Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Kako Pobediti U Igranju Saha Sa Samo 2 Figure Lmao

Download as txt, pdf, or txt
Download as txt, pdf, or txt
You are on page 1of 3

How to Make Viral AI Avatar for Faceless YouTube Channels

By: Alex Iverstone


Video tutorial link: https://youtu.be/kjT_rdZWfIo

Step 1. AI Avatar Generation


1. Go to https://leonardo.ai/
2. look for the "Image Generation" option.
3. select the "Leonardo Diffusion XL" model for high-quality results.
4. Choose the aspect ratio 16:9
5. enter a detailed prompt. For example, try "a young man sitting in the middle
of a tidy home desk, front-facing, with a window at the back showing the city
skyline."
6. Feel free to add creative descriptions, specifying details like the gender,
the color of the shirt and the background.
7. Also copy and paste the following negative prompt: “(((2 heads))), duplicate,
blurry, abstract, disfigured, deformed, cartoon, animated, toy, figure, framed, 3d,
cartoon, 3d, disfigured, bad art, deformed, poorly drawn, extra limbs, close up,
b&w, weird colors, blurry, watermark, blur haze, 2 heads, long neck, watermark,
elongated body, cropped image, out of frame,draft,deformed hands, twisted fingers,
double image, malformed hands, multiple heads, extra limb, ugly, poorly drawn
hands, missing limb, cut-off, over-saturated, grain, lowères, bad anatomy, poorly
drawn face, mutation, mutated, floating limbs, disconnected limbs, out of focus,
long body, disgusting, extra fingers, gross proportions, missing arms, (((mutated
hands))),(((bad fingers))) cloned face, missing legs”
8. We can also explore different models for diverse styles. For instance, try
the "Anime Pastel Dream" model for an anime-inspired look.
9. Remember to adjust the aspect ratio again since it resets when you change the
model.
10. If the results aren't perfect, keep tweaking the prompt until you get the
character you desire. You have 150 free credits with Leonardo, and they reset every
day. so if you run out, try again tomorrow.

2 bonus tips
Tip 1:
1. click on ‘image guidance’ and leverage an existing image to guide your
generations.
2. we can then reduce the guidance strength so the image varies a bit more from
the reference image.
Tip 2:
1. use the canva editor and micro-edit the image.
2. we can do so by moving the selection box and then drawing a mask over the
area
3. select the number of generations and the model.
4. then type laptop in the prompt box and click generate
5. go through the 4 options and choose one the best option.

Upscale image:
Upscaler: https://affiliates.pixelbin.io/8gsm3lfiju2y-7czmi
1. head over to this AI tool.
2. Upload our image, click 4 x, download

Step 2. Human Touches


Fiverr: https://go.fiverr.com/visit/?bta=903731&brand=fiverrcpa
1. Simply search for "Photoshop," filter by your budget, and discover numerous
talented individuals.
Canva: https://www.canva.com/
2. Upload your image and click "Edit Photo."
3. Apply the Fresco filter to enhance the overall quality of the image.
4. separate the foreground from the background with this tool.
5. mess around with the sliders,

Step 3. Voice-over
ElevenLabs: https://try.elevenlabs.io/7zgd8ld67lcr
1. explore the default sounds available
2. choose a voice from the Voice Library, a public collection of voices.
3. Navigate through the filters above the library to find a voice best suited
for your channel. You can sort by trending and even select the gender, age, and
accent.
4. if you want a unique AI voice that no one else has, head over to voicelab,
you can clone your own voice by clicking on add voice and uploading a file of
yourself speaking for over 1 minute.
5. return to Speech Synthesis in Eleven-Labs. Pick your chosen voice, type your
desired text in the box, and click "Generate." Take a moment to listen and ensure
it's perfect before pressing "Download.”

Step 4. Avatar Animation


HeyGen: https://heygen.com/?sid=rewardful&via=alex-i
1. Begin by creating a landscape video. Head to the Photo Avatar section, upload
your image.
Image compressor: https://imagecompressor.com/
Image resizer: https://imageresizer.com/
2. use an image compressor and an image resizer to reduce the file size and
dimensions without compromising quality.
3. upload it again to Heygen. Click on the Audio Script section, upload the
Eleven-Labs voiceover, and then click submit.
4. you might notice a watermark in the corner. you can crop it out later or use
a free watermark remover tool such as this one

Alternative tool
Vidnoz: https://www.vidnoz.com/?insur=en_yt_AlexIverstone&a_aid=658cdbe3748e0
1. click "Create Video," start blank, click avatar then choose "Talking Photo,"
and upload your photo. Uncheck the "Remove Background" tool, then click generate.
2. Next, click "Full Photo," resize, upload the Eleven-Labs voiceover, and click
generate again.

My strategy to maximise the limited minutes in the free plan.


Instead of showing the avatar throughout an entire 10 minute video, instead I only
show the avatar for around 3 seconds every minute or at some key phrases.

Bonus Tricks
1. in capcut you can use simple built in animations such as shake 1. shake 2.
and zoom out.
2. capcut offers some built-in effects such as leak 2. black noise. and noise 2.
i’m actually giving out a pack of overlays containing 50 cinematic effects i’ve
collected, such as rain. film. and old tv.
to receive this overlay pack for free, simply like this video, subscribe, and
comment ‘Overlay’ and i’ll personally send it to you, allowing you to take your
avatar animations to the next level.

Video tutorial link: https://youtu.be/kjT_rdZWfIo

Other cool stuff


Turn 1 long form video into multiple viral shorts instantly: https://www.opus.pro/?
via=AlexIverstone
Cool editing stuff I use: https://1.envato.market/alex
Make viral captions instantly: https://submagic.co/?fpr=alexiverstone code:
Iverstone10 for 10% off

* This information should not be construed as investment advice; it is provided for


entertainment purposes only. This video contains affiliate links which means If you
buy using one of them, I'll receive a small commission without additional cost to
you. The video is accurate as of the date of publication, but this may change in
the future.

You might also like