode menu is set to Select nodes in the ModelBuilder toolbar.
9. If you don’t want the shapes to cast or receive shadows, uncheck cast shadow or receive
shadow.
For more information on shadows, see Casting Shadows.
10. Proceed to Positioning Shapes below.
USER GUIDE
1688
Positioning Shapes | Quick Start
Positioning Shapes
To position shapes:
1. If the Align Mode isn’t already active, click in the ModelBuilder toolbar to activate it.
2. The Viewer should also automatically be locked to your input Camera node. If this isn’t the case,
select the Camera from the dropdown menu in the top right corner of the Viewer and click .
The positioning handles only appear when the Viewer is locked to the input camera.
3. Use the Align dropdown menu on the top of the Viewer to choose whether you want to position
the shape by transforming the entire shape or individual vertices:
• Object transform - ModelBuilder tries to transform the whole object. This guarantees that flat
polygons stay flat and the angle between polygons doesn’t change, but it's less flexible because
you're always changing the whole object at once.
The original card. The card after transforming

the entire object.
• Vertex position - ModelBuilder transforms each vertex separately. While this can be a very
convenient way to make the geometry match up exactly, it can also lead to non-flat polygons
and other problems.
The original card. The card after transforming

a vertex position.
USER GUIDE
1689
Positioning Shapes | Quick Start
4. Move a few vertices so that they match the shape of the object you want to model. For example, if
you are modeling a building using a cube, drag the corners of the cube so that they line up with
the corners of the building.
When you click on a vertex, you’ll notice a zoom window in the top left of the Viewer. This allows
you to accurately position vertices without zooming the entire Viewer. To adjust the size and
magnification factor for the zoom window, use the Zoom dropdown menus on top of the Viewer.
As you do this, ModelBuilder transforms the shape to fit your adjusted vertex positions. It also
sets keys on the 2D positions of the vertices. Any vertices that have keys set on them turn purple
in the Viewer.
5. Go to a different frame and move one of the blue vertices to match its new location in the src
footage. When you do so, a purple line appears. The correct location for the vertex on the current
frame should lie somewhere on this line; however, you can drag the vertex away from the line if
necessary.
Positioning just one vertex is usually enough to get the object position right, but not the rotation
or scaling.
6. In order to lock off the rotation and scaling of the object, move a few more vertices to their
correct location.
7. Play back through the sequence. The object should stay roughly in the right place relative to the
src footage. If you see any vertices starting to drift, move them to their correct location.
8. Proceed to Editing Shapes.
USER GUIDE
1690
Editing Shapes | Quick Start
Editing Shapes
When editing ModelBuilder shapes, there are two types of editing actions:
• Single-step actions that happen instantly, such as extruding and merging faces.
• Multi-step actions (such as beveling and subdiving faces) that have extra parameters you can set.
All multi-step actions behave in the same way:
• When you start the action, you can see a preview in the Viewer even though the action has not
been applied yet. A colored border appears around the Viewer window to indicate that the
action is in progress.
Tip: You can change the color of the border in the Preferences. Press Shift+S on the Node
Graph to launch the Preferences dialog, go to the Appearance tab, and use the Highlight
button to select a new color. Note that he brightness of the color is adjusted to prevent it
from clashing with the Viewer background.
• The parameters for the action appear at the top of the Viewer. As you adjust them, the preview
in the Viewer is updated.
• To cancel the action, you can press Esc on the Viewer.
• To accept the current parameter values and complete the action, you can press Return (or
change the selection mode, start a new action, press the +, -, or duplicate buttons, add a new
shape, change into Align mode, or close the ModelBuilder properties).
To edit your shapes:

1. You can edit vertices, edges, faces, and entire objects. For more information, see:
• Editing Vertices
USER GUIDE
1691
Editing Shapes | Quick Start
• Editing Edges and Edge Loops

• Editing Faces, and
• Editing Objects.
2. When you’re happy with your edits, proceed to Exporting Shapes to Separate Geometry Nodes.
USER GUIDE
1692
Editing Vertices | Quick Start
Editing Vertices
To edit vertices:
1. Activate Edit Mode by selecting from the ModelBuilder toolbar.
2. Set the selection mode menu to Select vertices in the ModelBuilder toolbar and select one
or more vertices on the object.
3. Edit your selection as necessary:
• To translate, rotate, or scale the selected vertices, drag the transform handles that appear on
them. To move the pivot point for the transform handles, press Ctrl/Cmd+Alt while dragging on
the center of the transformation overlay.
Tip: If necessary, you can also use the controls at the top of the Viewer to set the initial
position and alignment of the transform handles (that is, the position used whenever you
change the selection). For more information, see Setting the Initial Transform Action Center.
• To carve the selected vertices, right-click on them and select carve. A colored border appears
around the Viewer window to indicate that the action is in progress. Click on an edge or vertex
to begin a carve. All of the surrounding faces are highlighted in red. Click anywhere inside a
highlighted face, or on an edge or vertex of a highlighted face, to carve an edge between the
previous vertex and the place you just clicked.
A highlighted face. A new edge carved inside

the highlighted face.
To carve out a freehand edge, hold down Ctrl/Cmd+Shift and drag.
• To extrude the selected vertices, select extrude from the right-click menu and drag. This
stretches the selected vertices into three-dimensional polygons. ModelBuilder creates faces to
complete the object, for example a rectangular pyramid from a vertex.
USER GUIDE
1693
Editing Vertices | Quick Start
A selected vertex. Extruding the selected vertex.

• To bevel the selected vertices, select bevel from the right-click menu.
A colored border appears around the Viewer window to indicate that the action is in progress.
At the top of the Viewer, use relative inset to define how far back along the surrounding edges
to start the bevel from: 0.0 means no distance and 1.0 means at the opposite end of the edge.
The default is 0.1, meaning the beveling starts 1/10th of the way back along each edge.
Set round level to the number of times to repeat the initial bevel, effectively rounding the
edges. A value of 0 just does the initial bevel, a value of 1 bevels the output of the initial bevel; a
value of 2 bevels the bevel of the initial bevel, and so on.
When you’re happy with the values you’ve entered, press Return to apply the bevel.
A selected vertex. Beveling the selected vertex.

Beveling is like extruding the selected vertices, except that the resulting polygons have smooth
edges and corners. You may want to use beveling to add realism to your model, as real-world
objects rarely have perfectly sharp corners.
• To delete the selected vertices, select delete vertices from the right-click menu (or press
Delete).
Note: You can only delete simple vertices that have no more than two edges connected to
them.
USER GUIDE
1694
Editing Edges and Edge Loops | Quick Start
Editing Edges and Edge Loops

To edit edges and edge loops:
2. Set the selection mode menu to either Select edges or Select edge loops in the
ModelBuilder toolbar and select one or more edges or edge loops on the object.
Tip: An edge loop is a string of connected edges across the model. Often, the last edge
meets again with the first edge, forming a loop. In ModelBuilder, the loop follows the middle
edge in every intersection that has an even number of edges. The loop ends when it
encounters an intersection with an odd number of edges.
Edge loops can help you produce more natural deformations for organic models that are
animated. They work best when you’ve got a mesh that’s mostly made up of four-sided
polygons, known as quads.

• To translate, rotate, or scale the selected edges or edge loops, drag the transform handles that
appear on them. To move the pivot point for the transform handles, press Ctrl/Cmd+Alt while
dragging on the center of the transformation overlay.
Selected edges. Translating the selected edges.
USER GUIDE
1695
• To carve the selected edges, right-click on them and select carve. A colored border appears
Highlighted faces. Carving a new edge inside

the highlighted faces.
• To extend your current edge selection along loops, select select edge loop from the right-click
menu. This is the same as activating Select edge loops and clicking on an edge.
Original selection. Select edge loop.

• To extend your current selection to form a ring around the face or the entire shape, choose
select edge ring from the right-click menu.
Original selection. Select edge ring.
USER GUIDE
1696
If the starting edge is part of the boundary for an entire shape, this selects all edges along that
boundary.

Otherwise, it selects all edges around the face that the starting edge belongs to.

• To extrude the selected edges or edge loops, select extrude from the right-click menu and drag.
This stretches the selected edges or edge loops into a three-dimensional polygon.
A selected edge. Extruding the selected edge.

• To bevel the selected edges or edge loops, select bevel from the right-click menu.
USER GUIDE
1697
Selected edges. Beveling the selected edges.

Beveling is like extruding the selected edges, except that the resulting polygons have smooth
• To divide the selected edges or edge loops into segments of equal length, select subdivide. A
colored border appears around the Viewer window to indicate that the action is in progress. Use
the segments control on top of the Viewer to set the number of segments, and press Return to
complete the action.
• To delete the selected edges or edge loops, select delete edges from the right-click menu (or
press Delete).
USER GUIDE
1698
Editing Faces | Quick Start
Editing Faces
To edit faces:
2. Set the selection mode menu to Select faces in the ModelBuilder toolbar and select one or
more faces on the object.
• To translate, rotate, or scale the selected faces, drag the transform handles that appear on
them. To move the pivot point for the transform handles, press Ctrl/Cmd+Alt while dragging on
the center of the transformation overlay.
• To carve the selected faces, right-click on them and select carve. A colored border appears

• To extend your current selection along loops, choose select face loop from the right-click
menu. A face loop is a string of connected faces across the model. Often, the last face meets
again with the first face, forming a loop.
To create a face loop, you need to select at least two neighboring faces, so that ModelBuilder
knows which direction you want to expand the selection in. If the selected faces are above and
USER GUIDE
1699
below each other, ModelBuilder expands the selection vertically. If the selected faces are to the
left and right of each other, ModelBuilder expands the selection horizontally. If you have
multiple groups of selected faces, they all get expanded out.
Original selection. Select face loop.

Face loops only work with four-sided polygons, known as quads. If a face loop reaches a face
with any other number of sides, the loop ends there.
• To extrude the selected faces, select extrude from the right-click menu (or press Return) and
drag. This stretches the selected faces into a three-dimensional polygon. ModelBuilder creates
faces to complete the object.
A selected face. Extruding the selected face.

If you select extrude from the right-click menu again and extrude the same faces, ModelBuilder
creates another three-dimensional polygon next to your first one.
• To merge adjacent faces, select merge from the right-click menu.
• To bevel the selected faces, select bevel from the right-click menu.
USER GUIDE
1700
A selected face. Beveling the selected face.

Beveling is like extruding the selected faces, except that the resulting polygons have smooth
• To invert the selected face's normals, select flip face normals from the right-click menu. This
can be useful, for example, if editing operations (such as a sequence of extrusions) have
resulted in normals that point the wrong way.
Tip: To see the normals, press S on the Viewer to display the Viewer Settings, go to the 3D
tab, and activate the show primitive normals button .
Before flip face normals. After flip face normals.

• To turn the selected face into a triangle fan, select tessellate > triangle fan from the right-click
menu. This adds a new vertex at the center of the face and joins each of the vertices around the
perimeter of the face to it. It can be useful if you want to quickly add more detail to your mesh.
A selected face. The same face with
USER GUIDE
1701
a tessellated triangle fan.

• To delete the selected faces, select delete faces from the right-click menu (or press Delete).
USER GUIDE
1702
Editing Objects | Quick Start
Editing Objects
To edit objects:
2. Click the Select object button in the ModelBuilder toolbar and select an object.
3. Edit the object as necessary:
• To translate, rotate, or scale the selected object, drag the transform handles that appear on it.
To move the pivot point for the transform handles, press Ctrl/Cmd+Alt while dragging on the
center of the transformation overlay.
• To carve the selected objects, right-click on them and select carve. A colored border appears

• To mirror the selected object along its x axis, right-click on the object and select mirror x.
• To mirror the selected object along its y axis, right-click on the object and select mirror y.
• To mirror the selected object along its x axis, right-click on the object and select mirror z.
• To invert the selected object's normals, right-click on the object and select flip face normals.
This can be useful, for example, if your camera is located inside the selected object.
USER GUIDE
1703
Editing Objects | Quick Start
Tip: To see the normals, press S on the Viewer to display the Viewer Settings, go to the 3D
tab, and activate the show primitive normals button .
Before flip face normals. After flip face normals.

• To delete the selected object, select delete objects from the right-click menu (or press Delete).
USER GUIDE
1704
Setting the Initial Transform Action Center | Quick Start
Setting the Initial Transform

Action Center
When you translate, rotate, or scale vertices, edges, faces, or objects, you can change the initial
position from where the action originates (that is, where the transform handles appear whenever you
change the selection). To do so:
1. Make sure Edit Mode is active in the ModelBuilder toolbar.

2. To change the default position of the transform handles relative to your current selection, set
handle pos at the top of the Viewer to:
• selection average - The transform handles appear at the average location of all selected items.
• object center - The transform handles appear at the center of the bounding box for the object
that the selected items are part of.
• object surface - The transform handles are snapped to the surface of the object, as close as
possible to the average position in the selection. This is the default behavior.
USER GUIDE
1705
Note: The object center and object surface options only apply when you have selected
items on a single object. If you have selected items on more than one object, selection
average is always used.
3. To change the orientation of the transform handles relative to your current selection, set handle
align at the top of the Viewer to:
• world axes - The transform handles are oriented so that the x, y, and z arrows align with the
world x, y, and z axes.
• object axes - The transform handles are oriented so that the x, y, and z arrows align with the
object's local x, y, and z axes.
USER GUIDE
1706
• surface normal - The transform handles are oriented so that the y arrow aligns with the
object's surface normal and the x and z axes are at an arbitrary orientation. This is the default
behavior.
Note: The object axes and surface normal options only apply when you have selected
items on a single object. If you have selected items on more than one object, world axes is
always used.
4. To prevent the position and alignment of the transform handles from being changed when you
change your selection, enable locked at the top of the Viewer.
This makes it easier to rotate a face relative to one of its edges, for example: instead of having to
Ctrl/Cmd+Alt+drag the transform handles into the right place by eye, you can set handle pos to
selection average, select the edge, and enable locked. When you then select the face, the
transform handles stay where they are.
Selecting an edge and Rotating a face around the

enabling locked. previously selected edge.
USER GUIDE
1707
Applying Textures | Quick Start
Applying Textures
There are a couple of ways to texture your ModelBuilder models in Nuke:
• If your 3D model closely matches the original 2D footage, you can project the 2D footage onto the
geometry. See Projecting Textures onto Your Shapes.
Projecting the original 2D footage onto the geometry.

• If you have added new objects over the top of your 2D footage (for example, added a new window to
a building or extra bins alongside a street), you can't project your 2D footage over those objects
because they were never in the original footage. Instead, you need to supply a texture from
somewhere else and map it over the surface of your 3D object. See UV Unwrapping.
USER GUIDE
1708
Applying Textures | Quick Start
Texturing an added cube object separately.
USER GUIDE
1709
Projecting Textures onto Your Shapes | Quick Start
Projecting Textures onto Your

Shapes
If your ModelBuilder model closely matches the original 2D footage, you can simply project the 2D
footage onto the geometry. To do so:
1. At the bottom of the ModelBuilder properties panel, set the bake menu to Projection and click
Bake.
This creates a projection at the current texture frame: a Project3D node with a FrameHold node
set to lock the input image and camera to the texture frame. The Project3D node is also
connected to the mat input of an ApplyMaterial node.
2. To apply the projection to the geometry, connect the ModelBuilder node (or a geometry node
created by baking Selected geometry) to the ApplyMaterial node’s unnamed input and view the
result.
USER GUIDE
1710
Projecting Textures onto Your Shapes | Quick Start
Tip: To texture your geometry on more than one frame, you can bake Projection on several
frames and use Roto nodes to mask things out before combining the Project3D nodes using
a MergeMat node. Then, connect the MergeMat node to the ApplyMaterial node’s mat input
and ModelBuilder to its unnamed input. Connect a Viewer to the output of ApplyMaterial
and view the results.
Tip: The options in the bake menu are implemented in Python, and you can also use Python
to add your own entries to the menu. To see how the built-in options have been
implemented, have a look at the modelbuilder.py module in the nukescripts folder of your
installation (for more information on the location of this file, see Viewing More Examples).
Then, to create your own menu options, edit the modelbuilder.py file and use the
populateBakeMenu function to add entries where it says "# Add your own entries here, if
desired".
USER GUIDE
1711
UV Unwrapping | Quick Start
UV Unwrapping
If you've used ModelBuilder to add new objects over the top of your 2D footage (for example, added a
new window to a building or extra bins alongside a street), you can't project your 2D footage over
those objects because they were never in the original footage. Instead, you need to supply a texture
from somewhere else and map it over the surface of your 3D object.
An extra cube has been added to the Here, the cube has been
scene. It cannot be textured using textured separately.
the original footage.
To map a texture image over the surface of your 3D object, you first need to flatten your 3D object
out into 2D space through a process called UV unwrapping. This generates UV coordinates for each
vertex in the object, allowing you to map points on the surface of your 3D object to pixels in your
texture image.
A 3D cube flattened out into 2D space.
USER GUIDE
1712
UV Unwrapping | Quick Start
Previewing a texture image while editing the cube's

UV coordinates.
Tip: UVs are simply 2D coordinates that tell 3D applications how to apply a texture to a
model. The letters U and V were chosen because X, Y, and Z were already used to denote the
axes of objects in 3D space. The U coordinate represents the horizontal axis of the 2D
texture, and the V coordinate the vertical axis.
Quick Start
The process of UV unwrapping is roughly the following:
1. Define where to cut the model in order to flatten it out into 2D space. See Creating Seams on Your
Model.
2. Unwrap the model and preview the results. See Unwrapping Your Model and Previewing Its UVs.
3. Edit the generated UVs as necessary. See Editing UVs.
4. Apply a texture to your model. See Applying Textures.
USER GUIDE
1713
UV Unwrapping | To Create Seams
Creating Seams on Your Model

To prepare your model for UV unwrapping, you need to mark some seams. The seams tell
ModelBuilder where it's allowed to cut the model in order to flatten it out into 2D space.
Tip: If you don't have enough seams marked, the unwrapping will be very poor with lots of
overlapping faces as well as stretching and distortion. On the other hand, too many seams
can make it difficult to texture the object, as the gaps in the seams can easily become
noticeable on the textured object. If possible, it's a good idea to hide seams in areas where
they will not be seen.
To Create Seams
1. If you have created several objects using the same ModelBuilder node, click in the node
properties to hide all the other objects except the one you want to generate UVs for.
2. In the ModelBuilder toolbar, click to activate UV mode.

3. Repeat one or more of the following until you've got a good set of seams:
• To mark an edge as a seam, right-click on the Viewer and choose edge selection. Select an edge
on your model, right-click, and choose mark as seam.
• To mark an edge loop as a seam, right-click on the Viewer and choose edge loop selection.
Select an edge loop on your model, right-click, and choose mark as seam.
USER GUIDE
1714
UV Unwrapping | Unwrapping Your Model and Previewing Its UVs
• To mark all edges around a face as seams, right-click on the Viewer and choose face selection.
Select a face on your model, right-click, and choose mark as seam.
Tip: You can also select a face loop and mark all edges around it as seams. To do so, select
at least two faces on your model, right-click, and choose select face loop. Then, right-click
on the Viewer again and choose mark as seams.
• To remove an edge, edge loop, or face from the set of seams, select it in the Viewer, right-click,
and choose unmark as seam.
• To clear all seams, right-click on the object and choose clear all seams.
Any edges that have been marked as seams are displayed in red in the Viewer.
4. Proceed to Unwrapping Your Model and Previewing Its UVs below.
Tip: The easiest way to create a good set of seams for buildings (or any other roughly cubic
or cylindrical object) is to 1) select the top faces and mark them as seams, 2) select the
bottom faces and mark them as seams, and 3) select edges along the side of the model that
connect the top to the bottom and mark them as seams.
Unwrapping Your Model and Previewing Its UVs

Once you're happy with your set of seams, you are ready to unwrap the model.
To Unwrap a Model and Preview Its UVs
1. Right-click on the Viewer and choose object selection.

2. Right-click on your model and choose unwrap.
3. Press Return to complete the unwrap.
USER GUIDE
1715
UV Unwrapping | Unwrapping Your Model and Previewing Its UVs
ModelBuilder generates UV values for each of the vertices on your model. The unwrapped model
appears in the UV preview window in the top left of the Viewer.
4. To display a 2D texture image as a background in the UV preview window, connect the image to
ModelBuilder's tex input and make sure the Preview dropdown in the ModelBuilder properties
(or on top of the Viewer) is set to tex input. To also see the tex input on the model in the 3D
Viewer, set display to either textured or textured+wireframe in the ModelBuilder properties.
5. To zoom in or out of the UV preview window, press Ctrl/Cmd+Alt and drag.

6. To pan the UV preview window, press Ctrl/Cmd and drag.
7. To change the size of the UV preview window, click and drag on its lower right corner.
8. To hide the UV preview window, disable uv window at the top of the Viewer.
9. Proceed to Editing UVs below.
Tip: If you are not happy with the generated UVs, you can right-click on your model and
select clear existing uvs. If necessary, you can also change the seams and unwrap your
model again. Each time you do, ModelBuilder replaces the UVs you had with the newly
generated UVs.
USER GUIDE
1716
UV Unwrapping | Editing UVs
Editing UVs
You can tweak the controls at the top of the Viewer or edit the UVs manually in the UV preview
window.
1. Adjust the following controls at the top of the Viewer:
• iterations - The number of times the unwrapper applies its rules about how to move the UV
coordinates towards good locations. This gives you a speed vs. quality trade-off: more iterations
give a better result but take longer to unwrap. Note that the results can only be improved up to
a point, and where that point is depends on the complexity of your model.
• threshold - The unwrapper stops before reaching the maximum number of iterations if it sees
that the improvement from iteration to iteration is negligible. The threshold control tells it how
much change is considered negligible. If the amount of change between iterations is below the
threshold, the unwrapper stops. This is another way to trade off quality for speed. You generally
use this if you want a higher quality result: if you've increased iterations but the unwrapper is
stopping before reaching the maximum, you can lower the threshold to make it keep running.
• separation - The number of pixels to leave between each patch in the unwrapping. If you don't
have enough pixels between your patches, you can get color bleeding through from neighboring
patches when you use the texture. A higher value means more widely spaced patches, but also
more wasted space in your texture image.
As you edit these controls, the UV preview window updates to allow you to see the effect of your
changes.
2. If necessary, you can also edit the UVs manually in the UV preview window:
• To edit a single vertex, drag it to a new location.
• To edit several vertices together, select them and use the transform jack that appears to
translate, rotate, or scale the selection. The transform jack is the same one used elsewhere in
Nuke, so all the controls work the same way. For example, you can click and drag on the
handles to scale out from the center, or Ctrl/Cmd+click and drag to scale out from the opposite
edge.
USER GUIDE
1717
UV Unwrapping | Applying Textures
3. Proceed to Applying Textures.
Applying Textures
At this point, it may be that you've got several objects in the scene that you still want to project your
camera footage onto (see Projecting Textures onto Your Shapes), but there's one object that you want
to texture differently. There are two ways to do this:
• You can export your model as a separate geometry node and use an ApplyMaterial node to texture
it. This can be a convenient way to work if your geometry is finished and locked off, so you never
need to go back and change it. But what you lose with this is the "live" view: if you do need to go
back and edit the object, you have to export it again. See Method 1.
• Alternatively, you can add an ApplyMaterial node directly after your ModelBuilder node and tell
ApplyMaterial to ignore all the geometry that doesn't match the filter you give it. See Method 2.
Method 1
1. In the ModelBuilder properties, select the object that you want to texture differently.
2. At the bottom of the properties panel, set the bake menu to Selected geometry and click Bake.
ModelBuilder creates a geometry node for the selected object.

3. From the toolbar, select 3D > Shader > ApplyMaterial to create an ApplyMaterial node.
USER GUIDE
1718
4. Connect the geometry node you created in step 2 to the unnamed input of the ApplyMaterial
node.
5. Then, connect your 2D texture image to the mat input of the ApplyMaterial node.
The ApplyMaterial node applies the texture from the mat input onto your 3D geometry object. (To
be able to see this in the 3D Viewer, you may have to hide the object in the ModelBuilder
properties.)
6. Connect the ApplyMaterial node to your Scene node and use a ScanlineRender node to render all
the objects connected to that scene.
USER GUIDE
1719
Method 2
1. From the toolbar, select 3D > Shader > ApplyMaterial to create an ApplyMaterial node.
2. Connect the unnamed input of the ApplyMaterial node to your ModelBuilder node and the mat
input to your 2D texture image.
By default, ApplyMaterial applies the texture from the mat input onto all objects in your
ModelBuilder node.
3. To only apply the texture onto a particular object, open the ApplyMaterial properties and set filter
to name. This allows you to tell ApplyMaterial to ignore any geometry that doesn't match the
filter.
4. To set the filter, click the choose button. In the dialog that opens, select the object you want to
apply the texture to and click OK.
Tip: You can also Ctrl/Cmd+click or Shift+click to select multiple objects.
USER GUIDE
1720
For more information on ApplyMaterial, see Applying a Material Using the ApplyMaterial Node.
5. Connect the ApplyMaterial node to your Scene node and use a ScanlineRender node to render all
the objects connected to that scene.
USER GUIDE
1721
USER GUIDE
1722
Exporting Shapes to Separate Geometry Nodes | To Export Shapes to Separate Geometry Nodes:
Exporting Shapes to Separate

Geometry Nodes
If necessary, you can export shapes on your ModelBuilder model to separate geometry nodes. This
allows you to operate on one part of the scene separately from the rest. You can use the geometry
nodes in the same way as any other geometry nodes in Nuke.
To Export Shapes to Separate Geometry Nodes:

1. In the ModelBuilder properties, select any shapes you want to export to a separate geometry
node.
2. Make sure the bake menu under Export is set to Selected geometry and click Bake.
ModelBuilder creates a geometry node for the selected items in the scene.
Tip: The options in the bake menu are implemented in Python, and you can also use Python
to add your own entries to the menu. To see how the built-in options have been
implemented, have a look at the modelbuilder.py module in the nukescripts folder of your
installation (for more information on the location of this file, see Viewing More Examples).
Then, to create your own menu options, edit the modelbuilder.py file and use the
populateBakeMenu function to add entries where it says "# Add your own entries here, if
desired".
USER GUIDE
1723
Creating 3D Particles
Nuke's Particle node set is a solution for creating particles in a 3D environment. You can create things
like smoke, fog, falling snow, explosions, and bubbles - the possibilities are endless. You can use the
various Particle nodes for emitting, manipulating, and displaying limitless types of particles in your
3D scene.
Quick Start
1. Create a ParticleEmitter node, and connect it to a Viewer.
2. Connect a source of emission and a particle representation to the emit and particle inputs of the
ParticleEmitter. For more information, see Connecting Particle Nodes.
3. Modify your particles’ lifetime, velocity and other basic properties in the ParticleEmitter properties
panel. For more information, Emitting Particles.
4. Connect other particle nodes to the ParticleEmitter’s output. See Adjusting the Speed and
Direction of Particles, Modifying the Particles’ Movement and Adjusting Particle Simulation
Settings.
5. If necessary, cache your particle simulation in order to read it back in without the need for
recalculation. For more information, see Caching Particles.
Rain created using particles. Snow created using particles
USER GUIDE
1724
Connecting Particle Nodes |
Connecting Particle Nodes

In order to create particles, the minimum setup you need is the ParticleEmitter node. To connect your
Particle nodes:
1. Click the Particles menu in the Toolbar and select the ParticleEmitter node.
2. Connect it to a Viewer node.
3. Connect a 3D geometry object, or PositionToPoints point cloud, to the emit input of the
ParticleEmitter. A 3D object from which to emit the particles is optional: if you don’t use one, the
particles are emitted along the y axis from a point of origin.
4. To specify the appearance of your particles, connect an image or a geometry in the particle input
of the ParticleEmitter. This image or geometry is then multiplied and used as representations of
each of your particles. If you want to use more than one representation, connect further image or
geometries to the other numbered particle inputs. ParticleEmitter picks one of these at random
for each particle.
5. If you have another particle system that you’d like to connect to your new ParticleEmitter, you can
connect it to the merge input. You can also merge particle systems with the ParticleMerge node
(see Merging Particle Streams).
6. Now you’re ready to start modifying your particles to look the way you want. To do this, you can
pick any of the various particle nodes in the Toolbar’s Particle menu and connect them to the
ParticleEmitter node’s output, or to other particle nodes. Each of them has its own effect on the
particles and a set of controls you can make adjustments with.
Particles emitted from Sphere geometry.
USER GUIDE
1725
Emitting Particles |
Emitting Particles
The ParticleEmitter node is the first and only required node in a minimum particle setup. Once you’ve
created a ParticleEmitter, connected it to a Viewer and clicked play on the timeline, you’ll see the
default set of particles emitting (from a 3D geometry or point cloud, if you’ve connected one). You can
then adjust the ParticleEmitter controls to change the way the particles appear:
1. Set the channels in which particles are emitted. Channels a and b are arbitrary names for
channels which are useful if you want different particle force nodes to have an effect on separate
channels.
2. Use the start at field to pre-roll or delay the point at which the first particles are emitted. For
example, particles could imitate snow that has fallen already, instead of displaying the first flakes
falling down, by using a negative start at frame.
3. Select the emission order and rate for the particles. Set:
• emit from - select points, edges, faces or bbox (bounding box) to specify from which part of
the 3D object, or point cloud, particles are emitted.
emit from > points emit from > edges
emit from > faces emit from > bbox

• emit order - select the order that the particles are emitted:
• randomly - produces a random emission order.
• uniformly - emits all particles at the same time.
USER GUIDE
1726
• in order - emit particles as a multiple of the emission rate. For example, an emission rate of 2
could emit from two points, edges, or faces at a time.
Particles emitted in order. Particles emitted uniformly.

• randomize type - select whether or not particles are emitted randomly. Select no random
direction for a specific direction based on the emit object’s normals, randomized direction for
a randomly selected initial direction, or randomize outwards for a randomly selected direction
(dependent on the emit from selection):
• bbox - particles move randomly away from the center.
• points, edges, and faces - particles move randomly away from the origin, but at no more than
90 degrees from the nearest normal.
• emission rate - select the number of particles emitted per frame. This is an exact number, and
it is affected by the rate channel. If the channel is at a varying value, the emission rate also
increases or decreases.
• only emit from selected points - specify whether the particles are emitted from the object, or
selected vertices of the object using a GeoSelect node. You can also emit from selected points in
a PositionToPoints point cloud using any normals information present.
See 3D Selection Tools for more information.
Vertices selected using Emitting particles only from

an upstream GeoSelect node. the selected vertices.
• rate variation - specify the range of variation for emitting particles. If you set this to zero,
particles are emitted at a constant even level.
• rate channel - select a specific channel to which you want to emit particles. Unchecking this is
the same as selecting none, and in that case particles are emitted to all the channels. For
USER GUIDE
1727
example, if you are emitting from a Card node that has a Ramp texture, it emits particles from
the light parts of the ramp at a higher rate (values closer to 1) than from the dark parts (values
closer to 0).
4. Set the particles’ color and channels. Enter:
• color - select a color for your particles. Use this if you’re not using the particle input.
• color from texture - check to tint the particles with the colors from the geometry’s texture. If
you leave this unchecked, the particles get color only from their own texture.
• channels - select the channels you want to apply particles to. By default, particles are emitted to
channel a. Channels a and b are arbitrary names for channels which are useful if you want
different ParticleEmitter nodes or other particle force nodes to have an effect on separate
channels. An example of this might be if you want two ParticleEmitter nodes emitting particles,
one to a channel, the other to b channel, so further down in your particle stream you can apply
an effect to a specific set of particles.
5. Select how long you want the particles to exist. Set:
• max lifetime - specify the maximum life span for a particle in frames.
• max lifetime range - adjust the range within which your particles’ lifetime varies. If you set this
to 0, all particles have the same lifetime.
• lifetime channel - select a channel of the input geometry texture that you want to use to
modulate the lifetime. For example, if you are emitting particles from a Card node with a Ramp
texture in the lifetime channel, particles emitted from the lighter parts of the ramp (values
closer to 1) would have a lifetime value closer to that set in the max lifetime than particles
emitted from the dark parts (values closer to 0).
• halflife - select the number of frames over which the number of the particles emitted is halved.
6. Adjust the velocity and rotation for your particles. Set:
• velocity - specify the initial speed at which you want your particles to move.
• velocity range - adjust the range within which you want your particles’ velocity to vary. If you set
this to 0, the velocity doesn’t vary between particles.
• velocity channel - select a channel of the input geometry texture that you want to use to
modulate the velocity. For example, if you are emitting from a Card node that has a Ramp
texture, it emits particles from the light parts of the ramp at a higher velocity (values closer to 1)
than from the dark parts (values closer to 0).
• rotation velocity - adjust the initial speed at which each particle rotates around its individual Y-
axis. The Y-axis points to the direction the particles were initially emitted, and then stays
unchanged (unless you use the ParticleMotionAlign node to change its direction). Rotational
velocity is most useful when you’re emitting particles from a geometry object.
• rotation velocity range - adjust the scale of a random variation in the rotation velocity value.
Value of 0 means the rotation velocity value is as set, value of 1 means it’s very random.
USER GUIDE
1728
• rotation velocity channel - Select a channel of the input geometry texture that you want to use
to modulate the speed of rotation. For example, if you are emitting from a Card which has a
Ramp texture, the rotation velocity of the particles emitted from the light part of the Ramp
(values closer to 1) is greater than that of those emitted from the dark parts (values closer to 0).
• transfer velocity - adjust this to transfer the velocity of the initial emitter to the particles. If you
set this to 0, no velocity is transferred to the spawned particles. At value 1, full velocity is
transferred from the originating particle.
• transfer window - adjust the time, in frames, to look forward or backward in order to
determine the velocity that should be transferred to the particles.
7. Modify the size and mass of the particles. Set:
• size - specify the initial size of each particle.
• size range - specify the range within which your particle size varies.
• size channel - select a channel of the geometry texture that you want to use to modulate the
size of the particles. For example, if you are emitting particles from a Card node which has a
Ramp texture in the size channel, the size of the particles emitted from the lighter part of the
Ramp (values closer to 1) is greater than that of those emitted from the darker parts (values
closer to 0).
• mass - adjust the initial mass of each of your particles. The mass of the particles only becomes
relevant when you apply a force to your particles, such as one generated by the
ParticleDirectionalForce node.
• mass range - adjust to produce a random variation in the mass value. Simply put, 0 value
means the particles’ mass is the value specified by the mass control, whereas a value of 1 means
it’s very random.
• mass channel - select a channel of the input geometry texture that you want to use to modulate
the mass. For example, if you are emitting from a Card which has a Ramp texture in the mass
channel, the particles from the light part of the Ramp (values closer to 1) has a higher mass
value than from the dark parts (values closer to 0). With this control, you can emit particles with
different masses from different areas based on the input geometry’s texture.
• spread - adjust the extent to which you want your particles to spread in different directions
during their lifetime. By default, this forms a cone around the direction of emission. If you set
this to zero, each particle has a straight trajectory.
8. Modify the way your particles are affected by the ParticleEmitter’s inputs. Adjust:
• input order - if you’re using more than one particle input, you can select which particle input
Nuke should select when creating particles. Select Randomly to pick one of the inputs
randomly, or in order to rotate the inputs in numerical order.
• start at - select which frame of the particle input should be the representation of each new
particle. Select first to pick the first frame of the particle input for each new particle. Select in
order to pick consecutive frames from the input for each new particle. Select current to pick
USER GUIDE
1729
the frame where the particle was emitted. Select random to pick a random input frame for each
new particle
• limit to range - limit the particle output to the representation’s frame rate, looping the frame
range when in order or current is selected.
• advance - use this to determine if a particle should animate after being emitted. Select
constant to keep the same representation throughout the particle’s lifetime. Select in steps - to
animate the representation frame by frame. Select randomly to animate the representation one
random frame after another.
9. Vary the results of your range controls with the random seed field. Enter the random number
used in the range controls (such as max lifetime range) to achieve slightly different effects.
Tip: When you set large values for any of the particle controls it might take Nuke a few
moments to calculate the result. In such a case a progress bar appears with a Cancel button
you can use to cancel the calculation if necessary.
USER GUIDE
1730
Spawning Particles with ParticleSpawn |
Spawning Particles with

ParticleSpawn
If you’re looking to have your existing particles emit even more particles, you should turn to
ParticleSpawn.
1. Connect the ParticleSpawn node to your particle stream (the ParticleEmitter output, for example).
All the particles emitted now start spawning more particles.
2. Adjust the ParticleSpawn controls. Most of the ParticleSpawn controls are identical to those in the
ParticleEmitter node (see Emitting Particles), with only a few exceptions:
• transfer velocity - adjust this to transfer the velocity of the initial emitter to the particles. If you
set this to 0, no velocity is transferred to the spawned particles. At value 1, full velocity is
transferred from the originating particle.
• conservation of mass - check if you want the mass of spawned particles to be removed from
the mass of the original particle. If the mass of a particle is zero at the end of a frame, it gets
deleted.
• conservation of momentum - check this to subtract the momentum of the spawned particles
from the original particle, in correspondence with Newton’s third law of motion.
• align velocity to direction of motion - check to align velocity with the direction of the particles’
motion.
• inherit color - check to take the particle color from the originating particle. Otherwise the color
is determined by the color control.
Spawned particles.
USER GUIDE
1731
Adjusting the Speed and Direction of Particles | Applying Gravity to Particles
Adjusting the Speed and

Direction of Particles
Applying Gravity to Particles

When applying gravity to particles, as opposed to our familiar gravity, Nuke doesn’t restrict you to a
certain direction but works in any or all of the x, y and z directions. You can add gravity either by:
• Using the ParticleDirectionalForce to apply a directional force. Just connect it to your particle
stream, and adjust the strength of the force in the x, y, and z directions by entering x, y, and z values
in the strength fields.
OR
• Using the ParticleGravity node. When you connect the ParticleGravity node to your particle stream,
an arrow appears in the Viewer, which you can then use to determine the direction and velocity of
the gravity. The bigger and longer the arrow, the stronger the effect. Instead of adjusting the arrow,
you can also use the controls in the properties panel:
• from - enter the point of origin for the gravity effect on the x, y, and z axis. This determines
from which direction the force appears to come, indicated by the base of the arrow in the
Viewer.
• to - enter direction for the gravity effect on the x, y, and z axis. This is indicated by the point of
the arrow in the Viewer.
Aligning Particles
To align your particles’ motion, direction, and orientation, you can use two nodes:
• You can add the ParticleMotionAlign node in your particle stream to realign all the particles along
their direction of motion. This is useful if your particles seem too rigid in their movement.
• Add the ParticleLookAt node to determine a 3D point that all the particles are looking toward. To
specify this point, adjust the position control. The x, y and z coordinates specify the point that the
particles are looking at.
Controlling Particle Speeds

The ParticleSpeedLimit node limits the particles to a specified minimum and maximum speed.
USER GUIDE
1732
Adjusting the Speed and Direction of Particles | Attracting Particles to a Specific Point
1. Connect it to your particle stream.

2. In the properties panel, adjust:
• minimum - the minimum speed at which each particle can travel.
• maximum - the maximum speed at which each particle can travel.
Attracting Particles to a Specific Point

With the ParticlePointForce, you can attract particles to or repel them from a certain point in the 3D
space.
1. Connect the node to your particle stream.
2. Adjust the ParticlePointForce controls:
• strength - set the strength of the force attracting or repelling particles. Negative values cause
attraction, positive values repulsion.
• falloff - choose how quickly strength of the attraction falls off with respect to the distance by
selecting the type of falloff, none for no falloff, inverse for inverse falloff, or inverse square for
inverse falloff squared.
• radius - set the radius of influence. Outside of this radius, no particles are affected by point
force.
• position - set the position of the point that attracts or repels particles. You can use an animated
or a still Axis node expression-linked to these fields, or you can just enter a position value
manually.
Particles attracted to a 3D point.
USER GUIDE
1733
Modifying the Particles’ Movement | Bouncing Particles off Objects
Modifying the Particles’

Movement
Bouncing Particles off Objects

With the ParticleBounce node, you can make your particles bounce off the shape of a 3D object
instead of traveling through it. Connect this node to your other particle nodes, and adjust the
ParticleBounce controls:
1. To set your particles to bounce off the outside of the bounce object (specified by the object
control), select a mode in the external bounce mode dropdown: none to apply no external
bounce effect, bounce to bounce the particles, kill to end the life of the particles as they bounce.
2. Select a channel where a particle should be assigned to when a bounce is detected in the new
channels dropdown. Setting this to none doesn’t perform any channel assignment.
3. In the bounce field, specify the strength of the external bounce effect, and adjust friction to
control the amount of friction for external bounce effect.
4. To set your particles to bounce off the inside of the bounce object (specified by the object
control), use the internal bounce mode dropdown:
• none - to apply no internal bounce effect.
• bounce - to bounce the particles.
• kill - to end the life of the particles as they bounce.
5. Select a channel where a particle should be assigned to when a bounce is detected in the new
channels dropdown. Setting this to none doesn’t perform any channel assignment.
6. In the bounce field, specify the strength of the internal bounce effect, and adjust friction to
control the amount of friction for the internal bounce effect.
7. Select the object you want to use for the bounce effect in the object dropdown:
• plane, sphere or cylinder - add a Nuke standard primitive to use as the bounce surface for the
particle system. You can use these shapes to quickly test your particle effects.
• input - add custom geometry attached to the geometry input.
8. Use the transform controls to fine-tune your geometry's position to achieve the desired bounce
result.
USER GUIDE
1734
Modifying the Particles’ Movement | Adding Drag to Particles
Particles bouncing off a plane.
Adding Drag to Particles

With the ParticleDrag node you can apply drag on your particles. This gradually slows them down
over time.
1. Connect the ParticleDrag node to your particle stream.
2. Adjust the ParticleDrag controls:
• drag - increase to add a drag effect to your particles, slowing and stopping their movement as
they are making distance from the particle center.
• rotational drag - increase to add a drag effect to your particles’ rotation, slowing and stopping
their rotation.
Tip: You can also enter negative values to apply a reverse drag effect and speed the particles
up.
Adding Turbulence Motion on Particles

The ParticleTurbulence node applies Perlin noise to the particle movement in the x, y and/or z
directions.
1. Connect a ParticleTurbulence node to your particle stream.
2. Adjust the ParticleTurbulence controls:
USER GUIDE
1735
Modifying the Particles’ Movement | Adding Spiral Motion to Particles
• strength - set the strength for the turbulence effect on the x, y and z axes.
• scale - set the scale of the effect, or the size of the area affected, on the x, y and z axes.
• offset - set the offset applied to the effect, or the offset of the area affected, on the x, y and z
axes.
Adding Spiral Motion to Particles

The ParticleVortex node applies a circular force to the particles and attracts them to an imaginary
line, thus creating a whirlpool of particles.
1. Connect the ParticleVortex node to your particle stream.
2. An arrow appears in the Viewer, which you can drag to determine direction and velocity of the
vortex effect. The bigger and longer the arrow, the stronger the effect. Alternatively, you can use
the from and to controls in the properties panel.
3. In the ParticleVortex properties panel modify the parallel effect with the parallel control. This
accelerates the particles in the direction of the imaginary vortex center line. If you set this to 0, no
parallel force is applied, and positive and negative values determine the direction of the force.
Adjust the parallel falloff to choose how quickly strength of the parallel force falls off with
respect to the distance by selecting the type of falloff, none for no falloff, inverse for inverse
falloff, or inverse square for inverse falloff squared.
4. With the tangential slider you can force the particles to circulate the vortex center line. Adjust
tangential falloff to choose how quickly strength of the tangential force falls off with respect to
the distance by selecting the type of falloff, none for no falloff, inverse for inverse falloff, or
inverse square for inverse falloff squared.
5. Use the radial slider to adjust the force that attracts the particles to (positive values), or repels
(negative values), them from the center line. Adjust radial falloff to choose how quickly strength
of the radial force falls off with respect to the distance by selecting the type of falloff, none for no
falloff, inverse for inverse falloff, or inverse square for inverse falloff squared.
Tip: If you want to create a helix of particles, you can turn up both the parallel and
tangential values. This creates a particle vortex in a shape of a corkscrew.
USER GUIDE
1736
Modifying the Particles’ Movement | Adding a Wind Effect to Particles
ParticleVortex on a sphere.
Adding a Wind Effect to Particles

With the ParticleWind, you can simulate a wind blowing on your particles.
1. Connect the ParticleWind node to your particle stream. An arrow appears in the Viewer, which you
can then use to determine direction and velocity of the wind. The bigger and longer the arrow, the
stronger the wind effect. Alternatively, you can use the from and to controls in the properties
panel.
2. In the ParticleWind controls, check air resistance to enable a drag effect on the wind.
3. Adjust the drag slider to increase or decrease the simulated air resistance.
USER GUIDE
1737
Adjusting Controls Common to Several Particle Nodes | Particle Rendering Controls
Adjusting Controls Common to

Several Particle Nodes
Many of the particle nodes share a number of controls, such as rendering and transform controls.
The following covers the use of these.
Particle Rendering Controls

The first controls on the properties panels of all the particle nodes have to do with how the particles
are output to the Viewer and rendered out. To read more about the display, selectable and render
controls, see Object Display Properties.
Particle Transform Controls

Several of the particle nodes have a set of transformation controls in their properties panels. You can
use these controls for example to translate, rotate and skew the force that the node applies on your
particles. For more information on how transform controls work, see Transforming Geometry,
Cameras, and Lights.
Condition and Region Controls

Various particle nodes have controls on the Conditions and Region tabs in the properties panel.
Conditions tab
On the Conditions tab, use the following controls to restrict the way in which the specific node affects
your particles:
• probability - set the probability that this node affects your particles. If you set this to zero, the node
won’t affect any particles, and if the value is 1, the node affects every particle.
• min age - set this to limit the effect of this node only to particles above this minimum age. The age
of the particle is its lifetime normalized between 0 and 1.
• max age - set this to limit the effect of this node only to particles below this maximum age. The age
of the particle is its lifetime normalized between 0 and 1.
• random seed - enter an integer to change the results of generated randomness in your particles.
You can achieve slightly different effects by changing this number.
USER GUIDE
1738
Adjusting Controls Common to Several Particle Nodes | Condition and Region Controls
• channels - specify which particle channels the effect of this node should be applied to. Channels a
and b are arbitrary names for channels which are useful if you want different ParticleEmitter nodes
or other particle force nodes to have an effect on separate channels.
Region tab
1. Using the region control you can select the region which you want to use to confine the particle
effect to. For example, if you select a sphere, only particles inside that sphere-shaped region is
affected by the particle effect. Select none to apply no confining region, or the appropriate shape
for a region between sphere, box, half-space and cylinder.
2. You can also check invert region to only affect the particles outside the region specified.
3. Then move on to adjusting your region with the transform controls. For more information, see
Transforming Geometry, Cameras, and Lights.
USER GUIDE
1739
Adjusting Particle Properties Using Curves | Condition and Region Controls
Adjusting Particle Properties

Using Curves
With the ParticleCurve, you can apply a curve to particle properties (such as size or mass) to change
them over time.
1. Connect the node to your particle node stream.
2. Adjust the curves in the ParticleCurve properties panel. The x axis represents the lifetime of the
particles.
• r - adjust the curve for the red channel.
• g - adjust the curve for the green channel.
• b - adjust the curve for the blue channel.
• a - adjust the curve for the alpha channel.
• size - adjust the curve for the size of the particles.
• mass - adjust the curve for the mass of the particles.
A size curve modifying the particles to grow

in size toward the end of their lifetime.
Note: If you’re using an image or a 3D object as your ParticleEmitter’s particle input, the
ParticleCurve might not alter the colors of the particles as expected.
3. If you want, you can adjust the curve for your particles’ alpha channel so that each particle fades
to invisibility toward the end of its lifetime.
USER GUIDE
1740
Adjusting Particles Using Expressions | Condition and Region Controls
Adjusting Particles Using

Expressions
With the ParticleExpression node, you can adjust your particles by setting expressions on their
attributes. Using expressions gives you a vast variety of ways of adjusting how your particles behave.
You can use a similar expression syntax as you would elsewhere in Nuke, with the exception that
some functions that work in normal Nuke expressions aren't available in particle expressions and
vice versa.
The main difference between Nuke’s Expression node and ParticleExpression is that particle
expressions can return a 3D vector instead of just a single floating point number. If a particle
expression returns a single number N in a field that expects a vector (such as velocity or acceleration)
it is converted into a vector with N for each of its components. For more information about the
functions you can use with ParticleExpression, see Particle Expression Functions.
1. Connect the ParticleExpression node to your particle stream.
2. In the ParticleExpression controls you can use four temporary expression fields. With these, you
can set up any expression on a particle attribute and then give it a temporary name. This
temporary name can then be used in the following fields to refer to the corresponding temporary
expression. This can be useful if you need to use a long expression in several fields. By default,
the per-particle box is ticked to make the expressions affect each particle individually. Uncheck
the box to apply the expression to all particles at once. For more information on expressions, see
Expressions.
3. You can also set expressions on a set of attribute fields:
• color - set an expression to edit the color of the particles.
• opacity - set an expression to edit the opacity of the particles.
• size - set an expression to edit the size of the particles.
• mass - set an expression to edit the mass of the particles.
• accel - set an expression to edit the acceleration of the particles.
• force - set an expression to edit the force of the particles.
• pos - set an expression to edit the position of the particles. For example, enter sin (age * 10) * 5
to emit particles randomly on a single static line.
• vel - set an expression to edit the velocity of the particles.
• onlyonnew - check this next to each attribute field to make the expression only affect new
particles and ignore any existing ones.
USER GUIDE
1741
Adjusting Particles Using Expressions | Particle Expression Functions
Particle Expression Functions

Here are some functions you can use with the ParticleExpression node:
Function Purpose Related Functions
abs(f) Returns the absolute value of f. See also: fabs.
acos(f) Returns the arc cosine of f. The result See also: cos, cosh, asin, atan2.
is the angle in radians whose cosine
is f.
age The age of the particle, in frames. -
asin(f) Returns the arc sine of an angle. The -

result is the angle in radians whose
sine is f.
atan(f) Returns the arc tangent of an angle. -

The result is the angle in radians
whose tangent is f. Can be called with
one or two arguments; if called with
two arguments it's equivalent to
atan2.
atan2(x,y) Returns the principal value of the arc See also: sin, cos, tan, asin, acos,
tangent of y/x, using the signs of the atan, hypot.
two arguments to determine the
quadrant of the result.
ceil(f) The ceiling of f, rounding any -

fractional part up.
color The color of the particle. This is a 3D -

vector value, where x() is the red
component, y() is green and z() is
blue.
cos(f) Returns the cosine of angle f. The See also: sin, tan, asin, acos, atan,
angle is in radians. hypot.
USER GUIDE
1742
cosh(f) Returns the hyperbolic cosine of f. -
exp(x) Returns the value of e (the base of -

natural logarithms) raised to the
power of x.
fabs(f) A synonym for abs(f). -
floor(f) The floor of a number, rounding any -

fractional part down.
fmod(x, y) Floating point modulus function. -

fmod(x, y) returns the remainder
after dividing x by y.
hypot(x,y) The Euclidean distance function. -

hypot(x, y) returns the length of the
hypotenuse of a right-angled triangle
where the other sides have length x
and y respectively.
id The index number for each particle. -
int(f) Convert floating point number f to an See also: trunc(f).

integer, discarding any fractional
part.
life The maximum lifetime of a particle, -

in frames.
log(x) Returns the natural logarithm of x. -
log10(x) Returns the base 10 logarithm of x. -
mag(v) Returns the magnitude (length) of the -

3D vector v.
mass The mass of the particle. Used when -

applying a force to a particle.
new Returns 1 if the particle has just been -

created, 0 otherwise.
USER GUIDE
1743
norm(v) Normalise the 3D vector v to have a -

length of 1.0 while pointing in the
same direction.
opacity A number between 0.0 and 1.0, where -

0.0 is fully transparent and 1.0 is fully
opaque.
pos The position of the particle. This is a -

3D vector.
pow(x, y) Returns x raised to the power of y. -
pow2(f) Returns the square of f, i.e. f raised to -

the power of 2.
random Returns a random number. -
randomv Returns a random vector. The vector See also: uniform samplesphere.
directions are uniformly distributed
around the unit sphere.
rint(f) Round the floating point number f to -

an integer.
sin(f) Returns the sine of the angle f. The -

angle is in radians.
sinh(f) Returns the hyperbolic sine of f. -
size The size of the particle. -
sqrt(f) Returns the square root of f. f must -

be greater than or equal to zero.
tan(f) Returns the tangent of angle f. The -

angle is in radians.
tanh(f) Returns the hyperbolic tangent of f. -
trunc(f) A synonym for int(f). -
uniform A synonym for randomv. -
USER GUIDE
1744
samplesphere
v(x, y,z) Create a vector from three separate -

numbers.
vel The velocity of the particle. This is a -

3D vector.
x(v) Get the x component of the 3D vector -

v.
y(v) Get the y component of the 3D vector -

v.
z(v) Get the z component of the 3D vector -

v.
USER GUIDE
1745
Adjusting Particle Simulation Settings | Particle Expression Functions
Adjusting Particle Simulation

Settings
If you want to adjust how many steps of particle simulation take place per animation frame, you can
use the steps per frame control in the ParticleSettings properties panel. Sometimes simulations
cannot generate enough accuracy only calculating once per frame, and the resulting particle
movement can appear jagged. Steps per frame forces the simulation to assess the movement of the
particles multiple times per frame to enable a more analogue movement. You should enter the
lowest value you can, as having an unnecessarily high value can slow down your particle calculations.
USER GUIDE
1746
Merging Particle Streams | Particle Expression Functions
Merging Particle Streams

If you have more than one set of particle nodes, and you want to combine them into one stream,
ParticleMerge is your node. You can merge as many particle streams as you need into a single
ParticleMerge.
1. Connect a ParticleMerge node to other particle nodes at any point in the particle stream.
2. Attach another particle stream (or a ParticleEmitter node) to another one of ParticleMerge’s
numbered inputs and you’re all set.
USER GUIDE
1747
Controlling Particles by Channel | Particle Expression Functions
Controlling Particles by Channel

The ParticleToGeo node allows you to control particles in a simulation by channel, giving you the
ability to isolate certain particles at any point in the Node Graph. For example, you might want to
freeze certain particles within a simulation while allowing others to emit as normal or apply a particle
effect to the particles in a single channel.
ParticleToGeo allows you to control geometry or sprite particles by channel, but in the case of sprite
particles, you can also influence their alignment using the align mode control.
The following example uses a very simple particle system containing three colored particles in
separate channels and adds effects to one channel using ParticlesToGeo. The other particles are
unaffected.
USER GUIDE
1748
Split off the channel you're interested in using ParticleToGeo and MergeGeo nodes. Set the
ParticleToGeo node's channels control to the channel you want to affect. In this example, channel c
with the Sphere geometry is the affected particle.
To add a new texture to the geometry, you can use the AddMaterial node to apply a material to the
affected channel.
USER GUIDE
1749
USER GUIDE
1750
Caching Particles | Particle Expression Functions
Caching Particles
The ParticleCache node allows you to store the geometry simulation for a particle system to file. It can
then be read back in different sessions of Nuke or on different machines without the need for
recalculation.
This allows a particle system to be produced by an artist and then used by a render farm without
recalculation, speeding up render times.
Note: The ParticleCache node doesn't replace the particle system. It just stores the
simulation to disk and still relies on the particle system being connected in the same way
with the same inputs. If anything in the particle system changes, the ParticleCache node
detects this and shows an error to alert you to changes that have potentially been made
upstream without your knowledge.
To cache your particle simulation:

1. Once you are happy with your particle simulation, select Particles > ParticleCache to create a
ParticleCache node.
2. Place the ParticleCache node at the bottom of a single particle system or multiple merged particle
systems.
USER GUIDE
1751
Caching Particles | Particle Expression Functions
Note: You cannot place a ParticleCache node in the middle of a string of particle nodes or
beneath a Scene node connected to two separate particle streams.
3. In the ParticleCache properties, click the folder icon next to the file field and navigate to the
directory where you want to cache the particle simulation. After the directory path, enter a name
for the cache files and include the frame number variable (for example, ####) in the name. Click
Open.
ParticleCache uses the .nkpc file extension.
Note: ParticleCache may need to render up to 100 sub-frames. To account for this, it adds
decimals to the file name's frame number. For example, if the file name in the file field is
particle_cache.####.nkpc, ParticleCache may generate files called particle_
cache.0001.01.nkpc, particle_cache.0001.02.nkpc, and so on.
4. If you have nodes downstream that generate motion blur, the particle system may need to
request frames outside the normal frame range. If this is the case, increase the padding value in
the ParticleCache properties to set the number of extra frames added to the start and end of the
ParticleCache render.
5. Click Render.
ParticleCache renders your particle simulation out to file frame by frame.
6. To use the cache data, enable read from file.
If you get a "Particle cache data not found" error, return to step 4 and increase the padding value.
USER GUIDE
1752
PrmanRender
PrmanRender is a render node that works together with Pixar’s PhotoRealistic RenderMan® Pro
Server software to give you an even better quality render result. PrmanRender is an alternative to
Nuke’s ScanlineRender with additional features for rendering 3D scenes.
Setting Up RenderMan Pro Server and PrmanRender

In order to use the PrmanRender node, you need to have Pixar’s PhotoRealistic RenderMan Pro
Server 20, or earlier, installed and licensed on your machine (for brevity, we call it RenderMan from
now on). To do this:
1. If you're using Mac OS X 10.8 (Mountain Lion) or above, make sure X11 is installed on your system.
Unlike previous versions of the operating system, 10.8 and above do not have X11 installed by
default. For more information, see http://support.apple.com/kb/HT5293.
2. Follow the instructions in the RenderMan installation guide to get RenderMan working on your
machine. This is in most cases enough to get you going with using both RenderMan and the
PrmanRender node in Nuke. Note that RenderMan specifically needs two environment variables
set in order to work with Nuke:
• RMANTREE - This needs to point to the location of your RenderMan distribution.
• Depending on which operating system you’re using either DYLD_LIBRARY_PATH (on Mac), LD_
LIBRARY_PATH (on Linux) or PATH (on Windows). This environment variable needs to point to
"%RMANTREE%/lib". For more information on setting environment variables, see Environment
Variables.
3. To make sure your PrmanRender node is working in Nuke, start the application from the terminal
(for more information, see the Installation and Licensing chapter).
4. Create a PrmanRender node in Nuke (3D > RenderMan > PrmanRender) and connect it to your
nodes. You can try the following combination of nodes as a test: Checkerboard > Cube >
PrmanRender > Viewer.
USER GUIDE
1753
Using The PrmanRender Node |
Using The PrmanRender Node

The PrmanRender node can render nearly everything that you previously used ScanlineRender for,
but with PrmanRender, you have control over aspects like shadows and reflections in your render
result.
Connect the PrmanRender node to your Scene node, Camera node and any optional inputs, in the
same way you would connect ScanlineRender. For more information about connecting a
ScanlineRender node, see Setting Up a Scene.
On the PrmanRender tab, you can select which aspects you’d like to render out by checking
shadows, reflections, refractions or dof (depth of field). For more information, see Shadows,
Reflections, Refractions and Depth of Field.
You can select your projection mode in the projection mode dropdown:
• Perspective - objects in front of the camera have the illusion of depth defined by the camera’s
focal-length and aperture.
• Orthographic - objects are viewed using a parallel projection.
If necessary, also adjust:

• overscan - to set how many pixels are rendered over the right/left and top/bottom of the frame, if a
subsequent operation requests this.
• ambient - to add global ambient lighting.
USER GUIDE
1754
Render Quality |
Render Quality
On the Sampling tab, you can adjust controls that affect your render quality. Adjust:
• ray trace max depth - to set the maximum depth of the view rays PrmanRender uses to trace your
scene and make render calculations.
• pixel samples - to set the number of samples to render per pixel. Having more samples increases
your render quality, but also increases render time.
• filter - to select a texture sampling filtering algorithm. For more information on the different
options, see Choosing a Filtering Algorithm.
• antialiasing filter - to select an antialiasing filter: box, triangle, catmull-rom, sinc, gaussian,
mitchell, separable-catmull-rom or blackman-harris.
• antialiasing filter size - to select the size of the antialiasing filter.
• shading rate - to set the shading calculation for primitives. This value, along with pixel samples,
directly affects your rendering time and the final quality of your results. A small shading rate value
means your render takes more time, but the quality is very high. A large value on the other hand
means your render is faster, but the final quality is not as good.
USER GUIDE
1755
Shadows, Reflections, Refractions and Depth of Field |
Shadows, Reflections, Refractions

and Depth of Field
On the PrmanRender tab, you can select if you want to render shadows, reflections, refractions and
depth of field, or all of them into your result. All these effects are calculated using a retracing method
that is based on drawing rays from the camera to the object. Check the box for:
• shadows - to add shadows to your render. You can adjust the parameters for your shadows on your
Light node’s Shadows tab. For example, if you adjust the sample width, the shadows are softer. For
more information, see Casting Shadows.
• reflections - to add reflections to your render. You can adjust the parameters for your reflections
on your Reflection node’s properties panel. See Using the Reflection Node.
• refractions - to add refractions to your render. You can adjust the parameters for your refractions
on your Refraction node’s properties panel. See Using the Refraction Node.
• dof - to add depth of field to your render.
USER GUIDE
1756
Motion Blur Parameters |
Motion Blur Parameters

On the Sampling tab, you can adjust controls that affect motion blur. Adjust:
• motion blur samples - to set the number of samples to render per pixel when motion blurring.
• shutter - Enter the number of frames the shutter stays open when motion blurring. For example, a
value of 0.5 would correspond to half a frame. Increasing the value produces more blur, and
decreasing the value less.
• shutter offset - Select when the shutter opens and closes in relation to the current frame value
when motion blurring:
current frame, enter a negative value. For example, a value of -0.5 would open the shutter half a
• randomize time - adjust this to add randomness to the distribution of samples in time so they
don’t produce regularly spaced images. The larger the value, the larger the time difference between
the samples.
• shutter opening - to set the shutter opening behavior. Select:
• none - to use the default shutter opening method, resulting in instantaneous timing to open
and close.
• linear - to set the shutter to open and close in linear intervals.
• bezier - to set the shutter to open and close more gradually, according to a Bezier curve.
USER GUIDE
1757
Shader Parameters |
Shader Parameters
On the Shader tab you can select which channels are affected by motion vectors and output vectors.
• Click motion vectors and select the type of vectors you’d like to render:
• off - no motion vector information is rendered.
• velocity - store the velocity of every single pixel in the motion vector channels.
• distance - for every pixel, store the distance (in pixels) between samples in the motion vector
channels.
• Adjust motion vector channels to select which channels you want the motion vectors to be output
to.
• Check output vectors if you want to render output vectors.
• Select which channels to apply surface points and surface normals from the surface point and
surface normal dropdowns.
USER GUIDE
1758
RIB Parameters |
RIB Parameters
With the RIB (RenderMan Interface Bytestream) parameters, you can choose to filter the information
Nuke generates for RenderMan, set your arguments to it, and output your own RIB file. On the RIB
tab:
• filter - check this to filter the information on your scene, generated by Nuke for RenderMan. In order
to do this, Nuke calls a Python function called nukescripts.renderman.filterRIB. Filtering can make
the render startup slightly slower as a temporary RIB file is created for each render.
• arguments - specify your arguments for filtering. This string is passed by Nuke’s Python filter
function as extra arguments to RenderMan. If you want to use your own filter, you can also replace
Nuke’s Python function, and have your arguments passed directly to your own Python function. For
example, you could set the filter arguments to "-rif myfilter.so" to load your own RenderMan
Interface filter.
Tip: For further details on filtering, have a look at your RenderMan documentation.
A RIB file is a RenderMan-compatible ASCII file with information that Nuke generates when rendering
your footage. To output a RIB file:
1. In the file field under output, specify the file name and location for your RIB file.
2. Click Execute.
USER GUIDE
1759
Using the ModifyRIB Node |
Using the ModifyRIB Node

You can use the ModifyRIB node to insert RIB (RenderMan Interface Bytestream) statements into your
script to modify a RIB stream before it’s passed to the PrmanRender node. This can be useful in
situations where you might want to adjust the shading on the surface of an object, replace it, or
perform a variety of transformations and adjustments to multiple objects in a scene. For example, to
replace an object in your script with basic geometry from RenderMan:
1. Add a Scene node to the basic script that you created at the beginning of the chapter and connect
two Light nodes and a Camera.
2. Click 3D > Modify > RenderMan > ModifyRIB and insert the node between the Cube and the
Scene.
3. Next to the archive field, uncheck use. This activates the statements field.
Note: You can load a RIB archive by leaving use checked and clicking the Select file button
to locate it.
4. In the ModifyRIB control panel select replace from the operation dropdown menu and enter the
following RIB statement in the statements field:
Sphere 0.25 -0.25 0.25 360
You should now see a basic sphere in the Viewer where the cube was.
5. To change the color of the sphere and add a basic surface shader to the sphere, enter the
following statements underneath:
Color 0.4 0.6 0.1
Surface "wood"
This changes the color of the sphere to green and applies a wood-like finish to the surface.
Note: If you’re copying statements directly from a RIB file, only copy statements from
between WorldBegin and WorldEnd as copying the entire contents of the file may result in
instability issues.
For more information on RIB statements, please refer to the documentation provided with
RenderMan.
USER GUIDE
1760
Using the Reflection Node |
Using the Reflection Node

Reflection is the familiar physical phenomenon where an image of an object is cast back from a
particular kind of surface, such as glass or water. Using PrmanRender node, you can replicate this
effect in your render result of 3D objects, and using the Reflection node, you can adjust the controls
for creating the reflection effect. PrmanRender uses raytracing to create this effect and you can use
the Reflection node to adjust the result.
Note: Nuke's RayRender node can also use the Reflection node to control reflections.
1. Create a Reflection node by clicking 3D > Shader > RenderMan > Reflection.
2. Connect the node to the PrmanRender node and set the controls to adjust your reflection:
• reflection color - sets the color of the reflection.
• value - sets the intensity of the reflection.
USER GUIDE
1761
Using the Refraction Node |
Using the Refraction Node

Refraction is the familiar physical phenomenon of light traveling differently through different
materials and thus reflecting differently off objects behind that material. For example, if you have a
glass of water with a straw in it, the part of the straw that’s not in water appears to be in a different
angle to the part which is in the water. This is due to water bending the light waves. PrmanRender
uses raytracing to create this effect and you can use the Refraction node to adjust the result. Without
the PrmanRender node and RenderMan software though, the Refraction node has no effect.
1. Create a Refraction node by clicking 3D > Shader > RenderMan > Refraction.
2. Connect the node to the PrmanRender node and set the controls to adjust your refraction:
• refraction index - slide to change the type of refraction.
• value - sets the intensity of the refraction.
USER GUIDE
1762
Using F_DeFlicker2
When working in film, you sometimes have to deal with shots that have a luminance flicker. This
chapter concentrates on removing flicker using F_DeFlicker2.
Introduction
F_Deflicker2 is used to remove flicker. It is particularly suitable for removing flicker that is localized
and dependent on the geometry of the scene (that is, flicker that is not present across the whole of
the image), such as that caused by an unsynchronized fluorescent light in a shot. It works by
calculating the gain between the current frame and each frame in a small analysis range surrounding
it. It then tries to adjust the gain so that it varies smoothly over this frame range. This means it is
better at reducing fast flicker than flicker which varies slowly over the image sequence, as the latter
will already appear smooth over the small frame range and F_DeFlicker2 will leave it largely
untouched.
The algorithm used by F_DeFlicker2 can introduce blurring in areas where there is rapid motion. This
problem could be alleviated by using local motion estimation before deflickering the frames.
However, this process is complicated by the fact that the presence of flicker can adversely affect the
results of the motion estimation. F_DeFlicker2 therefore adopts a two-stage approach to this
problem. First, the normal deflickering process is performed. Then, the motion vectors for the
sequence are calculated on the resulting deflickered frames, and applied to the original frames in
order to align them. The deflicker calculation is then performed on the aligned frames to give the
final result. To use this approach, turn on Use Motion in F_DeFlicker2.
Note: Because F_DeFlicker2 looks at input frames outside the current frame when
performing its calculation, it can be a computationally expensive plug-in. As such, using
more than two instances of F_DeFlicker2 in a node tree will dramatically increase render
times. It is strongly advised therefore, that you render each instance out separately.
USER GUIDE
1763
Quick Start |
Quick Start
To remove flicker from a sequence:
1. Select Image > Read to load the sequence you want to remove flicker from.
2. Select FurnaceCore > DeFlicker2 to apply DeFlicker2. View its output.
3. If you’re not happy with the results, adjust the DeFlicker2 parameters. The available parameters
are described below.
USER GUIDE
1764
Parameters |
Parameters
The parameters for F_DeFlicker2 are described below.
• DeFlicker Amount - Use this to reduce flicker without removing it entirely; smaller values mean
more will be left behind.
• Block Size - To find where a certain pixel is located in the analysis range, the deflicker algorithm
looks for a block of pixels centered around that pixel. Block size defines the width and height of
these blocks (in pixels). On rare occasions, a large block size can produce data that’s lacking in
detail. This is because a small feature can fit into a large block, causing the motion estimation to
concentrate on the background motion and ignore the small feature. A small value, instead, can
produce a noisy motion field, as there aren’t enough constraints in a small block to fit the motion
accurately. In most cases, however, the default value is small enough so that details aren’t lost, and
the smoothing step of the algorithm ensures the motion field isn’t too noisy. Therefore, this value
very rarely needs editing.
• Use Motion - Turn this on to do a second deflicker pass using motion-compensated frames. This
can improve results in areas where there is fast motion, where the initial deflicker pass can
introduce blurring.
• Vector Detail - Determines the density of the motion vectors used when Use Motion is turned on.
The maximum value of 1 will generate one vector per pixel. This will produce the most accurate
vectors but will take longer to render. A value of 0.5 will generate a vector at every other pixel.
• Analysis Range - The number of frames searched each side of the current frame when calculating
the flicker. Higher values may give better results, but can also bring in erroneous information and
take longer to process.
USER GUIDE
1765
Using F_ReGrain
This chapter looks at adding grain to sequences using F_ReGrain.
Introduction
F_ReGrain is used to add grain to a sequence. It has been designed to sample an area of grain from
one image and then to generate unlimited amounts of this grain with exactly the same statistics as
the original. This new grain can then be applied to another image.
The figure on the left shows an enlarged and exaggerated sample of grain from Kodak 320 film stock.
F_ReGrain was used to sample the original Kodak 320 stock and synthesize a plate of grain. The result
is shown in the figure on the right. Note that the grain characteristics closely match the original.
Kodak 320. F_ReGrain.
Similarly, below the figure on the left is a sample from Kodak 500 film stock and the figure on the
right shows this replicated using F_ReGrain.
Kodak 500. F_ReGrain.
USER GUIDE
1766
Quick Start | Adding Sampled Grain
Quick Start
You can sample grain from an image and apply it to another or select from a variety of pre-sampled,
standard grain types.
Adding Sampled Grain

To add grain to a sequence, do the following:
1. Select Image > Read to load the sequence you want to add grain to. Then, load the image you
want to sample grain from.
2. Make sure you are working at full resolution and not proxy resolution. F_ReGrain will not work at
proxy resolution. (See Proxy Resolutions.)
3. Select FurnaceCore > F_ReGrain.
4. Connect the sequence that you want to have grain to F_ReGrain’s source (Src) input. Then, connect
the sequence you want to sample grain from to the Grain input. View the output from F_ReGrain.
5. Set Grain Type to From Grain Clip.
6. Position the on-screen sample region over an area of the Grain sequence just containing grain
and no picture detail. See the figure below.
This shows two possible selection regions

that contain no edge detail and little
luminance variation.
It is important to get your selection right. You should avoid any image detail or even a plain area
that has luminance variations underneath the grain. The better this initial selection, the better the
result will be. If you can’t find a decent sample area on the current frame, then try other frames
from the same film stock. The default size of the sample area should be enough to gather
USER GUIDE
1767
Quick Start | Using Pre-Sampled, Standard Grain Types
information about the grain characteristics of your image. However, you may need to change its
size and shape to fit over a plain area free of image detail.
Warning: There is a minimum size of this sample area below which the statistical analysis
of the grain will be unreliable. If the sample area you select is too small, you will see a
warning message which prompts you to select a larger region. (See Proxy Resolutions.)
7. View the output of F_ReGrain to judge the results. The output will now contain the Src image with
grain from the Grain image applied. Both the size and the luminance of the new grain can be
manually tweaked using Grain Size and Grain Amount respectively. It helps to view the Grain
input while editing the parameters of F_ReGrain.
The grain is sampled on a single frame which is set when you adjust the sample area (or by manual
adjustment of the Analysis Frame parameter). Although it is sampled on only one frame, the
algorithmically created grain will change from frame to frame but mirror the characteristics of the
sample grain.
Using Pre-Sampled, Standard Grain Types

If you don’t have an image to sample grain from, you can also select from a variety of pre-sampled,
standard grain types. Do the following:
1. Select Image > Read to load the sequence you want to add grain to.
2. Make sure you are working at full resolution and not proxy resolution. F_ReGrain will not work at
proxy resolution. (See Proxy Resolutions.)
3. Select FurnaceCore > F_ReGrain.
4. Connect the sequence that you want to have grain to F_ReGrain’s source (Src) input.
5. Set Grain Type to Preset Stock.
6. Try the different grain types using the Preset Stock dropdown menu. 2K, 4K, aperture corrected,
and non aperture corrected stocks are included. Individual color channels can be selected and
adjusted using the Advanced parameters.
USER GUIDE
1768
Response | Using Pre-Sampled, Standard Grain Types
Response
In its default setting, F_ReGrain adds the same amount of grain over the whole image. However, the
amount of grain on an image is normally a function of luminance. Various parameters in the Grain
Response group allow you to adjust how the amount of grain added varies with luminance:
• Pressing Sample Grain Response will cause the variation of the amount of grain with luminance to
be calculated from the Grain input, and switching on Use Sampled Response will apply these
curves to the grain added to the Src sequence.
• To view the sampled response curves, switch on Draw Response; an example is shown in the figure
below.
• The amount of grain added to the lowlights, midtones and highlights of the image can be adjusted
using the Low Gain, Mid Gain and High Gain parameters. The effect of adjusting these can also be
seen on the response curves.
This shows an example of the grain response

with luminance. The x axis represents
luminance and the y axis the amount of grain.
USER GUIDE
1769
Checking the Result | Using Pre-Sampled, Standard Grain Types
Checking the Result

To test that the new grain is the same as the old grain, set Output to Grain Plate.
This generates a sheet of grain with the same luminance level as the mean of the sample region. The
sample region with the original grain is also displayed. It should be impossible to differentiate
between the two regions. The figure on the left shows a good selection area giving a good test plate of
grain in the figure on the right.
Good selection area ... ... producing a good test

plate of grain, free of
artifacts.
Below, the figure on the left shows a poor selection area since it contains image detail. The figure on
the right shows the resulting test plate which clearly highlights the problem.
Bad selection area ... ... producing a poor result.
USER GUIDE
1770
Proxy Resolutions | Using Pre-Sampled, Standard Grain Types
Proxy Resolutions
Grain manipulation at proxy resolution should be avoided as the results are unreliable. The grain
selection area may be too small at proxy resolution to give a good result, and making this area larger
may drag in unwanted detail from the image. If you try to use F_ReGrain at proxy resolution, we
simply pass the image through untouched and issue the following warning:
Cannot work at proxy scale.
We decided that this was preferable behaviour to doing poor grain replication at proxy resolution.
You can, of course, crop the input clip and work with that rather than the proxy. There is a minimum
size for the selection box, which is about 37x37 at base resolution. If the box you select is smaller,
you will get this warning along the top of the viewer:
Sample box is too small - please select a larger sample of the grain.
USER GUIDE
1771
Parameters | Using Pre-Sampled, Standard Grain Types
Parameters
The parameters for this plug-in are described below:
• Grain Type - Selects whether the grain is sampled from the Grain image (From Grain Clip) or from
a set of standard stocks.
• Preset Stock - grain characteristics are sampled from a supplied film stock. 2K, 4K, aperture
corrected and non aperture corrected stocks are supplied. Although standard stocks are
included, it is recommended where possible that you sample from the film stock you are trying
to match.
• From Grain Clip - samples and reconstructs the grain characteristics from the Grain input.
• Preset Stock - The film stock the grain characteristics are sampled from when Grain Type has been
set to Preset Stock. Common Fuji and Kodak stocks are supplied. The exposure can be under, over,
or, if left blank, non aperture corrected. The size is either 2K or 4K pixels. For example, FUJIF500 2K
refers to the grain characteristics sampled from a 2K plate of Fuji Film 500 film stock non aperture
corrected.
• Grain Amount - Adjusts the brightness of the grain. Setting this to 0 means no grain is added.
• Grain Size - Adjusts the size of the grain granules. The larger the value, the bigger and softer the
granules.
• Output - Sets whether to render the result or a test image.
• Result - shows the Source image with the grain applied.
• Grain Plate - shows a test image with the grain applied. This test image is composed from a
section of the input image surrounded by a uniform solid color sampled from the image with
the grain applied. If the inner area is indistinguishable from the outer area, then you have a
good grain sample.
Analyse - This is a push button which will trigger an analysis of the input clip. Press this button if the
input clip from which the grain was analyzed has changed but you do not want to move the analysis
region to trigger re-analysis. Whenever the input clip changes, you will see the following warning in
the Viewer:
The clip from which the grain was analyzed has changed. Press Analyse or move the analysis region to re-
analyze grain.
• Analysis Region - A selection box that marks the region of image used to analyze the grain when
Grain Type is set to From Grain Clip. This part of the frame must contain no image detail, only
grain.
• Analysis Region BL - controls the position of the bottom left corner of the analysis region.
• Analysis Region TR - controls the position of the top right corner of the analysis region.
• Analysis Frame - sets the frame to sample the grain from.
USER GUIDE
1772
• Grain Colour Space - This tells F_ReGrain what color space the grain sample clip was in when the
grain originated. Setting this correctly ensures that the grain is not exaggerated by any color space
conversions prior to sampling.
• Cineon
• sRGB
• Linear
• Advanced - The parameters under Advanced allow detailed adjustment of the grain.
• Process Red - Switch this on to process the red channel.
• Red Amount - sets the brightness of the grain in the red channel.
• Red Size - adjusts the size of the grain granules in the red channel.
• Process Green - Switch this on to process the green channel.
• Green Amount - sets the brightness of the grain in the green channel.
• Green Size - adjusts the size of the grain granules in the green channel.
• Process Blue - Switch this on to process the blue channel.
• Blue Amount - sets the brightness of the grain in the blue channel.
• Blue Size - adjusts the size of the grain granules in the blue channel.
• Grain Response - The parameters under Grain Response allow the amount of grain added to be
varied as a function of the image luminance.
• Apply Grain In - This controls what color space the grain sample is re-applied to the image.
Generally, this should be set to Grain Colour Space to ensure the most accurate recreation.
You may want to override this though for some looks or special cases.
• Cineon / sRGB / Linear - The grain sample will be applied in the specified space.
• Grain Colour Space - The Grain sample will be applied in the color space set in the Grain
Colour Space dropdown menu, in the Grain Sample section.
• Low Gain - adjusts the gain of the grain in the lowlights.
• Mid Gain - adjusts the gain of the grain in the midtones.
• High Gain - adjusts the gain of the grain in the highlights.
• Use Sampled Response - switch this on to scale the brightness of the grain as a function of the
luminance of the Grain image.
• Sampled Response Mix - this control is usually set to 1. Decreasing it reduces the effect of the
response curves until, at 0, they have no effect on the output. This parameter is only available if Use
Sampled Response is on.
• Sample Grain Response - press this to update the response curves from the current frame.
Multiple presses accumulate the grain response rather than resetting every time. This parameter is
only available if Use Sampled Response is on.
USER GUIDE
1773
• Reset Grain Response - press this to reset the grain curves to their default (flat) response. This
parameter is only available if Use Sampled Response is on.
• Draw Response - overlays the response curves on the bottom left corner of the viewer. This
parameter is only available if Use Sampled Response is on.
USER GUIDE
1774
Color Space in FurnaceCore Plug-ins | Color Space in F_ReGrain
Color Space in FurnaceCore Plug-

ins
Some of the algorithms in the FurnaceCore tool set are sensitive to the color space of the source
footage. If the footage is not in the expected color space, you may get poor results from some of the
plug-ins.
Like Nuke, FurnaceCore expects all footage to be in Linear space. Nuke converts all footage to Linear
upon import, so unless you have changed the colorspace in your Read nodes (or in the node tree by
using a Colorspace node before a FurnaceCore plug-in), your footage should be Linear by the time it
reaches the plug-in anyway.
If you know that the input to a plug-ins isn’t Linear, you should use a Colorspace node before and
after the plug-in to convert to and from Linear for processing. This should ensure optimal results.
Color Space in F_ReGrain

Because color space transformations can distort the look of grain, and by default, Nuke converts your
footage to linear, if you are sampling grain from your own plate, you need to make sure you tell the
plug-in what space the plate was in originally, so the sample isn’t distorted.
This is accomplished by setting the Grain Colour Space dropdown menu in the F_ReGrain controls to
the right space. For example, if you were sampling grain from a film scan, you would want to set this
to Cineon. If you had footage from a digital video camera, this would most likely be sRGB. F_ReGrain
automatically sets the right color space when using one of the pre-sampled grain clips, which are in
sRGB.
F_ReGrain also works best when applying grain in the same color space that the sampled grain
originally existed. There is a dropdown menu in the Response section of the F_ReGrain controls
which allows you to match or override this. It defaults to applying the grain in the same space as the
sample.
USER GUIDE
1775
Using F_WireRemoval
This chapter looks at the removal of wires from images using F_WireRemoval.
Introduction
Many effects movies feature complicated stunts that require an actor to be suspended by wires for
their safety. These wires need to be digitally removed.
There are many ways of doing this, including painting the wires out frame by frame and replacing
large parts of the background to cover the wires. The method used depends on the type of image
under the wire. F_WireRemoval is particularly good at replacing the painting method but it also
includes tools to assist in clean plating when new backgrounds have to be tracked into place.
Three wires. Two of the wires removed

using F_WireRemoval.
USER GUIDE
1776
Background | Clean Plates
Background
Clean Plates
The use of clean plates in wire removal is very common and gives good results in certain situations.
Consider a scene shot with an actor suspended from wires and then the same scene shot again
without the actor. This second sequence is called the clean plate. The wires from the first shot can be
painted out using pixels from the clean plate leaving the actor suspended in thin air.
Shooting a clean plate if the camera is locked off is easy. If the camera moves, then motion control
rigs can be used to exactly reproduce the first pass. But it doesn’t always work, particularly if the
scene is shot against backgrounds that don’t look the same on the second pass, such as clouds, sky,
or smoke. Motion control rigs are also expensive and that makes them a rarity. Often, a clean plate is
not shot during the filming and the compositor is left to create one by merging together
unobstructed pixels from many frames. This single frame can then be tracked into place to cover the
wires.
FurnaceCore
FurnaceCore’s wire removal plug-in should make the process of wire removal much easier. It is
particularly good at removing wires over heavily motion blurred backgrounds or wires over smoke,
dust, or clouds. It can be used to remove each wire in a sequence or to quickly create a clean plate
which can then be tracked into place.
The reconstruction of the background behind the wire can be done spatially, in other words, using
only information from the current frame. Alternatively, motion estimation techniques can be used to
warp frames from before and after onto the current frame so that, where available, background
pixels from these frames can be used to improve the repair. Our reconstruction methods are unique
in that they remove the wire without removing and reapplying the grain. They are also tuned to
remove long thin objects, leaving other objects untouched. For example, if transient foreground
objects cover the wire, they will be left alone.
USER GUIDE
1777
Reconstruction Methods | FurnaceCore
Reconstruction Methods
Our four reconstruction methods are:
• Spatial
• Temporal With Static Scene
• Temporal With Moving Scene
• Clean Plate
The spatial method takes the background information from adjacent pixels in the current frame, and
the clean plate method takes the information from a separate clean plate input. The other methods
are temporal and attempt to get background information from frames either side of the current
frame. Where this information is not available, for example, because the wire covers part of the same
region in one or more of the other frames, the spatial method will be used for the repair.
The spatial method is fastest. It uses a slope-dependent filter that interpolates across the wire at the
most likely angle, given the image behind the wire. It works well when the wire is over backgrounds
such as sky, clouds, or smoke, and can also cope with some backgrounds where there is a noticeable
gradient, such as the edge of a roof, building, or car. If this fails and the wire is moving relative to the
background, you should try one of the temporal methods. These look at frames before and after the
current one to find likely pixels with which to repair the region and use the spatial method in areas
where there are none available.
Where traditional clean plates are possible or give better results than using F_WireRemoval to repair
the wire on each frame, you can supply a clean plate as the fourth input to the plug-in. It will then
automatically match-move it to repair each frame. If the overall luminance of the background is
different to that of the clean plate or varies during the sequence, turn on Luminance Correct. The
luminance of the background pixels used for the reconstruction will then be adjusted before the
repair is made. Of course, various FurnaceCore tools can be used to create a clean plate, including an
F_WireRemoval pass on a single frame where the repair has been successful.
USER GUIDE
1778
Tracker | FurnaceCore
Tracker
F_WireRemoval incorporates a tracker which can automatically track a moving wire through a clip.
This tracker has its own control panel, which will float inside the NukeX viewer if you have checked
Show On Screen Controls in the WireRemoval controls. Below is a screenshot that illustrates this.
Screenshot with tracker panel.
You can resize the panel by dragging its edges. For details of the rest of the controls, see the section
on Tracker Controls on Tracker Controls. All the controls are also available in the WireRemoval
properties.
USER GUIDE
1779
Quick Start | FurnaceCore
Quick Start
To remove a wire from an image sequence, do the following:
1. Select Image > Read to load the clip that needs a wire removed.
2. Select FurnaceCore > F_WireRemoval. Connect your image sequence to F_WireRemoval’s
Source input. View the output from F_WireRemoval.
3. Make sure Output is set to Source, and use the on-screen tools to draw a region that defines the
position and shape of the wire to be removed. For more information, see Positioning the On-
Screen Wire Tool.
4. If the wire you want to remove is moving, start the wire tracker by clicking on the Track Forwards
button, so that it follows the wire during the clip. For more information, see Tracker Controls.
5. From the Repair dropdown menu, choose a repair method that removes the wire. If you choose
to use the Clean Plate method, connect a clean plate image to F_WireRemoval’s CleanPlate
input. For more information on the available methods, see Reconstruction Methods.
6. To see the output, set Output to Repair.
7. If necessary, increase Overall Width to expand the width of the repair region until the wire
disappears.
USER GUIDE
1780
Positioning the On-Screen Wire Tool | FurnaceCore
Positioning the On-Screen Wire

Tool
To position the on-screen wire tool, do the following:
1. In the F_WireRemoval controls, set Output to Source. This way, you will be able to see the wire
you’re trying to remove but won’t have to wait for F_WireRemoval to repair the wire every time you
change wire tool’s position.
2. Choose the number of points that are needed to describe the wire: two for straight lines, three for
simple curves, or five for more complex shapes. This can be done using the Type dropdown in
the properties panel, or by toggling the Number of Points button in the on-screen tracker panel.
Tip: To show the on-screen tracker panel, check Show On Screen Controls in the
WireRemoval properties.
3. Position the on-screen wire tool so that it roughly fits the wire you want to remove. To move the
entire tool, drag its central line to a new location. To move an individual wire point on the tool,
drag its center. To move an individual point in either a horizontal or vertical direction only, drag
the horizontal or vertical line on the cross that marks the point.
4. Press the Snap To button in the WireRemoval properties or the on-screen tracker panel. This
will find the edges of the wire and adjust the removal region to fit it more closely. It does this by
locating the areas of highest gradient, which should correspond to the edges of the wire.
5. In practice, the edges of the wire will be slightly soft, so the resulting region might not cover the
whole width of the wire. To correct for this, you can adjust the Overall Width parameter. This
expands the region outwards, over all key frames, to ensure the entire width of the wire is
covered.
USER GUIDE
1781
Tracking | FurnaceCore
Tracking
F_WireRemoval incorporates a tracker that tracks the region to be repaired through the image
sequence.
To use the tracker, do the following:

1. Drag the on-screen wire tool to position the repair region on top of the wire on one frame. This
sets a user keyframe for the current frame, enabling tracking.
2. To start the tracker, press the Track Forwards button .

If you haven’t set a user keyframe (step 1), the tracker cannot be started and F_WireRemoval
displays the following message:
Please select a user key frame before tracking.
3. When the track is ready, check that the wire has been correctly tracked across the other frames
and adjust the track if necessary. The tracking controls in the tracker panel are outlined under
Tracker Controls. Any points initially positioned off the image will remain off the image. This way,
the tracker will not lose the wire at the edges of the image.
4. If after tracking the wire you adjust the position of the on-screen wire tool manually on one
frame, you may want to create a smooth transition between the user keyframe you created and
the track keyframes around it. The easiest way to do this is to go on the previous and next frame
on the timeline and press the Delete track keyframe button . This deletes the track keyframes
around the frame you adjusted and forces NukeX to interpolate smoothly around the adjustment.
Similarly, if there is too much jitter in the track between some frames, you can delete the track
keyframes between these frames. This way, NukeX will interpolate smoothly from one frame to
another.
If it is not possible to track the wire automatically, the wire can be manually keyframed. Adjusting the
on-screen wire tool on each frame will automatically set user keyframes. Alternatively, they can be set
by pressing the Create User Keyframe button in the tracker panel.
USER GUIDE
1782
User Keyframes and Track Keyframes | FurnaceCore
User Keyframes and Track

Keyframes
User keyframes are keyframes you set manually. When a user keyframe has been set on the current
frame, this is indicated by a red key that appears on the on-screen wire tool. Track keyframes are
keyframes set by F_WireRemoval’s tracker.
User keyframes override track keyframes on the same frame. This has the following consequences:
• Once you set a user keyframe, the track keyframe on that frame will be lost.
• When you track a wire, existing user keyframes will not be replaced by track keyframes.
USER GUIDE
1783
Indicators on the On-Screen Wire Tool | FurnaceCore
Indicators on the On-Screen Wire

Tool
You may see the following indicators appear on the on-screen wire tool:
Control Description
The current frame is inside the range of frames to track the wire over.
The current frame is not inside the range of frames to track the wire over.
The current frame is the first frame in the range of frames to track the wire over.
The current frame is the last frame in the range of frames to track the wire over.
A user key frame has been set on the current frame.
USER GUIDE
1784
Tracker Controls | FurnaceCore
Tracker Controls
The following controls appear in the on-screen tracker panel that you can display in the Viewer by
checking Show On Screen Controls in the WireRemoval properties.
Control Description
Toggle display mode - Toggle the display mode for the on-screen wire tool: show the
points and lines, show just the points, or hide both points and lines.
Number of points - Changes the number of points used to describe the wire.
Create user key frame - Creates a user keyframe.
Delete user key frame - Deletes a user keyframe.
Snap to wire - Finds the edges of the wire and snaps the edges of the region onto
them.
Track backwards - Plays backwards through the sequence tracking from frame to
frame.
Step backward - Tracks backwards one frame.
Step forward - Tracks forward one frame.
Track forwards - Plays forwards through the sequence tracking from frame to frame.
Smart track - Tracks from beginning to end of frame range in an intelligent order.
USER GUIDE
1785
Tracker Controls | FurnaceCore
Control Description
Delete track key frames backwards - Deletes track keyframes backwards through the
sequence until either a user keyframe or the beginning of the sequence is reached.
Delete track key frame and step backward - Deletes a track keyframe and steps
backwards one frame.
Delete track key frame - Delete the current track keyframe.
Delete track key frame and step forwards - Deletes a track keyframe and steps
forwards one frame.
Delete track key frames forwards - Deletes track keyframes forwards through the
sequence until either a user keyframe or the end of the sequence is reached.
Delete all track key frames - Deletes all track keyframes from the sequence.
Delete all track and user key frames - Deletes both track keyframes and user
keyframes.
The corresponding controls are also available in the WireRemoval properties.
USER GUIDE
1786
Parameters | FurnaceCore
Parameters
The parameters for this plug-in are described below.
Tracker buttons - most of the on-screen tracker panel buttons also appear here. For more
information, see Tracker Controls.
Type - this parameter controls the number of points on the on-screen wire tool. Choose the number
of points needed to describe the wire you wish to remove.
• Two Points - choose this if your wire is straight.
• Three Points - choose this if your wire is a simple curve.
• Five Points - choose this if your wire has an s-shaped curve.
On-Screen Wire - the display mode for the on-screen wire tool.
• Show - show both points and lines.
• Hide - hide both points and lines.
• Points only - only show the points.
Show On Screen Controls - use this to show or hide the tracker panel in the Viewer.
Output - sets the output mode for the plug-in.

• Source - output the untouched source image. Use this output mode to position the on-screen wire
tool over the wire you wish to remove.
• Repair - output the repaired source image, with the wire removed from under the on-screen tool.
• Wire Matte - renders a matte for the wire. This may be useful if the wire has been tracked but
cannot be repaired using F_WireRemoval and other techniques have to be used.
• Repair Matted - output the repaired source image and a matte in the alpha channel. If you want,
you can adjust your image further manually using the matte.
Track Range - the range of frames to track the wire over.

• Specified Range - use the Track Start and Track End parameters to specify the range over which
to track the wire.
• Source Clip Range - track the wire over the entire range of the Source clip.
Track Start - use this to specify the start of the tracking range when Track Range has been set to
Specified Range.
Track End - use this to specify the end of the tracking range when Track Range has been set to
Specified Range.
Repair - sets the algorithm used to remove the wire from under the grain.
USER GUIDE
1787
• Spatial - this method uses a slope dependent filter that interpolates across the wire at the most
likely angle, given the image behind the wire. It uses information from the current frame only.
• Temporal With Static Scene - this method uses local motion estimation to align frames from
before and after onto the current frame. If the wire is not in the same place with respect to the
background in these two frames, it can then use background information from them to fill in the
background behind the wire in the current frame. Where no such information is available, for
example, if the wire covers part of the same region in all three frames, the spatial repair method
above will be used. This is useful for sequences where the wire is moving and where the motion in
the rest of the scene is non-uniform, for example, if the camera has been locked to place but there
are objects moving in the area surrounding the wire.
• Temporal With Moving Scene - also aligns frames from before and after onto the current frame,
but uses global motion estimation. Again, it gets background information from these two frames
where it can and uses the spatial repair method to fill in the rest. This is useful for sequences where
the wire is moving and the motion in the rest of the scene is fairly uniform, for example, if the entire
scene is moving in the same direction as a result of a camera pan.
• Clean Plate - choose this method if you have a clean plate you wish to use for the repair, or if F_
WireRemoval does not do a good job of removing the wire from each frame. Connect the clean
plate or single frame with the wire removed to the CleanPlate input. F_WireRemoval will then align
this frame to the frame to be repaired in order to reconstruct the background behind the wire.
Filter Size - this is a trade off between the amount of grain removed and the blurring of image along
the direction of the wire. If the wire you are trying to remove has detail within it (for example, a steel
wire in which the twisted threads are reflecting light), then the algorithm may leave these alone,
thinking that they are grain. In this situation, you should decrease the filter size. A value of zero will
definitely remove the artifacts but also the grain, which would then have to be put back using
FurnaceCore’s F_ReGrain.
Temporal Offset - the time offset of the additional frames to use for the Temporal With Static
Scene or Temporal With Moving Scene methods. This determines which two frames before and
after the current frame are used to fill in the background behind the wire. For example, if this is set to
2 and the current frame is 30, frames 28 and 32 are used.
Luminance Correct - turn this on for methods other than Spatial repair where there are global
luminance shifts between one frame of the sequence and the next, or between a frame of the
sequence and a clean plate you are using for the repair. With Luminance Correct turned on, F_
WireRemoval will adjust the luminance of the background pixels it uses to the correct level before
doing the repair.
Lum Block Size - change this value if luminance correction is not working as well as it should. The
luminance correction uses overlapping blocks and matches the luminance within those blocks;
changing the size of the blocks will change the regions covered by each block and could result in a
better match.
USER GUIDE
1788
Points - Parameters to set the position of the points used to define the wire.
Point 1 - the position of the start point on the wire.
Point 2 - the position of the point on the wire between the start point and the mid point (this is only
active if Type is Five Points).
Point 3 - the position of the mid point on the wire.
Point 4 - the position of the point on the wire between the mid point and the end point (this is only
active if Type is Five Points).
Point 5 - the position of the end point on the wire.
Start Width - the width of the wire at point 1 of the on-screen wire tool.
End Width - the width of the wire at point 5 of the on-screen wire tool. This allows you to make your
repair region wider at one end than the other, which is useful for cases where there is motion blur
on the wire.
Overall Width - increase this parameter to expand the width of the repair region along its entire
length, and for all key frames.
USER GUIDE
1789
Global Motion Estimation
FurnaceCore has three effects based on global motion estimation (GME).
Introduction
FurnaceCore effect based on global motion estimation calculate a four-corner pin, which finds the
best fit of one image onto another, and then apply that pin. These effects differ in how that corner pin
is calculated and to which image that pin is applied. This chapter describes the general idea behind
these plug-ins; for more detailed information on each plug-in, please read their individual chapters.
These effects are:

• F_Align - which lines up two shots of the same scene, by finding a corner pin from each source clip
frame to the corresponding reference clip frame. For example, you can use this to align two
separate but similar steady cam shots of the same scene.
• F_RigRemoval - which removes unwanted objects (rigs) from image sequences without accurate
rotoscoping or keying to produce a clean plate.
• F_Steadiness - which removes motion from a single clip, by calculating a pin that either locks all
frames in the clip to a single reference frame, or smooths that motion out over a window of frames.
For example, you can use this to remove camera shake.
USER GUIDE
1790
What is Global Motion Estimation? | Limitations of GME
What is Global Motion

Estimation?
Global motion estimation (GME) is a technique that attempts to map one image onto another with a
simple four-corner pin. This differs from local motion estimation (LME), which attempts to find where
each individual pixel in the image is in the other image. GME is much cheaper to compute than LME,
but gives you less information about the image. Nevertheless, it is still very powerful for a variety of
applications.
Using the plug-ins’ parameters, you can tell the GME engine what type of motion to expect. This can
be a combination of any of:
1. translation - which allows the four corners to translate by the same amount,
2. rotation - which allows the corners to rotate about their center,
3. scale - which allows the size of the area defined by the corners to change,
4. perspective - which allows the angles at the corners to change, so that the area defined by them
is no longer a rectangle.
The more types of motion you allow, the more expensive the motion estimation becomes. For many
scenes, rotation and translation are sufficient.
The GME effects have an accuracy control, which controls the amount of work Foundry’s GME engine
does to calculate the global motion. Typically, the higher this is, the better the estimation, but the
more expensive it is.
Limitations of GME
As stated above, global motion estimation simply calculates a four-corner pin to transform one image
onto another. This means that GME can’t be used to match two images where there is heavy parallax,
very complicated foreground motion, changing objects, and so on.
The best way to think of what GME can do is that if you can do it with a four-corner pin, it can; if you
can’t, it can’t. However, GME will take the pain out of hand matching pins frame by frame.
USER GUIDE
1791
Global Motion Estimation Effects | Using Them
Global Motion Estimation Effects

F_Align and F_Steadiness work in a similar way, which is distinct from the way F_RigRemoval works.
These two effects calculate a four-corner pin for each frame and save it into the corner pin
parameters. These pins are then used during the render to move the source image.
Using Them
These effects analyze images over a range of frames to figure out their four-corner pins. This is done
in response to the user pressing the Analyse button in the effects control panel. During analysis, the
effect will run through a range of frames adding keys to the corner pin parameters. These corner pins
are then applied to the source clip to render a frame.
F_Steadiness has a separate analysis pass that happens interactively; it then uses the previously
computed and key-framed corner pins during render. This speeds up its operation, as the analysis
step only has to be done once, and once it has been done this plug-in is very quick to render.
However, F_Align, which only ever needs the two current frames from each clip, can compute the
corner pin on the fly (but not keyframe it!). This leads to a slightly different mode of operation for the
following effects:
• F_Steadiness
• needs to have an analysis run before it renders useful output,
• will always use the value in the corner pin parameters when rendering the output image.
• F_Align
• no need to have the analysis run to render output, but doing so will give you a key-framed
corner pin,
• during render, it will use the value of the corner pin parameters only if there is a keyframe there,
otherwise, it will analyse on the fly during render. This means that analysis will speed up later
renders, as just rendering the corner pin is much cheaper than calculating it.
Some parameters to the effect control how the effect performs GME analysis, and some only affect
the rendering. If you ever modify one of these parameters, then any analysis you may have
performed will be out of date. To let you know, an overlay warning will be posted whenever this
happens. You don’t have to re-analyze and your renders will still look at the keyed corner pins.
If you have not modified a parameter that affects analysis (the warning overlay will let you know),
pressing Analyse will only re-analyse a frame if there is no key on the corner pin at that time. This
avoids redundant re-analysis if you have interrupted the analysis or extended the analysis range.
However, if you want to force re-analysis, press Clear Analysis and all keys will be deleted.
USER GUIDE
1792
Global Motion Estimation Effects | Using Them
F_Steadiness and F_Align have an analysis region rectangle parameter which is used to specify which
area of the reference image should be analysed during GME. So, for example, with F_Steadiness set
to Lock Mode, this is the area inside the lock frame that a match will be sought for. The
documentation for each plug-in describes exactly how to use the analysis region.
USER GUIDE
1793
Controls | Parameters That Affect Analysis
Controls
The controls common to all GME plug-ins are described below. They are grouped into two sections:
• Ones that determine how analysis is carried out. See Parameters That Affect Analysis.
• Ones that control the rendering of the output. See Parameters That Affect Rendering.
Parameters That Affect Analysis

The following parameters affect the analysis of the four-corner pin.
Analyse - This is a push button which triggers an analysis of the input clips and calculate a corner pin.
Interrupting the analysis does not delete the corner pin keys that have already been calculated.
Render During Analysis - If set, this toggle causes the effect to update the time line and render a
freshly analyzed frame so you can see the progress of the effect. Doing so slows down the analysis
somewhat, so toggle this off to speed up the general analysis.
Clear Analysis - Pressing this push button deletes all keyframes from the corner pin parameters,
allowing you to force a re-analysis if you feel the need to.
Analysis Range - This controls the range of frames any analysis will be run over. It can be one of:
• Specified Range - which looks at the parameters Analysis Start and Analysis Stop for the range of
frames to analyze,
• Source Clip Range - which automatically determines the range of frames to analyze from the length
of the input clip.
Analysis Start - The first frame to analyze from if AnalysisRange is set to Specified Range.
Analysis Stop - The last frame to analyze from if Analysis Range is set to Specified Range.
Scale - A toggle that indicates whether the calculated corner pin can include a scaling factor.
Rotate - A toggle that indicates whether the calculated corner pin can include rotations.
Translate - A toggle that indicates whether the calculated corner pin can include translations in x and
y.
Perspective - A toggle that indicates whether the calculated corner pin can include perspective
transforms.
USER GUIDE
1794
Controls | Parameters That Affect Rendering
Analysis Region - This is the region analyzed to calculate the four-corner pin. This is especially useful
when doing any form of frame locking, in which case, go to the lock frame, look at the reference clip
and position the box over the area you want locked.
• Analysis Region BL - controls the position of the bottom-left corner of the analysis region.
• Analysis Region TR - controls the position of the top-right corner of the analysis region.
Accuracy - This controls the time/accuracy trade off in the GME engine. The higher this is, the slower
it goes, but you have a better likelihood of a good result.
Parameters That Affect Rendering

These following parameters control how a GME effect renders the four-corner pin. Some of them are
set during the analysis pass.
Filtering - This controls the quality of the rendering.

• Low - uses nearest neighbor filtering. This gives low quality but is quick to render.
• Medium - uses a bilinear filter. This gives good results and is quicker to render than high filtering.
• High - uses a sinc filter to interpolate pixels giving a sharper repair. This gives the best results but
takes longer to process.
Invert - If set, then the inverse of the calculated four-corner pin is used during render.
Four Corner Pin - The corner pins calculated during the analysis pass:

• Bottom Left - the lower left corner pin.
• Bottom Right - the lower right corner pin.
• Top Left - the upper left corner pin.
• Top Right - the upper right corner pin.
USER GUIDE
1795
Widgets | Parameters That Affect Rendering
Widgets
All the Analyzing GME effects have two on-screen widgets: one to provide feedback and one to set up
the analysis region.
Analysis Region Widget - This is a rectangle widget which you use to set the analysis region over the
reference image.
Four Corner Widget - This is a widget that shows the state of the four-corner pin that has been
calculated. You can change it by grabbing any of the corners and tweaking the shape of the pin. To
give you more feedback as to what frames have been analyzed, it will be drawn solid if there is a key
in the corner pin at the frame being displayed; otherwise, it will be drawn dashed.
USER GUIDE
1796
Using F_Align
This chapter looks at how to use F_Align to register (line up) two shots that are of the same scene, but
have slightly different camera motion and foreground objects. This can be useful, for example, for
doubling up the crowd size by lining up and comping together two shots of the same scene, or
locking your freshly generated clean plate to the original.
Introduction
F_Align takes two sequences that were shot of the same scene and lines them up spatially. It uses
Global Motion Estimation (GME) to calculate a four-corner pin so that each frame in one shot (the
source input) will be aligned with the corresponding frame in a second shot (the reference input). The
result is the source image which has been repositioned to line up with the reference image.
Source image. Notice the Reference Image.

position of the background.
Output from F_Align. The output of F_align

The source image has been comped together with the
repositioned so that the reference image.
background lines up with
the reference image.
To be able to align the sequences, F_Align analyzes them for global motion. This analysis can be
triggered for the complete sequence, specified frame range, or a single frame when you press the
Analyse button in the F_Align controls. Alternatively, it can be done on the fly for a single frame when
USER GUIDE
1797
Using F_Align |
you move to a new frame on the timeline. The advantage of pressing Analyse is that during the
analysis, F_Align stores the calculated four-corner pin as key-framed parameters. When you then
render the output of the plug-in later, F_Align can use these key frames without having to calculate
them again.
If you analyze on the fly, you won’t have direct access to the calculated corner pins. Any re-rendering
will also be significantly slower, as the ’on the fly’ calculations will have been lost and F_Align will have
to analyze again.
If at any stage you modify the effect in such a way to invalidate the key-framed analysis (for example
by changing the Accuracy parameter), a warning will be posted and the effect will analyze on-the-fly
during render, ignoring the keyed analysis.
The on-screen widget and the Analysis Regionparameters are used to control which section of the
Referenceframe is being matched to each Source frame. Typically, leaving the region at its default is
good enough. However, a heavy mismatch in foreground detail may make it necessary to change the
region to a section that is shared between shots.
The transformation in F_Align is concatenated with other NukeX transform nodes. This means that if
you add a row of F_Align and NukeX transform nodes to a tree, their functions are combined.
Because the image is only re-sampled once, there is no loss of image quality and processing time is
decreased. However, as in NukeX, certain nodes, including color correction nodes, can break the
concatenation.
For more of an overview of Global Motion Effects and a description of the common way of working
many of these effects have, please see Global Motion Estimation.
USER GUIDE
1798
Quick Start | Analyzing with the Analyse button
Quick Start
This section gives a very brief outline of how to use the plug-in. It covers both analyzing using the
Analyse button and analyzing on the fly.
Analyzing with the Analyse button

To align two shots and store the results as keyframed parameters, do the following:
1. Find two shots that are of the same scene, but have slightly different camera motion and
foreground objects. Select Image > Read to load both these shots.
2. Select FurnaceCore > F_Align. Connect the shot you want to reposition to the source (Src) input
of F_Align and the shot you want to match to the reference (Ref) input. View the output from F_
Align.
The source will be immediately repositioned so that it aligns to the reference shot without any
need for analysis.
You will see the following banner in the overlay:
No valid key framed analysis found. Analyzing during render.
3. Depending on the exact difference between the two shots, you may need to enable the Scale
and/or the Perspective toggles to get a decent alignment.
4. You may also need to reposition the Analysis Region depending on the differences in foreground
detail. However, leaving it at the default position works well for most shots.
5. Click on the Analyse button.
F_Align will now start analyzing each frame in the shot, figuring out the four-corner pin and
writing it as key frames to the corner pin parameters, Bottom Left, Bottom Right, Top Left, and
Top Right. You will find these parameters under Advanced > Four Corner Pin.
F_Align will update the timeline at each frame, and you will see the aligned image render in the
output. If you don’t want this to happen, uncheck Render During Analysis.
If you interrupt the analysis, the pins it has keyed until that point will be retained.
6. Play or scrub through the aligned frames. The rendering will be faster as F_Align will no longer
need to analyze on the fly. However, if you scrub to a frame where a corner pin has not been
keyed, F_Align will re-analyze that frame on the fly.
7. To see how closely the two clips have aligned, press M on the Node Graph to insert a Merge node.
Connect the Merge node’s A input to F_Align and the B input to the reference image. View the
output from the Merge node. Then, adjust the mix slider in the Merge controls to dissolve
between F_Align’s output and the reference clip you wanted to match.
USER GUIDE
1799
Quick Start | Analyzing On The Fly
Analyzing On The Fly

To align two shots and calculate the alignment on the fly, do the following:
1. Find two shots that are of the same scene, but have slightly different camera motion and
foreground objects. Select Image > Read to load both these shots.
2. Select FurnaceCore > F_Align.
3. Connect one of the shots to the source (Src) input of F_Align and the other to the reference (Ref)
input. View the output from F_Align.
The Src will be immediately repositioned so that it aligns to the Reference shot without any need
for analysis.
No valid key framed analysis found. Analyzing during render.
4. Depending on the exact difference between the two shots, you may need to enable the Scale
and/or the Perspective toggles to get a decent alignment.
5. You may also need to reposition the Analysis Region depending on the differences in foreground
detail. However, leaving it at the default position works well for most shots.
6. To see how closely the two clips have aligned, press M on the Node Graph to insert a Merge node.
Connect the Merge node’s A input to F_Align and the B input to the reference image. View the
output from the Merge node. Then, adjust the mix slider in the Merge controls to dissolve
between F_Align’s output and the reference clip you wanted to match.
USER GUIDE
1800
Parameters | Analyzing On The Fly
Parameters
Analyse - This is a push button which will trigger an analysis of the input clips and calculate a corner
pin. Interrupting the analysis will not delete the corner pin keys that have already been calculated.
Render During Analysis - If set, this toggle will cause the effect to update the time line and render a
freshly analyzed frame in the Viewer so you can see the progress of the effect. Doing so will slow
down the analysis somewhat, so toggle this off to speed up the general analysis.
Clear Analysis - Pressing this push button will delete all keyframes from the corner pin parameters,
• Specified Range - which will look at the parameters Analysis Start and Analysis Stop for the range
of frames to analyze,
• Source Clip Range - which will automatically determine the range of frames to analyze from the
length of the input clip.
• Current Frame - the analysis occurs only on the current frame. This is useful for correcting any
errors that may have occurred while analyzing the entire clip.
Analysis Start - The first frame to analyze from if Analysis Range is set to Specified Range.
y.
transforms.
Analysis Region BL - controls the position of the bottom left corner of the analysis region.
Analysis Region TR - controls the position of the top right corner of the analysis region.
USER GUIDE
1801
Parameters | Analyzing On The Fly
Advanced - The lesser used refinement controls.
Accuracy - This controls the time/accuracy trade off. The higher this is, the slower the analysis, but
you have a better likelihood of a good result.
Filtering - Sets the filtering quality.

• Low - low quality but quick to render.
• Medium - uses a bilinear filter. This gives good results and is quicker to render than high
filtering.
• High - uses a sinc filter to interpolate pixels giving a sharper repair. This gives the best results
but takes longer to process.
Invert - if set, then the inverse of the calculated four-corner pin is used during render.
Bottom Left - the lower left corner pin.
Bottom Right - the lower right corner pin.
Top Left - the upper left corner pin.
Top Right - the upper right corner pin.
USER GUIDE
1802
Using F_RigRemoval
This chapter looks at the removal of unwanted objects (rigs) from image sequences without accurate
rotoscoping or keying to produce a clean plate.
Introduction
In this context, we define a rig as a foreground element in a sequence that moves over a background
element. The plug-in will only work satisfactorily if it is possible to model the background motion by a
global 3D transform. For example, if the background contains multiple objects moving in different
directions, the results will be poor. Typically, good results will only be achieved in situations where a
skilled artist could generate, and track in, a clean plate in order to repair the sequence. However, this
plug-in should make the process quicker and easier. The rig removal algorithm works by estimating
the background motion between successive frames, ignoring the foreground object, and then using
the motion information to look forward and backward in the sequence in order to find the correct
piece of background to fill in the missing region. The size of the missing region and the speed of the
background dictate how far away from the current frame it is necessary to search to find the correct
information.
Before with bird. After applying F_RigRemoval.
Before with taxi. After applying F_RigRemoval.
USER GUIDE
1803
Quick Start |
Quick Start
To remove an unwanted object from an image sequence, do the following:
1. Select Image > Read to load the sequence with an unwanted object.
2. Select FurnaceCore > F_RigRemoval and connect your image sequence to F_RigRemoval’s Src
input. View the output of F_RigRemoval.
3. On each frame, define the area that will be repaired. You can do this in the following three ways:
• If the image sequence has an embedded alpha channel, then you can use that alpha to define
the area to be repaired. To do so, set Rig Region to Src Alpha.
• Using F_RigRemoval’s RigMask input, you can feed in a matte sequence to define the area to be
repaired. To use this matte, set Rig Region to Rig Mask Alpha (or one of the other Rig Mask
options).
• You can use the on-screen rectangle to define the area to be repaired. To do so, set Rig Region
to Box. Position the on-screen rectangle on top of the unwanted object. To key-frame the
position to follow the object, select Set key from the animation menus next to Rig Region BL
and Rig Region TR in the Rig Region Box parameter group. Move to a new frame and reposition
the on-screen rectangle. A new key frame is set automatically. Repeat as necessary until the
rectangle covers the object on every frame you want to remove the object from.
Whichever method you choose, the region does not need to be the exact foreground region, but
just a rough outline. However, you should avoid making it unnecessarily large as this will increase
rendering time.
4. Having defined the region to repair throughout the clip, set Frame Range to the number of
frames that the plug-in needs to analyze forwards and backwards to find enough data to repair
the sequence. On the first frame, this will be quite time consuming as the algorithm needs to
estimate the motion between each pair of frames. Subsequent frames will be much quicker.
5. If it has not been possible to replace all the foreground pixels, either because Frame Range was
set too low or the background information does not exist anywhere within the sequence, the
pixels will be displayed in red. Try to adjust Frame Range until no red pixels are visible and then
render the sequence.
Below, in the figure on the left, we are using a box to define the pixels to replace, and Frame
Range is set to zero. Increasing this value, as shown in the figure on the right, gathers pixels from
other frames and improves the result. To completely remove the red pixels, you’d need a Frame
Range value of 5.
USER GUIDE
1804
Quick Start |
Frame Range = 0. Frame Range = 3.

6. View the results.
Original image. The output of

F_RigRemoval.
Tip: F_RigRemoval is fairly slow to process, which can make the initial keyframing of the
rectangular region frustrating.
Sometimes, the easiest way to adjust the region is to load up F_RigRemoval, and to view the
source so the effect is not processed, but the parameters are visible. Then, animate the
rectangular region over the foreground object you’re trying to remove throughout the
sequence. When you’re happy with the region position, click back onto F_RigRemoval’s
output and wait for it to update. Slowly increase the Frame Range parameter until the
whole region is repaired, then check the output on other frames.
Note: F_RigRemoval uses frames from a wide range around the current frame. Feeding the
output of one F_RigRemoval node into another will greatly increase the memory over head,
as the second F_RigRemoval node will require the first F_RigRemoval node to calculate all its
frames before passing the result on. Therefore, it is strongly recommended that you process
out the first result and then use the processed result with the second F_RigRemoval node.
USER GUIDE
1805
Occlusions |
Occlusions
The algorithm used in F_RigRemoval is unable to differentiate between multiple foreground objects. If
there is another foreground object in the sequence that moves through the background region that is
being used in the repair, this second foreground object will also be cut up and used, resulting in an
incorrect repair. To try and assist in these situations, it is possible to mark regions of the image as not
to be used for repair by setting their alpha value to mid gray. This will ensure that spurious bits of
other foreground objects do not appear in the repair.
In the figure below, we are trying to remove the woman in the center of the screen as she walks from
left to right down the street. At this frame, a man walks in the opposite direction and her feet and his
head overlap.
Original shot.
Below, the figure on the left shows the normal matte for the woman, and the figure on the right
shows the result of using this in F_RigRemoval. Note that the man’s head interferes with the repair
and the reconstruction of the pavement is particularly bad, probably due to the man becoming the
dominant motion source.
The normal matte. The output.
To fix this, we can adapt the matte to include a mid gray area over the man. This tells the rig removal
algorithm to ignore that area in the repair. This matte is shown below in the figure on the left, and the
result is shown in the figure on the right. Note that the repair on the pavement is improved, and the
man is simply clipped rather than being used in the repair.
USER GUIDE
1806
Occlusions |
The matte with a mid-gray The output.

area.
USER GUIDE
1807
Parameters |
Parameters
Rig Region - Defines the area to repair.

• Box - repair the area inside a rectangular box, controlled by the box parameters below or the on-
screen box.
• Src Alpha - repair the region defined by the alpha of the source input.
• Src Inverted Alpha - repair the region defined by the inverted alpha of the source input.
• RigMask Luminance - repair the region defined by the luminance of the Rig Mask input.
• RigMask Inverted Luminance - repair the region defined by the inverted luminance of the Rig
Mask input.
• RigMask Alpha - repair the region defined by the alpha of the Rig Mask input.
• RigMask Inverted Alpha - repair the region defined by the inverted alpha of the Rig Mask input.
Frames Searched - Sets whether to search forwards, backwards, or in both directions to find missing
data.
• Forward and Backward - searches before and after the current frame.
• Forward - searches frames after the current frame.
• Backward - searches frames before the current frame.
Frame Range - Sets the number of frames the algorithm should look forwards and backwards in the
sequence to find the missing data. If you are getting red pixels, then increase this value. See Quick
Start .
Frames Used in Range - If Frame Range has to be set to a large number to make an effective repair,
the rendering time can be prohibitive. Frames Used in Range can speed up the repair by not using
every frame to fill the foreground region, effectively skipping frames. However, this may reduce the
quality of the result.
• All Frames - use every frame in the searched frame range to construct the repair.
• Half of Frames - use every other frame in the searched frame range to construct the repair.
• Quarter of Frames - use every fourth frame in the searched frame range to construct the repair.
• 10% of Frames - use every tenth frame in the searched frame range to construct the repair.
• Max 25 Frames - use no more than 25 frames from the searched frame range to construct the
repair. This option can be useful if Frame Range has been set to a very large number.
Max Rig Movement - To avoid perspective changes, F_RigRemoval searches for the missing data
inside an area immediately around the rig region. Max Rig Movement defines the width of this area
(in pixels). Fast movement in the Src footage requires a higher value than slow movement. However,
USER GUIDE
1808
Parameters |
because the area used for the repair may be from a different part of the image, high values can
introduce perspective problems.
Rig Region Box - The rectangular area used to define the repair when Rig Region is set to Box.
Rig Region BL - controls the position of the bottom left corner of the rig region.
Rig Region TR - controls the position of the top right corner of the rig region.
Advanced - The lesser used refinement controls.

• Medium - uses a bilinear filter. This gives good results and is quicker to render than high
filtering.
• High - uses a sinc filter to interpolate pixels giving a sharper repair. This gives the best results
but takes longer to process.
Luminance Correct - Switch this on to correct for luminance changes from information taken from
other frames. This is particularly important if the lighting changes throughout the sequence.
Perspective Correct - Switch this on to correct for minor perspective changes.
Overlap Correct - The repair is built up using slices of information from other frames in the
sequence. These slices can be overlapped and blended to give a more natural looking repair. This
parameter controls how much the regions overlap. Increasing this parameter too much will degrade
image sharpness.
Repair Fail Marker Opacity - Sets the level of transparency of the red pixels used to show where
the repair has failed.
Preserve Alpha - Switch this on to preserve the original alpha channel. By default, this is switched
off, and the alpha channel is set to white where the repair has failed and black everywhere else.
USER GUIDE
1809
Using F_Steadiness
This chapter looks at how to stabilize a shot using F_Steadiness.
Introduction
F_Steadiness uses Global Motion Estimation (GME) to calculate a four- corner pin, so that camera
motion within a single shot can be smoothed out over a range of frames or removed by locking to a
specific frame.
F_Steadiness needs to analyze the input clip before it can render useful output. This analysis is done
when you press Analyse in the F_Steadiness controls. During the analysis, F_Steadiness keyframes a
four-corner pin which will stabilize the clip in subsequent renders. Without having performed an
analysis pass, F_Steadiness will not do anything useful on render.
F_Steadiness can work in two ways. These are:

1. Smooth - A range of frames around each frame is analyzed for motion and an average of that
motion used to calculate the corner pin. Use this to keep the overall camera motion, but to
smooth out sharp bumps and kicks.
2. Lock - A lock frame is specified and steadiness attempts to register each individual frame to that
lock frame. Use this to completely remove camera motion from the sequence.
In lock mode, each frame in the clip must share a substantial amount of the scene with the lock
frame, so you can’t lock each frame in a 360 degree pan to the first one. However, in smooth mode,
because F_Steadiness is only working on a small window of frames around the current frame, you
can use it in shots which change completely over time.
The analysis region is used to control which section of the reference frame is being matched to each
source frame. In lock mode, the reference is the lock frame, so you should position the analysis
region when looking at the lock frame. In smooth mode, it looks at the incremental differences
between frames, in which case you should place the analysis region in the area you want to appear
with smooth motion.
The transformation in F_Steadiness is concatenated with other NukeX transform nodes. This means
that if you add a row of F_Steadiness and NukeX transform nodes to a tree, their functions are
combined. Because the image is only resampled once, there is no loss of image quality and
USER GUIDE
1810
Using F_Steadiness |
processing time is decreased. However, as in NukeX, certain nodes, including color correction nodes,
can break the concatenation.
For an overview of Global Motion Estimation, and a description of the common way Global Motion
Effects work, please see Global Motion Estimation.
USER GUIDE
1811
Quick Start | Smoothing Out Camera Motion
Quick Start
This section gives a very brief outline of how to use the plug-in.
Smoothing Out Camera Motion

To keep the overall camera motion but to smooth out sharp bumps and kicks, do the following:
1. Find a shot that has some camera shake in it and select Image > Read to load it.
2. Select FurnaceCore > F_Steadiness to apply F_Steadiness, and view its output.
No valid key framed analysis found, please press Analyse
3. In the Advanced F_Steadiness controls, make sure Mode has been set to Smooth.
4. Click on the Analyse button.
F_Steadiness will now start analyzing each frame in the shot, figuring out the smoothing corner
pin and writing it as key frames to the corner pin parameters, Bottom Left, Bottom Right, Top
Left, and Top Right.
After a brief pause (while F_Steadiness calculates the transforms in the frame range), F_Steadiness
will update the timeline and you will see the steadied image render in the viewer.
If at any point you interrupt the analysis, the pins it has calculated until that point will be retained.
5. Play or scrub through the stabilized frames.
6. If you want to make the result smoother, increase the Smoothing parameter in the Advanced
parameter group and the corner pin will be recalculated immediately to give a smoother shot.
You don’t need to re-analyze the sequence if you do this; as F_Steadiness has kept the raw inter-
frame transforms cached away, all it needs to do is re-write the keys for the average smoothing
pin.
Locking To A Frame
To completely remove camera motion from a sequence, do the following:
1. Find a shot that has camera shake but where all frames share scene information and select
Image > Read to load it.
2. Select FurnaceCore > F_Steadiness to apply F_Steadiness, and view its output.
No valid key framed analysis found, please press Analyse.
3. In the Advanced F_Steadiness controls, set Mode to Incremental Lock.
USER GUIDE
1812
Quick Start | Locking To A Frame
4. Choose a frame somewhere in the middle of the sequence that you want to lock to, and set the
Lock Frame parameter to that frame.
5. Scrub back and forth through the shot and look for a region of the shot that is shared by all
frames and doesn’t change much (for example, avoid a region with people walking in front of it).
6. Whilst looking at the lock frame, position the on-screen widget for the Analysis Region over that
region.
7. Hit Analyse.
The effect will start to analyze, working first forward from the lock frame, then backwards from it,
until all frames to be analyzed are done.
The timeline will immediately update to give you feedback in the viewer.
USER GUIDE
1813
Parameters | Locking To A Frame
Parameters
Analyse - This is a push button which will trigger an analysis of the input clip and calculate a corner
pin. Interrupting the analysis will not delete the corner pin keys that have already been calculated.
Render During Analysis - If set, this toggle will cause the effect to update the time line and render a
freshly analyzed frame so you can see the progress of the effect. Doing so will slow down the analysis
somewhat, so toggle this off to speed up the general analysis.
Clear Analysis - Pressing this push button will delete all key frames from the corner pin parameters,
• Specified Range - which will look at the parameters Analysis Start and Analysis Stop for the range
of frames to analyze,
• Source Clip Range - which will automatically determine the range of frames to analyze from the
length of the input clip.
Analysis Start - The first frame to analyze from if Analysis Range is set to Specified Range.
y.
transforms.
Analysis Region BL - controls the position of the bottom left corner of the analysis region.
Analysis Region TR - controls the position of the top right corner of the analysis region.
Advanced - the lesser used refinement controls.
USER GUIDE
1814
Mode - This parameter controls whether F_Steadiness is smoothing the shot while keeping the
overall camera motion, or locking the shot to a single frame to completely remove camera motion. It
can be set to:
• Incremental Lock - in which case, it will calculate the pin that takes each frame to the lock frame.
This calculates the pin by working from the lock frame out to each frame, calculating the GME
between each frame incrementally and accumulating it to create the corner pin.
• Absolute Lock - this also calculates a pin that takes each frame to the lock frame. However, it does
so by doing GME directly from the frame in question directly to the lock frame.
• Smooth - in which case, we are smoothing the shot for a range of frames described by the
Smoothing parameter. This is the default value. It can be used to keep the overall camera motion,
but to smooth out sharp bumps and kicks.
Incremental locks work better in some situations, whilst absolute locks work better in others.
However, absolute locks, when they work, are often more accurate. If, for example, you want to get a
lock between frame 0 and frame 100, try using absolute lock first. If absolute lock doesn't work, it can
be because frame 0 and frame 100 are too different. In this case, incremental lock can help if there is
a gradual spatial progression from frame 0 to frame 100.
Smoothing - This controls the range of frames to average motion over when Mode is set to Smooth.
The default value is 10.
Lock Frame - This controls the frame which will be locked to when Mode is set to either of the lock
modes. By default, this is set to 0. Note that what is frame 0 for F_Steadiness is frame 1 in NukeX.
Therefore, if you are looking at frame 3 in the Viewer and want to use that frame as the lock frame,
you need to enter 2 as the Lock Frame value.
Accuracy - This controls the time/accuracy trade off. The higher this is, the slower the analysis, but
you have a better likelihood of a good result.

• Medium - uses a bilinear filter. This gives good results and is quicker to render than high filtering.
• High - uses a sinc filter to interpolate pixels giving a sharper repair. This gives the best results but
takes longer to process.
Invert - If set, then the inverse of the calculated four-corner pin is used during render. This works
best with the lock modes, and can be used to track static locked-off plates back into a shot.
Auto Scale - To smooth out or remove camera motion, F_Steadiness translates and rotates the
frames in the source clip. This leaves black pixels around the image. The Auto Scale parameter lets
you fill in the black gaps at the edges by scaling the output image up. A value of 1 (the default) scales
the image up until no black is visible, whereas a value of 0 disables scaling and leaves the black edges
USER GUIDE
1815
untouched. Auto Scale uses the minimum scale necessary to remove the black gaps at the edges to
preserve as much detail as possible.
Bottom Left - the lower left corner pin.
Bottom Right - the lower right corner pin.
Top Left - the upper left corner pin.
Top Right - the upper right corner pin.
USER GUIDE
1816
Using the BlinkScript Node
The BlinkScript node runs Foundry's Blink framework enabling you to write your code once and run it
on any supported device. This is achieved through code translation, in which the Blink code is turned
into specific code for each target device. Code is generated and compiled on-the-fly, allowing you to
switch between devices at will.
About BlinkScript
BlinkScript runs a Blink "kernel" over every pixel in the output, where a Blink kernel is similar to a C++
class, but with some special parameter types and functions. Through translation, the code in the
BlinkScript node can be turned into normal C++ or SIMD code for the CPU, or OpenCL for the GPU.
The Blink framework streamlines plug-in development workflow significantly, as you no longer have
to exit Nuke to compile your code.
Note: GPU acceleration requires an NVIDIA GPU and drivers for CUDA 4.2 or above. To use
the GPU when rendering from the command line, add --gpu to your command.
If your computer enters sleep mode, the CUDA drivers cannot recover and you must restart
Nuke to use GPU acceleration.
Note: Nuke supports AMD GPUs on late 2013 Mac Pro 6,1 and mid 2015 Mac Pro 11,5,
running OS X 10.9.3 'Mavericks', or later.
Note: BlinkScript is disabled in the Non-commercial version of Nuke.
You can publish kernels in Group nodes which can then be saved as gizmos, if required. Published
kernels can be encoded to protect your IP using BlinkScript's built-in kernel protection. Protected
kernels are not readable when the published node is saved to a script.
Warning: BlinkScript is very flexible, as there are no restrictions on the code you can write
within a kernel. As a result, code compiled from the Kernel Source can cause Nuke to crash,
so please use caution!
USER GUIDE
1817
Using the BlinkScript Node |
For more information, see Nuke's Help menu under Documentation > Guide to Writing Blink
Kernels or navigate to https://learn.foundry.com/nuke/developers/110/BlinkKernels/.
USER GUIDE
1818
Quick Start |
Quick Start
To get started with BlinkScript:
1. Connect the BlinkScript node to your sequence or image. See Connecting the BlinkScript Node for
more information.
2. Edit the default InvertKernel, load an example kernel, or write your own from scratch. See
Loading, Editing, and Saving Kernels for more information.
3. Set the standard Kernel Parameters, those that are present in all kernels, such as Use GPU if
available and format. See Setting Kernel Parameters for more information.
4. Set the available memory and height-per-tile limits to control performance. See Changing
Performance Settings for more information.
5. Publish your kernel and create a gizmo, if required. See Publishing and Protecting Your Kernels
USER GUIDE
1819
Connecting the BlinkScript Node |
Connecting the BlinkScript Node

1. Select Other > BlinkScript to create a BlinkScript node in the Node Graph.
2. Connect the src input to your sequence or image.
Note: BlinkScript supports as many inputs as you create in your kernel, but the default
InvertKernel only has a src input.
3. Connect a Viewer to the output of the BlinkScript node.

The Viewer displays your inverted image.
USER GUIDE
1820
Loading, Editing, and Saving Kernels | Loading Kernels
Loading, Editing, and Saving

Kernels
Kernel management is taken care of on the first tab of the BlinkScript properties panel. Double-click
the BlinkScript node in the Node Graph to display its properties.
Note: Loading, editing, and saving kernels is only available if you have a NukeX license.
Loading Kernels
1. Enter the file-path in the Kernel File field or click the folder icon to browse to the kernel's
location.
Tip: BlinkScript kernels use the .rpp file extension.
2. Click Load.
The selected kernel is compiled and read into the Kernel Source field.
USER GUIDE
1821
Loading, Editing, and Saving Kernels | Editing Kernels
Editing Kernels
You can edit existing kernels or write your own from scratch by clicking Clear and entering code in
the Kernel Source field. Bear in mind that you won't see any results in the Viewer until you click
Recompile.
Tip: You can change how text in the Kernel Source appears using the controls in the
Preferences > Panels > Script Editor tab. After saving your preferences, close and then re-
open the Kernel Source in the Properties panel to apply the changes.
The first line in a kernel is always a declaration, similar to a C++ class and derived from
ImageComputationKernel, which describes a kernel used to produce an output image.
Note: BlinkScript in Nuke only works with ImageComputation kernels. It does not work with
reduction or rolling kernels. Reduction and rolling kernels can, however, be used in the Blink
API as part of an NDK plug-in written in C++.
In the case of the default InvertKernel:

kernel InvertKernel : ImageComputationKernel<eComponentWise>
Parameters for the kernel are declared in the param section, in the same way as you would declare
member variables in a C++ class. If your kernel doesn’t require any parameters, you can omit this
section.
For example, the InvertKernel has a single parameter, multiply:

param:
float multiply; //This parameter is made available to the user.
When a kernel is compiled inside the BlinkScript node, controls are generated for each of the kernel’s
parameters and added to the Kernel Parameters tab. In the case of InvertKernel, the node has just
one custom parameter, Multiply.
Kernel parameters can be C++ built-in types such as float or int. Vector and matrix parameters are
also supported. The following table contains the control types that the Kernel Source can expose on
the Kernel Parameters tab in the properties panel.
USER GUIDE
1822
Loading, Editing, and Saving Kernels | Editing Kernels
Control Type Description
Bool_knob Parameters of type bool are represented by a Bool_knob, or single checkbox.
Int_knob Parameters of type int generate an Int_knob with a single numeric input box.
Right-clicking in the input box displays a menu allowing you to animate the
value.
MultiInt_knob Parameters of type int2, int3, int4, or int[ ] generate a MultiInt_knob, with
multiple numeric input boxes. As with the Int_knob, right-clicking in the input
boxes displays a menu allowing you to animate the values.
Float_knob Parameters of type float generate a Float_knob with a numeric input box, linear
slider, and animation menu button. The slider ranges from 0 to twice the default
value of the parameter, or 0 to 1 if no default value is set.
XY_knob Parameters with two floating-point values, types float2 or float[2], are
interpreted as positions in 2D space and generate an XY_knob with an adjustable
handle in the Viewer.
XYZ_knob Parameters with three floating-point values, types float3 or float[3], are
interpreted as positions in 3D space and generate an XYZ_knob with an
adjustable handle in the 3D Viewer.
AColor_knob Parameters with four floating-point values, type float4 or float[4], are interpreted
as colors with alpha, and generate an AColor_knob. This initially displays a single
entry box and slider, a button to split to four entry boxes, a button to display the
Nuke color wheel, and a swatch showing the current color with an eyedropper
for sampling in the Viewer.
Array_knob Parameters with nine floating-point values, including float3x3, are displayed as
an Array_knob with a 3x3 grid. Parameters with sixteen floating-point values,
including float4x4, are displayed as an Array_knob with a 4x4 grid.
MultiFloat_knob Parameters with numbers of floating-point values not already listed above, such
as float[5], generate a MultiFloat_knob with a numerical entry box for each
value, a single linear slider, and an animation menu button. The slider ranges
from 0 to 1.
The rest of the kernel is up to you, within certain guidelines listed in Nuke's Help menu under
Documentation > Guide to Writing Blink Kernels or at
https://learn.foundry.com/nuke/developers/110/BlinkKernels/. There are also several example kernels
to get you started.
USER GUIDE
1823
Loading, Editing, and Saving Kernels | Saving Kernels
Saving Kernels
1. Enter the file path in the Kernel File field or click the folder icon to browse to the intended
location.
Tip: BlinkScript kernels use the .rpp file extension.
2. Click Save.
The selected kernel is saved to the specified location. There is no compile step during saving as
the kernel is compiled when loaded.
USER GUIDE
1824
Setting Kernel Parameters | GPU or CPU?
Setting Kernel Parameters

Some BlinkScript controls are used by all kernels. Rather than write these individually in each
instance, the BlinkScript properties panel includes Kernel Parameters and Settings tabs, providing
easy access to the standard kernel controls.
The Kernel Parameters tab also contains any controls exposed by specific kernel commands. For
example, the default InvertKernel exposes the Multiply control. See Editing Kernels for more
information on exposing controls.
The Publish button also resides on the Kernel Parameters tab, but we'll get into that in more detail
in Publishing and Protecting Your Kernels.
GPU or CPU?
The read-only Local GPU field displays the graphics card installed locally, if there is one available. You
can toggle the GPU on and off using the Use GPU if available control. If you have multiple GPUs
available, select the required device using the Preferences > Performance > Hardware > default
blink device dropdown.
Note: Selecting a different GPU requires you to restart Nuke before the change takes effect.
Tip: Even if there is no GPU available locally, you can still enable Use GPU if available.
BlinkScript then uses the GPU as and when one becomes available.
You should also enable this if you wish to render from the command line with the --gpu
option.
preferences, GPU processing is shared between the available GPUs for extra processing speed. See
Windows, Mac OS X and macOS, or Linux for more information on the GPUs Nuke supports.
Additionally, enable Vectorize on CPU to use SIMD acceleration on the CPU where possible. See Help
> Documentation for more information on kernel execution.
USER GUIDE
1825
Setting Kernel Parameters | Specifying the Output Format
Specifying the Output Format

BlinkScript's output defaults to the union of all its inputs, but you can specify a different output
format by enabling Specify output format and selecting the required ratio from the format
dropdown.
Tip: Enabling Specify output format also adds the format control to the kernel once it has
been published. See Publishing and Protecting Your Kernels for more information.
Changing Performance Settings

The performance controls on the Settings tab are common to all Blink kernels and don't affect the
output, but can improve the processing speed:
• Percentage of GPU memory to use - sets the upper limit on the amount of memory BlinkScript
can allocate.
Higher values than the default 50% can improve render times, but may introduce errors if the GPU
is in use by other tasks.
• Percentage of image height per tile - processing is done in tiles, where each tile is the full width of
the image. Adjusting the percentage of the image height processed in any one tile can affect
performance.
Higher values generally improve performance, providing that there is enough GPU memory
available for processing large tiles, whereas lower values produce faster feedback to the Viewer.
USER GUIDE
1826
Publishing and Protecting Your Kernels | Changing Performance Settings
Publishing and Protecting Your

Kernels
When you're happy with your kernel, you can publish it for use elsewhere in the script or in an
entirely separate Nuke script. Published kernels are wrapped in a Group node, so you can save them
as gizmos if required.
Blink kernels can also be encoded using the built-in kernel protection. Protected kernels are not
readable when the published node is saved to a script.
Note: Kernel protection offers a fairly low level of security and might not be sufficient to
protect valuable source code.
Certain BlinkScript functions are only available with a NukeX license. The following table describes the
various levels of access for the Nuke family.
Feature
Create node Load, edit, and Adjust controls Publish

Product save
Nuke
Assist
Nuke
NukeX
Nuke
Studio
To publish a kernel:
1. Double-click the BlinkScript node to open the properties panel and click the Kernel Parameters
tab.
2. If you intend to encode your kernel, enable Protect kernel.
In a Nuke script, protected kernels appear as shown in the example below.
USER GUIDE
1827
Publishing and Protecting Your Kernels | Changing Performance Settings
3. Click Publish.
Nuke creates a Group node using the name of the kernel as a label. For example, using the
WipeKernel supplied in the documentation produces the following expanded group.
If you want to save the group as a gizmo, go to the group's properties panel Node tab and then
click export as gizmo. See Creating and Accessing Gizmos for more information.
USER GUIDE
1828
Limitations and Known Issues | RGBA Only
Limitations and Known Issues
RGBA Only
Processing is currently limited to the RGBA channels only. To work with other channels, use a Shuffle
node to shuffle them into RGBA before connecting to the BlinkScript node.
GPU
Complicated kernels can cause driver timeouts on the GPU if they take too long to execute. The lower
the specification of your GPU, the more likely this is to happen.
Nuke supports AMD GPUs on late 2013 Mac Pro 6,1, mid 2015 MacBook Pro 11,5, and later 2016
MacBook Pro 13,3, running OS X 10.9.3 'Mavericks', or later While, in some cases the bit-wise equality
between GPU and CPU holds, for some operations there are limitations to the accuracy possible with
this configuration.
Crashing and Infinite Loops

Since the BlinkScript node allows arbitrary code to be written and run, it is possible to either crash or
lock up Nuke. The BlinkScript node does not check your code to make sure it is sensible before
running it, so we advise you to use caution!
Known Issues
The following known issues exist in BlinkScript's current form:
OpenCL on OS X 10.8 (Mountain Lion)

Running OpenCL kernels on Mac OS X 10.8.2 and 10.8.3 can be unreliable. We recommend updating
to 10.8.4 or later.
USER GUIDE
1829
Written Tutorials
Welcome to the written tutorials!
Introduction
If you’ve gone through the Getting Started section - which we highly recommend - you already know
something about Nuke. These tutorials show how to pull everything together through a series of
practical examples.
You can also go through these tutorials using Nuke Non-commercial, but some features used in the
tutorials may be disabled in the non-commercial version. For more information, see About Nuke
Non-commercial.
USER GUIDE
1830
The Projects
These tutorials include the following projects:
• Tutorial 1: Compositing Basics explains the Nuke user interface, project workflow, and basic
compositing tasks.
• Tutorial 2: 2D Point Tracking demonstrates how to track image patterns, stabilize footage, lock
down images for clean plates, and match-move.
• Tutorial 3: Keying and Mattes shows you how to pull mattes with standard keying tools and Nuke's
own image-based keyer.
• Tutorial 4: 3D Integration shows how you can use Nuke's 3D workspace to help your 2D
compositing.
USER GUIDE
1831
Installing the Project Files | To Create the Tutorial Directory (Windows)
Installing the Project Files

Before you continue, download the tutorial project files from our website and move them to a
directory you’ll create, called “Nuke_Tutorials”. It’s up to you where you put your tutorial files, but
here’s our recommendations below depending on your operating system. Whatever you do, you’ll
need to remember where you put these files.
Tip: If you’re using a Mac or Linux system, log in under administrator privileges to avoid
issues with permissions when installing the files.
To Create the Tutorial Directory (Windows)

1. On the Windows desktop, double-click the My Computer icon to open a file browser.
2. Double-click on Local Disk (C:) and open the C:\Documents and Settings\All Users\Application
Data folder.
3. Click the right-mouse button over the displayed directory and choose New > Folder.
4. Name the folder: Nuke_Tutorials.
To Create the Tutorial Directory (Mac or Linux)

1. Open a shell or terminal window.
2. At the command line, enter mkdir ~/Nuke_Tutorials to create the tutorial directory under your
user or “home” directory.
To Download and Install the Project Files

1. Click Nuke Tutorials to download the project files to your local computer.
2. Extract the downloaded files and move (or copy) them to the Nuke_Tutorials directory you
created earlier.
You’re now ready to start the first tutorial with Nuke.
USER GUIDE
1832
Tutorial 1: Compositing Basics
Hello! This tutorial is your introduction to Nuke, where you’ll create a simple composite and breeze
through most of the windows, on-screen controls, and other user interface items.
Introduction
We’ve heard rumors that many people would rather march through icy rain than review an
introductory tutorial on digital compositing. Certainly, that’s not you. When you finish this lesson
you’ll have a good understanding of the Nuke workflow and should feel confident about approaching
the other tutorials.
Your first composite in Nuke.
Before you get into the project, we have some administrative things to do - such as defining a few
application preferences and project settings. We know this sort of thing is not terribly exciting, but it
is terribly important, so please be patient and we’ll get through it as quickly as possible.
Note: If you haven’t already downloaded and moved the tutorial project files to the Nuke_
Tutorials directory you created, turn to Installing the Project Files for instructions.
USER GUIDE
1833
Starting Nuke | To Launch Under Windows
Starting Nuke
The Nuke icon may appear on your desktop. If so, double-click it to launch the application. Otherwise,
start Nuke with one of the methods described below, assuming you have installed Nuke to the default
location.
To Launch Under Windows

• From the Start menu, choose All Programs > The Foundry, and then select Nuke11.3v5.
To Launch Under Mac

• Open the /Applications/Nuke/ folder and double-click the Nuke11.3v5 icon.
To Launch Under Linux

• Open the /usr/local/Nuke11.3v5/ folder and double-click the Nuke11.3v5 icon.
Tip: If you’re operating under Linux, you can also launch Nuke from the command line of a
terminal. Simply navigate to the Nuke directory and enter the name of the Nuke application.
A clean copy of the main Nuke window appears. Divider lines organize the window into different
panes. Each pane has one or more pages of content, separated by tabs at the top of the pane. The
Toolbar appears at the left edge of the main window.
USER GUIDE
1834
Starting Nuke | To Launch Under Linux
By default, the panes are setup to display the Viewer, the Node Graph/Curve Editor, and Properties.
You’ll create the script for this project inside the Node Graph page on the Node Graph/Curve Editor
pane. We’ll talk about each of these on-screen controls when you need them for the project.
USER GUIDE
1835
Using the Toolbar | To Launch Under Linux
Using the Toolbar

The Toolbar includes the options you can use to build your project, such as importing images,
layering images, drawing shapes and masks, applying color correction, and so on. Each Toolbar icon
displays a menu of operators or nodes that you can select. Roll the mouse pointer over the Toolbar
and you’ll see pop-up tool tips that identify each icon.
USER GUIDE
1836
Using the Menus | To Launch Under Linux
Using the Menus

The Nuke menu bar appears at the top of your screen, outside the main window. This menu begins
with the options File, Edit, Workspace, and so on. When instructed to do so, make selections from
the menu bar, or click the right mouse button to choose from a pop-up version of the menu bar.
The Nuke menu bar.
The “right-click” menu is highly contextual. Its options change according to the location of the mouse
pointer. Right-click over the Node Graph, for example, and you’ll see the options from the menu bar
and the nodes you can insert from the Toolbar. Right-click over the Viewer pane and you’ll see a
menu of Viewer options.
The "right-click" menu.
USER GUIDE
1837
Using the Menus | To Launch Under Linux
Try the right-click menu when you can’t find appropriate controls or menu options. Many features are
hidden in the menu until you’re in the situation where you need to use them.
Note: Nuke’s menu bar, at the top of the screen, is organized a little differently between the
operating systems, but the right-click menu contains the same options, regardless of the
system you’re using to run Nuke.
USER GUIDE
1838
Customizing Your Workspace | To Launch Under Linux
Customizing Your Workspace

Nuke gives you several options for customizing the window layout. It’s time for you to claim your copy
of Nuke and make it your own! You don’t need to customize the layout for this lesson, but why not try
it now for your own personal amusement? Here are some things you can do to reorganize the
window layout:
• Drag a divider line between panes to change the size of the panes.
Resizing a pane.
• To divide a pane, click on the content menu (the checkered box at the upper-left corner of each
pane), and choose Split Vertical or Split Horizontal.
Splitting a pane.
• To discard a pane, click on the content menu and choose Close Pane.
Closing a pane.
• To add a new tabbed page to a pane, click on the content menu and choose one of content options,
such as New Viewer or Curve Editor.
• Click on the “x” inside a tab to discard a tabbed page.
USER GUIDE
1839
Customizing Your Workspace | To Launch Under Linux
Closing a tab.
• To move a tabbed page, drag the tab to another pane inside the main window.
• To tear-off a page as a floating window, drag the tab outside the borders of the main window, or
simply Ctrl+click (Mac users Cmd+click) on the tab name.
• Drag a floating window into a pane, inside the main window, to convert it to a tabbed page.
• From the menu bar, choose Workspace > Save Workspace to save the current layout. Choose
Workspace > Restore Workspace x to apply a previously-saved layout.
USER GUIDE
1840
Saving Files and File Backup | Defining File/Saving Options
Saving Files and File Backup

We assume you already know how to save files (Hint: choose File > Save Comp As). In addition, Nuke
includes an autosave feature, which helps recover project files after a system failure. Yes, we know
that never happens to you, but in the unlikely event that it does, you won’t lose your work when you
have autosave enabled.
Defining File/Saving Options

1. Click the right mouse button over the Node Graph pane, and choose Edit > Preferences.
Notice the autosave filename directory is set to:
[firstof [value root.name] [getenv NUKE_TEMP_DIR]/].autosave
You don’t need to make a change; this simply tells Nuke to store automatic backup files in the
same directories as your project files or the path supplied by the NUKE_TEMP_DIR environment
variable (for more information on environment variables, see Configuring Nuke).
Now, how often would you like Nuke to generate an automatic backup while you’re working? Every
five minutes?
2. Change the force comp autosave after option to 300 seconds, to generate an automatic backup
every five minutes.
3. Click Save Prefs to keep the changes and then Close to return to the main window.
If you close this dialog box without clicking the Save button, then the changes affect only the
current session of Nuke.
USER GUIDE
1841
Saving Files and File Backup | Recovering Back-Up Files
Recovering Back-Up Files

You may ask, “How do I recover a back-up file in the event of a system or power failure?” Good
question! When you relaunch Nuke, you’ll see a message that asks if you want to recover the
.autosave file for the project that was last open. Click Yes and Nuke opens the back-up file.
Tip: The .autosave files can still be useful, even when you properly exit Nuke, because they
are not deleted from the directory. You can, for example, rename an .autosave file to create
an archive of the previous version of your project file.
Sometimes you may see the recovery message even though you have not experienced a system
failure. This happens when you exit Nuke without saving the changes to a project file, and Nuke
recognizes that the time stamp on the .autosave file is later than the Nuke project file you’re trying to
open. In this case, you decide which version of the project file you want to open.
Turning off Automatic Back-Up

Okay. You’re reading this, so we assume you’re a freewheeling rebel who possibly enjoys the risk of
losing your work. It’s an adrenaline thing. Or perhaps you prefer to do everything yourself, manually,
and you have a secret obsession for saving your files. Whatever the reason, you can disable the
autosave features by setting the intervals for both “autosave idle” and “force autosave” to zero
seconds. That’s all you need to do. Good luck.
USER GUIDE
1842
Setting Up the Project | To Set Up Your Project
Setting Up the Project

When you start a new project, you need to define project settings for length or frame range, the
number of frames per second for playback, and the output format. These options appear on the
Project Settings dialog box.
To Set Up Your Project

1. Click the right mouse button over the Node Graph, and then choose Edit > Project Settings.
USER GUIDE
1843
Setting Up the Project | To Set Up Your Project
2. In the frame range fields, enter a range of 1 to 28. This is the length of the shot we create for the
project.
3. Enter 24 as the frames per second (fps).
4. Click the full size format dropdown menu and choose PC_Video 640 x 480.
5. Close the Project Settings control panel.
Note: On the Project Settings control panel, the Color tab includes options that ensure
color integrity for your display and output devices. You don’t need to change the LUT for
these tutorials, but we recommend that you research and set these options for your own
projects.
Until now, everything you’ve done is standard procedure for a new project. You used the menu bar to
access several features during the setup process, and now you’ll use the Nuke toolbar to insert nodes
and create a compositing tree.
USER GUIDE
1844
Working with Nodes | Inserting Nodes
Working with Nodes

A node is simply one of the building blocks for the list of operations you want to complete. A node tree
is a diagram that shows the order in which the operations are performed. Do the following to add a
few nodes and start your node tree. The result creates the background for the project.
Inserting Nodes
To insert nodes:
1. On the Toolbar, click the first icon to display a menu for nodes that are in the Images category.
2. Select Constant from the menu to insert this node into the Node Graph pane.
When you insert a new node, its control panel also displays with parameters that let you define
what the node produces. In this case, the Constant node creates a solid color backdrop.
3. In the Constant control panel, click on the color wheel to open the Color Picker.
USER GUIDE
1845
4. Drag the color sliders and the cursor inside the wheel to choose a light color, something
appropriate for the “horizon” of the composite background. Then, close the color wheel window.
At this point, you should probably rename “Constant” to something more descriptive.
5. Inside the control panel, click on the Constant name. You can now edit the name, so type
Background and press Enter.
From here onward, we’ll call this node the “Background” node.
6. Close the control panel for the Background node. When you need to reopen it, just double-click
the node and the control panel reappears.
7. Click on the Background node to select it. Then, click the right mouse button and choose Draw >
Ramp.
USER GUIDE
1846
8. Drag the tail of the arrow from the Viewer1 node to the center of the Ramp1 node. You’ll see the
output of the Background node and the ramp controls displayed in the Viewer window.
9. Click the Color tab inside the control panel for Ramp1. Then choose a dark color that blends well
with the color you selected for the Background node.
10. Click the Ramp tab in the control panel to reactivate the overlay controls. Then, drag the p0 and
p1 control points to adjust the spread and angle of the ramp over the background.
USER GUIDE
1847
11. When you’re happy with the results, close the Ramp1 control panel to remove the overlay.
USER GUIDE
1848
Connection Tips | Inserting Nodes
Connection Tips
Most nodes have input and output connectors that are used to establish the order in which the
operations are calculated.
Connectors on a node.
Try the following to connect nodes after you insert them into the Node Graph:
• Drag an input or an output connector onto another node to establish a connection.
• Select a node, press the Shift key and select a second node. Then press Y to connect the first node
to the output of the second node.
USER GUIDE
1849
• Select a node, press the Shift key and select a second node. Then press Shift+Y to connect the
second node to the output of the first node.
• Select a node and press Ctrl/Cmd+Shift+X to extract the selected node from the tree.
• For nodes that have two inputs, select the node and press Shift+X to swap the A/B inputs.
USER GUIDE
1850
• Drag the mask connector to the node that provides the image you want to use as the mask for the
selected node.
USER GUIDE
1851
Importing Image Sequences | To Read the Images
Importing Image Sequences

For this project, you need to import a few image sequences for the foreground elements and a
background plate.
To Read the Images

1. Click on a blank space in the Node Graph. This ensures none of the nodes are selected.
2. Click the right mouse button over the Node Graph and choose Image > Read (or press R over the
Nuke window).
Tip: Pressing R with an existing Read node selected, opens the file browser at the location
specified by that node.
A file browser appears. This is where you select the image file you want to import. When you
browse through your directories from this window, Nuke displays sequentially-numbered files as
one item in the directory.
3. Browse to the Nuke_Tutorials/CompBasics/ directory.
4. Add a bookmark to this directory. Right-click over the list, on the left side of the file browser
window, and choose Add from the menu.
USER GUIDE
1852
5. Type a name for the bookmark or keep the default, which is the directory name. Then click OK.
6. Open the engine_rgba directory, select the engine.v01.####.exr image sequence, and click
Open.
Nuke retrieves the image sequence and displays it as a thumbnail on the node. The Read control
panel displays the resolution and the frame range for the image.
Note: Nuke reads images from their native format, but the Read node outputs the result
using a linear color space. If necessary, you can change the Colorspace option in the Read
node’s control panel, or insert a Color > Colorspace node to select the color scheme you
want to output or calculate.
7. Drag a marquee (hold down the left mouse button while dragging) around the Background and
Ramp nodes to select them. Then drag them to the right to make room for additional nodes.
USER GUIDE
1853
8. Choose Image > Read from the right-click menu to import another image sequence. Use the file
browser to select the image sequence stored in Nuke_Tutorials/CompBasics/smoke_
left.wh/smoke_left.####.rgba.
9. Add one more Read node and retrieve the image sequence stored in Nuke_
Tutorials/CompBasics/smoke_right.wh/smoke_right.####.rgba
10. Arrange the nodes, as shown above, to allow some room to create the connections for the node
tree.
USER GUIDE
1854
Navigating Inside the Windows | Panning Your View
Navigating Inside the Windows

The Node Graph panel can seem very small, especially when your node tree grows. True, you already
know how to resize and tear-off the windows, but sooner or later you may run out of display real
estate. It’s time to learn some navigation controls that can help you work in the Node Graph (and
other windows) in Nuke. Try the following navigation controls:
Panning Your View

• Windows/Linux: While pressing the Alt key and the left mouse button, drag the mouse pointer
across the Node Graph.
• Mac: While pressing the Option (alt) key and the left mouse button, drag the mouse pointer across
the Node Graph.
As you drag the mouse, you pan your view of the Node Graph.
Zooming or Magnifying Your View

• Windows/Linux: While pressing Alt and the middle mouse button, drag the mouse pointer across
the Node Graph.
• Mac: While pressing Option (alt) and the middle mouse button, drag the mouse pointer across the
Node Graph.
Drag to the right and you’ll zoom-in. Drag to the left and you’ll zoom-out.
• Keyboard zoom-in/out. Tap the plus (+) key to zoom-in. Tap the minus key (-) to zoom-out.
Using the Node Graph Overview

• When the node tree extends beyond the borders of the window, a navigation box appears in the
lower-right corner of the Node Graph. Drag the shaded rectangle inside the box and you’ll quickly
pan to another view of the node tree.
USER GUIDE
1855
Navigating Inside the Windows | Framing the View in the Window
Framing the View in the Window

• Press the letter F on your keyboard to fit the entire contents of the node tree within the borders of
the Node Graph.
The navigation controls for the Node Graph also work inside the next window on our agenda, the
Viewer.
USER GUIDE
1856
Working with Viewers | Framing the View in the Window
Working with Viewers

The postage stamps on the nodes - those little pictures, often called thumbnails - show what each
node passes onto the next node in the tree. Although quite lovely, they won’t do for real compositing
work. You need to open a Viewer window to see the full picture.
You can open several Viewers at once. In addition, you have up to 10 pages, or buffers, for each
Viewer window; these allow you to toggle between different views along the node tree.
When you start Nuke, you see a default Viewer node in the Node Graph. You can easily drag the
connection arrow from a node onto the Viewer to display the node’s output. You can open additional
Viewers by choosing Viewer > Create New Viewer from the menu bar or by pressing Ctrl+I.
USER GUIDE
1857
Displaying the Images in a Viewer Window | Framing the View in the Window
Displaying the Images in a Viewer

Window
To display the images in a Viewer window:
1. Drag the connector from the Viewer node onto the Read node for the engine.v01 clip.
Here’s an alternate method: Select the engine.v01 clip node and then press 1 to connect to the
Viewer node. Nuke displays the node’s output in the Viewer window.
2. Press the Alt key (Mac users press Option) and the left mouse button, and drag the mouse
pointer across the Viewer window to pan.
3. Press Alt (Mac users press Option) and the middle mouse button, and drag to zoom in/out. You
can also use the “zoom” dropdown menu at the top of the Viewer to magnify the view.
4. Press F to fit the current image into the borders of the Viewer window.
USER GUIDE
1858
Displaying the Images in a Viewer Window | Framing the View in the Window
This image has different channels of information you can view. The “RGB” label appears at the top
because the Viewer now shows the result of the red, green, and blue channels.
5. To view individual color channels, press R (red), G (green), B (blue) or A (alpha). As you press each
keyboard shortcut, the label at the top of the Viewer reflects the displayed channel.
6. Press one of the channel keyboard shortcuts again to return to the “RGB” display, or choose RGB
from the Viewer’s channel dropdown menu.
In addition to the standard color channels for red, green, blue, and alpha, this image also includes
channels for specular highlights, reflections, and other masks.
7. To view additional channels, press A to display the alpha channel, and then select the
lightingpasses.reflection channel from the Viewer channel dropdown menu.
You now see the reflection mask from the image file.
8. Select rgba.alpha from the Viewer channel dropdown menu to reset this as the preferred channel
when you press the A key.
9. Press A again to toggle the display and show all color channels.
USER GUIDE
1859
Viewing Multiple Inputs | Framing the View in the Window
Viewing Multiple Inputs

To view multiple inputs:
1. Select the Read node for the smoke_left clip, and press 2 at the top of your keyboard or on the
numeric key pad.
This creates a second connection to the Viewer from the selected node. When the cursor is over
the Viewer, you can press a number on the keyboard to pick the connection you want to view.
2. Move the mouse pointer over the Viewer and press 1 to display the engine.v01 clip. Press 2 to
display the result of the smoke_left node.
In this manner, you can connect multiple images to the same Viewer and then switch between the
images.
3. Select each of the other nodes and press a number to establish a connection to the Viewer.
4. Move the mouse pointer over the Viewer and press the numbers on your keyboard to display
each of the connected nodes.
As you switch between the different views, the images may appear to be the same size. However,
if you look in the lower-right corner of the Viewer, you’ll see the images have different resolutions.
USER GUIDE
1860
Viewing Multiple Inputs | Framing the View in the Window
These images have different resolutions.
Nuke allows multiple resolutions in one composite, but you need to conform these images to match
the project resolution. This allows the elements to be properly aligned in the composite.
USER GUIDE
1861
Reformatting Images | To Conform Images to the Project Format
Reformatting Images
Elements created within the Nuke script, such as Background and Ramp, automatically inherit the
global format and that’s how you want it for this project. The imported images, however, do not
conform to the project settings and must be reformatted.
To Conform Images to the Project Format

1. Click the Read node for the engine.v01 clip to select it.
2. Click the right mouse button and choose Transform > Reformat.
3. Repeat steps 1 and 2 for all the Read nodes in the Node Graph.
4. Move the mouse pointer over the Viewer, and press the keyboard numbers (1, 2, and 3) to switch
between the connected images.
Each image should now conform to the project format.
If you change the delivery format in the project settings, then all elements set to “root.format” also
change to the new project settings. If you neglect to reformat images when you read them into the
project, the images retain their original format, independent of the project settings.
USER GUIDE
1862
Using Proxies and “Down-res” | To Activate Proxy Mode
Using Proxies and “Down-res”

Proxies are low-resolution versions of the final image you intend to create. For many compositing
tasks, the low-res version can help you work faster. Then, when you’re ready to create the final
output, switch proxy mode off and return to the full-res version.
Full resolution. Proxy resolution.
Nuke can generate proxies on-the-fly, according to the scale or format of your images. You select the
method under Edit > Project Settings.
To toggle the proxy resolution defined under Project Settings, you use the “proxy” button on your
Viewer. Alternatively, you use the “down-res” button to lower the display resolution of individual
Viewers. The down-res button works both in the full-res and proxy mode.
To Activate Proxy Mode

1. Click the right mouse button over the Node Graph and choose Edit > Project Settings.
2. Make sure the Viewer window is open.
3. Press the keystroke to toggle Proxy mode, Ctrl+P.
A label inside the Viewer indicates that you are now in proxy mode.
4. Move the mouse pointer over the Viewer, and press the plus (+) key several times to zoom-in.
5. Press Ctrl+P a few times to toggle between hi-res and proxy mode.
6. Before you continue, press Ctrl+P to switch back to full resolution.
To Activate “Down-Res”
1. Choose 1:4 from the “down-res” dropdown menu to change the display resolution to 25% of full
resolution.
With a reduced resolution, Nuke requires less time to calculate and display your images.
2. Change the “down-res” setting back to 1:1, which is 100% of the active resolution.
USER GUIDE
1863
Using Proxies and “Down-res” | To Activate “Down-Res”
If you turned off proxy mode, you should be back to full resolution. If proxy mode is turned on,
the display resolution is 100% of the proxy.
USER GUIDE
1864
Compositing Images | To Composite Two Nodes
Compositing Images
The Merge nodes create a composite with two or more images, using various compositing
algorithms. In this example, we’ll do a very simple “A over B” composite to layer the foreground image
over the background.
You can insert a compositing node from the Toolbar or menus, but we’ll show you a keyboard
shortcut that bypasses both of these. The trick is the select both nodes you want to composite and
then press a keyboard shortcut to assign a compositing node.
To Composite Two Nodes

1. Select the Reformat1 node, attached to engine.v01. This provides the foreground image for the
first compositing operation.
2. Press the Shift key and select the Ramp1 node. Both “engine.v01” and “Ramp1” nodes should be
selected.
3. Press the letter M to insert a Merge node.
The first node you selected is attached to the A input on the Merge node, as the foreground input.
The second node you selected is attached to B, the background input. If necessary, you can swap
the A and B inputs of a merge node by pressing Shift+X.
In the Merge node control panel, the operation parameter determines the compositing algorithm
used to generate the result of the two inputs - the selected operation becomes the name of the
4. Rearrange the nodes, so that the node tree looks similar to this:
USER GUIDE
1865
5. For the next layer, select the Reformat3 node, attached to smoke_right. Then hold down the
Shift key and select Ramp1.
6. Press M to insert a Merge node and composite one image over the other. This composites the
“smoke_right” image over the background.
7. The default compositing algorithm, “Over,” isn’t what we need here. In the Merge2 control panel,
click on the operation dropdown menu and select screen.
8. In the Merge2 properties panel, drag the mix slider and change its value to 0.30 to reduce the
amount of the image supplied by the A input.
9. An additional Merge node is required. Select Reformat2 for smoke_left. Hold down the Shift key
and select the Over node (the first Merge node you inserted).
USER GUIDE
1866
10. Press M to composite the two nodes. In the Merge3 control panel, change the mix slider to 0.75.
The result of your composite should look similar to the example below.
USER GUIDE
1867
Color-Correcting Images | To Composite Two Nodes
Color-Correcting Images
Color-correction and filters can help you integrate the elements for a better composite. In our
example, you want to limit the correction to the foreground element only, so you’ll insert a color
correction node before the Merge nodes.
1. Select the Reformat1 node. Then, right-click over the Node Graph and choose Color > Exposure.
This inserts the Exposure1 node.
2. Suppose you want to adjust the value of the red color channel. Move the mouse pointer over the
Viewer window and press R to display the red channel.
3. In the Exposure1 control panel, uncheck the box for gang sliders. This allows you to adjust
individual color channels.
4. Drag the red slider to adjust the color values. When you are finished, press R over the Viewer to
display all channels.
The Exposure node worked as expected, but the result is less than spectacular. The color change
is too uniform. If only there were a way to limit - or, in fact, mask - the color correction, perhaps
we’d see a better composite. Hmm...
USER GUIDE
1868
Masking Effects | To Create and Apply a Bezier Mask
Masking Effects
You can apply masks to limit how each of these nodes affects the images. The following shows how to
create a Bezier mask to limit color-correction.
To Create and Apply a Bezier Mask

To create and apply a Bezier mask:
1. Click on a blank space in the Node Graph, so that nothing is selected in the node tree.
2. From the Toolbar, choose Draw > Roto to insert a Roto node.
3. Click inside the Viewer window to draw a Bezier shape over the image, like this:
4. To refine the shape, click on a point to select it and then drag to adjust its position.
5. To create sharp corners, select a point, right-click and choose Cusp.
6. To add points to the shape, simply select the Add Points tool and click on the shape’s outline.
7. When you’re satisfied with the shape, drag the mask connector from the Exposure1 node to the
output of the Roto node.
In the Exposure1 control panel, the mask channel option is now set to the rgba.alpha channel of
the node that is connected to the mask input. In this case, this is the alpha channel of the Roto
node.
USER GUIDE
1869
Creating Flipbook Previews | To Generate a Flipbook
Creating Flipbook Previews

On the Viewer window, the timeline buttons let you play the project, but if you pay attention to the
frames-per-second (FPS) field at the top of the Viewer window, you may notice that Nuke doesn’t
provide real-time playback. This is because Nuke renders on-the-fly to display images in the Viewer.
It’s fast, but also limited by the amount of memory and computer-processing power available to you.
The Flipbook feature provides better real-time preview, because it is prerendered for the Flipbook
viewer. Keep in mind that the Flipbook feature renders a preview that matches the active resolution; if
you’re in proxy mode, for example, that’s the resolution you’ll get in the flipbook.
Note: The Flipbook feature renders temporary files in the directory you specified for disk
cache under Nuke > Preferences. You’ll also find an option there that allows you to limit
the amount of disk space the flipbook feature uses.
To Generate a Flipbook
1. Select the Over node at the bottom of your node tree.
2. From the menu bar, choose Render > Flipbook selected.
3. Enter 1-28 as the number of frames to preview and click OK.
4. When the flipbook is ready to view, a Flipbook Viewer is launched. Click the Play button to view
the results.
5. Close the Flipbook Viewer to return to your project.
Rendering Final Output

When you’re ready to render the results of your composite, you insert a Write at the bottom of the
node tree, and specify the path name for the rendered images. Although we’ll use just one here, you
can place several Write nodes in your script, anywhere you like, to render output from different
places in the tree. When the render order is important, use the render order option in the Write
nodes to specify the order in which multiple renders should be executed.
To Render the Result of Your Composite

1. Select the last Over node at the bottom of the node tree.
2. Right-click and choose Image > Write to add a node for output.
USER GUIDE
1870
Creating Flipbook Previews | Rendering Final Output
3. In the control panel for the Write node, click the file folder icon.
4. Browse to the Nuke_Tutorials directory.
5. Click the “new folder” icon, in the upper-left corner of the browser, and type Rendered as the
name for the new folder. Click OK.
6. Select the folder you just created.
You should see the “Nuke_Tutorials/Rendered/” path name displayed at the bottom of the
browser.
7. At the end of the “Nuke_Tutorials/Rendered” path name, type first_comp.####.exr as the name
for the rendered image sequence, and then click Save.
8. Choose Render > Render All to render the images, or simply click the Render button inside the
Write control panel.
9. Nuke prompts you to specify the frames to render. Enter 1-28 as the frame range and click OK.
USER GUIDE
1871
A status window appears that shows the progress of your render. When the render is complete, you’ll
find the sequential images in the “Nuke_Tutorials/Rendered” directory. To check the results, simply
insert a new Read node, point to the new image sequence, and then generate a flipbook with the
Read node selected.
Using the Nuke Frame Number Variable

What’s that ”####” bit in the filename, you say? That’s the variable that tells Nuke where to place the
sequential numbers or frame numbers. You only type one name to represent the image sequence,
but Nuke creates one image file for each frame in your shot.
So, in this case, you entered “first_comp.####.exr” but Nuke renders these files for frames 1 through
5: “first_comp.0001.exr,” “first_comp.0002.exr,” “first_comp.0003.exr,” “first_comp.0004.exr,” and
“first_comp.0005.exr.” You can change the number of hash marks in the variable - ##, ###, ##### - to
change the number of padded digits for the frame numbers.
An alternative way of marking frame numbers is the Printf (%0d) notation. In this case, the same
frame numbers would look like this: “first_comp.%04d.exr”. Instead of the number of hash marks,
with the printf notation you would change the number before d to adjust the number of padded
digits, for example “%03d” or “%05d”. You can choose which style you want to use by setting
sequence display mode option on the Appearance tab of the Preferences.
Image Formats
If you don’t specify a file format inside the Write node control panel, Nuke uses the format specified
by the file name extension you typed. For example, in this tutorial, you used the “.exr” extension to
tell Nuke to save the images as OpenEXR files.
Rendering with the Active Resolution

When you execute a render or a flipbook, Nuke assumes you want to render the active resolution.
When you’re in full-res mode, for example, Nuke renders full-resolution images to disk. When you’re
in proxy mode, Nuke assumes you want to render the proxy resolution - defined in the Project
Settings window - to the path and file name you specified as the proxy file name in the Write node. If
the proxy field is empty or pointing to an invalid path, Nuke returns an error.
It’s easy to toggle to proxy mode and then forget your images are rendered in the lower resolution.
Before you execute a render, it’s always a good idea to check which resolution is active. In the Viewer,
the label at the lower-right corner of your image indicates whether you are in full-res or proxy.
USER GUIDE
1872
Rendering Multiple Channels

When you insert a Write node, Nuke assumes that you need only the RGB channels in the final
render. In many cases, this is acceptable because you won’t need the alpha channel or other channels
from the node tree when you deliver final shots to your clients. However, sometimes you need to
render intermediate files - such as mattes, projection elements, or subcomps - and include all the
channels in your node tree.
For example, rather than manage several elements for an animated character, you could combine the
character animation, the lighting passes, alpha channel, and a depth mask in one image sequence on
disk. This makes it easier to manage elements in the final composite and simplifies the artist’s
workflow.
To output all channels, change the Write node’s channels dropdown menu from rgb to all, select the
OpenEXR file format, and then execute the render. Currently the OpenEXR format (.exr) is the only file
format that supports unlimited channels.
USER GUIDE
1873
Epilogue | Rendering Final Output
Epilogue
In this tutorial, you set up a new project and created a simple composite. You learned how to use (or
at least, locate) practically every Nuke window and tool, and you rendered out the result of your
composite. You’re finished! Go home!
Well... there might be a few more things you want to know. After this tutorial, you should feel
comfortable with the Nuke user interface, so put on your explorer hat and review the other tutorials.
There’s no specific order from here, so look through the following pages until you find what interests
you.
USER GUIDE
1874
Tutorial 2: 2D Point Tracking
This tutorial teaches you how to use Nuke's Tracker node for tracking, stabilizing, and match-moving.
Introduction
Every filmmaker knows the challenges of putting together a vision. You may not have the money to
build post-apocalyptic Montreal, but you might have enough to create it in post. You may have
brilliant performances by your actors - but not together in the same shot. Fortunately, you can
composite the best takes. Your battle sequence with 5 A-list actors, 100,000 extras and 57 elephants,
comes back from the lab with scratches on the negative. You can fix it. You can. A savvy production
team knows how to leverage digital technology to make it possible, and Nuke’s tracking tools are
indispensable for these situations.
As you may know, tracking is the process of recording the location of features as they move through
the scene. The result is stored as 2D coordinates on the image plane. Once you have the tracking
data, you can use the movement to perform a variety of useful tasks, such as stabilizing the footage,
applying the movement to other elements in your composite, and improving the accuracy of roto
mattes.
Tracking image features.
An important aspect of the tracking process involves carefully reviewing your footage before you
attempt to track. Play through your sequences several times and look at the direction of movement
for the features you want to track. Note potential problems with motion blur, obscuring objects, or
frames where the features are hidden or move off screen.
USER GUIDE
1875
Tutorial 2: 2D Point Tracking |
Tip: Nuke can often compensate for problem footage, but tracking works best when you can
identify distinct features throughout the length of the shot.
USER GUIDE
1876
One-Point, Two-Point, Three-Point, Four |
One-Point, Two-Point, Three-

Point, Four
Before we get into the first example, let’s review a few tracking concepts. You can track as many
features or patterns as required with the Tracker node in Nuke. How do you decide whether to track
one, two, or more features? It depends on what you want to do with the data and the level of accuracy
you need in the result. Here are some general guidelines:
One track: Two tracks:

X and Y position only. X, Y, and Z-rotation.
Three tracks:
X, Y, Z-rotation, & scale.
• One-point tracking - Track one feature’s horizontal (x-axis) and vertical (y-axis) position, with little
or no perspective change on the image. You can apply this information to move other elements in
the composite or apply the inverse to stabilize the image.
• Two-point tracking - Track horizontal and vertical position for two features. The feature positions,
relative to each other, indicate whether the image is rotating clockwise or counter-clockwise (z-axis
rotation). In some cases, two tracking points are sufficient to calculate the scaling of the features as
well.
• Three-point tracking - Track horizontal and vertical position for three features. Provides all the
benefits of two-point tracking with an additional set of tracking data for more accuracy on z-rotation
and scaling.
USER GUIDE
1877
One-Point, Two-Point, Three-Point, Four |
• Multi-point tracking - Again, all the benefits of fewer tracks with additional sets of tracking data.
Three-point is usually sufficient for most 2D tracking needs, but multi-point makes it possible to
distort and match-move another element into the points of the features you track, for example,
using four tracks and a CornerPin2D node.
USER GUIDE
1878
Open the Tutorial Project File | To Open the Project File
Open the Tutorial Project File

In this tutorial, you work from a project file that already includes the node trees. Each tree is setup for
the examples that follow.
To Open the Project File

1. Launch the Nuke application and choose File > Open Comp from the menu bar.
2. In the file browser, navigate to your Nuke_Tutorials/Tracking/ folder, select the tracking_
tutor.nk project file and click Open.
3. It should show some nodes in error. Don’t worry! It can’t find the tutorial files. So before doing
anything else you have to tell this script where to find these tutorial images. Double-click on the
NoOp node in the top left corner of your Node Graph. It’s called Tutorial_Path. Double clicking
brings up a Properties panel on the right. Enter the path to the tutorial files in the Tutorial Project
Directory. Use the file browser as that’s often easier than typing it in. You should then see tutorial
images appear.
4. Move the mouse pointer over the Node Graph, and press F to frame the entire contents of the
project file.
The examples in this project file are grouped with colored boxes, called backdrops, and each contains
a node tree for the tutorial examples that follow.
Tip: Backdrops let you organize groups of nodes, like those shown in this project file. Select
Other > Backdrop from the Toolbar and drag the backdrop title bar to move it. Drag the
backdrop corner to resize it. Any nodes surrounded by the borders of the backdrop move
with the backdrop when you drag its title bar.
USER GUIDE
1879
Tracking a Single Feature | Setting a Tracking Anchor
Tracking a Single Feature

In this first example you’ll learn how to set up a tracking anchor and then track a single feature, which
is the most basic 2D tracking operation. After you achieve a solid track for one feature, you can build
on that and track other features as needed.
Setting a Tracking Anchor

1. In the project workspace for the tracking_tutor.nk file, locate the node tree labeled Tracking an
Image.
2. Click on the LondonEye Read node to select it.
3. Play through the sequence several times, using RAM cache or a Flipbook of your choosing, to
review the footage.
4. Look at the features in the image and notice the amount and direction of movement as the clip
plays.
5. Choose Transform > Tracker from the Toolbar to attach a new Tracker node to the LondonEye
Read node.
6. Connect the Viewer node to the Tracker node, as shown.

7. In the Viewer, scrub the time slider to frame 1, to make sure you’re at the beginning of the shot.
8. Double-click the Tracker node to display its Properties panel.
9. Click add track to create a tracking anchor in the Viewer.
10. Drag the tracking anchor over the tower spire, as shown. Use the zoom window in the Viewer to
help you with positioning.
USER GUIDE
1880
Tracking a Single Feature | Auto-Tracking vs. Keyframe Tracking
11. Click on the pattern box (inner box) of the tracking anchor, and adjust its size to contain the
feature.
12. Click the search area (outer box) of the tracking anchor, and adjust its size to enclose the amount
of space you think the feature may move between frames.
Large search areas require more calculation time, so keep it as small as possible. However, when
the search area is too small, the feature may move outside the box and you’ll lose the track. If you
aren’t sure how large to make the search area, go back and review the flipbook of your image.
Auto-Tracking vs. Keyframe Tracking

After placing a tracking anchor, you’re ready to calculate your track. Nuke’s Tracker provides two
calculation methods:
• Automatic tracking - ideal for simple tracks, there are no extra preparation steps once you’ve set
your tracking anchors.
• Keyframe tracking - a more involved method, requiring you to set keyframes on the sequence in
order to calculate tracks. Keyframe tracking may be the better option for more complicated patterns
and movement.
Using Auto-Tracking
1. Enable show error on track paths by clicking the traffic light icon in the Viewer tools.
This color codes tracks showing their pattern matching error values, green for a good match
through to red for a poor match.
2. In the Tracker Properties panel, select the track you wish to calculate in the Tracks list.
3. Select the type of movement the track is intended to output: translation, rotation, or scaling. In
this simple example, you only need to select Translation.
USER GUIDE
1881
4. At the top of the Viewer, click the track forward button to generate the track.
If you only need a certain frame range, use the button and enter the required range.
When Tracker has finished, you’ll see the track curve with color-coded points along the curve.
Note: If tracking fails, try resizing the pattern or search boxes, as described in Setting a
Tracking Anchor and retracking.
5. Use the next frame and previous frame buttons on the timeline to step through the track to
verify its accuracy.
The track is fairly solid in this example. However, some images don’t track as easily as this one.
What are your options? You could retrack Using Keyframe Tracking as described below, or you
could edit your existing track in the Curve Editor, see Editing Track Data on page 1884.
USER GUIDE
1882
Note: Bear in mind that a red keyframe doesn’t necessarily mean that the tracking result is
poor, only that Tracker couldn’t reliably match the pattern from one keyframe to the next.
Using Keyframe Tracking

1. In the Tracker Properties panel, select the track you wish to calculate in the Tracks list.
2. Select the type of movement the track is intended to output: translation, rotation, or scaling. In
this simple example, you only need to select Translation.
3. Scrub through the sequence a few frames and adjust the position of the tracking anchor by
dragging the anchor to the location of the pattern. You can use the zoom window to fine-tune
your positioning. Continue on through the sequence as required.
At each frame, a new keyframe window is added to the right of the zoom window. The keyframe
closest to the current playhead frame is highlighted in orange.
It’s a good idea to place more keyframes around areas of complexity or greater movement and
fewer on straight forward translation. Generally speaking, a greater number of keyframes
produces a better track, but at the expense of processing time.
4. When you’re satisfied with your keyframes, make sure your track is selected in the Tracks list and
then click to track all keyframes.
You can also force the selected tracks to recalculate between the two nearest keyframes by
clicking in the Viewer toolbar.
5. Use the next frame and previous frame buttons on the timeline to step through the track to
verify its accuracy.
Note: Bear in mind that a red keyframe doesn’t necessarily mean that the tracking result is
poor, only that Tracker couldn’t reliably match the pattern from one keyframe to the next.
6. Proceed to Editing Track Data if the track is still not satisfactory.
USER GUIDE
1883
Tracking a Single Feature | Editing Track Data
Editing Track Data

1. In the Tracker Properties panel, select the tracks you want to view.
2. Click the animation button next to the Tracks list, and choose Curve editor.
3. Click the track_x and track_y items in the Curve Editor tree and you’ll see values recorded for
each of the parameters during the tracking process.
4. Select both curves by holding down the Shift key and clicking the track_x and track_y curves,
under tracks.1.
5. Press the F key to “frame” the curves in the Curve Editor.
• To adjust a value, select a point and drag it up or down.
• To change the frame for a particular point, select it, hold down the Ctrl key and drag the point
left or right.
6. Let’s assume you want to smooth a curve by applying a filter to the values. Draw a marquee - drag
while pressing the left mouse button - around a section of the curve to select multiple points.
USER GUIDE
1884
Tracking a Single Feature | Editing Track Data
7. Click the right mouse button to display a menu of Curve Editor options. Choose Edit > Filter and
enter 2 as the number of times to filter the key frames.
Nuke averages the location of each point based on the values of the surrounding points,
smoothing the curve.
8. Close the Curve Editor window and then play the result in the Viewer.
Those are the basics for tracking and editing the results for a single feature. In the next example,
we’ll make it a little harder - tracking a feature that moves out of view.
USER GUIDE
1885
Tracking Obscured Features | To Track a Feature That Moves off Screen
Tracking Obscured Features

At the end of the previous example, you may have noticed the track was dropped at frame 58 when
the feature moved off the screen. When features move out of frame, or become obscured by other
elements in the image, you can use the track offset feature to pass the tracking operation to another
feature in the image. Nuke then attempts to continue the track along its current course.
To Track a Feature That Moves off Screen
Note: This example uses auto-tracking, but the offsetting principle is the same for keyframe
tracking.
1. In the project workspace, locate the node tree Tracking Obscured Features.
2. Double-click the Tracker2 node to open its control panel. This node tracks one of the chimneys in
the clip you used from the previous example.
3. Attach a Viewer to the Tracker2 node and scrub the timeline until you see the tracked feature
move out of frame.
As you can see, track1 accurately tracks its feature through most of the clip - until it moves off the
screen at frame 44. This is where the problem starts.
4. Press the plus key (+) on your keyboard a few times to zoom in on the Viewer.
Examine the sequence to find an alternate feature that stays in view during the length of the clip.
5. At frame 44, press the Ctrl/Cmd key and drag the track1 anchor to the first chimney on the right
of the building.
USER GUIDE
1886
Tracking Obscured Features | To Track a Feature That Moves off Screen
A line connects the new feature to the original feature indicating the offset, and the Tracks list is
updated to show the x and y offset values.
6. In the Tracker control panel, press the button to continue the off screen track using the
offset feature.
How is this useful? Well, now you can use the track data to matchmove an element - a trail of
chimney smoke, for example - that locks to the feature even after it moves off the screen.
7. The track is now complete, so you can clear the offset by clicking in the Viewer tools.
8. Deselect track 1 in the Tracks list to prevent it from being recalculated.
9. Before you continue, close all Tracker control panels that are currently open.
The offset doesn’t change the track location. Instead, it allows Nuke to continue the track with the
assumption that the offset feature remains at the same relative distance to the original feature. Later
in this chapter, you’ll see how to use this tracking data to composite another element to match the
background plate.
USER GUIDE
1887
Stabilizing Elements | To Track and Stabilize
Stabilizing Elements
Stabilization is the process of removing motion - camera-shake, for example - and locking down the
element for your composite. A one-point track provides enough information to stabilize horizontal
and vertical motion along the image plane. A two-point track lets you stabilize horizontal and vertical
motion, and remove rotation in the image, as well.
To Track and Stabilize
Note: This example uses auto-tracking, but the stabilizing principle is the same for
keyframe tracking.
1. Locate the node tree labeled Stabilizing Elements.

2. You’ll see a copy of the same LondonEye Read node that we’ve been using for the other
examples. Click on it to select it.
3. Choose Transform > Tracker and then attach a Viewer to the new Tracker3 node. Double-click
the node to open up the Properties panel.
4. In the Properties panel, click add track twice to create two tracking anchors.
5. For each track in the Tracks list, check the boxes for T (translate), R (rotate) and S (scale).
6. In the Viewer, scrub to the end of the sequence and adjust the size and position of each tracking
marker for the features shown in the image below.
7. Select both tracks in the Tracks list.

8. At the top of the Viewer, click the track backward button to generate the tracks.
USER GUIDE
1888
Now you have positional data for two tracks, and you can use this information to remove the
unwanted movement in the image.
9. In the Tracker3 Properties panel, click the Settings tab and open up the Auto-Tracking sub-menu.
10. Set the warp type to Translate/Rotate/Scale so that Tracker expects all three transforms.
Note: If you’re using keyframe tracks, there’s no need to set the warp type.
11. Click the Transform tab and choose stabilize from the transform list.
Note: When you’re tracking simple translation using a single track, you can use stabilize 1-
pt for faster calculation.
12. In the Viewer, click play to see the results.
USER GUIDE
1889
As the clip plays, you’ll see the features remain locked to the same position within the
compositing frame.
Tip: After you track and stabilize footage, you can add a Transform > Transform node after
the Tracker3 node to adjust the position and the rotation of the stabilized image for a final
composite.
USER GUIDE
1890
Match-Moving Elements | To Match-Move an Element
Match-Moving Elements
Match-moving is the opposite of stabilization. The intent is to record and use the motion in an image
and apply it to another element. In the following example, you’ll use the tracker to match-move and
composite a mask image onto the performer in a background plate.
To Match-Move an Element
Note: This example uses auto-tracking, but the match-move principle is the same for
keyframe tracking.
1. Find the node tree labeled Matchmoving Elements.

2. Drag the time slider to the beginning of the timeline. Select the ColorCorrect1 node and then
choose Transform > Tracker.
3. Attach a Viewer to the new Tracker4 node, create a tracking anchor, and position the track 1
anchor over the performer’s right ear.
4. Adjust the size of the pattern box and the search area as shown.
5. In the Properties panel, check the boxes for T (translate), R (rotate) and S (scale) for track 1.
6. Adjust the size and position of the second tracking anchor, as shown below.
USER GUIDE
1891
Match-Moving Elements | To Match-Move an Element
7. Create another tracking anchor, track 2, and check the boxes for T (translate), R (rotate) and S
(scale) on this track.
8. Select both tracks in the Tracks list and click track forward at the top of the Viewer to generate
the tracks. Edit the tracks as described in Tracking a Single Feature if necessary.
9. Once you have two solid tracks on the performer, make a copy of the Tracker4 node by selecting it
and pressing Ctrl+C to copy it.
10. Select the Transform1 node and press Ctrl+V to paste the Tracker node copy (Tracker5).
11. Connect the Viewer to the Over node. Your node tree should now look similar to this:
12. In the Tracker5 Properties panel, click the Transform tab and choose match-move. Then close
the Tracker5 Properties panel.
13. Click play in the Viewer or render a flipbook and you should see the Mardi Gras mask transform
to match the movement of the performer.
If you see jitter in the movement, you can edit the track data in the Curve Editor to smooth out the
data. You can also add values to the smooth T, R, and S controls on the Transform tab to filter the
tracks.
USER GUIDE
1892
Epilogue | To Match-Move an Element
Epilogue
In this tutorial, you worked with several examples for the Tracker node. You learned how to record
the locations for multiple features and you applied the tracking data for other tasks in the composite,
such as stabilization and match-moving.
USER GUIDE
1893
Tutorial 3: Keying and Mattes
This tutorial introduces you to keying in Nuke. You will learn how to use the Primatte and IBK nodes.
Introduction
Keying is one of those fundamental compositing skills. You can’t composite anything until you have
mattes pulled for the elements you want to layer together. It’s nice to say you could just push a button
to complete this task, but as you probably know, one keying operation seldom produces an
acceptable matte. Image quality, lighting conditions, subject motion, colors - even camera moves -
affect the steps required to get a clean matte for your composite.
Keying Footage in Nuke.
So how do you get a clean matte in Nuke? The best approach is to understand the strengths of each
keying tool and combine them as needed. This tutorial shows how to pull keys in Nuke and how to
layer the results with channel operations, merge nodes, and roto shapes.
USER GUIDE
1894

The project file for this tutorial includes several node trees for the keying operations described in this
chapter.

2. In the file browser, navigate to your Nuke_Tutorials/Keying/ folder, select the keying_tutor.nk
project file and click Open.
3. Double-click on the Tutorial_Path node, located on the left side of the script, to open its control
panel.
4. In the Tutorial_Path control panel, click the “file folder” button. Browse to the location where you
installed the tutorial project files, and then click Open to select the location.
After you select the correct path, the error messages should clear from the Read nodes, and the
thumbnails in the script update with the correct images.
5. Close the Tutorial_Path control panel. Then, choose File > Save Comp As to save a copy of the
project file.
project file.
The green arrows (lines) show the links between the Tutorial_Path node and the Read nodes.
7. If you wish, press Alt+E to hide the expression arrows.
The Tutorial_Path node saves the location of the project files on your computer, so you don’t need
to repeat this for future sessions.
USER GUIDE
1895
Keying with Primatte | To Pull a Key with Primatte (Method 1)
Keying with Primatte

The Primatte keyer includes a quick “Auto-Compute” option that evaluates your image and
determines a good baseline key. From there, you can easily tweak the settings and generate an
acceptable matte.
The two examples in this section show how to pull a key with the Auto-Compute option (method 1),
and also how to manually sample a color from the screen background and build your key from there
(method 2).
To Pull a Key with Primatte (Method 1)

1. In the project file, locate the node tree labeled “Keying with Primatte,” and make sure a Viewer is
attached to the Reformat1 node.
2. Choose Keyer > Primatte to insert the keyer between the foreground image and the Viewer.
3. Drag the bg connector from Primatte1 to the Reformat2 node, which supplies the background
image for this example. The fg connector should be attached to Reformat1.
4. Move the time slider to frame 50, and click the Auto-Compute button inside the Primatte1
control panel.
USER GUIDE
1896
That’s it. You’re done... well, nearly done. We need a “free-floating” goldfish, but the reflections in
the aquarium glass clearly indicate “captivity.”
A garbage matte easily removes the reflections, and you’ll learn how to do that later in the section
on rotoscoping. For now, let’s keep working with Primatte.
As you’ve seen, Primatte’s auto-compute option can quickly pull keys on certain images. However,
you should also know how to pull and tweak keys manually. You might, for example, need more
control over the transparency of the fins on the goldfish.
To Pull a Key with Primatte (Method 2)

1. Continuing from the previous example, open the Primatte1 control panel.
2. Click the “undo” button at the top of the control panel to step back to the previous state of the
Primatte1 node. Or, you can also delete the current Primatte1 node and insert a new one.
3. Scroll down through the Primatte options and set the keying operation to Select BG Color.
4. The current color chip should display the eyedropper icon. If it doesn’t, click on the color chip to
toggle the eyedropper.
5. Hold down the Ctrl+Shift keys (Mac users, hold down Command+Shift) and drag - or scrub - over
a portion of the greenscreen in the image displayed in the Viewer.
USER GUIDE
1897
This returns an average color-pick of the sampled pixels. If you want a color pick from a single
pixel, press Ctrl or Command and click once over the greenscreen. After you pick, you can clear
the red square by Ctrl- or Command-clicking again.
6. Press A over the Viewer to toggle to the alpha channel display. Looks like the aquarium is not as
clean as we thought. Our color pick gave us a fairly noisy key, so let’s clean it up.
Now you’ll sample a few areas of the image to “push” selected pixels to one of three areas: the
transparent matte, the opaque subject, or the semi-transparent part of the matte.
7. In the Primatte1 control panel, change the keying operation to Clean BG Noise.
8. Press Ctrl+Shift or Command+Shift and drag a small square over the dark area in the lower-right
corner of the image.
This second color sample cleans the background by “pushing” the selected pixels into the
transparent area of the matte. You probably need a few more samples to get a better key.
9. Scrub a few small areas in the background, focusing on the gray pixels until the matte improves.
USER GUIDE
1898
The background doesn’t need to be solid black. We’re just trying to get a good separation between
our foreground subject and the greenscreen background.
10. Change the keying operation to Clean FG Noise. This time, sample areas of gray pixels inside the
goldfish.
One or two small samples should be enough. The color pick pushes the selected pixels to the
opaque part of the matte.
You want to keep the gray pixels inside the fins to retain a semi-transparent matte in these areas.
If you go too far, you can always press the undo button in the control panel to step back to the
previous action.
11. Press A again over the Viewer to toggle to all color channels. Your image should look similar to
the example shown below. You may see some detail dropping out from the fins.
12. Change the keying operation to Restore Detail, and scrub over the fins to bring back some of
the edge detail.
USER GUIDE
1899
You may get different results than those shown here, depending on the pixel values you sample
from the image.
Use Restore Detail to push the selected pixels back toward the opaque part of the matte. Use the
Make FG Transparent operation to fine-tune the semi-transparent area.
You could go back and forth, between cleaning the background and foreground, but this usually
produces a matte with “crunchy” edges. The goal is to find the balance between foreground and
background that produces an acceptable matte for your subject.
Later in this chapter, you’ll use the rotoscoping tools to clean-up this matte and combine this with
the image from the next example.
USER GUIDE
1900
Image-Based Keying | To Pull a Key with IBK
Image-Based Keying
Many keying tools, like Primatte, use a color-pick as the baseline for the matte extraction process and
then require the artist to tweak the matte from that baseline. Nuke’s image-based keyer (IBK) uses the
pixel values of the compositing images, instead of a color-pick, to generate the best matte for the
image you want to extract. It works by generating a processed screen image that preserves the color
variations of the blue- or greenscreen and using this - rather than a single color - to pull the key. This
generally gives good results and speeds up the keying process when working with uneven blue- or
greenscreens.
Image-based keying requires two nodes in Nuke. First, you insert an IBKColour node to process the
screen image, which is preset to work with either greenscreen or bluescreen. This node generates the
processed screen image that preserves the color variations in your blue- or greenscreen. Then, you
insert an IBKGizmo node to generate the matte using the processed screen image, the original image,
and also the background image for the composite.
To Pull a Key with IBK

1. In the keying_tutor.nk project file, locate the node tree labeled, “Image-based Keying”.
2. Right-click over the Reformat3 node and choose Keyer > IBKColour. Drag the IBKColourV3_1
node to the right.
3. Click an empty spot in the Node Graph to deselect all nodes. Then, right-click and choose Keyer >
IBKGizmo.
4. From IBKGizmoV3_01 node, connect fg (foreground) to the Reformat3 node. Connect c (color
screen) to the IBKColourV3_1 node.
USER GUIDE
1901
5. Connect bg from IBKGizmoV3_1 to the Reformat4 node, which supplies the background for the
comp.
6. Connect the Viewer to the IBKGizmoV3_1 node, and your node tree should look similar to this:
7. Open the control panel for IBKColourV3_1 and change the screen type to green.
8. Open the control panel for IBKGizmoV3_1, and change its screen type to C-green.
You should see an acceptable matte, shown in the screen capture below, on frame 50.
This is a very good start for this image.

9. Connect the Viewer to the IBKColourV3_1 node. You’ll see the processed screen image, which is
essentially a Gaussian-filtered high-contrast key.
USER GUIDE
1902
10. Choose Merge > Merge (or press M over the Node Graph) to insert a Merge (over) node.
11. Connect IBKGizmoV3_1 to the A input of the Merge (over) node. Then connect the B input to the
Reformat4 node.
The color of this greenscreen is completely out of the region of the acceptable industry standard,
but IBK does a good job anyway, by smoothing the screen and using the result to essentially
create a difference matte with the foreground.
Tip: IBK has presets for green and blue screens, but you can also do a color-pick for any
screen color inside the IBKGizmo node.
If you zoom-in on the image, you’ll see small areas near the subject’s hair, where the matte is
compromised.
USER GUIDE
1903
12. Connect the Viewer to IBKColourV3_1 and you’ll see color artifacts around the edges of the matte.
When you look at the smoothed screen produced by IBKColour, you should see only values of
your screen color and black.
13. In the IBKColourV3_1 control panel, lower the darks, g (green) setting to -0.08. (If you were keying
a bluescreen image, you would lower the “b” value for “darks.”).
This fixes most of the problems at the hairline, but destroys the good key for the lower portion of
the image.
Tip: The artifacts in the IBKColour image appear as specific color shades: light green, dark
green, light red, dark red, light blue, and dark blue. To remove these, simply adjust the
appropriate controls for the artifacts you want to remove: lights/g, darks/g, lights/r, darks/r,
lights/b, and darks/b.
USER GUIDE
1904
14. In IBKColourV3_1, raise the darks, g value to -0.03. Then change the lights, g value to 0.75. This
corrects the artifacts of the screen image.
15. Now, change the patch black setting to 1.5 to restore the edge detail of the hairline.
16. Connect the Viewer to the IBKGizmoV3_1 node. Press A and you’ll see the current alpha channel
produced by the IBK system.
The displayed alpha image shown is correct for the IBK. If the intensity of the noise in your alpha
channel is greater than the example show above, you may need to adjust - in very small
increments - the dark and light values for the color channels in the IBKColour node.
17. Press A again over the Viewer to toggle back to display all color channels, and scrub through the
timeline to check the matte at various points in the clip.
18. If you haven’t already done so, save your project under a new file name to save the changes you’ve
made to project.
USER GUIDE
1905
Rotoscoping | To Draw a Garbage Matte
Rotoscoping
In this example, we’ll return to our first keying example to apply a garbage matte and clean-up the
aquarium image.
To Draw a Garbage Matte

1. Go back to the node tree from the first example, and connect the Viewer to the Primatte1 node.
Drag the time slider to frame 50.
2. Click an empty spot on the Node Graph to deselect all nodes. Then, right-click and choose Draw >
RotoPaint.
3. At this point, you don’t need to connect the RotoPaint1 node to anything, but its control panel
must be open, and the first tab, RotoPaint, should be active.
4. Inside the Viewer, you’ll see the goldfish image. Click the Bezier tool in the RotoPaint toolbar
on the left side of the Viewer. Then in the Viewer, click four points around the goldfish to create a
roto shape. You can drag+click to draw a point and adjust its curve at the same time.
USER GUIDE
1906
Tip: As long as the RotoPaint1 control panel is open, you can view and edit the roto shape.
You can press Q over the Viewer to toggle the display overlay, if necessary. Click the right
mouse button over any point to select options for the roto shape.
Because this is a garbage mask, we want to edit the shape to remove elements from the glass
aquarium.
5. Drag the points and adjust the tangents - the handles on each of the points - to refine the roto
shape.
Now we need to animate the garbage matte to follow the motion of the fish.
6. In the RotoPaint tool settings panel, on top of the Viewer, the autokey option should be active. If
not, click the box for this option.
7. Move the time slider to frame 1 and click the Transform tab in the RotoPaint control panel. Then
select the entire Bezier shape in the stroke/shape list (at the bottom of the control panel) or by
clicking one of the points in the shape using the SelectAll tool. A transform jack appears.
8. Drag the center point of the transform jack, and move it over the current position of the goldfish.
9. Go to end of the timeline, to frame 60. Drag the shape once more to adjust for the movement of
the goldfish.
USER GUIDE
1907
If your Bezier shape is similar to the one shown above, then you probably don’t need more than
the three key frames at frames 1, 50, and 60.
However, you may want to scrub through the timeline and make adjustments.
10. Scrub to frame 60 on the timeline and you’ll see the roto gets a little close to corner-line that we
want to remove from the aquarium glass.
11. Click on an empty spot in the Viewer to deselect all points. Then, press Ctrl/Cmd and click on the
point near the goldfish’s nose, to temporarily break the point’s tangent handle.
12. Adjust the handles to create a peak at the fish’s nose.
Now, for good measure, let’s create a feathered edge for this particular point.
13. With the point selected, drag the feather handle away from the fish to create a feathered edge for
this point, at this frame.
So you’ve drawn and animated the roto shape. Let’s wire it into the node tree to mask out the
“garbage.”
14. Drag the bg connector off the Primatte1 node to disconnect it from the Reformat2 node.
15. Choose Merge > Merge from the right-click menu. Connect Primatte1 to the A input on the over
node. Connect Reformat2 to the B input.
USER GUIDE
1908
16. Connect the RotoPaint1 node to the mask connector on the over node. This effectively removes
the aquarium reflections in the image.
You might want to scrub through the timeline to see if there are places where you need to adjust
the roto shape.
If you want to take this a little further, you can now add the goldfish to the composite from the
second example.
17. Select and drag the Merge (over) node and the RotoPaint1 node below the node tree you used
for the IBK example.
18. Drag the Viewer node over, as well, and keep it connected to the Merge (over) node.
19. Drag the B connector from the Reformat2 node and connect it to the over node in the IBK node
tree.
USER GUIDE
1909
The Viewer shows the result. Of course, you might want to add a Transform node after the first
Merge (over) node, to size and position the goldfish. Otherwise, this project is completed.
USER GUIDE
1910
Epilogue | To Draw a Garbage Matte
Epilogue
Keying is rarely a simple matter of picking the screen color you want to remove. To get the very best
mattes, you often need to combine several techniques and you’ve learned several in this chapter.
You’ve pulled mattes with Primatte and Nuke’s Image-based Keyer, and you’ve used the rotoscoping
tools to cleanup a matte and control the parts of the image you want to use in the composite.
USER GUIDE
1911
Tutorial 4: 3D Integration
This tutorial teaches you the basics of using Nuke's 3D workspace.
Introduction
Nuke’s 3D workspace creates a powerful compositing environment within your project script. This
workspace combines the advantages of cameras, lighting, and a three-axis (x, y, and z) environment,
with the speed of node-based compositing. You pipe 2D images into the 3D space, setup a camera,
animate your scene, and then render the results back to the 2D composite.
Compositing in the 3D workspace.
USER GUIDE
1912
The Basic 3D System | The 3D Viewer
The Basic 3D System

The 3D workspace is defined by a group of nodes in your script. The most basic setup includes a
Camera node, a Render node, a Scene or Geometry node, and nodes that provide the 2D images you
want to pipe into the 3D compositing space.
The basic 3D node tree: 2D image,

geometry, scene, render, and camera.
The 3D Viewer
Once you have the 3D node structure, you can use any Viewer in Nuke as a gateway to the 3D
compositing space. Choose 3D from the view dropdown menu, or press the Tab key over the Viewer
to toggle between the 2D and 3D views.
On the view dropdown menu, you’ll also see orthographic views - rtside, lfside, top, bottom, front,
back - which provide non-perspective views into the scene. In three-quarter perspective, it can be
difficult to accurately place objects on the axes, and the non-perspective views make it easier to line
things up.
USER GUIDE
1913
The Basic 3D System | The Geometry or Scene Node
The Geometry or Scene Node

Every 3D system needs a piece of geometry - a card, a sphere, a cube, something - to receive an
image or clip that the camera can “see.” One is all you need, but you can setup complex systems with
a large amount of 3D data. When you have two or more objects for a 3D system, you need a Scene
node to create a “place” where the camera (and the ScanlineRender node) can see all the objects at
once.
The Camera Node

The Camera node creates your view into a scene. It has several controls to help you match the
properties of a physical camera. You can animate its position or import animation or tracking data to
matchmove your 3D scene with a background plate. A 3D system can have multiple cameras
connected to the Scene node, to create different views on a 3D scene.
Tip: Only one camera can be connected to the ScanlineRender node to generate the output,
but you can insert multiple ScanlineRender/Camera node pairs to generate output from
various perspectives.
The ScanlineRender Node

The last node, ScanlineRender, sends the results of your 3D scene back into your composite as a 2D
image. It’s always 2D in, 3D manipulation, and then 2D back out, which is why this is often called “2-
and-a-half-D.”
USER GUIDE
1914
The Basic 3D System | The ScanlineRender Node
The scanline render node converts the image back to 2D.
The image created by the ScanlineRender node is the same resolution as your project settings. When
you need to render a specific resolution, use the optional bg pipe. Connect a Constant node with the
resolution you want and that defines the output of the ScanlineRender node.
So now you know the basic 3D setup for your compositing script. Let’s take a test drive.
USER GUIDE
1915

The “3Dinteg_tutor.nk” project file includes the node trees for the first part of this chapter.

2. In the file browser, navigate to your Nuke_Tutorials/3DInteg/ folder, select the 3dinteg_tutor.nk
project file and click Open.
3. Locate the Tutorial_Path node, on the left side of the script, and double-click it to open its control
panel.
4. Click the “file folder” button. Browse to the location where you installed the tutorial project files,
and then click Open to select the location.
After you select the correct path, the error messages should clear from the Read nodes, and the
thumbnails in the script update with the correct images.
5. Close the Tutorial_Path control panel. Then, choose File > Save Comp As to save a copy of the
project file.
project file.
The green arrows (lines) show the links between the Tutorial_Path node and the Read nodes.
7. If you wish, press Alt+E to hide the expression arrows.
The Tutorial_Path node saves the location of the project files on your computer, so you don’t need to
repeat this for future sessions with this project file.
USER GUIDE
1916
Setting Up a 3D System | To Set Up a 3D Node Tree
Setting Up a 3D System
Let’s start with the basics. In this first example, you’ll create a basic 3D node tree, map an image to a
3D card, manipulate it, and then render the result back out to the 2D composite.
To Set Up a 3D Node Tree

1. In the “3Dinteg_tutor.nk” project file, locate the backdrop node labeled “Setting Up a 3D System.”
You’ll see a Read node with the image you’ll use for this example.
2. Right-click over the nuke_sign.jpg node, and choose 3D > Geometry > Card.
This attaches a “Card1” node. Let’s see what it looks like in 3D.
3. Attach a Viewer to the Card1 node and Nuke switches the Viewer to 3D.
Wow, that’s amazing. It looks exactly like the 2D Viewer. How can anyone tell the difference?
Check the lower-left corner of the Viewer and you’ll see an orientation marker for the three axes
in 3D. You’ll also see that “3D” is displayed on the view dropdown menu.
USER GUIDE
1917
Setting Up a 3D System | To Set Up a 3D Node Tree
That sign is a little darker than expected, isn’t it? Actually, you can’t see the image yet because the
default view of the 3D workspace is at the origin or center of the space. Perhaps zooming-out may
improve the view.
4. Press the Alt key (Windows /Linux) or the Option key (OS X), and drag with the middle mouse
button to zoom or “dolly.” Drag to the left and you’ll zoom out.
Hey, look. There’s the Nuke emblem. In the 3D Viewer, the “pan” and “zoom” controls are exactly
the same as what you’ve used for the node tree and the 2D Viewer, but let’s try “tumbling” to get a
better view.
5. Alt- or Option-drag with the right mouse button to rotate around the origin point of the 3D
workspace. You now see the 3D grid and the image mapped to the card.
When an image is connected directly to a Card node like this, it is applied as a flat or “planar”
map. The size of the card adjusts to the dimensions of the image.
6. Click on the card and you select the node in the node tree and also the card inside the 3D
workspace.
7. Use the mouse (and the Alt key) to navigate through the workspace. Go ahead, pan, dolly, and
rotate at will. Then, press F over the Viewer to frame the 3D view.
Tip: If you don’t like the standard navigation controls, open the Preferences control panel
(Shift+S), select the Viewers tab and change the 3D control type to Maya, Lightwave, or
Houdini.
USER GUIDE
1918
Setting Up a 3D System | To Position Objects in the Scene
8. Click on an empty spot in the Node Graph to deselect all nodes. Let’s add the other nodes you
need.
9. Right-click on the Node Graph and choose 3D > Camera. Keep its control panel open so you can
manipulate the camera in the Viewer.
10. Right-click and choose 3D > ScanlineRender to insert a render node, and then connect the nodes
as shown below.
11. Connect the Viewer to the ScanlineRender node, and you have the most basic 3D system in
Nuke.
12. Press Tab over the Viewer to change to the 2D view. You won’t see the Nuke emblem - hey, where
did it go? We saw it before.
13. Press Tab again to switch back to 3D. You’ll see the default camera position is too close to view
the card. Let’s move things around to get an image for 2D.
To Position Objects in the Scene

1. Alt- or Option-drag with the middle mouse button to dolly out and show more of the 3D
workspace.
2. Select the camera. You can do this by clicking the camera object in the Viewer or clicking the
Camera1 node in the Node Graph.
3. Drag the transform handles to move the camera away from the card, along the z-axis.
USER GUIDE
1919
As you drag the camera, look at the camera’s control panel. You’ll see the x/y/z transform values
reflect the camera’s current position.
4. Press and hold Ctrl (Mac users press Command) over the Viewer and the transform handles
change to rotation rings.
5. Drag the green ring to rotate the camera around the Y-axis. Notice the x/y/z rotation values in the
control panel reflect the angle of the rotation.
The blue handle “rolls” or rotates on the Z-axis, and the red handle rotates on X.
6. Now, select the card object and move it away from the camera.
Keep the control panel open for the Card node. As with the Camera node, the transform handles
disappear from the Viewer when you close the control panel.
7. Drag the card’s transform handles to position it in the 3D workspace. If you wish, press the Ctrl
key (Mac users press Command) over the Viewer and rotate the card.
USER GUIDE
1920
8. Press Tab over the Viewer to switch between the 2D and 3D views to see the image the
ScanlineRender node produces.
9. Before you continue to the next example, close all the control panels that are currently open.
In this example, it doesn’t matter where you move the camera or card. In reality, however, you
often need to use specific values, which you can enter directly in the control panels.
You can also import camera data or animation curves - did you notice the import chan file
button in the camera’s control panel? - and apply them to the objects in the workspace.
USER GUIDE
1921
Making a Scene | To Set Up a Scene
Making a Scene
We mentioned earlier the Scene node creates a place where multiple objects may be seen by the
camera and the render node. If you only have a single object, you don’t need a Scene node, but
where’s the fun in that? Scene nodes make it possible to really tap into Nuke’s ability to handle huge
amounts of 3D information, and you should know how to use it.
To Set Up a Scene
1. Inside the “Setting Up a 3D System” node tree, drag a selection around the nuke_sign.jpg node
and the Card1 node to select them.
2. Press Ctrl+C (Mac users press Cmd+C) to copy the selected nodes or choose Edit > Copy from the
right-click menu.
3. Press Ctrl+V (Mac users press Cmd+V) or choose Edit > Paste to insert a copy of the nodes. Press
Ctrl+V or Cmd+V again to insert a second copy, and then arrange the nodes as shown below.
4. There are multiple cards now, and you need a Scene node to create a space where the rendering
node can see all cards at once.
5. Click on an empty space in the Node Graph to deselect all nodes. Right-click and choose 3D >
Scene to insert the “Scene1” node.
6. Drag the Scene1 node onto the obj/scn connector to insert the Scene1 node between Card1 and
ScanlineRender1.
USER GUIDE
1922
7. Connect the obj/scn connector from the ScanlineRender1 node to the Scene1 node. Connect
each Card node to the Scene1 node.
8. Double-click on the Card1 node to open its control panel. In the Viewer, you’ll see the transform
handles for the first card.
9. Move and rotate the card to a different position. For this example, it doesn’t matter where you
place it.
10. Open the control panel for Card2 and move its card to a different place in the scene.
11. Open the Card3 control panel and move that card, also.
You could switch to the 2D view to see the result, but why not just look through the camera? Next
to the view dropdown menu, you see the button that locks the 3D view to the camera.
USER GUIDE
1923
12. From the view dropdown menu, choose 3D (V) to switch to a 3D perspective view. Then click the
“lock view to 3D camera” button.
13. Turn off the “lock 3D view to selected camera” button. You won’t need it during the rest of this
tutorial.
USER GUIDE
1924
Merging and Constraining Objects | To Set Up a Scene
Merging and Constraining Objects

You can merge objects and move them together as a group. To do so, you need to insert MergeGeo
and TransformGeo nodes after the objects. The MergeGeo node first merges the objects together,
after which you can use the controls of the TransformGeo node to move the merged objects in the 3D
space. You can also use the TransformGeo node to constrain objects, as you may notice later in this
tutorial.
To merge the three card objects together, right-click on the Card1 node and select 3D > Modify >
MergeGeo. This inserts a MergeGeo node between Card1 and Scene1. Disconnect the Card2 and
Card3 nodes from the Scene node and connect them into the MergeGeo node. Then, right-click on
the MergeGeo node and select 3D > Modify > TransformGeo. Your node tree should now look like
the following:
On the TransformGeo nodes, you see multiple connectors. The connector without a label should be
attached to a geometry object or a MergeGeo node. The other connectors act as constraints on the
connected object’s position.
When a camera or object is connected to the optional look connector, the TransformGeo node
adjusts the rotation so that the object’s z-axis always “points” to the camera or object.
USER GUIDE
1925
Merging and Constraining Objects | To Move the Merged Objects Together
The axis connector can be used to link the current object to the position, rotation, and scale of a
special 3D object called the Axis node. If you’ve worked with other 3D applications, you know the Axis
node as a “null” or “locator” object.
You are still working with the “Setting Up a 3D System” node tree. The following steps show how you
can move the merged nodes, and also how to make objects “look” at the camera and other objects.
To Move the Merged Objects Together

1. Click on the TransformGeo1 node to select it. Its control panel should also be open and you’ll see
its transform handles in the Viewer.
2. Drag the handles to move all the cards merged with the MergeGeo node.
3. Press the Ctrl or Command key and drag the rings to rotate the cards as a group.
4. In the TransformGeo1 control panel, drag the uniform scale slider to increase the size of the
entire group of cards.
To Make Objects ‘Look’ at the Camera

1. Drag the look connector from the TransformGeo1 node onto the Camera1 node.
USER GUIDE
1926
Merging and Constraining Objects | To Make Objects ‘Look’ at the Camera
In the Viewer, you’ll now see the TransformGeo1 node is constrained to the location of the
camera.
2. Select and move Camera1 in the Viewer window. As you do so, the three cards controlled by the
TransformGeo1 node rotate to “look at” the camera location.
Why is this useful? Let’s assume you have a 2D matte painting mapped to a card in your scene.
The “look” option ensures that the plane of the painting always faces the camera, regardless of the
camera position, and maintains the illusion depicted by the painting.
Before you move on, disconnect the TransformGeo node’s look connector from the camera.
USER GUIDE
1927
Animating a Scene | To Animate the Camera
Animating a Scene
The little scene you’ve created would be more interesting with a camera move. You can animate both
cameras and objects; in each control panel, you’ll see an Animation button next to each parameter
you can animate over time.
To Animate the Camera

1. In the Viewer, drag the time slider to frame 1 on the timeline.
2. Let’s switch to an overhead view to move the camera. Choose top from the view dropdown menu.
3. Double-click on the Camera1 object (either inside the Viewer or on the Node Graph) to open its
control panel.
4. Move the camera to the right and rotate it to “look” at the center of the 3D workspace.
5. Click on the animation button next to the camera’s translate parameters, and choose Set key.
The background of the parameter boxes change color to show that a keyframe is now set for
these values at the current frame in the timeline. Now, you need to set a keyframe for the rotation
values.
USER GUIDE
1928
6. Next, click the animation button next to the camera’s rotate parameters, and choose Set key.
7. In the Viewer, scrub to the end of the timeline. Then, move the camera to the left side and rotate
it to face center. This automatically sets keys for the translate and rotate parameters, for the last
frame.
8. Drag through the timeline and the camera moves between the positions recorded by key frames.
Yawn. With only two key frames, the camera moves in a straight line, from start to finish. Let’s edit
the animation curves to make this more interesting.
9. Click the animation button next to the camera’s translate parameters and choose Curve editor.
This opens the Curve Editor window in the same pane as the Node Graph. The outline at the left
side of the Curve Editor displays a list of the parameters you’ve animated, and the curves show the
values plotted over time. Each dot shows the value recorded for a value on a keyframe.
Yes, we know. They don’t look like curves - yet. You only set two key frames, remember? You can
press Ctrl+Alt (Mac users press Cmd+Alt) and click on any line to add more key frames. However,
let’s assume you want to create a smooth arc between the first and last recorded position of the
camera. Rather than set more key frames, let’s just change the shape of the curve.
10. You want to control the shape of the camera path along the z - the distance between the origin
point and the camera, so click the translate / z parameter in the Curve Editor.
Click on the first point of the translate/z curve to select it, and drag the tangent handle upward, as
shown below.
USER GUIDE
1929
From this point forward, the curve increases the distance of the camera on the z-axis, which you’ll
now see in the Viewer.
11. Click on the last point of the translate/z curve to select it, and drag it upward also to finish the
desired shape. This eases the distance back toward the value of the keyframe at the end of the
timeline.
Select the camera in the Viewer and you should see the gradual slopes of the curve create a more
interesting arc for the camera move.
Switch to the 3D (V) perspective view and scrub through the timeline to see the new camera move.
Your version may look a little different than this, depending on the positions and rotations you
defined, but you get the idea.
12. If you wish, you can set additional key frames to refine the camera path. Hold down Ctrl+Alt or
Command+Option and click on the z-axis curve in the Curve Editor, to add new points to the
curve, and then adjust their positions.
13. Before you continue, click the Node Graph tab to hide the Curve Editor and return to the node
trees.
14. Close all the control panels that are currently open.
USER GUIDE
1930
Working with Geometry | To Add Primitive Objects to the Scene
Working with Geometry

In the previous example, you worked with the card object. Nuke also includes primitive geometry,
which can be used as set-extension geometry or placeholders for other elements you plan to add to
scene.
To Add Primitive Objects to the Scene

1. In the “3Dinteg_tutor.nk” project file, locate the node tree labeled “Working with Geometry.”
We’ve already supplied the 3D node tree with a camera for you, so you need to add the geometry
objects, and also create a “scene” where they can co-exist.
2. Right-click over the Node Graph and choose 3D > Scene. Connect Scene2 to ScanlineRenderG.
3. Connect a Viewer to the ScanlineRenderG node and switch to the 3D perspective view.
4. Right-click and choose 3D > Geometry > Cube.
The default cube primitive appears at the center of the 3D workspace. Let’s reduce the number of
subdivisions on the cube.
5. In the Cube1 control panel, change the rows parameter to 4. Change the columns parameter to 4,
also.
USER GUIDE
1931
6. Connect the Cube1 node to the Scene2 node. Now let’s adjust the shape of the cube.
7. Reduce the height of the cube by dragging the top-center point down.
8. From the view dropdown menu, choose the front view to see a non-perspective view of the cube.
These non-perspective views can help you size and position objects with more accuracy than you
might get in a perspective view.
Mm... the cube is actually below the x-axis. Let’s move it up, but check the values in the Cube1
control panel.
9. Drag the top of the cube until the t (top) value in the Cube1 control panel is about 0.3. Drag the
bottom of the cube to align it with the x-axis.
10. It looks like you don’t need 4 divisions on the sides of the cube, so change the number of rows to
2 in the Cube1 control panel.
USER GUIDE
1932
Now let’s add a few more primitives - a cylinder and a sphere.

11. Right-click on the Node Graph and choose 3D > Geometry > Cylinder. Connect the Cylinder1
node to the Scene2 node.
12. Change the view to 3D (V) and zoom out a little to see the whole cylinder.
13. In the Cylinder1 control panel, set the rows to 1, the columns to 20.
14. Set the radius to 0.35 and the height to 1.5. Also check the box for close top.
USER GUIDE
1933
So now you have a cylinder in the scene.

15. Choose front from the view dropdown menu and move the cylinder up to rest on top of the cube.
16. Now add a sphere. Choose 3D > Geometry > Sphere. In the Sphere1 control panel, set both the
rows and columns to 15, and change the radius to 0.35.
17. Make sure the Sphere control panel is open and move the sphere object to cap the top of the
cylinder.
USER GUIDE
1934
18. Select the 3D from the view dropdown menu and rotate the view around the objects in your
scene.
At this point, they have no surface properties, so you’ll need to connect a 2D image from the Node
Graph to each object.
19. In the Node Graph, connect the concrete.jpg to each of the objects.
USER GUIDE
1935
Lighting and Surface Properties | To Define Surface Attributes to Objects
Lighting and Surface Properties

Nuke includes lighting tools to enhance the existing lighting in the plates and images you include in a
3D scene. Also included are fundamental surfacing tools to control the attributes of the objects in the
3D workspace.
These tools are not designed to replace the use of true 3D lighting and surfacing, but they can
definitely help you punch up the scene and quickly tweak the settings without sending elements back
to the 3D application.
Nuke’s lighting objects introduce lighting and surface attributes into your scene. When the scene has
no lighting objects, then all surfaces are illuminated with the same properties and level of
“brightness.”
In the following steps, you’ll first add nodes that define surface properties for the objects, and then
you’ll add the light objects to illuminate them.
To Define Surface Attributes to Objects

1. Click on an empty place in the Node Graph to deselect all nodes. Then, choose 3D > Shader >
BasicMaterial.
2. Drag the BasicMaterial1 node onto the connector between concrete.jpg and Sphere1.
In the Basic Material control panel, you can see parameters to define the amount of light
emission, diffusion, and specular properties of the surface. You can mask these properties by
connecting images to the mapS (specular), mapE (emission), and mapD (diffuse) connectors, but
this is not required for this example.
USER GUIDE
1936
Lighting and Surface Properties | To Add Light Objects to a Scene
3. Set the light emission control to 0.25. Set diffuse to 0.18, and specular to 0.75.
4. Adjust the min shininess and max shininess values to adjust the quality of the specular
highlights. Once again, nothing seems to happen! That’s because you haven’t yet added a light into
the scene. All surfaces are illuminated with the same level of brightness, so there are no specular
highlights or other controllable light properties visible. It’s not very exciting, but don’t worry - you
get to add a light into the scene soon.
5. Make two copies of the BasicMaterial1 node and attach a copy before Cylinder1 and before
Cube1.
To Add Light Objects to a Scene

1. Choose 3D > Lights > Spot and connect the light node to the Scene2 node.
2. In the Spotlight1 control panel, rename the light to Keylight.
3. Switch to the top view and drag x-axis handle (red) to move the light to the left.
4. Drag the z-axis handle (blue) to move the light closer to the bottom edge of the screen. Then,
press the Ctrl or Command key and rotate the light to face the pillar object.
USER GUIDE
1937
Lighting and Surface Properties | To Add Light Objects to a Scene
5. Switch to the 3D (V) view and rotate the view so you can see both the Keylight and the pillar
geometry.
6. Drag the y-axis handle (green) to move the light up above the pillar.
7. Press the Ctrl or Command key and rotate the light down to shine on the pillar.
That’s the basic setup for lighting and surfaces, but there are other tools for more complex setups.
Refer to 3D Compositing for more information on the 3D lighting and surfacing tools.
USER GUIDE
1938
Epilogue | To Add Light Objects to a Scene
Epilogue
In this chapter, you learned how to setup a 3D workspace for your composite, and how to work with
3D geometry and lighting. These are all prerequisite skills for more advanced topics, such as camera
projections, match-moving, and set replacement.
USER GUIDE
1939
Appendices
This section contains supplemental reference information that you may need when using Nuke.
Organization of the Section

The section consists of the following appendices:
• Appendix A: Preferences lists the preferences supported by Nuke.
• Appendix B: Keyboard Shortcuts lists the keyboard shortcuts you can use for quicker and easier
access to Nuke’s features. You can also open a list of keyboard shortcuts from the application by
selecting Help > Timeline or Comp Keyboard Shortcuts.
• Appendix C: Supported File Formats lists the image, video, and audio file formats Nuke supports.
• Appendix D: External Software lists third party libraries used in Nuke.
USER GUIDE
1940
Appendix A: Preferences
The Available Preference Settings

The Preferences dialog is divided into the following sections:
General Settings for auto-saving and path substitutions.
Project Defaults General project settings, and settings for color management.
Performance Settings for caching, hardware, localization, and threads/processes.
Behaviors Settings for start up, file handling, export options, scripting, node
behaviors, and more.
Panels Settings for the interface appearance, file browser, control panels, nodes,
Viewers, script editors, and scopes.
General
General
Autosave
force project autosave Set the number of seconds after which to automatically save your project.
after <300> seconds Disable this by setting it to zero.
idle comp autosave Define how long (in seconds) Nuke waits before performing an automatic
after <5> seconds backup after you have left the system idle (that is, haven’t used the mouse
or the keyboard). If you set the value to 0, automatic backups are disabled.
force comp autosave Define how long (in seconds) Nuke waits before performing an automatic
after <30> seconds backup regardless of whether the system is idle. If you set the value to 0,
forced automatic backups are disabled.
autosave comp Sets the file name of the autosaved project. If this is not set, it defaults to
filename [firstof[value root.name] [getenv NUKE_TEMP_DIR]/].autosave
USER GUIDE
1941
Appendix A: Preferences | Project Defaults
General
Path Substitutions
path substitutions Allows you to remap file paths in order to easily share projects across
different operating systems. When the application encounters a file path,
any text in the OSX/Linux column is replaced with the text in the Windows
column, or vice versa.
For example, if you enter /Volumes/networkmount in the OSX/Linux

column and Z: in the Windows column:
• On Mac and Linux, any file paths that start with Z: are converted to start
with /Volumes/networkmount.
• On Windows, any file paths that start with /Volumes/networkmount are
converted to start with Z:.
To be able to enter text in either column, you need to click on the + button
below to add a row to the table.
+ Adds a row under path substitutions.
- Deletes the selected row(s) under path substitutions.
Project Defaults
Note: You must restart the application for changes to Project Defaults preferences to be
applied.
Channel Management
Channel Management
Channel Warning Sets the total number of channels required in a script to trigger the
Threshold Channel Warning. Nuke only supports 1023 uniquely named channels per
script.
The Channel Count is displayed in the bottom-right of the interface, next

to the Localization Mode indicator.
USER GUIDE
1942
Channel Management
Note: Nuke does not remove unused channels until you close
and reopen a script, so the Channel Count does not decrease
when you remove Read nodes from the Node Graph.
Color Management
OpenColorIO config
OpenColorIO config Sets the OpenColorIO configuration to use, if you don’t intend to use the
file nuke-default settings.
If you select custom from the dropdown, enter the file path of the
configuration file or click Choose to use the browser.
Note: Nuke also includes an environment variable method for

setting a config file. See Environment Variables in the Nuke Online
Help for more information.
Default Color Transform
working space Sets the colorspace files should be converted to, on import, and from,
during render - it's the colorspace used by Nuke under the hood.
viewer Sets the default LUT applied to Viewers.
thumbnails Sets the default LUT applied to thumbnails when ever they are generated.
8 bit files Sets the default LUT applied to the specified ingested file type.
16 bit files
log files
floating point files
Nuke Script Project Settings
USER GUIDE
1943
Color Management
color management Sets whether Nuke uses the LUTs read from the configuration specified or
the Nuke native LUTs during export. Selecting OCIO makes the relevant
OCIO LUTs available to the Read and Write nodes in scripts on a per
project basis.
All configurations except nuke-default automatically switch this control to

OCIO.
General - these preferences only apply to new scripts and projects. To affect the current project,
use the Project Settings.
Project
project directory Sets the project directory used by new projects. You can change the project
directory for the current project in the Project Settings.
Hrox Directory Click to set the project directory to the location of the .hrox file using the
[python {nuke.script_directory()}] expression.
export directory Sets the directory used by timeline exports:

• Use Project Directory - use the directory specified by the project
directory preference.
• Use Custom Directory - use the directory specified by the custom
export directory control.
custom export Sets the export directory used by new projects when the export directory
directory control is set to Use Custom Directory.
Sequence
output resolution Use this to set the output resolution in the Timeline environment for new
projects. By default, clips in the sequence are not reformatted to fit this
format, they retain the source clip resolution. You can adjust the
reformatting for new clips added to a sequence using the Clip settings
below or by selecting clips in the timeline and adjusting the settings in the
Properties tab.
USER GUIDE
1944
frame rate Select the frame rate for new projects in the Timeline environment.
start timecode Use this to define the start timecode for new projects. For shots, this
overrides the timecode defined in the media.
time display You can use this to select the display format for times. You can select
either Timecode or Frames.
drop frame Use this to choose whether timecodes from this sequence are displayed in
drop frame times or not.
Drop Frame is a timecode display option that leaves out two frames from
the 30 fps timecode sequence every minute (except every 10th minute) so
that long running NTSC sequences are accurate to a real-time clock (NTSC
frame rate is 3000/1001, or approximately 0.01% slower than 30fps).
Note: Enabling Drop Frame is a Timecode display feature only -

the source media remains a continuous stream of frames.
Clip
clip reformat Sets how new clips added to a sequence are formatted. The default, None,
displays shots at the source clip resolution, whereas To Sequence
Resolution reformats the shot to the output resolution using the resize
type control.
resize type When clip reformat is To Sequence Resolution, determines how the shot
is resized to fit the output resolution. This control is disabled when clip
reformat is set to None.
center When clip reformat is To Sequence Resolution, disabling center places

the clip at the bottom-left of the output resolution, rather than the center.
Poster Frame
poster frame Sets a preset poster frame for all new source clips or select Custom to
select a relative frame offset as the poster frame:
• First - new source clips display the first frame of the file on disk as the
USER GUIDE
1945
poster frame.
• Middle - new source clips display the middle frame of the file(s) on disk
as the poster frame.
• Last - new source clips display the last frame of the file(s) on disk as the
poster frame.
• Custom - the poster frame number is derived from the number of
frames in the clip starting at 0. Use the frame offset control to set the
relative offset.
For example, a .dpx sequence myClip.####.dpx 1001-1500 can have a

custom poster frame between 0 and 499.
frame offset When poster frame is set to Custom, enter the relative frame offset from
the first available frame in the file(s) on disk.
Views - these preferences only apply to new scripts and projects. To affect the current project, use
the Project Settings.
Views
+/- Click to add and remove views from the views list.
↑↓ Click to move views up and down the views list. Moving a view changes the
position of the corresponding Viewer button.
View Lists the views in the script or project and the color associated with the
view when Use color in UI is enabled.
hero Sets the principal view selected on script or project load.
Set up views for stereo Click to automatically add left and right views to scripts or projects.
Use colors in UI? When enabled, the colors specified in the views matrix are applied to the
interface. For example, if you click Set up views for stereo and enable
this control, any UI items representing the left and right views are colored
red and green.
USER GUIDE
1946
Appendix A: Preferences | Performance
Performance
Caching
Timeline Disk Caching
directory path The directory where all .exr files from timeline disk caching are stored by
default. The caching directory should be on a local disk, preferably with the
fastest access time available. It’s also important to leave enough space for
the maximum disk cache size (defined below).
limit to (GB) This allows you to set the maximum amount of space (in GB) to use for the
timeline disk cache. Set to zero for unlimited size. Values lower than zero,
leave that amount of space free. The currently in use and currently free
fields display how much free cache remains from the total specified.
If this limit is reached during caching, a dialog displays options to free up

disk space.
EXR compression Sets the type of compression applied when writing files to the disk cache:
• DWAB
• Zip (1 scanline)
DWAB compression produces smaller cache files more quickly, but can be
lossy compared to Zip compression.
clear cache Clear All removes all cached files from the root directory specified in the
directory path control. A dialog displays a list of files for confirmation
before the files are deleted.
Comp Disk Caching
temp directory The comp disk cache saves all recent images displayed in the Viewer for
fast playback. Using this control, you can specify where you want Nuke to
save these images. Select a local disk (for example, C:/temp), preferably
with the fastest access time available. It’s also important to leave enough
space for the maximum disk cache size (defined below).
comp disk cache size Specifies the size of Nuke's disk cache, independent of the playback cache.
(GB) It is the maximum amount of disk that Nuke can allocate for caching.
When the limit is reached, Nuke attempts to free disk space before using
USER GUIDE
1947
Caching
any more.
The environment variable NUKE_DISK_CACHE_GB overrides this setting.
Note: Setting this control to 0 GB also disables Nuke's RAM cache

(the orange bar in the timeslider) as both caches rely on files
written to disk.
rotopaint cache size Specifies the size of Nuke's RotoPaint tile cache. It stores tiles for the
(GB) output image of each RotoPaint node, enabling you to paint on top of
existing RotoPaint without having to re-render the strokes underneath.
If you run out of RotoPaint disk cache, response times may suffer when
painting onto tiles containing lots of strokes.
Memory Caching
playback cache size (% Specifies the percentage of system RAM used for the timeline Viewer
of system RAM) playback cache. The entire amount is allocated, even if you've only got a
few frames in the Viewer.
Recently used frames are retained in the memory to avoid relying on the
disk buffering system. The cache is freed when you switch to the
compositing Viewer and reallocated when you switch back to the timeline
Viewer.
Tip: On low-end machines, minimizing this may improve

application responsiveness at the expense of smooth playback.
free timeline playback When enabled, any frames cached to RAM (the white bar in the timeline
RAM cache when Viewer) are discarded when you switch to the Node Graph within Nuke
switching to the node Studio, freeing the RAM for use in Nuke.
graph
Note: When you switch back to the timeline, the cached files are
re-cached, which can be time consuming.
USER GUIDE
1948
Caching
comp cache size (% of Specifies the percentage of system memory available for comp caching. It
system memory) is independent of the playback cache and is the maximum amount of
memory that Nuke can use for caching. When the limit is reached, Nuke
attempts to free memory before using any more.
comp playback cache Specifies the percentage of the comp cache available for comp playback.
size (% of comp cache) This cache holds data displayed in Nuke's compositing Viewer, the result
from the node tree you're viewing. Results from further up the tree are
sometimes cached inside the comp cache as well, such as the output from
a node which is needed by more than one downstream node.
comp paint cache size Specifies the percentage of the comp cache available for comp paint. The
(% of comp cache) comp playback cache and comp paint cache are limits for the total
memory that is used, so it's possible to have a combined size greater than
100%.
How the memory is shared between the two depends on what you're
doing, and you won't necessarily be filling both up at the same rate. For
example, if you're doing a lot of paint work, you might want to allow the
paint cache to fill up more than 50% of the available memory if the
playback cache isn't using its full share.
Audio Waveforms
waveform memory Sets the amount of memory available for storing timeline audio
(MB) waveforms.
Application in Background
pause timeline Viewer When enabled, pause timeline Viewer caching when the application is in
when the application the background.
goes to the
background
clear timeline Viewer When enabled, the timeline Viewer cache is cleared when the application
cache when the goes into the background.
application goes to the
background
Note: This preference is only available when pause timeline
Viewer when the application goes to the background is
USER GUIDE
1949
Caching
enabled.
pause comp Viewer When enabled, pause comp Viewer caching when the application is in the
when the application background.
goes to the
background
clear comp Viewer When enabled, the comp Viewer cache is cleared when the application
cache when the goes into the background.
application goes to the
background
Note: This preference is only available when pause comp
Viewer when the application goes to the background is
enabled.
Undo Caching
undo history size Allows you to set the amount of RAM, in MB, to use for the undo history. If
this limit is exceeded, older items are discarded.
minimum undo events Use this to set the amount of undo events. This setting always applies,
even if it breaches the undo history size limit.
Expression
Expressions Re-Evaluation
Mode Sets when expressions in Nuke scripts are re-evaluated:

• Always - expressions in the script are re-evaluated after every change to
the Node Graph. This can cause performance issues in large, expression-
heavy scripts.
• Lazy - expressions are only re-evaluated when required for GUI or render
updates. This option can speed up interactive performance in large,
expression-heavy scripts.
USER GUIDE
1950
Hardware
Audio
audio device The audio device control allows you to select an existing audio device for
playout from a list of automatically detected devices. You can disable
playout on a device by selecting Disabled.
RED Rocket
use red rocket You can select the use red rocket checkbox to speed up decoding RED
media. If you’re using R3D Rocket graphics card, note that using it is likely
to only be considerably faster when you’re reading in at full resolution. If
you’re reading in at half resolution, for instance, using without the R3D
Rocket card enabled may be faster. This is because the R3D Rocket
graphics card is designed to be fast when reading in multiple frames at the
same time. This is not how it works internally, and therefore reads with the
R3D Rocket card disabled may sometimes be faster when working in lower
resolutions (< 4K widths). Also, note that the R3D Rocket card always
produces better results than when downsampling. Also, the R3D Rocket
card can only be used by one application at a time, so if you are viewing
multiple scripts at once, you may only be able to use the R3D Rocket card
in one.
GPU
expand 3 to 4 You can use this to expand images cached for playback from 3 to 4 color
channels channels per pixel. Some graphics hardware performs better at loading
images to video memory with 4 channels per pixel, than it does with 3.
Enabling this option improves playback performance on such hardware, at
the expense of reducing the number of frames that it's possible to cache.
If you are seeing poor playback performance, enabling this option may
help. However, if you are seeing acceptable playback performance with this
option disabled, then leaving it disabled increases the number of frames
that may be cached for smooth playback.
Note: You must restart Nuke for this option to take effect.
GPU texture cache Use this to set the maximum amount of GPU memory to use for caching
USER GUIDE
1951
Hardware
size (MB) textures.
default blink device You can use this to set the default blink device to use the CPU, or choose
which GPU to use for GPU accelerated plug-ins, such as, ZDefocus,
Convolve, Denoise, BlinkScript, Kronos, MotionBlur, and so on. Any
changes to this setting only take effect after you have restarted the
application.
enable multi_GPU If have multiple GPUs of the same type installed, you can enable this
support preference to share work between the available GPUs for extra processing
speed. This is a global preference and is applied to all GPU enabled nodes.
See Windows, Mac OS X and macOS, or Linux for more information on the
GPUs Nuke supports.
Localization
System
mode Sets the overall localization mode:

• on - checks for updates to source clips and Read nodes with localization
policy set to On or From auto-localize path and localizes those files
automatically.
• manual - checks for updates to source clips and Read nodes with
localization policy set to On Demand and prompts you to update them
manually.
• off - no source clips or Read nodes are localized, regardless of the their
localization policy.
Note: The current localization mode is displayed in the status

bar at the bottom-right of the interface.
read source files when When enabled, source clips and Read nodes referencing out of date
localized files are out localized files automatically revert to reading the entire sequence from the
of date original source files. Source files are read in the following circumstances:
USER GUIDE
1952
Localization
• With Localization mode set to on:

• A localized source clip or Read node with localization policy set to on
demand is detected to be out of date.
• A localized source clip or Read node with localization policy set to on
or from auto-localize path is detected to be out of date and it is
queuing to be automatically localized.
• With Localization mode set to manual:
• A localized source clip or Read node with localization policy set to
on, on demand, or from auto-localize path is detected to be out of
date.
When disabled, the out of date localized files are read until you update
them manually.
hide out of date When read source files when localized files are out of date is enabled,
progress bar you can enable this control to hide the progress/status bar on Read nodes
that are reading from the original source files.
pause localization on When enabled, localization does not start automatically when you open a
script/project open script or project. Enabling this option can help to open scripts and projects
more quickly.
Inputs
localization policy Sets the localization policy for all new source clips and Read nodes:
• on - always localize source clips and Read nodes with this policy
automatically.
• from auto-localize path - localize these source clips and Read nodes
automatically if they reside in the auto-localize from directory.
• on demand - only localize these source clips and Read nodes when you
manually update them.
• off - never localize these source clips or Read nodes.
Paths
auto-localize from Enter the location of the files you need automatically localized, unless
otherwise specified in the Read node’s cache locally control or in the bin
right-click, Localization Policy menu. Commonly this would be your
USER GUIDE
1953
Localization
remote working folder. If you leave this field blank, automatic local file
caching doesn’t take place.
localize to Enter the file path where all the localized files are automatically stored.
Localizing files allows for faster reloading for files that are stored in a
location that is slow to access (such as a network drive).
On Windows, files saved to the localize to directory replace \\ (double back

slashes) and : (colon drive signifiers) with underscores so that the file path
works as expected between operating systems. For example:
Storage
limit to (GB) This allows you to set the maximum amount of space (in GB) to use for
localized files. Set to zero for unlimited size. Values lower than zero, leave
that amount of space free. The currently in use and currently free fields
display how much free storage remains from the total specified.
If this limit is reached during localization, a dialog displays options to free

up storage space.
Network
check for updated files When files are localized, specifies the time interval (in minutes) before
every Nuke checks for updated versions of the files.
Appearance
progress bar Sets the colors used to represent the localization states of source clips and
Read nodes.
Threads/Processes
Playback
USER GUIDE
1954
Threads/Processes
default number of Sets the number of threads to use per reader. If your source files are
threads per reader located on high performance local drives, increasing the number threads
can significantly improve read times.
CPU intensive operations, such as .jpg decoding, can also be improved by

increasing the number of threads per reader.
override number of Allows you to override the default number of decode threads used,
threads per reader dependent on file format.
Use the plus button to add an entry to the table and then select the file
format using the dropdown menu. Double click the Number of threads
column to set the required number of decode threads for that format.
OpenEXR helper Sets the number of helper threads to use for OpenEXR only. The default,
threads to use zero, automatically sets the number of helper threads used.
Arri helper threads to Sets the number of helper threads to use for ARRI only. The default, zero,
use automatically sets the number of helper threads used.
Note: The OpenEXR and ARRI helper thread preferences are independent of the threads
per reader and override table per format settings.
QuickTime decoders Sets the number of background processes available to handle QuickTime
to use file I/O. You must restart the application for this preference change to take
effect.
Note: Using too many decoders can affect performance,

depending on the available hardware.
Note: You must restart Nuke for this setting to take effect.
Rendering
render using frame When enabled, the Frame Server is always used for rendering.
server (Nuke)
USER GUIDE
1955
Threads/Processes
Note: Local Frame Server processes use ports 5558-5662.
frame server render Allows you to increase the number of minutes a render process can stay
timeout unresponsive before ending. If you're experiencing Render application
timed out messages with process-heavy scripts, you can try increasing this
value.
focus background When enabled, rendering using the Frame Server automatically opens the
renders Background Renders panel, or if it is already open, shifts focus to the
panel.
frame server Set the number of slave Nuke processes to run for the frame server.
processes to run
export renders You can select from several render options:

• limit renderer (more responsive ui) – Select this option to make the
user interface more responsive during transcodes. It tells Nuke to use 2
threads to transcode and to use 25% of RAM to cache. Using this option is
likely to result in slower transcodes.
• no renderer limits (fastest transcoding) – Select this option to ensure
that transcodes happen as quickly as possible. This option may result in a
less responsive user interface during transcodes.
• customize render limits – Select this option to manually configure the
number of threads used and cache memory available when
transcoding files.
number of threads Sets the number of threads that Nuke uses when transcoding. Lower
numbers allow the Timeline environment's interface to be more
responsive. Higher numbers allow faster transcodes. This setting is passed
to Nuke using the -m option.
cache memory (GB) Use this to set the number of gigabytes of RAM that Nuke uses for its cache
USER GUIDE
1956
Appendix A: Preferences | Behaviors
Threads/Processes
settings. Lower numbers may improve the Timeline environment's

interface responsiveness, while higher numbers may improve the speed of
the transcodes. This setting is passed to Nuke with the -c option.
background renders Sets when background renders occur:

• don't auto-start background renders – Comps on the timeline are not
rendered in the background automatically.
• start background renders on Comp save – Comps on the timeline are
rendered in the background automatically when they are saved.
• start background renders on Comp create, Comp save and Comp
version change – Comps on the timeline are rendered in the
background automatically when first created, when saved, and when a
new version is selected.
Downsize Filtering
8-bit images Customizes the downsize filtering behavior by bit-depth. The default (1x)
retains the original image size. You can select 2x to halve the image size, or
10-, 12- and 16-bit 4x to quarter the image size.
integer images
The Viewer image quality dropdown affects the decode rate and resolution
16-bit float images of clips displayed in the Viewer. Lower resolutions decode faster and vice
32-bit images versa.
Behaviors
Documentation
documentation source Sets the help source for the Properties ? button:
• local – use Nuke's built-in HTML help server. Nuke's local help server also
searches the NUKE_PATH, .nuke, and .nuke/Documentation directories
for HTML files with the same name as the requested node, such as
Blur.html.
• foundry – This uses Foundry's Online Help, the most up-to-date version
of the documentation.
• custom – Select this to point to your own help server.
USER GUIDE
1957
Documentation
auto port When enabled, assign a free port automatically.
local port Specify a local documentation server port manually. This is usually >= 1024.
You can also set this to 0 to automatically assign the port.
range Specify a range of ports to attempt with the local documentation server.
File Handling
scan for file sequence When enabled, identify and import the file range of media that is dropped
range on drop into Bin into the bin. When disabled, no range is detected and only a single frame
view is ingested. (This does not affect container formats, such .mov and .r3d.)
automatically rescan When enabled, incrementing a source clip or shot’s version past the end of
versions when moving the previously discovered versions list, forces a rescan to update the
off end of the version versions list. See Using Versions for more information.
list
frame number style Sets the sequence display mode to be used in the file browser.
assume fixed width When enabled, assume frames have a fixed width number. With this
frame numbers in file selected, frame numbers need to be padded with zeros to a fixed length,
sequences otherwise frame numbers without preceding zeros are assumed to belong
to sequences with no padding. This is important as the sequence identifier
specifies a unique file name for each and every frame. For example:
Files With fixed frame With fixed frame

width width
assumed NOT assumed
sequence 1.18.exr, sequence 1.##.exr / sequence 1.#.exr /

sequence 1.19.exr, and sequence 1.%02d.exr sequence 1.%d.exr
sequence 1.20.exr
default red clip video Sets the default red clip decode mode for new projects. You can choose
decode mode from FullPremium, HalfPremium, HalfGood, QuarterGood, EighthGood,
or SixteenthGood.
USER GUIDE
1958
File Handling
Note: Changing this preference does not change the default

decode setting for existing projects.
alembic files
always load abc files When enabled, all .abc files are imported as a single node, without
as all-in-one displaying the Alembic import scenegraph.
Nodes
new Merge nodes When enabled, inserting a new Merge node automatically connects the A
connect A input input.
autokey roto shapes When enabled, keyframes are added automatically to Roto shapes when
they are adjusted.
when Viewer is closed When enabled, Viewer nodes are deleted when you close the associated
delete its node Viewer.
Tab Search Menu
Weighting When enabled, nodes in the Tab menu are weighted so that more
commonly used nodes appear at the top of the list.
When disabled, the node list is ordered as it appears in the nodes toolbar
or alphabetically if you start typing.
Favorites When enabled, nodes that you have marked as favorite in the Tab menu
appear at the top of the list.
When disabled, the node list is ordered as it appears in the nodes toolbar
or alphabetically if you start typing.
Clear Weighting Click to reset any weighting information collected by Nuke.
Clear Favorites Click to reset any nodes marked as favorite.
USER GUIDE
1959
OFX Plug-ins
allow trial mode in When enabled, OFX plug-ins that offer a trial mode render in that mode, if
OFX plugins a license cannot be found. When disabled, OFX plug-ins that can't get a
license appear in an error state.
Positions
show menus with When enabled, opening contextual menus positions them with the most
previous item under recently used item under the pointer.
the cursor
Scripting
script command When enabled, the dialog that appears when you press X in the Node
dialog defaults to Tcl Graph defaults to Tcl, rather than Python.
Startup
startup workspace Sets which workspace to display on startup. You can choose from
Compositing, Conforming, Editing, Finishing, Reviewing, and Timeline.
You can also choose to save a customized workspace, which would also be
available from this list.
show splash screen at When enabled, display the splash screen on startup.
startup
show startup dialog When enabled, display the dialog on startup.
restore workspace When enabled, restore the selected saved workspace when opening
when opening projects projects.
Analytics
USER GUIDE
1960
Startup
submit usage statistics When enabled, certain usage statistics are collected from the machine on
which you license Nuke, NukeX, Nuke Studio, Hiero, and HieroPlayer.
When disabled, we don't collect any usage data from your machine.
Note: The port number used to communicate with Foundry is

443, the same one used for uploading crash reports.
Timecode
R3D file timecode Sets the source timecode for RED files. You can choose from Default From
File, Absolute Timecode, or Edge Timecode.
other media timecode Sets the timecode source for file-per-frame media (such as .dpx). You can
choose from File Header or Frame Number. If File Header is selected
and a timecode exists, then the timecode is used. Otherwise it defaults
back to using the frame number from the file name.
max valid timebase Sets the maximum image header timebase above which the value is
(fps) clamped.
Image files are often created with application specific timebase values in
the header description. This can lead to reading in spuriously high frame
rates and the clamp aims to prevent this from happening.
If your clips do have extremely high frame rates, increase this value as
necessary to avoid clamping.
EDL style spreadsheet When disabled, the srcOut and dstOut values in the spreadsheet use the
timecodes film convention, representing the last frame of the cut.
When enabled, the srcOut and dstOut values in the spreadsheet use the
video convention, representing the frame directly after the cut.
USER GUIDE
1961
Appendix A: Preferences | Panels
Panels
Appearance
Font Change the type, weight, angle, and size of the font used on Nuke’s user
interface.
UI Colors - right-click on any color button and select Set color to default to revert changes.
Background Change the background color of most user interface elements (menus,
toolbars, panes, properties panels, Viewers, and pop-up dialogs).
Base Change the color of input fields, the input pane of the Script Editor, and
the left side of the Curve Editor.
Highlight Change the color of the highlighting that appears when you hover the
cursor over a control, select a file or folder in the File Browser, or scrub to
a new frame on the timeline.
Highlighted Text Change the color of any highlighted text (for example, text you select in
node properties).
Label Change the color of labels and text on the application interface. Note that
this does not set the color of the labels on nodes in the Node Graph.
Button Change the color of buttons and dropdown menus.
Animated Change the color that indicates a control has been animated.
Keyframe Change the color that indicates a keyframe has been set.
Disk cached frames Change the color of the disk cached frames on the Viewer timeline.
RAM cached frames Change the color of the RAM cached frames on the Viewer timeline.
Playhead Change the color of the frame marker on the Viewer timeline.
In/Out Markers Change the color of the in and out frame markers on the Viewer timeline.
Curve Editor / Dope Sheet - right-click on any color button and select Set color to default to
revert changes.
no. of curves visible Sets the maximum number of curves visible in the Curve Editor.
USER GUIDE
1962
Appearance
background Change the background color of the Dope Sheet tab.
unselected key Change the color used for an unselected key on the Dope Sheet.
part-selected key Change the color used for a part-selected key on the Dope Sheet.
selected key Change the color used for a selected key on the Dope Sheet.
timeline Change the color used for the timeline on the Dope Sheet.
control text Change the color used for the control text on the Dope Sheet. These
indicate the frame number of a key when you select one.
control text shadow Change the color used for the shadow of the control text on the Dope
Sheet.
time label Change the color used for the time labels on the Dope Sheet. These
indicate frame numbers.
current frame Change the color used for the current frame on the Dope Sheet. This is a
vertical line that indicates the current frame on the timeline.
project frame range Change the color used for the project frame range on the Dope Sheet.
These are two vertical lines indicate your frame range.
Control Panels
max nodes in Use this to set the maximum number of panels that can be open in the
properties bin Properties pane.
reopen acts like new When this is enabled, double-clicking a node that has been open before,
panel places the panel in the same place as a new panel. If this is disabled, the
panel appears in its previous position.
expand / collapse If this is enabled, the node selection in the Node Graph determines which
panels in Properties control panels are expanded (all unselected nodes automatically have their
bin to match selection panels collapsed). This does not apply to floating control panels.
input button action Use this to define node input button action, which is located in the top-left
of the node properties panel. For example, you can set this to center a
USER GUIDE
1963
Control Panels
selected input of a node in the Node Graph.
max items channel Use this to set the maximum number of channels or layers that are
menu displayed in a single sub-menu of the main channel control.
Color Panel
color picker button Sets the type of color picker displayed when you click a color picker button
opens in the properties panel:
• in-panel color picker - opens a color wheel and sliders in the properties
panel.
• floating color picker - opens a color wheel and sliders in a floating
panel.
Tip: Holding Ctrl/Cmd and clicking the color picker button opens
the alternate color picker to the one specified in the Preferences.
File Browser
start file browser from When enabled, new file browsers open at the last location used. When
most recently used disabled, new file browsers open at the current working directory.
directory
Node Colors
autocolor Deselect this checkbox to ignore individual node/soft effect color settings
and, instead always use the All Others color.
Shade Nodes Select this checkbox to apply a slight gradient shading to nodes.
<node name or type> The nodes listed here have been given a default color. You can change this
by clicking the assigned color to open the color menu, and selecting a new
one.
USER GUIDE
1964
Node Colors
All Others Use this to select the color to use as default for all nodes not otherwise
specified above, or all nodes if autocolor is disabled.
Text Use this to select the color used for node label text.
Selected Use this to choose the highlight color applied to any selected nodes.
Selected Input Use this to choose the color used for node inputs' label text when selected.
GL Color Use this to select the color to draw nodes' Viewer controls in. For example,
the Ramp node's gradient control.
Node Graph
autolabel When disabled, nodes only show the filename or node name - most of the
code in autolabel.py is disregarded.
For example, the Blur node does not display the affected channels when
this control is disabled.
highlight running When enabled, highlight any nodes whose output is currently being
operators calculated.
postage stamp mode When displaying a thumbnail render of the node’s output on its surface
(either using the PostageStamp node or the postage stamp control on the
Node tab of each node), you can select one of two modes:
• Current frame - The postage stamp is always updated to match the
current frame.
• Static frame - The postage stamp displays a fixed frame. To specify the
USER GUIDE
1965
Node Graph
frame to use, open the node’s controls, go to the Node tab, and adjust
static frame.
Note: If the frame number you use is outside the frame range for
the node, it is clamped to the first or last frame in the range
accordingly.
node name When a node is selected and the node’s name is too long to fit inside the
background node, a background is drawn behind the name to improve legibility. Use
this control to set the intensity of the background, from 0 (no background)
to 1 (fully opaque background).
label font Sets the font for labels. You can use the B and I, to the right of the font
dropdown, to bold or italisize the selected label font.
tile size (WxH) Sets the size of nodes in the Node Graph using the width and height.
snap to node When enabled, nodes snap into positions (while dragging them) that line
up horizontally and vertically with their inputs and outputs.
grid size (WxH) Sets the grid size using width and height.
snap to grid When enabled, nodes snap into positions (while dragging them) that line
them up with the grid.
show grid When enabled, display the grid using the overlay color.
snap threshold When snap to grid is enabled, use this to set the maximum number of
pixels to jump by, when snapping nodes to the grid or other nodes.
Colors
Node Graph Sets the color of the Node Graph background.
Overlay Sets the color of the selection marquee when you lasso nodes.
Elbow Sets the color of the dots created when you 'elbow' a connection pipe by
USER GUIDE
1966
Node Graph
holding Ctrl/Cmd.
Bounding Box Warning
Highlight When the bounding box warning is enabled, sets the color of the warning
displayed on nodes that exceed the threshold.
Line When the bounding box warning is enabled, sets the color of the dotted
line around the Highlight color on nodes that exceed the threshold.
enable When enabled, nodes that force the bounding box past the format size are
marked in the Node Graph:
• red rectangle with dotted stroke - the indicated node creates a
bounding box greater than the format.
• dotted stroke without the red rectangle - the bounding box size is
greater than the format at the indicated node, but the bounding box size
has been set by an upstream node.
threshold % Sets the threshold past which the bounding box warning is displayed.
Arrow
<directional arrows> You can select one of the directional arrows (up, down, left, or right) to
change the color it is displayed in. Click the arrow to open the color menu
and select a new color.
deep arrows Sets the color of arrows carrying deep data. Click the button to open the
color menu and select a new color.
expression arrows Sets the color of expression arrows, if enabled. Select the enable checkbox
to display expression arrows.
link knob arrows Sets the color of arrows indicating that a node contains linked knobs, that
is, knobs that are in use in another node's Properties panel. Select the
enable checkbox to display link knob arrows.
clone arrows Sets the color of clone arrows, if enabled. Select the enable checkbox to
display clone arrows.
<arrow components> Sets arrow and arrow head, lengths and widths. You can also use the
numeric field next to each component to enter a specific value.
USER GUIDE
1967
Node Graph
allow picking of Select this checkbox to be able to pick up and move connected arrow
connected arrow heads.
heads
allow picking of arrow When enabled, press Ctrl (Cmd on a Mac) on the Node Graph to display
elbows to create Dots yellow “elbows” on the Node Graph arrows and then click on these to
insert Dot nodes.
If you Ctrl/Cmd+Shift+click on an elbow, the new Dot node is branched off

to a new arrow rather than inserted in the existing arrow.
When disabled, adding Dot nodes in this manner is not possible.
drag-to-insert only Select this checkbox to restrict the arrow hotspot, for inserting nodes, to
work near middle of the middle of the arrow.
arrows
size of dots Use this slider to set the size of the Dot nodes. You can also enter a
specific value in the numeric field to the left of the slider.
Project Items
shade project items When enabled, additional shading is applied to source clips and shots in
the Project bin and timeline.
USER GUIDE
1968
Project Items
Item Labels
project bin Click to change the color of labels in the Project and timeline panels. A
color wheel displays allowing you to select the required color.
timeline
auto-adjust contrast When enabled, label colors are automatically adjusted if a potential color-
clash is detected.
Item States
offline Click the buttons to change the color of shots and comps in the timeline
panel. A color wheel displays allowing you to select the required color.
error
freeze
comp not rendered
comp out of date
comp rendered
Item Colors
display in project When enabled, the specified item colors, or defaults, are displayed in the
panel Project panel.
display in sequence When enabled, the specified item colors, or defaults, are displayed in the
panel timeline panel.
spreadsheet color When enabled, the specified item colors applied to rows in the
rows spreadsheet.
USER GUIDE
1969
Project Items
project Click the buttons to change the color of items in the Project and timeline
panels. A color wheel displays allowing you to select the required color.
bin
sequence
source
audio
comp
file types Allows you to add custom color-coding by file extension. Click the
button and then select a file type from the extension dropdown. Double-
click the color swatch to display a color wheel allowing you to select the
required color.
Any source clip or shot with that extension is then colored accordingly in
the interface.
Scopes
black point Use the slider or the entry box to select the black point value.
white point Use the slider or the entry box to select the white point value.
luma/chroma Use this to select the video standard to use when converting from RGB to
encoding luma and chroma for scope display.
Include viewer color Select this checkbox to include applied Viewer color transforms (gain,
USER GUIDE
1970
Scopes
transforms gamma, and LUT) in scope data. If this checkbox is disabled, all Viewer
transforms are ignored.
Note: If disabled, rendering may become slow as image

calculation may be needed.
Force full frame Select this checkbox so that the Viewer always requests full frame data
when a scope is displaying data for that Viewer. If this checkbox is
disabled, the scopes only display data for the current area requested by
the Viewer, rather than the whole image.
Script Editor
font Use this to select the font to use in the Script Editor.
Note: This control also changes the font in the BlinkScript Kernel
Source field.
indent You can use this control to set the indent value to use when scripting.
save and restore script Disable this checkbox if you prefer that the contents of the Script Editor is
editor history not saved and restored between sessions of Nuke.
echo python Select this checkbox to print any Python commands executed by Nuke
comments to output itself to the Script Editor output window.
window
Note: Note that not everything you do results in a command

being echoed, because many of Nuke's internal functions are not
executed using Python commands.
clear input window on Disable this checkbox if you want the most recent script to remain in the
successful script input window after execution.
execution
USER GUIDE
1971
Timeline
show frame end When enabled, an extra line is drawn on the timeline to the right of the
marker playhead, indicating the end of the current frame.
visible range follows When enabled, the timeline scrolls with the playhead, constantly updating
playhead the view. When disabled, the playhead is allowed to move off screen.
Audio Tracks
half waveforms When enabled, audio tracks on the timeline display only the rectified
waveform. When disabled, the full waveform is displayed.
Viewer (Comp)
Defaults
new Viewers go to When enabled, new Viewers are placed in their own window instead of
own window docking in existing windows.
prevent auto zoom for When enabled, new Viewers are not auto-zoomed to the current Viewer's
new Viewers zoom level.
apply LUT to color When this is checked, look-up tables (LUTs) are only applied to the red,
channels only green, and blue channels.
When this is NOT checked, LUTs are applied to all channels.
viewer buffer bit depth Use this to select the OpenGL buffer depth, and enable use of the GPU for
(byte) the Viewer process and input process.
• byte – Converts to 8-bit with error diffusion.
• half-float – Converts to 16-bit (half) float. In this mode, the GPU can be
used to apply Viewer effects like gamma and the LUT in a Viewer process.
• float – Uses a full 32-bit floating point texture. This may be slow on
selected cards. In this mode, the GPU can be used to apply Viewer effects
like gamma and the LUT in a Viewer process.
You can choose a default value for this setting in the Preferences, or by
using knobDefault() in a startup script.
USER GUIDE
1972
Viewer (Comp)
use GPU for Viewer When this is checked, the Viewer applies its effects (such as the Viewer
when possible Process node) in the GPU when possible. However, in some cases, like
when monitor output is enabled or gl buffer depth is set to byte in the
Viewer settings, effects (such as gain and gamma) must still be computed
on the CPU.
use GPU for inputs Normally, the Viewer only attempts to run its own effects (such as the
when possible Viewer Process node) on the GPU. However, when this is checked, any
nodes connected to the Viewer are also computed on the GPU when
possible. Note that this cannot be done for all nodes because not all nodes
have a GPU implementation.
If nodes are computed on the GPU, the color values displayed in the
Viewer are inaccurate. This is because they show the color from the last
node computed in the CPU prior to transferring the image into the
graphics card.
disable GPU Viewer Check this to disable dithering in the Viewer when you’re using half-float
dithering depth. Uncheck to allow dithering at all times.
Note: When the Viewer AB mode is changed to wipe or stack, the state of the GPU
acceleration controls is stored, GPU acceleration is turned off, and GPU acceleration is
disabled.
When the Viewer AB mode is changed back to default, GPU acceleration is re-enabled and
the state of the GPU acceleration controls is restored.
no incomplete stereo When this is checked, the Viewer only displays one view of a stereo project
for new viewers until both views have been rendered. This is to prevent disorienting effects
when watching the results.
When this is not checked, the Viewer displays both stereo views, even if the
render of either is incomplete.
Settings
flip stereo interlaced Select this checkbox to flip the left and right stereo views when the Viewer
views is set to Stereo Mode > Interlaced in the right-click/context menu.
USER GUIDE
1973
Viewer (Comp)
texture size Use this to define the OpenGL texture size in Nuke's Viewers. It affects the
2D Viewer preview texture size, such as the preview displayed when
dragging transform or SplineWarp handles, and all textures in the 3D
Viewer. You can choose from 256x256, 512x512, 1024x1024, and
2048x2048.
Note: The larger the texture size, the sharper it is, but larger
texture sizes need more time and memory. Setting this
preference to a high value can impact 3D Viewer performance.
texture mode Use this to choose how textures are handled in the Viewer. You can choose
from the following options:
• Multiframe – Cache each frame of a texture; this gives animated textures
and fast playback after each frame is cached, but uses a lot more
memory. It also enables you to have multiple frames of a texture visible
at one time (for example, on particles).
• Classic – Textures are not updated in the Viewer during playback; this
gives the fastest playback speed.
Viewer (Monitor Out)
use video legal range When enabled, automatically limit monitor output to the legal range when
for monitor out swapping between the Timeline environment and Compositing
environment.
Viewer (Sequence) or (Flipbook)
playback mode Use this to set the Viewer playback mode:

• Play All Frames - the default setting, plays all frames in real-time
(dependent on hardware).
• Skip Frames - plays frames in real-time skipping where necessary to
USER GUIDE
1974
Viewer (Sequence) or (Flipbook)
maintain the frame rate.

• Play All Frames, Buffering - plays all frames by buffering and playing
frames back as they become available.
guides You can use this to choose to show overlays in the image area. You can
choose from:
• Title Safe – Indicates where text should be entered to be visible.
• Action Safe – Indicates the area in which to place actions so that they are
visible.
• Format – Displays the size of the format over the Viewer.
fullscreen display Use this to select which display to use for fullscreen mode. This setting
takes effect the next time fullscreen mode is entered.
see through missing Select this checkbox to see through missing media in the timeline,
media displaying the first displayable media in the underlying tracks.
background Use this to select the Viewer background. You can select black, or gray
(using the slider to determine the grayscale), or checkerboard (using the
slider to determine the size of the squares).
frame increment Use this to set the default number of frames skipped by the Viewer skip
controls, and the timeline Nudge More commands.
filtering mode Use this to determine the filtering used during rendering in the Timeline
environment. You can select Auto, Nearest neighbour, or Linear.
Auto uses ths same automatic selection as in the Compositing

environment. This does not affect exports or rendering in the Compositing
environment.
Audio
default latency Use this to adjust the default timing offset (in milliseconds) between audio
adjustment (ms) and video to apply to new Viewers. Positive values make audio play earlier
relative to video; negative values make audio play later. To convert from
video frames to ms, divide 1000 ms by the video frame rate. For example:
• at 25fps, a video frame is 1000/25 = 40ms, or
• a 1.5 video frame delay = 1.5 * 40ms = 60ms.
default volume Use the slider or numeric field to set the default volume.
USER GUIDE
1975
Viewer Handles
Colors
2D • bg – Change the background color of the 2D Viewer.

• fg – Change the color of borders and text in the 2D Viewer.
3D • bg – Change the background color of the 3D Viewer.

• fg – Change the color of borders and text in the 3D Viewer.
• sel –Change the color of the selected vertices or faces of an object in the
3D viewer.
Splines
line width Sets the width of splines drawn in RotoPaint and SplineWarp.
draw shadows When enabled, shadows are added to splines drawn in RotoPaint and
SplineWarp to make the overlay easier to see.
general • Expression color – Change the default color of the control points when
an expression is set.
• Focus color – Change the default color of the control points when
focused.
roto • Points - Change the default color of the points on RotoPaint shapes and
strokes.
• Curves - Change the default color of the rotoshape and stroke curves in
RotoPaint and Roto.
• Transform - Change the default color of the RotoPaint transform jack.
• Locked - Change the default color of RotoPaint points and curves when
locked or otherwise unmodifiable.
splinewarp • A Sourcecolor - Change the default color of SplineWarp’s A source

curves.
• B Sourcecolor - Change the default color of SplineWarp’s B source
curves.
• draw source stippled - Check to change source curves from solid to
stippled.
• A Destinationcolor - Change the default color of SplineWarp’s A
USER GUIDE
1976
Viewer Handles
destination curves.
• B Destinationcolor - Change the default color of SplineWarp’s B
destination curves.
• draw destination stippled - Check to change destination curves from
solid to stippled.
• Correspondencecolor - Change the default color of the SplineWarp
correspondence lines.
• Boundarycolor - Change the default color of the SplineWarp boundary
curves.
• Hardboundarycolor - Change the default color of the SplineWarp hard
boundary curves.
Note: To set a color back to its default, right-click on the button and select Set Color to
Default.
Controls
middle button pans Check this to use the middle-mouse button to pan in the Viewer, Node
Graph, the Curve Editor, and the Dope Sheet.
left-middle to zoom Check this to use the left and the middle-mouse button together to zoom
in the Viewer, Node Graph, the Curve Editor, and the Dope Sheet.
show transform Check this to disable the OpenGL preview when manipulating the handles
preview of 2D nodes, such as Transform and CornerPin.
3D control type Select the navigation control scheme you want to use in the 3D Viewer:
Nuke, Maya, Houdini, Lightwave, or Modo.
2D handle size Adjust the size of the square control handles that appear on the Viewer for
some operations, such as transformations, warps, and Bezier and B-spline
shapes.
USER GUIDE
1977
Viewer Handles
By default, this value is set to 5. You can also set the pickable area size of
the square control handles in the numeric field or slider to the right of the
2D handle size control.
3D handle size Adjust the size of the square control handles that appear when you’re, for
instance, selecting vertices on a 3D object in the 3D view. By default, this
value is set to 5. You can also set the pickable area size of the square
control handles in the numeric field or slider to the right of the 3D handle
size control.
icon size Adjust the size of the 2D transformation overlay, 3D camera, 3D object
normals, and 3D axis on the Viewer. By default, this value is set to 50.
icon scaling Adjust how much the scale of display affects the size of the 2D
transformation overlay, 3D camera, and 3D axis. When this is set to 0,
these controls are always drawn the same size, regardless of the zoom
level. When the value is set to 1, the controls scale with the displayed
image or 3D scene when you zoom in or out.
Intermediate values mix this so that the controls do scale, but not as much
as the image does. This gives an optical illusion that you are zooming in or
out without making the controls unusably small or large.
object interaction Set how fast mouse movements rotate and translate 3D axis and cameras.
speed The lower the value, the finer the movements. The default value is 1.
camera interaction Set how fast mouse movements tumble and roll the 3D view in the Viewer.
speed The lower the value, the finer the movements. The default value is 1.
USER GUIDE
1978
Appendix B: Keyboard Shortcuts
Keyboard shortcuts, or hotkeys, provide quick access to the features of Nuke. The following tables
show these keystrokes.
Conventions
The following conventions apply to instructions for mouse-clicks and key presses.
• LMB means click or press the left mouse button.
• MMB means click or press the middle mouse button
• RMB means click or press the right mouse button.
• When you see the word “drag” after a mouse button abbreviation (i.e., “MMB drag”), this tells you to
press and hold the mouse button while dragging the mouse pointer.
• Keystroke combinations with the Ctrl, Alt, and Shift keys tell you to press and hold the key and then
type the specified letter.
For example, “Press Ctrl+S” means hold down the Ctrl key, press S, and then release both keys.
Note: On Mac, replace the Ctrl key with the Cmd key.
Note: Keystrokes in the tables appear in upper case, but you do not type them as upper
case. If the Shift+modifier does not appear before the letter, just press the letter key alone.
Note: This section assumes you are using the default keyboard and mouse-button
assignments. If the mouse buttons do not work for you as described here, try resetting the
mouse control type back to the standard Nuke setting (Preferences > Panels > Viewer
Handles > 3D control type > Nuke).
USER GUIDE
1979
Appendix B: Keyboard Shortcuts | Global
Global
Keystroke(s) Action
Backspace/Delete Delete selected clips or folders
F12 Clear buffers and playback cache
MMB drag Virtual slider (number fields)
Space bar (short Expand the focused panel to the full window
press)
Space bar (long Raise the right-click menu

press)
Alt+S Make the application or floating window fullscreen.
Alt+` Show Curve Editor.
Ctrl+A Select all
Ctrl+C Copy selected item(s)
Ctrl+D Duplicate selected item(s)
Ctrl+F# Save current window layout. The # represents a function key number, F1
through F6
Ctrl+LMB on panel Float panel

name
Ctrl+N Create a new project or script, depending on environment
Ctrl+O Open a project or script, depending on environment
Ctrl+Q Exit the application
Ctrl+S Save current project or script, depending on environment
Ctrl+T Cycle through tabs in the current pane. Note that this does not work if the
focus is on the input pane of the Script Editor
Ctrl+V Paste the contents of the clipboard
USER GUIDE
1980
Appendix B: Keyboard Shortcuts | Nuke Studio's Timeline Viewer
Keystroke(s) Action
Ctrl+W Close the current project or script, dependent on environment
Ctrl+X Cut selected item(s)
Ctrl+Z Undo last action
Shift+Esc Close the current tab
Shift+F1..F6 Change workspace
Shift+S Open the Preferences dialog
Alt+Shift+1..6 Open a recent project or script, depending on environment
Ctrl+Shift+A Select none
Ctrl+Alt+` Goto next pane
Ctrl+Shift+[ Goto next tab
Ctrl+Shift+] Goto previous tab
Ctrl+Shift+S Save current project or script, depending on environment, and specify name
(Save As)
Ctrl+Shift+Z Redo last action
Ctrl+Alt+Shift+` Goto previous pane
Nuke Studio's Timeline Viewer

Keystroke(s) Action
- Zoom Out
+ Zoom In
A Alpha
B Blue
USER GUIDE
1981
Keystroke(s) Action
C Razor selected
E Clipping Warning
End Go to End
F Zoom to Fit
F12 Clear playback cache
G Green
H Zoom to Fill
Home Go to Start
I Mark In point
J Play Backward
K Pause
L Play Forward
Left Arrow Frame Backwards
O Mark Out point
PgDown Next Layer
PgUp Previous Layer
Q Show Overlays
R Red
Return Swap A/B Inputs
Right Arrow Frame Forwards
V Display the version selector
W Wipe
Y Luma
USER GUIDE
1982
Keystroke(s) Action
Alt+Down Arrow Version down the selected clip
Alt+I Clear In Point
Alt+O Clear Out Point
Alt+Shift+Left Arrow Previous Tag
Alt+Shift+Right Arrow Next Tag
Alt+U Clear In/Out Points
Alt+Up Arrow Version up the selected clip
Ctrl+/ Show Timeline Editor
Ctrl+F Full Screen
Ctrl+LMB Color picker (single pixel)
Ctrl+RMB Deselect sampled pixels
Ctrl+S Save current project
Shift+C Razor all under playhead
Shift+I Go to In Point
Shift+Left Arrow Skip Backwards
Shift+O Go to Out Point
Shift+Right Arrow Skip Forwards
Ctrl+Alt+Down Arrow Go to min version
Ctrl+Alt+Up Arrow Go to max version
Ctrl+Shift+1 Zoom to Actual Size
Ctrl+Shift+2 Zoom to Half Size
Ctrl+Shift+F Full Quality 1:1
Ctrl+Shift+LMB Color picker (region of pixels)
USER GUIDE
1983
Appendix B: Keyboard Shortcuts | Nuke Studio's Timeline
Keystroke(s) Action
Ctrl+Shift+P Ignore Pixel Aspect
Ctrl+Shift+S Save current project and specify name (Save As)
Nuke Studio's Timeline

Keystroke(s) Action
, (comma) Nudge selected shot(s) left, where space is available
. (period) Nudge selected shot(s) right, where space is available
1 Display in the A input buffer
2 Display in the B input buffer
Alt and drag Ripple and duplicate the dragged shot
Alt then drag Duplicate the dragged shot
D Enable or disable the selected shot(s)
Down Arrow Next Edit
drag then Alt Activate Ripple mode while dragging shot
E Cycles between the Slip Clip and Slide Clip tools
Enter (numeric Edit playhead time

keypad)
F12 Clear playback cache
F5 Render all Comp containers
F7 Render selected Comp containers
LMB Select a clip including any linked tracks
M Insert the contents of a source Viewer into the timeline at the current playhead
position overwriting existing shots
USER GUIDE
1984
Appendix B: Keyboard Shortcuts | Nuke Studio's Timeline
Keystroke(s) Action
N Insert the contents of a source Viewer into the timeline at the current playhead
position and ripple existing shots downstream to accommodate the change
Numeric keypad Change selection between shots

arrows
Q Cycles between the available move tools: Multi, Move/Trim, and Select
R Cycles between the available edit tools: Ripple, Roll, and Retime
T Cycles between the available razor tools: Razor, Razor All, and Join
U Mark clip
Up Arrow Previous Edit
W Cycles between the available selection tools
Shift+drag clip Disable snap to transition when dragging clips
Alt+, (comma) Nudge selected shot(s) up, overwriting any clips on the tracks above
Alt+. (period) Nudge selected shot(s) down, overwriting any clips on the tracks below
Alt+D Display metadata for the selected track item(s)
Alt+Down Arrow Version down the selected shot
Alt+LMB Select a clip, ignoring linked tracks (for example, audio only)
Alt+Up Arrow Version up the selected shot
Ctrl+numeric Nudge selected shot(s), where space is available

keypad arrows
Ctrl+A Select all shots
Ctrl+I Import files
Ctrl+Return Open selected shot in the Viewer
Ctrl+T Add a dissolve between two selected shots
USER GUIDE
1985
Appendix B: Keyboard Shortcuts | 2D Compositing Viewer
Keystroke(s) Action
Shift+, (comma) Nudge selected shot(s) left by the increment amount set under the Viewer,
where space is available
Shift+. (period) Nudge selected shot(s) right by the increment amount set under the Viewer,
where space is available
Shift+Backspace Ripple delete
Shift+U Mark selection
Alt+Shift+/ Rename shots
Ctrl+Alt+A Select all in track
Alt+Shift+Down Go to min version

Arrow
Alt+Shift+Up Go to max version

Arrow
Ctrl+Alt+Return Open selected shot in a new Viewer
Ctrl+Shift+A Deselect all shots
Ctrl+Shift+E Open the Export dialog
Ctrl+Shift+I Import folders
LMB then Select all clips between the left-clicks

Shift+LMB
Shift+Alt+LMB Ignore linked tracks during selection
Ctrl+Alt+Shift+I Import EDL, XML, or AAF
2D Compositing Viewer
Keystroke(s) Action
- Zoom Out
USER GUIDE
1986
Keystroke(s) Action
, (comma) Decrease Gain
; (semi-colon) Previous view (multi-view)
. (period) Increase Gain
' (apostrophe) Next view (multi-view)
[ Toggle left toolbar
] Toggle right toolbar
` (backtick) Toggle floating viewers
+ Zoom In
1, 2, etc. Set Viewer inputs for A buffer
A Toggles between the Alpha channel and RBG
B Toggles between the Blue channel and RBG
Down Arrow Switch to the previous input for the A buffer
End Go to the out point in the time slider
F or MMB Zoom to Fit
G Toggles between the Green channel and RBG
H Zoom to Fill
Home Go to the in point in the time slider
I Mark In point
J Play Backward
K Pause
L Play Forward
Left Arrow Step backward one frame
M Toggles between the Matte channel and RBG
USER GUIDE
1987
Keystroke(s) Action
Numeric pad Nudge one frame backward or forward

left/right arrow
keys
O Mark Out point
P Pause Viewer refreshing
PgDown Next layer
PgUp Previous layer
Q Toggle overlays
R Toggles between the Red channel and RBG
Return Swap A/B input buffers
Right Arrow Step forward one frame
S Open Viewer settings
Note: You can't use the S keyboard shortcut to open Viewer properties
when Roto or RotoPaint properties are open.
Tab Toggle 2D/3D
U Update the Viewer
Up Arrow Switch to the next input for the A buffer
W Toggle wipe tool
Y Toggles between the Luminance channel and RBG
Alt+# (1, 2, etc.) Zoom out by 100%, 200% and so on
Alt+G Go to a specific frame
Alt+I Clear In point
Alt+LMB Pan
USER GUIDE
1988
Keystroke(s) Action
Alt+MMB and Zoom in/Out at the pointer location

drag
Alt+O Clear Out point
Alt+P Toggle Input Process
Alt+U Clear In and Out points
Alt+W ROI - Activate new region
Alt+Z Toggle lock/unlock zoom level
Ctrl+# (1, 2, etc.) Zoom-in by 100%, 200% and so on
Ctrl+Left Arrow Move the playhead backwards by 1/2 the distance from the playhead to the in
point (subsequent key strokes recalculate the distance before applying the
move)
Ctrl+LMB Color picker (single pixel)
Ctrl+P Toggle Proxy
Ctrl+Right Arrow Move the playhead forwards by 1/2 the distance from the playhead to the out
point (subsequent key strokes recalculate the distance before applying the
move)
Ctrl+RMB Deselect sampled pixels
Ctrl+S Save current script
Ctrl+U Toggle previewing output on an external broadcast video monitor
LMB + MMB and Zoom in and out with the click point set as the center of the Viewer
drag
MMB and RMB Toggle between the last specified range (MMB and drag) and Visible mode
MMB and drag Zooms the time slider to the range specified by the drag operation
(time slider)
MMB and drag Pans in the Viewer

(Viewer)
USER GUIDE
1989
Keystroke(s) Action
MMB scroll Zoom in and out of the time slider
Shift+[ Toggle top toolbar
Shift+] Toggle bottom toolbar
Shift+1, 2, etc. Set Viewer inputs for B buffer
Shift+Down Switch to the previous input for the B buffer

Arrow
Shift+Left Arrow Move the playhead backward by the amount of frames specified in the skip
frames control
Shift+numeric Nudge the playhead backward by one frame

pad left arrow
Shift+numeric Nudge the playhead forward by one frame

pad right arrow
Shift+Right Move the playhead forward by the amount of frames specified in the skip
Arrow frames control
Shift+Up Arrow Switch to the next input for the B buffer
Shift+W Enable or disable ROI
Alt+Shift+S Save script and increment version number
The script name must include _v# for the increment to work as expected. For
example, myScript_v01.nk
Ctrl+Shift+LMB Color picker (region)
Ctrl+Shift+P Toggle pixel aspect ratio
Ctrl+Shift+S Save current script and specify name (Save As)
USER GUIDE
1990
3D Compositing Viewer
Keystroke(s) Action
C 3D Top view
PgDown Next layer (color channel display)
PgUp Previous layer (color channel display)
S Display Viewer Settings
Tab Toggle 2D/3D viewer
V 3D Perspective view
W Toggle the Wipe tool
X 3D right-side view
Z 3D front view
Alt+LMB Translate viewer on y,z axis
Alt+MMB Zoom in/out (drag left/right)
Alt+RMB or Ctrl+LMB Rotate viewer on x,y axis
Ctrl+L Toggle Unlocked/Locked/Interactive Camera or Light
Shift+C 3D Bottom view
Shift+X 3D Left-side view
Shift+Z 3D Back view
Ctrl+Shift+LMB Rotate viewer on z axis
USER GUIDE
1991
Appendix B: Keyboard Shortcuts | Node Graph
Node Graph
Keystroke(s) Action
– Zoom-out
. (period) Insert Dot node
/ Search by node name or class
\ Snaps all nodes to grid
1, 2, 3, etc. With a node selected in the Node Graph, create/connect Viewer inputs
With no nodes selected in the Node Graph, cycle through connected views
+ Zoom-in
B Insert Blur node
Backspace/Delete Delete selected nodes
C Insert ColorCorrect node
D Disable/Enable node
Down Arrow Next node in tree
F Fit selected nodes to Node Graph panel or, if no nodes selected, fits all nodes to
Node Graph panel
F5 Render all Write nodes
F7 Render selected Write nodes
G Insert Grade node
I Display selected node information
J Jump to bookmarked node
K Insert Copy node
L Auto-place selected nodes
USER GUIDE
1992
Keystroke(s) Action
M Insert Merge node
MMB Fit all nodes to Node Graph panel
MMB and drag Pan
N Rename selected node
Numeric pad Navigate around the node tree in the direction specified. For example, pressing
arrow keys the left arrow key (4) selects the next node to the left of the current node in the
tree.
O Insert Roto node
P Insert RotoPaint node
Q Show named script info
R Insert Read node
Return Open properties for selected node(s)
S Display Project Settings
T Insert Transform node
Tab Tab node search menu
U Splay first (splay selected nodes to first selected node)
Up Arrow Previous node in tree
W Insert Write node
X Command entry mode
Y With two or more nodes selected, splay the inputs of the first node selected to
the outputs of subsequent node selections upstream
Note: If more than two nodes are selected that don't have multiple
inputs, Y ignores all nodes except the first two selections.
Alt+# (1, 2, etc.) Zoom-out %
USER GUIDE
1993
Keystroke(s) Action
Alt+B Duplicate and branch selected nodes
Alt+C Duplicate selected node(s)
Alt+D Toggles whether a node is always displayed in Dope Sheet or not
Alt+Down Arrow Version down the current Read/Write node file name
Alt+E Toggle expression links on or off
Alt+F Generate Flipbook for node
Alt+H Hide node inputs when not selected
Alt+I Display script information, such as the node count, channel count, cache usage,
and whether the script is in full-res or proxy mode
Alt+K Clone selected node(s)
Alt+N Create StickyNote
Alt+P Toggle postage stamp on or off
Alt+U With two or more nodes selected, splay first selected node to input A of all
subsequent selections
Alt+Up Arrow Version up the current Read/Write node file name
Alt+X Run a script from the file browser
Ctrl+# (0, 1, 2, Zoom-in %

etc.)
Ctrl+A Select all nodes in Node Graph
Ctrl+B Toggle node buffer for selected nodes
Tip: When enabled, the output from the nodes is cached so that it can
be re-read quickly. A yellow line displays under the nodes to indicate
that caching is enabled.
Ctrl+C Copy selected node(s)
USER GUIDE
1994
Keystroke(s) Action
Ctrl+create node Replace selected node with new node, such as Ctrl+B to replace the target with a
Blur node
Ctrl+D Disconnect upstream node
Ctrl+Down Arrow Move selected node downstream
Ctrl+F7..F10 Save locations 1 through 4
Ctrl+G Nest selected nodes in Group
Ctrl+I Open new Compositing Viewer
Ctrl+K Copy as clone
Ctrl+L Collapse selected nodes to a LiveGroup or, if no nodes are selected, create an
empty LiveGroup
Ctrl+LMB on Highlight all upstream nodes

node
Ctrl+P Toggle Proxy
Ctrl+Return/Enter Open a Group's sub-graph
Ctrl+Up Arrow Move selected node upstream
Ctrl+V Paste node(s) from the clipboard
Ctrl+X Cut selected node(s)
LMB + MMB and Zoom in and out with the click point set as the center of the Node Graph
drag
Shift+\ Snap selected node to grid
Shift+# (1, 2, 3, Connect selected node to Viewer B buffer

etc.)
Shift+A Insert AddMix node
Shift+create node Create node in new branch, such as Shift+B to create a Blur node in a new
branch
USER GUIDE
1995
Keystroke(s) Action
Shift+drag Duplicate selected arrow
Shift+F7..F10 Restore locations 1 through 4
Shift+U With two or more nodes selected, splay first selected node to input B of all
Shift+X Swap A/B inputs on node
Shift+Y With two or more nodes selected, splay the outputs of the selected nodes to the
inputs of the last node selected downstream
inputs, Shift+Y ignores all nodes except the last two selections.
Alt+LMB drag Pan
Alt+MMB drag Zoom-in/out
Alt+Shift+K Declone nodes
Alt+Shift+U With two or more nodes selected, splay last selected node to input A of all
previous selections
Alt+Shift+Up Version to up to the latest file (Read nodes only)

Arrow
Ctrl+Alt+A on Select all nodes connected to the target node's tree

node
Ctrl+Alt+G Replace Group node with nested nodes
Ctrl+Alt+LMB on Open node properties in floating window

node
Ctrl+Alt+V Paste knob values to a node of the same class as the copy operation
Ctrl+Shift+/ Search and Replace (Read and Write nodes)
Ctrl+Shift+B Toggle selected node(s) bookmark on or off
USER GUIDE
1996
Appendix B: Keyboard Shortcuts | Project/Tags/Versions Bin
Keystroke(s) Action
Ctrl+Shift+C Change node color
Ctrl+Shift+G Copy gizmo to group
Ctrl+Shift+V Paste a copied node into a new branch from an existing node
Ctrl+Shift+LMB Select all upstream nodes

on node
Ctrl+Shift+N Create a new script in a new session
Ctrl+Shift+P Create Precomp from selected nodes
Ctrl+Shift+X Extract selected nodes from tree
Ctrl+Alt+Shift+G Create Group from selected nodes
Ctrl+Alt+Shift+K Force clone
Project/Tags/Versions Bin
Keystroke(s) Action
D Hide version(s) of a clip. You can only hide versions when a clip is opened in the
Versions Bin
F8 Refresh clips
Alt+D Display metadata for the selected shot(s)
Alt+Down Arrow Version down the selected clip
Alt+F5 Rescan clip range
Alt+Up Arrow Version up the selected clip
Ctrl+B Create a new bin
Ctrl+N Create a new sequence
USER GUIDE
1997
Appendix B: Keyboard Shortcuts | Properties Panel
Keystroke(s) Action
Ctrl+Return Open selected clip in the Viewer
Ctrl+Y Create a new tag. You can only add tags to the Tags panel
Ctrl+ALt+Down Go to min version

Arrow
Ctrt+Alt+Return Open selected clip a new Viewer
Ctrl+Alt+Up Go to max version

Arrow
Properties Panel
Keystroke(s) Action
\ Snaps all nodes to grid
/ Search by node name or class
. (period) Create Dot node
1, 2, 3, etc. With a node selected in the Node Graph, create/connect Viewer inputsWith no
nodes selected in the Node Graph, cycle through connected views
B Insert Blur node
Backspace/Delete Delete selected nodes
C Insert ColorCorrect node
D Disable/Enable node
Esc Closes currently active or last selected Properties panel
F5 Render all Write nodes
F7 Render selected Write nodes
G Insert Grade node
USER GUIDE
1998
Keystroke(s) Action
I Display selected node information
J Jump to bookmarked node
K Insert Copy node
L Auto-place selected nodes
M Insert Merge node
O Insert Roto node
P Insert RotoPaint node
Q Show named script info
R Insert Read node
Return/Enter Chooses selected menu item
S Display Project Settings
T Insert Transform node
Tab Move to next control in the Properties panel
Up/Down Arrow Increment control values
U Splay first (splay selected nodes to first selected node)
W Insert Write node
X Command entry mode
Y With two or more nodes selected, splay the inputs of the first node selected to
the outputs of subsequent node selections upstream
inputs, Y ignores all nodes except the first two selections.
LMB and drag Copy current value from one control to another
USER GUIDE
1999
Keystroke(s) Action
MMB and drag Adjust control values using virtual slider (regular)
Alt+B Duplicate and branch selected nodes
Alt+C Duplicate selected node(s)
Alt+D Toggles whether a node is always displayed in Dope Sheet or not
Alt+E Toggle expression links on or off
Alt+F Generate Flipbook for node
Alt+H Hide node inputs when not selected
Alt+K Clone selected node(s)
Alt+LMB on close Close all open Properties panels

(x)
Alt+N Create StickyNote
Alt+P Toggle postage stamp on and off for selected Properties panel
Alt+U With two or more nodes selected, splay first selected node to input A of all
Alt+X Run a script from the file browser
Ctrl+A Select all nodes in Properties panel
Ctrl+B Toggle node buffer for selected nodes
Tip: When enabled, the output from the nodes is cached so that it can
be re-read quickly. A yellow line displays under the nodes to indicate
that caching is enabled.
Ctrl+C Copy selected node(s)
Ctrl+D Disconnect upstream node
Ctrl+F7..F10 Save locations 1 through 4
Ctrl+G Nest selected nodes in Group
USER GUIDE
2000
Keystroke(s) Action
Ctrl+I Open new Compositing Viewer
Ctrl+K Copy as clone
Ctrl+L Collapse selected nodes to a LiveGroup or, if no nodes are selected, create an
empty LiveGroup
Ctrl+LMB Reset slider to default
Ctrl+LMB on Close all properties panels except the one clicked

close (x)
Ctrl+P Toggle Proxy
Ctrl+V Paste node(s) from the clipboard
Ctrl+X Cut selected node(s)
Shift+A Insert AddMix node
Shift+# (1, 2, 3, Connect selected node to Viewer B buffer

etc.)
Shift+create node Create node in new branch, such as Shift+B to create a Blur node in a new
branch
Shift+F7..F10 Restore locations 1 through 4
Shift+U With two or more nodes selected, splay first selected node to input B of all
Shift+X Swap A/B inputs on node
Shift+Y With two or more nodes selected, splay the outputs of the selected nodes to the
inputs of the last node selected downstream
inputs, Shift+Y ignores all nodes except the last two selections.
Alt+LMB drag Adjust control values using virtual slider (magnitude dependent on cursor
position)
USER GUIDE
2001
Keystroke(s) Action
Alt+MMB and Adjust control values using virtual slider (fine)

drag
Ctrl+LMB drag Expression link controls
Shift+LMB drag Copy animation from one control to another
Shift+MMB and Adjust control values using virtual slider (coarse)

drag
Shift+Tab Move to previous control in the properties
Alt+Shift+K Declone nodes
Alt+Shift+U With two or more nodes selected, splay last selected node to input A of all
previous selections
Ctrl+Alt+A (on a Select all nodes connected to the target node's tree
node)
Ctrl+Alt+G Replace Group node with nested nodes
Ctrl+Alt+V Paste knob values to a node of the same class as the copy operation
Ctrl+Shift+/ Search and Replace strings in selected Read and Write nodes
Ctrl+Shift+A Close all open Properties panels
Ctrl+Shift+B Toggle selected node(s) bookmark on or off
Ctrl+Shift+C Change node color
Ctrl+Shift+G Copy gizmo to group
Ctrl+Shift+P Create Precomp from selected nodes
Ctrl+Shift+V Paste a copied node into a new branch from an existing node
Ctrl+Shift+X Extract selected nodes from tree
Ctrl+Alt+Shift+K Force clone
USER GUIDE
2002
Appendix B: Keyboard Shortcuts | Curve Editor/Dope Sheet
Curve Editor/Dope Sheet

Keystroke(s) Action
A Frame all keyframes
C Interpolation (Cubic)
F Frame all selected keyframes
H Interpolation (Horizontal)
K Interpolation (Constant)
L Interpolation (Linear)
LMB Select single point
LMB drag on blank Select region of points

space
LMB drag on point Move all selected points
LMB drag on Scale points inside selection region

selection box
LMB drag on Move all points in selection region

transform handle
MMB drag Draw box in area and zoom to fit area to curve editor panel
MMB or F Fit selection to window
R Interpolation (Catmull-Rom)
X Break selected control points' handles
Z Interpolation (Smooth [bezier])
Alt+` (backtick) Display Curve Editor
Alt+LMB drag Pan
Alt+MMB drag Variable zoom
USER GUIDE
2003
Appendix B: Keyboard Shortcuts | Script Editor
Keystroke(s) Action
Ctrl+A Select all curves
Ctrl+C Copy selected keys
Ctrl+E Copy Expressions
Ctrl+L Copy Links
Ctrl+LMB drag Move keyframes freely on the x and y axes
Ctrl+Shift (hold Hide points to click on selection box/transform handle

down)
Ctrl+V Paste curve
Ctrl+X Cut selected keys
Shift+LMB Add or Remove points from selection
Shift+LMB drag Draw box to add/remove points from selection
Alt+Shift+LMB drag Move single point
Ctrl+Alt+LMB Add point to current curve
Ctrl+Shift+C Copy selected curves
Ctrl+Alt+Shift+LMB Sketch points freely on current curve
Script Editor
Keystroke(s) Action
Tab Increase indentation
Ctrl+[ Load previous script
Ctrl+] Load next script
Ctrl+Backspace Clear output window
USER GUIDE
2004
Appendix B: Keyboard Shortcuts | Roto/RotoPaint
Keystroke(s) Action
Ctrl+Return/Enter Run script in editor
Shift+Tab Decrease indentation
Ctrl+Shift+[ Decrease indentation of selected text
Ctrl+Shift+] Increase indentation of selected text
Roto/RotoPaint
Keystroke(s) Action
Backspace Delete an item from curve list or Delete points/shapes
C Toggle Clone tool
D Toggle Dodge/Burn
Delete Remove selected point(s)
E Increase feather on selected point(s)
Esc Switch back to the current Select tool
I Pick color
N Toggle Brush/Eraser
Return (Bezier/Spline tools) Close shape
S (with Viewer mouse-over Cycle between the selected tool's modes

focus)
T (Clone tool) Toggle source as onion skin with transform jack
T (Select tool) Display transform box (points) or jack (shapes)
V Toggle Bezier/B-Spline/Ellipse/Rectangle tools
X Toggle Blur/Sharpen/Smear tools
USER GUIDE
2005
Appendix B: Keyboard Shortcuts | Roto/RotoPaint
Keystroke(s) Action
Z Smooth selected points
Ctrl+A Select all points
Ctrl+LMB (Bezier/Spline Sketch a Bezier or B-spline

tool)
Ctrl+LMB (on a point) Break tangent handle for selected point
Ctrl+drag (Clone/Reveal Set offset between source and destination

tools)
Ctrl+Shift (transform box) Drag transform box points to move them
Shift+LMB (Bezier tool) Create a sharp point on previous point
Shift+LMB (editing points) Bring up transform box for selected points
Shift+drag Change brush size

(Brush/Eraser/Clone/Reveal
tools)
Shift+drag (editing Move both tangent handles at the same time

Bezier/Spline points)
Shift+E Remove feather from selected points
Shift+Z Cusp selected points
Ctrl+Alt+LMB (on Splines) Add point to curve
Ctrl+Shift+drag (B-Spline) Increase/decrease tension of B-Spline shape
USER GUIDE
2006
Appendix C: Supported File
Formats
Notes
When importing and exporting files, remember the following:
• When you import images with a Read node (Image > Read), Nuke analyzes the contents of the file
to determine the format. The file name extension is not used to determine file format, which allows
flexibility with naming conventions in a production environment.
• Regardless of format, Nuke converts all imported image sequences to its native 32-bit linear RGB
colorspace.
• When you render new images from Nuke (Image > Write), you can use a file name extension to
specify format.
• To import and export geometry objects from Alembic, FBX, or OBJ files, use the ReadGeo node (3D >
Geometry > ReadGeo). To write them out again, use the WriteGeo node (3D > Geometry >
WriteGeo). To import cameras or transforms from Alembic or FBX files, use the Camera node (3D >
Camera). To import lights from an FBX file, use the Light node (3D > Lights > Light).
• To import deep images (either in DTEX or scanline OpenEXR format), use a DeepRead node (Deep >
DeepRead). To export deep images (in scanline OpenEXR format), use a DeepWrite node (Deep >
DeepWrite).
Supported File Formats

The following table lists the supported file formats. The extensions listed under Extension let you
specify the image format; use these as the actual file name extensions or the prefix to indicate output
format for the image sequences.
Format Bit Read/Write Extension Notes

Depths
Alembic n/a read and abc You can read meshes, point clouds, cameras,
USER GUIDE
2007
Appendix C: Supported File Formats |

Depths
write and transforms from Alembic files into a Nuke

scene using the ReadGeo, Camera, and Axis
nodes.
To write meshes and point clouds out again,

use the WriteGeo node.
Apple 8, 10 read and mov Adds support for Apple ProRes 4444 and
ProRes write Apple ProRes 422 on Mac, Linux, and
Windows using the mov64 reader.
Apple ProRes 4444 includes the SD, HD, 2K,

UHD, and XQ formats. Apple ProRes 422
includes the HQ, LT, and Proxy formats.
ARRIRAW 12 read only ari
AVI n/a read and avi AVI files can be supported by default or via
write Nuke’s reader/writer that is based on the
FFmpeg open source library. If you get an
error when using AVI files in Read nodes, you
may need to use the prefix mov64: before the
file path and file name, for example:
mov64:z:/job/FILM/IMG/final_comp_
v01.####.avi
When working with Write nodes, you can also

select mov64 from the file type dropdown
menu and use avi as the file extension.
On Windows, in order to support more

codecs, the AVI reader uses the DirectShow
multimedia architecture. When decoding .avi
files, DirectShow tries to find the appropriate
codec on the system. If the codec is not
available, DirectShow and Nuke are unable to
open the .avi file. Note that the 64-bit version
of Nuke can only use 64-bit DirectShow
USER GUIDE
2008

Depths
codecs. If you only have a 32-bit codec

installed, the 64-bit version of Nuke cannot
use it to open .avi files.
CIN 10 (log) read and cin

write
DNG 8, 12 read dng Includes RAW 2.5K CinemaDNG
DPX 8, 10, 12, read and dpx YCbCr encoded DPX files are not supported on
and 16 write the timeline.
DTEX 32 read only dtex To use DTEX files, you need Pixar’s
PhotoRealistic RenderMan® Pro Server 20, or
earlier, on your computer.
To read a DTEX file, use the DeepRead node.
FBX read and fbx You can read meshes, point clouds, cameras,
write lights, and transforms from FBX files into a
Nuke scene using the ReadGeo, Camera,
Light, and Axis nodes.
To write geometry out again, use the

WriteGeo node.
GIF 8 read only gif
Radiance 16 read and hdr, hdri This format stores an 8-bit mantissa for each
write of r, g, and b and an additional 8-bit exponent
that is shared by all three, which packs the
floating point RGB triplet into 32 bits per
pixel.
JPEG 8 read and jpg, jpeg Adjust compression levels using the quality
write slider in the Write node’s properties panel.
Maya IFF 8, 16 read only iff
USER GUIDE
2009

Depths
MXF n/a read only mxf

Note: Currently, only 'complete' MXF
files are supported.
Supported codecs include:

• Uncompressed 4:2:2 YCbCr 8-/10-bit
• Uncompressed 4:4:4:4 RGBA 8-/10-bit
• Uncompressed Avid 4:2:2 YCbCr 8-/10-bit
• Uncompressed Avid 4:4:4:4 RGBA 8-/10-bit
• JPEG2000
• Avid DNxHD (1080p and 720p 1920x1080 and
1280x720, 4:4:4:4 and 4:2:2) 36, 115, 120, 145,
175, 185, 220, 220x
• Sony Raw from the F65, F55, F5 and FS700
cameras. All formats that these cameras
provide: 4K, 2K, 1K, 0.5K and 0.25K
• Sony X-OCN from the VENICE, F55, and F5
cameras.
• ARRIRAW from the Alexa Mini.
OBJ read and obj

write
OpenEXR 16, 32 read and exr OpenEXR handles 16- and 32-bit float. This 16
write is also called "half float" and is different from
and the 16-bit integer that all the other formats
OpenEXR that support 16-bit use.
2.2 Nuke supports multi-part OpenEXR files. For
more information, see Notes on Importing
OpenEXR Files and Notes on Rendering
OpenEXR Files.
When working with deep data, Nuke supports

scanline OpenEXR files. For more information,
see Importing Scanline OpenEXR Files and
USER GUIDE
2010

Depths
Writing Deep Data.
EXR Compression
EXR file metadata contains a compression key/value pair detailing the compression type used to
write the .exr file. The value is expressed as the name of the compression type or an integer
referencing the compression used:
0 - no compression
1 - RLE compression, run length encoding
2 - Zip compression, one scan line at a time
3 - Zip compression, in blocks of 16 scan lines
4 - PIZ-based wavelet compression, in blocks of 32 scan lines
5 - PXR24 compression, lossy 24-bit float
6 - B44 compression, lossy 4-by-4 pixel block, fixed rate
7 - B44A compression, lossy 4-by-4 pixel block, flat fields are compressed more
8 - DWAA compression, lossy DCT based compression, in blocks of 32 scan lines
9 - DWAB compression, lossy DCT based compression, in blocks of 256 scan lines
PNG 8, 16 read and png (8-bit)

write
png16 (16-
bit)
PSD 8, 16 read only psd While Nuke reads standard Photoshop®

blend modes, it doesn't read Photoshop layer
comps or recognize group blend modes.
Photoshop layers are read into separate Nuke
layers and anything that doesn't map into that
is ignored.
QuickTime n/a read and mov QuickTime is only supported by default on
USER GUIDE
2011

Depths
write Windows and Mac. To use QuickTime files on

Linux, you need to use the prefix mov64:
before the file path and file name, for
example,
mov64:/z:/job/FILM/IMG/final_comp_
v01.####.mov
The mov64 writer supports the following

codecs:
• Apple ProRes 4444
• Apple ProRes 4444 XQ
• Apple ProRes 422 HQ
• Apple ProRes 422
• Apple ProRes 422 LT
• Apple ProRes 422 Proxy
• Avid DNxHD (1080p and 720p 1920x1080 and
1280x720, 4:4:4:4 and 4:2:2) 36, 115, 120, 145,
175, 185, 220, 220x
Note: Interlaced writing is not

supported. See Avid DNxHD Notes
below.
• Photo - JPEG
• MPEG-1 Video
• MPEG-4 Video
• PNG
• Animation
• Uncompressed 10-bit 4:2:2
Avid DNxHD Notes
The bit rates listed in the codec profile dropdown are the bit rates for 1080p at 29.97 fps EXCEPT for
36 (which is actually 45 Mbps @ 29.97fps). You should look at the codec format (422/444, 8/10-bit).
USER GUIDE
2012

Depths
Note: Nuke only supports 1080p and 720p. Non-HD resolutions are scaled to 1080p
before writing.
This leads to a set of 1080p bit rates:

• 1080p/29.97 440x, 220x, 220, 145, 45
• 1080p/60 N/A, N/A, 440, 290, 90 (same at 59.94)
• 1080p/50 N/A, N/A, 367, 242, 75
• 1080p/25 365x, 185x, 185, 120, 36
• 1080p/24 350x, 175x, 175, 115, 36 (same at 23.976)
At 720p, the codec profile dropdown has a different interpretation. The bit rate is taken as the bit
rate at 720p at 59.94fps. This leads to another set of bit rates:
• 720p/59.94 N/A, 220x, 220, 145, N/A
• 720p/50 N/A, 175x, 175, 115, N/A
• 720p/29.97 N/A, 110x, 110, 75, N/A
• 720p/25 N/A, 90x, 90, 60, N/A
• 720p/23.976 N/A, 90x, 90, 60, N/A
Note: Since the bit rates are for 1080p at 29.97 fps AND 720p at 59.94 fps (except for 36
Mbit which should read 45 Mbit). It is possible to calculate the bandwidth for all the other
frame rates by:
• BandWidth@1080p = fps/29.97 * NominalBandWidth, or

• BandWidth@720p = fps/59094 * NominalBandWidth
where NominalBandWidth is the bandwidth listed in the codec profile knob OR 45 if the
bandwidth listed is 36 Mbit. (Avid labels the codec profile names by the approximate
bandwidth.)
RAW n/a read only n/a DSLR raw data files, such as Canon .CR2 files.
These are only supported via the dcraw
USER GUIDE
2013

Depths
command line program, which you can

download from the dcraw website. Bit depth
and other specifications depend on the
device. Some devices may not be supported.
REDCODE 16 read only r3d Note that .r3d files may look different in Nuke
compared to various versions of RED
applications, like RED ALERT or REDCINE.
Unlike most other file formats Nuke reads,
the .r3d REDCODE files must be processed to
convert from a raw format to an RGB color
image. From time to time, a new version of
the RED SDK that Nuke uses improves this
processing and due to the timing of release
cycles, Nuke may sometimes be using a
different version than the RED applications.
SGI 8, 16 read and sgi, rgb,

write rgba (8-bit
sequences)
sgi16 (for
16-bit
sequences)
SoftImage 8 read and pic

® write
PIC
TIFF 8, 16, and read and tif, tiff (8- If utilized, the compression schema on
32 write bit imported TIFF sequences must be LZW®.
sequences)
tif16, tiff16
(16-bit
sequences)
ftif, ftiff
USER GUIDE
2014

Depths
(32-bit
sequences)
T 8 read and tga, targa

ruevision® write
TARGA
Wavefront 8 read only rla

®
RLA
XPM 8 read and xpm This is the text-based format in which Nuke’s
write interface elements are stored.
YUV 8 read and yuv This format does not specify resolution, so
write Nuke assumes a width of 720 pixels.
USER GUIDE
2015
Supported Camera Formats |
Supported Camera Formats

The following table lists supported ARRI, R3D, and Sony camera/sensor formats.
Camera/Sensor Nuke 11.1 Nuke 11.2 Nuke 11.3
ARRI (.ari and .mxf)
ALEXA Classic
ALEXA XT
ALEXA SXT(W)
ALEXA LF
ALEXA 65
ALEXA Mini
AMIRA
Note: Nuke, Nuke Studio, and Hiero do not support .mxf from the AMIRA camera.
R3D (.r3d)
MONSTRO 8K VV
HELIUM 8K S35
SCARLET-W 5K
RED RAVEN 4.5K
WEAPON DRAGON 6K
EPIC/SCARLET DRAGON
RED ONE MYSTERIUM 4K
MYSTERIUM-X
USER GUIDE
2016
Supported Camera Formats |
Camera/Sensor Nuke 11.1 Nuke 11.2 Nuke 11.3
MYSTERIUM-X MONOCHROME
SONY (.mxf)
F5
F55
F65
FS700
VENICE
USER GUIDE
2017
Supported Audio Formats |
Supported Audio Formats

The following table lists supported audio formats.
Note: Although Nuke supports the import and editing of multi-channel audio, during
playback audio is currently mixed to 48 KHz stereo output.
Format Name Extension
Audio only formats
All platforms
Wave wav
Audio interchange format aiff
Mac and Windows
MPEG 4 Audio m4a
Integrated audio formats
RED Audio r3d
QuickTime Audio mov
Note: QuickTime Audio is not supported on Linux OS.
USER GUIDE
2018
Appendix D: External Software
Third-Party Contributions
The following table lists third-party contributions included in Nuke, along with their licenses.
Contributor Description License
Pixar DTEX file Copyright © Pixar. All rights reserved.

reader
This license governs use of the accompanying software. If you use
the software, you accept this license. If you do not accept the
license, do not use the software.
1. Definitions
The terms "reproduce," "reproduction," "derivative works," and

"distribution" have the same meaning here as under U.S. copyright
law. A "contribution" is the original software, or any additions or
changes to the software.
A "contributor" is any person or entity that distributes its

contribution under this license. "Licensed patents" are a
contributor's patent claims that read directly on its contribution.
2. Grant of Rights
(A) Copyright Grant- Subject to the terms of this license, including

the license conditions and limitations in section 3, each contributor
grants you a non-exclusive, worldwide, royalty-free copyright license
to reproduce its contribution, prepare derivative works of its
contribution, and distribute its contribution or any derivative works
that you create.
(B) Patent Grant- Subject to the terms of this license, including the
license conditions and limitations in section 3, each contributor
grants you a non-exclusive, worldwide, royalty-free license under its
licensed patents to make, have made, use, sell, offer for sale,
USER GUIDE
2019
Appendix D: External Software | Third-Party Contributions
Contributor Description License
import, and/or otherwise dispose of its contribution in the software

or derivative works of the contribution in the software.
3. Conditions and Limitations
(A) No Trademark License- This license does not grant you rights to
use any contributor's name, logo, or trademarks.
(B) If you bring a patent claim against any contributor over patents
that you claim are infringed by the software, your patent license
from such contributor to the software ends automatically.
(C) If you distribute any portion of the software, you must retain all
copyright, patent, trademark, and attribution notices that are
present in the software.
(D) If you distribute any portion of the software in source code form,
you may do so only under this license by including a complete copy
of this license with your distribution. If you distribute any portion of
the software in compiled or object code form, you may only do so
under a license that complies with this license.
(E) The software is licensed "as-is." You bear the risk of using it. The
contributors give no express warranties, guarantees or conditions.
You may have additional consumer rights under your local laws
which this license cannot change.
To the extent permitted under your local laws, the contributors

exclude the implied warranties of merchantability, fitness for a
particular purpose and non-infringement.
USER GUIDE
2020
Third-Party Libraries and Fonts | Third-Party Library Versions
Third-Party Libraries and Fonts

Note: If, for any reason, you think Foundry is not entitled to use these libraries and fonts,
please visit support.foundry.com.
Third-Party Library Versions

The following table lists third-party libraries included in Nuke and their current version.
Version
Library Linux Mac Windows
AAF 44m1_v1 44m1_v1 44m1_v1
ACES 1.0.3 1.0.3 1.0.3
AJA 12.5.2.7 12.5.1 12.5.1
Alembic 1.6.1 1.6.1 1.6.1
ARRIRAW SDK 5.4.3.5 5.4.3.5 5.4.3.5
Blackmagic 10.8.3 10.8.3 10.8.3
bmx 20140528 20140528 20140528
Boost 1.61.0 1.61.0 1.61.0
Breakpad r1338 r1338 r1338
bzip 1.0.6 1.0.6 1.0.6
cIFFT_AMD 20131219 20131219 20131219
cppunit 1.12.1 1.12.1 1.12.1
Ctypes 1.0.2 1.0.2 1.0.2
CUDA 8.0 & 6.5 8.0 & 6.5 8.0 & 6.5
USER GUIDE
2021
Version
Curl 7.53.0 7.53.0 7.53.0
DirectX N/A N/A Jun10
DNG SDK 1.3 1.3 1.3
DNxHD and DNxHR 2.3.1 2.3.1 2.3.1
EuCon N/A 2.5.5 N/A
Expat 2.0.1 2.0.1 2.0.1
FBX 2017.0.1 2017.0.1 2017.0.1
FFmpeg 2.1.4 2.1.4 2.1.4
FFTW 3.1.3 3.1.3 3.1.3
FreeType 2.5.0.1 2.5.0.1 2.5.0.1
FTGL 2.1.3 2.1.3 2.1.3
GCC 4.8.2 - with bugfix 4.8.2 - with bugfix 4.8.2 - with bugfix
GLEW 1.5.8 1.5.8 1.5.8
googletest 1.5.0 1.5.0 1.5.0
gperftools 2.0 2.0 2.0
HDF 5.1.8.7 5.1.8.7 5.1.8.7
Intel Libraries 17.1.143 12.1.0258 17.1.143
Intel MKL 11.3.4 11.3.4 11.3.4
Intel TBB 4.4.6 4.4.6 4.4.6
JPEG 6b 6b 6b
Libexif 0.6.20 0.6.20 0.6.20
Libpng 1.4.8 1.4.8 1.4.8
USER GUIDE
2022
Version
libresample 0.1.3 0.1.3 0.1.3
libsndfile 1.0.25 1.0.25 1.0.25
Libtiff 3.9.4 3.9.4 3.9.4
LLVM 3.3 3.3 3.3
Oculus Rift SDK N/A N/A 0.4.4
OFX 1.2 1.2 1.2
OfxHostSupport 1.0 1.0 1.0
OpenCL 1.0 N/A 1.0
OpenColorIO 1.0.9 1.0.9 1.0.9
OpenEXR 2.2.0 2.2.0 2.2.0
OpenJPEG 1.5.0 1.5.0 1.5.0
OpenSSL 1.0.2g 1.0.2g 1.0.2g
OpenSubDiv 3.1.0x 3.1.0x 3.1.0x
OpenVDB 4.x 4.x 4.x
PoissonRecon V2 V2 V2
PortAudio v19_20111221 v19_20111221 v19_20111221
Primatte 2017.5.0 2017.5.0 2017.5.0
ProRes 2017-04-02 2017-04-02 2017-04-02
protobuf 2.5.0 2.5.0 2.5.0
psutil 2.0.0 2.0.0 2.0.0
Ptex 2.1.28 2.1.28 2.1.28
PySide 2.0 2.0 2.0
USER GUIDE
2023
Version
PyString 1.1.0 1.1.0 1.1.0
Python 2.7.13 2.7.13 2.7.13
PyZMQ 13.0.2 13.0.2 13.0.2
QuickTime N/A 7.3 7.3
QuickTimeHelper N/A 1.0 1.0
Qt 5.6.1 - modified 5.6.1 - modified 5.6.1 - modified
R3D SDK 7.0.6 7.0.6 7.0.6
RLM 12.2 12.2 12.2
SCons 2.1.0 2.1.0 2.1.0
Skein 1.1 1.1 1.1
sony-ArbitraryOutputAttr 1.0 1.0 1.0
sony-BinaryIO 1.0 1.0 1.0
sony-dirent N/A N/A 1.0
sony-dlfcn N/A N/A 1.9
sony-expatwrap N/A 1.0 1.0
sony-mpeg4decodersdk N/A (header-only) N/A (header-only) N/A (header-only)
sony-ptr N/A (header-only) N/A (header-only) N/A (header-only)
sony-pystring 1.0 1.0 1.0
sony-resolutiontable 1.0 1.0 1.0
sony-sceneGraphAttr 1.0 1.0 1.0
sony-smdk 4.13.0 4.13.0 4.13.0
sony-xmlIO 1.0 1.0 1.0
USER GUIDE
2024
Version
Tcl 8.4.16 8.4.16 8.4.16
TinyXML N/A (header-only) N/A (header-only) N/A (header-only)
Ultimatte 1.0 1.0 1.0
uriparser 0.8.0 0.8.0 0.8.0
uuid 1.0 N/A N/A
VXL 1.10.0 1.10.0 1.10.0
xmlrpcpp N/A N/A 0.7
ZeroMQ 3.2.5 3.2.5 3.2.5
zlib 1.2.5 1.2.5 1.2.5
For a full list of all third-party licenses and fonts used in Nuke, please see the Nuke Online Help,
Third-Party Libraries and Fonts, or click here.
USER GUIDE
2025

Nuke

Uploaded by

Document Informationclick to expand document informationpresentation

Document Informationclick to expand document information

Copyright:

Available Formats

Nuke

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nuke

Uploaded by

Copyright:

Available Formats

USER GUIDE

The EULA is available here: End User License Agreement (EULA)

Nuke™ is a trademark of The Foundry Visionmongers Ltd.

Digital Domain ® is a registered trademark of Digital Domain, Inc.

Primatte™ is a trademark of IMAGICA Corp.

Primatte™ patent is held by IMAGICA Corp.

Linux ® is a registered trademark of Linus Torvalds.

Windows ® is the registered trademark of Microsoft Corporation.

Houdini ® is a registered trademark of Side Effects Software, Inc.

Boujou is a trademark of 2d3 Ltd.

3D-Equalizer is a trademark of Science.D.Visions.

RenderMan ® is a registered trademark of Pixar.

Cineon™ is a trademark of Eastman Kodak Company.

Rev: Wednesday, July 24, 2019

Meet the Nuke Product Family 81

About Nuke Assist 84

Using Gizmos, Groups, Precomps, Knobs, and Python 86

About Nuke Non-commercial 87

Understanding the Workflow 88

Working with Multiple Image Formats 89

8-, 16-, and 32-Bit Image Processing 91

Render Farms and Frame Servers 92

Installation and Licensing 93

Minimum Hardware Requirements 94

Requirements for GPU Acceleration 95

Installing with the User Interface (UI) 97

Installing from the Command Line 98

Nuke Non-commercial 100

Licensing on Windows 101

Obtaining Licenses 101

Installing Licenses 102

Further Reading 105

Licensing Nuke Non-commercial on Windows 106

Uninstalling Apps on Windows 108

Mac OS X and macOS 109

System Requirements 109

Minimum Hardware Requirements 109

Requirements for GPU Acceleration 110

Installing on Mac 112

Installing with the User Interface (UI) 112

Installing from the Terminal 112

Launching on Mac 114

Nuke Analytics 114

Nuke Non-commercial 115

Obtaining Licenses 116

Installing Licenses 117

Further Reading 120

Licensing Nuke Non-commercial on Mac 121

Uninstalling Apps on Mac 123

System Requirements 124

Minimum Hardware Requirements 124

Requirements for GPU Acceleration 125

Installing on Linux 127

Installing from the Terminal 127

Installing Remotely from the Terminal 128

Launching on Linux 129

Nuke Analytics 129