H.264 VUI and HRD parameters

Video usability information, or VUI, sits at the end of the sequence parameter set. It does not change how pictures are decoded, but it holds the aspect ratio, colour, timing and buffering details a player needs, including the only frame rate an Annex B file can carry.

VUI parameters

Name	Coding	Condition	Description
`aspect_ratio_info_present_flag`	`u(1)`
`aspect_ratio_idc`	`u(8)`	flag is 1	see aspect ratio
`sar_width`, `sar_height`	`u(16)` × 2	`aspect_ratio_idc` is 255
`overscan_info_present_flag`	`u(1)`
`overscan_appropriate_flag`	`u(1)`	flag is 1	1 means the picture is suitable for overscan display
`video_signal_type_present_flag`	`u(1)`
`video_format`	`u(3)`	flag is 1	see video format
`video_full_range_flag`	`u(1)`	flag is 1	1 means full-range sample values
`colour_description_present_flag`	`u(1)`	flag is 1
`colour_primaries`, `transfer_characteristics`, `matrix_coefficients`	`u(8)` × 3	flag is 1	see colour codes
`chroma_loc_info_present_flag`	`u(1)`
`chroma_sample_loc_type_top_field`, `…bottom_field`	`ue(v)` × 2	flag is 1	0 to 5, the chroma sample position
`timing_info_present_flag`	`u(1)`
`num_units_in_tick`, `time_scale`	`u(32)` × 2	flag is 1	see timing
`fixed_frame_rate_flag`	`u(1)`	flag is 1
`nal_hrd_parameters_present_flag`	`u(1)`
HRD parameters		flag is 1	for the whole NAL stream
`vcl_hrd_parameters_present_flag`	`u(1)`
HRD parameters		flag is 1	for video data only
`low_delay_hrd_flag`	`u(1)`	either HRD flag is 1
`pic_struct_present_flag`	`u(1)`		1 means picture timing SEI carries `pic_struct`
`bitstream_restriction_flag`	`u(1)`
`motion_vectors_over_pic_boundaries_flag`	`u(1)`	flag is 1
`max_bytes_per_pic_denom`	`ue(v)`	flag is 1	0 to 16
`max_bits_per_mb_denom`	`ue(v)`	flag is 1	0 to 16
`log2_max_mv_length_horizontal`	`ue(v)`	flag is 1	0 to 16
`log2_max_mv_length_vertical`	`ue(v)`	flag is 1	0 to 16
`max_num_reorder_frames`	`ue(v)`	flag is 1	0 to 16
`max_dec_frame_buffering`	`ue(v)`	flag is 1	0 to 16

Aspect ratio

`aspect_ratio_idc`	Sample aspect ratio
0	unspecified
1	1:1
2	12:11
3	10:11
4	16:11
5	40:33
6	24:11
7	20:11
8	32:11
9	80:33
10	18:11
11	15:11
12	64:33
13	160:99
14	4:3
15	3:2
16	2:1
17–254	reserved
255	Extended_SAR: `sar_width`:`sar_height`

Video format

`video_format`	Source
0	component
1	PAL
2	NTSC
3	SECAM
4	MAC
5	unspecified
6, 7	reserved

Colour codes

These are the code points that H.264, H.265 and ITU-T H.273 share, as listed in FFmpeg’s pixel format header.

Code	colour_primaries
0	reserved
1	BT.709
2	unspecified
3	reserved
4	BT.470 System M
5	BT.470 System B, G; BT.601 625-line
6	SMPTE 170M; BT.601 525-line
7	SMPTE 240M
8	film, Illuminant C
9	BT.2020
10	SMPTE ST 428-1 (CIE 1931 XYZ)
11	SMPTE RP 431-2 (DCI P3)
12	SMPTE EG 432-1 (P3 D65)
22	EBU Tech. 3213-E

Code	transfer_characteristics
0	reserved
1	BT.709
2	unspecified
3	reserved
4	gamma 2.2 (BT.470 System M)
5	gamma 2.8 (BT.470 System B, G)
6	SMPTE 170M; BT.601
7	SMPTE 240M
8	linear
9	logarithmic, 100:1 range
10	logarithmic, 100 × √10 : 1 range
11	IEC 61966-2-4
12	BT.1361 extended colour gamut
13	IEC 61966-2-1 (sRGB)
14	BT.2020, 10-bit
15	BT.2020, 12-bit
16	SMPTE ST 2084 (PQ)
17	SMPTE ST 428-1
18	ARIB STD-B67 (HLG)

Code	matrix_coefficients
0	identity (RGB, or GBR order)
1	BT.709
2	unspecified
3	reserved
4	FCC
5	BT.470 System B, G; BT.601 625-line
6	SMPTE 170M; BT.601 525-line
7	SMPTE 240M
8	YCgCo
9	BT.2020 non-constant luminance
10	BT.2020 constant luminance
11	SMPTE ST 2085
12	chromaticity-derived, non-constant luminance
13	chromaticity-derived, constant luminance
14	BT.2100 ICtCp

Timing

One clock tick lasts num_units_in_tick ÷ time_scale seconds, and one field lasts a tick, so with fixed_frame_rate_flag set the frame rate is usually time_scale ÷ (2 × num_units_in_tick).

HRD parameters

Hypothetical reference decoder parameters describe the coded picture buffer, up to 32 sets.

Name	Coding	Description
`cpb_cnt_minus1`	`ue(v)`	0 to 31
`bit_rate_scale`	`u(4)`	scale of the bit rate values
`cpb_size_scale`	`u(4)`	scale of the size values
`bit_rate_value_minus1[i]`	`ue(v)`	repeated `cpb_cnt_minus1` + 1 times, with the next two
`cpb_size_value_minus1[i]`	`ue(v)`
`cbr_flag[i]`	`u(1)`	1 for constant bit rate
`initial_cpb_removal_delay_length_minus1`	`u(5)`
`cpb_removal_delay_length_minus1`	`u(5)`
`dpb_output_delay_length_minus1`	`u(5)`
`time_offset_length`	`u(5)`	width of the time offset in picture timing SEI