Download as DOC, PDF, TXT or read online from Scribd
Download as doc, pdf, or txt
You are on page 1/ 10
ISO/IEC 13818-1:201X/PDAM 3 (E)
Rec. ITU-T H. 222.0/Amd.3
INTERNATIONAL STANDARD ISO/IEC 13818-1:201X/PDAM 3 (E) Rec. ITU-T H. 222.0/Amd.3
ITU-T RECOMMENDATION INFORMATION TECHNOLOGY -- GENERIC CODING OF MOVING PICTURES AND ASSOCIATED AUDIO INFORMATION: SYSTEMS AMENDMENT 3 Transport of High Efficiency Video Coding (HEVC) Streams over ITU-T Rec H.222.0 | ISO/IEC 13818-1 1) Clause 1.2.2 Add the following references:
ITU-T Recommendation H.26x (201x), Advanced video coding for generic audiovisual services
ISO/IEC 23008-2, Information technology High efficiency video coding
2) Clause 2.1 Add the following definitions after 2.1.94: 2.1.95 HEVC video stream: A byte stream according to [HEVC] Annex B. 2.1.96 HEVC access unit: An Access unit as defined in [HEVC] with the constraints specified in 2.15.1. 2.1.97 HEVC 24-hour picture (system): A HEVC access unit with a presentation time that is more than 24 hours in the future. For the purpose of this definition, HEVC access unit n has a presentation time that is more than 24 hours in the future if the difference between the initial arrival time tai(n) and the DPB output time to,dpb(n) is more than 24 hours. 2.1.98 HEVC slice: A byte_stream_nal_unit with nal_unit_type equal to 1, 4 or 5 as defined in [HEVC]. 2.1.99 HEVC entropy slice: An HEVC slice with the syntax element entropy_slice_flag in the slice header equal to 1 as defined in [HEVC]. 2.1.100 HEVC tile: One or more consecutive HEVC slices which form the coded representation of a tile. NOTE If the coded representation of a tile is not congruent with an integer number of slices, this standard do not refer to this entity as an HEVC tile.
2.1.101 HEVC CRA picture: The coded representation of a picture containing only slices with nal_unit_type equal to 4 as defined in [HEVC]. 2.1.102 HEVC IDR picture: The coded representation of a picture containing only slices with nal_unit_type equal to 5 as defined in [HEVC]. 2.1.103 HEVC still picture (system) : An HEVC still picture consists of a HEVC access unit containing an IDR picture preceded by SPS and PPS NAL units that carry sufficient information to correctly decode this IDR picture. Preceding a HEVC still picture, there shall be another HEVC still picture or an End of Sequence NAL unit terminating a preceding coded video sequence. 2.1.104 HEVC video sequence (system) : coded video sequence as defined in Clause 3.25 in ITU-T Recommendation H.265 | ISO/IEC 23008-2.
ISO/IEC 13818-1:201X/PDAM 3 (E)
2.1.105 HEVC video sub-bitstream: A subset of the NAL units of a byte stream according to [HEVC] Annex B in their original order of which NAL unit header syntax elements match a specified criterion. 2.1.106 HEVC video temporal sub-bitstream: An HEVC video sub-bitstream of which the NAL unit header syntax element temporal_id as specified in [HEVC] matches a specified subset within the range from 0 to max_temporal_layers_minus1 as included in the sequence parameter set specified in [HEVC].
3) Clause 2.4.2.6 Replace the following 2 paragraphs: Replace: The delay of any data through the System Target Decoder buffers shall be less than or equal to one second except for still picture video data and ISO/IEC 14496 streams. Specifically: tdn(j) t(i) 1 second for all j, and all bytes i in access unit An(j). by: The delay of any data through the System Target Decoder buffers shall be less than or equal to one second except for still picture video data, ISO/IEC 14496 and ISO/IEC 23008-2 streams. Specifically: tdn(j) t(i) 1 second for all j, and all bytes i in access unit An(j). Replace: For ISO/IEC 14496 streams, the delay is constrained by td n(j) t(i) 10 seconds for all j, and all bytes i in access unit An(j). by: For ISO/IEC 14496 and ISO/IEC 23008-2 streams, the delay is constrained by td n(j) t(i) 10 seconds for all j, and all bytes i in access unit An(j).
4) Clause 2.4.2 Add the following immediately after 2.4.2.10 as a new subclause:
2.4.2.11, T-STD extensions for carriage of HEVC:
T-STD extensions and T-STD parameters for decoding of HEVC video streams are defined in 2.17.2 and 2.17.3.
5) Clause 2.4.3.5 In the section specifying the discontinuity_indicator, add at the end of the bulleted list introduced by For the purpose of this clause, an elementary stream access point is defined as follows:
HEVC: The first byte of an HEVC access unit that contains an IDR or a CRA picture. The SPS, PPS and APS parameter sets referenced in this and all subsequent HEVC access units in the coded video stream shall be provided after this access point in the byte stream and prior to their activation.
In the section specifying the elementary_stream_priority_indicator, add:
In the case of HEVC, this field may be set to '1' only if the payload contains one or more bytes from a slice with slice_type set to 2 and entropy_slice_flag set to 0. A value of '0' indicates that the payload has the same priority as all other packets which do not have this bit set to '1'.
6) Clause 2.4.3.7 In the section specifying the PTS (presentation time stamp), add:
Rec. ITU-T H. 222.0/Amd.3
ISO/IEC 13818-1:201X/PDAM 3 (E)
For HEVC video streams, if a PTS is present in the PES packet header, it shall refer to the first HEVC access unit that commences in this PES packet. In the section specifying the DTS (decoding time stamp), add: For HEVC video streams, if a DTS is present in the PES packet header, it shall refer to the first HEVC access unit that commences in this PES packet. To achieve consistency between the STD model and the HRD model defined in Annex C of HEVC, for each HEVC access unit the DTS value in the STD shall, within the accuracy of their respective clocks, indicate the same instant in time as the nominal CPB removal time tr in the HRD, as defined in Annex C of HEVC.
7) Clause 2.4.4.9 In Table 2-34, Stream type assignments, replace the following line: 0x27-0x7E
ITU-T Rec. H.222.0 | ISO/IEC 13818-1 Reserved
with: 0x27 0x28-0x7E
ITU.-T Rec Hxxx | ISO/IEC xxxxx-1 / HEVC video stream
ITU-T Rec. H.222.0 | ISO/IEC 13818-1 Reserved
8) Clause 2.6.1 Replace Table 2-45 by: Table 2-45 - Program and program element descriptors
descriptor_tag
TS
PS
Identification
n/a
n/a
Reserved
n/a
Forbidden
video_stream_descriptor
audio_stream_descriptor
hierarchy_descriptor
registration_descriptor
data_stream_alignment_descriptor
target_background_grid_descriptor
video_window_descriptor
CA_descriptor
10
ISO_639_language_descriptor
11
system_clock_descriptor
12
multiplex_buffer_utilization_descriptor
13
copyright_descriptor
14
15
private_data_indicator_descriptor
16
smoothing_buffer_descriptor
17
maximum_bitrate_descriptor
STD_descriptor
18
19-26
IBP_descriptor
27
MPEG-4_video_descriptor
28
MPEG-4_audio_descriptor
29
IOD_descriptor
30
31
FMC_descriptor
32
external_ES_ID_descriptor
Defined in ISO/IEC 13818-6
SL_descriptor
Rec. ITU-T H. 222.0/Amd.3
ISO/IEC 13818-1:201X/PDAM 3 (E)
33
MuxCode_descriptor
34
FmxBufferSize_descriptor
35
36
content_labeling_descriptor
37
metadata_pointer_descriptor
38
metadata_descriptor
39
metadata_STD_descriptor
40
AVC video descriptor
41
IPMP_descriptor (defined in ISO/IEC 13818-11, MPEG-2 IPMP)
42
AVC timing and HRD descriptor
43
MPEG-2_AAC_audio_descriptor
44
FlexMuxTiming_descriptor
45
MPEG-4_text_descriptor
46
MPEG-4_audio_extension_descriptor
47
Auxiliary_video_stream_descriptor
48
SVC extension descriptor
49
MVC extension descriptor
50
n/a
J2K video descriptor
51
MVC operation point descriptor
52
MPEG2_stereoscopic_video_format_descriptor
53
Stereoscopic_program_info_descriptor
54
Stereoscopic_video_info_descriptor
55
ODUpdate_descriptor
multiplexBuffer_descriptor
56
n/a
Transport_profile_descriptor
57
n/a
HEVC video descriptor
58
n/a
HEVC timing and HRD descriptor
59-62
n/a
n/a
Rec. ITU-T H.222.0 | ISO/IEC 13818-1 Reserved
63
Extension_descriptor
64-255
n/a
n/a
User Private
9) Clause 2.6.11 Add the following immediately after Table 2-54: Table 2-xx describes the alignment type for HEVC when the data_alignment_indicator in the PES packet header has a value of '1'. Table 2-xx HEVC video stream alignment values Alignment type 00
Reserved
01
HEVC access unit
02
HEVC slice
03
HEVC access unit or slice
04
HEVC tile
05
HEVC access unit or tile
06
HEVC slice or tile
07
HEVC access unit or slice or tile
08-255
Rec. ITU-T H. 222.0/Amd.3
Description
Reserved
ISO/IEC 13818-1:201X/PDAM 3 (E)
10) Clause 2.6
Add the following immediately after clause 2.6.95 as new subclauses:
2.6.96 HEVC video descriptor
For HEVC video streams, the HEVC video descriptor provides basic information for identifying coding parameters of the associated HEVC video stream, such as profile and level parameters.
2.6.97 Semantic definition of fields in HEVC video descriptor
profile_idc This 8-bit field indicates the profile, as defined in the HEVC specification. reserved_zero_8bits This 8-bit field shall be copied from the 8 bits immediately following the profile_idc syntax element in the sequence_parameter_set according to the HEVC specification. level_idc This 8-bit field indicates the level, as defined in the HEVC specification. temporal_layer_subset_flag This 1-bit flag when set to '1' indicates that the syntax elements describing a subset of temporal layers are included in this descriptor. HEVC_still_present This 1-bit field when set to '1' indicates that the HEVC video stream may include AVC still pictures. When set to '0', then the associated AVC video stream shall not contain AVC still pictures. HEVC_24_hour_picture_flag This 1-bit flag when set to '1' indicates that the associated HEVC video stream may contain HEVC 24-hour pictures. For the definition of an HEVC 24-hour picture, see 2.1.97. If this flag is set to '0', the associated HEVC video stream shall not contain any HEVC 24-hour picture. temporal_id_min This 3-bit field indicates the minimum value of the temporal_id syntax element in the NAL unit header, as defined in the HEVC specification, for the associated Elementary Stream. temporal_id_max This 3-bit field indicates the minimum value of the temporal_id syntax element in the NAL unit header, as defined in the HEVC specification, for the associated Elementary Stream.
2.6.98 HEVC timing and HRD descriptor
Rec. ITU-T H. 222.0/Amd.3
ISO/IEC 13818-1:201X/PDAM 3 (E)
For HEVC video streams, the HEVC timing and HRD descriptor provides timing and HRD parameters of the associated HEVC video stream.
Table X-2 HEVC timing and HRD descriptor
Syntax
No. Of bits
Mnemonic
8 8 1 6 1
uimsbf uimsbf bslbf bslbf bslbf
1 7
bslbf bslbf
32 32
uimsbf uimsbf
32
uimsbf
HEVC_timing_and_HRD_descriptor() { descriptor_tag descriptor_length hrd_management_valid_flag reserved picture_and_timing_info_present_flag if (picture_and_timing_info_present_flag == '1') { 90kHz_flag reserved if (90kHz_flag = = '0') { N K } num_units_in_tick } }
2.6.99 Semantic definition of fields in HEVC timing and HRD descriptor
hrd_management_valid_flag This 1-bit flag is only defined for use in transport streams. When the HEVC timing and HRD descriptor is associated to an HEVC video stream carried in a program stream, then the meaning of the hrd_management_valid_flag is not defined. picture_and_timing_info_present_flag This 1-bit flag when set to '1' indicates that the 90kHz_flag and parameters for accurate mapping to 90-kHz system clock are included in this descriptor. 90kHz_flag The 90kHz_flag when set to '1' indicates that the frequency of the HEVC time base is 90 kHz. N, K For an HEVC video stream the frequency of the HEVC time base is defined by the syntax element time_scale in VUI parameters, as defined in Annex E of [HEVC]. The relationship between the HEVC time_scale and the STC shall be defined by the parameters N and K in this descriptor as follows. time_scale = (N x system_clock_frequency) / K If the 90kHz_flag is set to '1', then N equals 1 and K equals 300. If the 90kHz_flag is set to '0', then the values of N and K are provided by the coded values of the N and K fields. NOTE 1 This allows mapping of time expressed in units of time_scale to 90 kHz units, as needed for the calculation of PTS and DTS timestamps, for example in decoders for HEVC access units for which no PTS or DTS is encoded in the PES header.
num_units_in_tick Coded exactly in the same way as the num_units_in_tick field in VUI parameters in Annex E of [HEVC]. The information provided by this field shall apply to the entire HEVC video stream to which the HEVC timing and HRD descriptor is associated.
11) Clause 2 Add the following after clause 2.16 as new subclause:
Rec. ITU-T H. 222.0/Amd.3
ISO/IEC 13818-1:201X/PDAM 3 (E)
2.17 Carriage of HEVC
2.17.1 Constraints for the transport of HEVC For HEVC video streams, the following constraints additionally apply: Each AVC access unit shall contain an access unit delimiter NAL Unit; NOTE 2 HEVC requires that an access unit delimiter NAL Unit, if present, is the first NAL Unit within an AVC access unit. Access unit delimiter NAL Units simplify the ability to detect the boundary between pictures.
The sequence parameter sets, picture parameter sets and adaptation parameter sets necessary for decoding an HEVC video stream shall be present within the elementary stream carrying the HEVC video sub-bitstream.
2.17.2 T-STD Extensions for single layer HEVC
When there is an HEVC stream with the syntax element max_temporal_layers_minus1 in the SPS equal to 0 in an ITUT Rec. H.222.0 | ISO/IEC 13818-1 program, the T-STD model as described in 2.4.2 is extended as illustrated in Figure X-1 and as specified below.
Figure X-1 T-STD model extensions for single layer HEVC
TB, MB, EB buffer management The following applies:
There is exactly one transport buffer TB for the received HEVC bitstream where the size TBS is fixed to 512 bytes.
There is exactly one multiplexing buffer MB for the HEVC elementary stream, where the size MBS of the multiplexing buffer MB is constrained as follows: MBS = BSmux + BSoh + 1200 MaxCPB[level] cpb_size where BSoh, packet overhead buffering, is defined as: BSoh = (1/750) seconds max{1200 MaxBR[level], 2 000 000 bit/s} and BSmux, additional multiplex buffering, is defined as: BSmux = 0.004 seconds max{1200 MaxBR[level], 2 000 000 bit/s} MaxCPB[level] and MaxBR[level] are taken from the level specification of HEVC. cpb_size is taken from the HRD parameters included in the HEVC stream.
Rx Rbx Ak(j) j tk(j)
is the transfer rate from the transport buffer TB to the multiplex buffer MB as specified below. is the transfer rate from the multiplex buffer to the elementary stream buffer as specified below. is the k-th access unit of the HEVC bitstream is an index to the HEVC access unit of the HEVC bitstream is the decoding time, measured in seconds, in the system target decoder of the k-th HEVC access unit Transfer from TB to MB is applied as follows:
Rec. ITU-T H. 222.0/Amd.3
ISO/IEC 13818-1:201X/PDAM 3 (E)
When there is no data in TB then Rx is equal to zero. Otherwise: Rx = bit_rate where bit_rate is 1.2 x BitRate[ SchedSelIdx ] of data flow into the CPB for the byte stream format and BitRate[ SchedSelIdx ] is as defined in Annex E of [HEVC] when NAL hrd_parameters() is present in the VUI parameters of the HEVC stream. NOTE 3 Annex E also specifies default values for BitRate[ SchedSelIdx] based on profile and level when NAL HRD parameters are not present in the VUI. .
Transfer from MB to EB is applied as follows:
If the HEVC_timing_and_HRD_descriptor is present with the hrd_management_valid_flag set to '1' for the HEVC stream, then the transfer of data from MB to EB shall follow the HRD defined scheme for data arrival in the CPB of elementary stream ESH as defined in Annex C [HEVC]. Otherwise, the leak method shall be used to transfer data from MB to EB as follows: Rbx = 1200 MaxBR[level] where MaxBR[level] is defined for the byte stream format in Table A.1 (Level Limits) in [HEVC] for the level of the HEVC stream. If there is PES packet payload data in MB, and buffer EB is not full, the PES packet payload is transferred from MB to EB at a rate equal to Rbx. If EB is full, data are not removed from MB. When a byte of data is transferred from MB to EB, all PES packet header bytes that are in MB and precede that byte are instantaneously removed and discarded. When there is no PES packet payload data present in MB, no data is removed from MB. All data that enters MB leaves it. All PES packet payload data bytes enter EB instantaneously upon leaving MB.
2.17.3 T-STD Extensions for layered transport of HEVC
When there is an HEVC stream with the syntax element max_temporal_layers_minus1 in the SPS greater than 0 in an ITU-T Rec. H.222.0 | ISO/IEC 13818-1 program, the T-STD model as described in 2.4.2 is extended as illustrated in Figure X-2 and as specified below.
Figure X-2 T-STD model extensions for layered transport of HEVC
TBi, MBi, EB buffer management The following applies:
Rec. ITU-T H. 222.0/Amd.3
ISO/IEC 13818-1:201X/PDAM 3 (E)
There is one transport buffer TBi for each received HEVC video sub-bitstream where the size TBS i is fixed to 512 bytes.
There is one multiplex buffer MBi for each received HEVC video sub-bitstream ES i, where the size MBS i of the multiplex buffer MBi is constrained as follows: MBSi = BSmux + BSoh + 1200 MaxCPB[level] cpb_size (measured in bytes) where BSoh, packet overhead buffering, and BSmux, additional multiplex buffering, are as specified in clause 2.15.2. MaxCPB[level] and MaxBR[level] are taken from the level specification of HEVC. cpb_size is taken from the HRD parameters included in the HEVC stream.
There is one elementary stream buffer SBi for each HEVC elementary stream ESi, where the total size EBS of all elementary stream buffers SBi is constrained as follows: EBS = cpb_size (measured in bytes)
Rxi Rbxi
is the transfer rate from the i-th transport buffer TBi to the i-th multiplex buffer MBi as specified below. is the transfer rate from the i-th multiplex buffer MB i to the i-th elementary stream buffer SBi as specified below. Transfer from TBi to MBi is applied as follows: When there is no data in TBi then Rxi is equal to zero. Otherwise: Rxi = bit_rate where bit_rate is as specified in clause 2.15.2.
Transfer from MBi to SBi is applied as follows:
If the HEVC_timing_and_HRD_descriptor is present with the hrd_management_valid_flag set to '1' for the HEVC video sub-bitstream, then the transfer of data from MB i to SBi shall follow the HRD defined scheme for data arrival in the CPB of elementary stream ESH as defined in Annex C [HEVC]. Otherwise, the leak method shall be used to transfer data from MBi to SBi as follows: Rbxi = 1200 MaxBR[level] where MaxBR[level] is defined for the byte stream format in Table A.1 (Level Limits) in [HEVC] for the level of the HEVC stream. If there is PES packet payload data in MB i, and buffer SBi is not full, the PES packet payload is transferred from MBi to SBi at a rate equal to Rbx i. If EB is full, data are not removed from MB i. When a byte of data is transferred from MBi to SBi, all PES packet header bytes that are in MB i and precede that byte are instantaneously removed and discarded. When there is no PES packet payload data present in MB i, no data is removed from MBi. All data that enters MBi leaves it. All PES packet payload data bytes enter SB i instantaneously upon leaving MBi.
At the output of the elementary stream buffer EB which consists of one sub-bitstream buffer SB i per HEVC video subbitstream ESi, access units are concatenated in ascending DTS order. j AUn(jn) jn tdn(jn)
is an index to the re-assembled access units
is the jn-th access unit of the n-th HEVC video temporal sub-bitstream is an index to the HEVC access unit of the n-th HEVC video temporal sub-bitstream is the decoding time, measured in seconds, in the system target decoder of the j n-th HEVC access unit of the n-th HEVC video temporal sub-bitstream
