- NUT-devel - lists.mplayerhq.hu

[nut]: r696 - docs/nut4cc.txt
by michael 29 Jul '26

29 Jul '26

Author: michael Date: Wed Jul 29 00:03:21 2026 New Revision: 696 Log: Add P010/P016/P210/P216 semi-planar YUV FourCCs Microsoft defines these as a plane of Y samples followed by a plane of interleaved Cb,Cr sample pairs, all 16bit words, the 10bit variants leaving their 6 least significant bits 0: https://learn.microsoft.com/en-us/windows/win32/medfound/10-bit-and-16-bit-… DirectShow capture devices advertise P010, so nut needs a name for it before it can be stored. Modified: docs/nut4cc.txt Modified: docs/nut4cc.txt ============================================================================== --- docs/nut4cc.txt Fri Nov 6 23:22:24 2020 (r695) +++ docs/nut4cc.txt Wed Jul 29 00:03:21 2026 (r696) @@ -143,6 +143,18 @@ PAL[2] 2bpp indexes (ordered from PAL[4] 4bpp indexes (ordered from the msb to the lsb in each byte), followed by a up to 16 entry table of 8:8:8:8 BGRA entries [NOT in AVI] PAL[8] 8bpp indexes (ordered from the msb to the lsb in each byte), followed by a up to 256 entry table of 8:8:8:8 BGRA entries [NOT in AVI] +Semi-planar YUV: a plane of Y samples followed by a plane of interleaved +Cb,Cr sample pairs, each sample a 16bit word. FourCC and sample layout are +Microsoft's, see +https://learn.microsoft.com/en-us/windows/win32/medfound/10-bit-and-16-bit-yuv-video-formats +but lines and planes follow each other without the padding their surface +definition permits. + +P010 Semi-planar YUV 4:2:0, 15bpp, 6 unused LSB per 16bit value, little-endian [NOT in AVI] +P016 Semi-planar YUV 4:2:0, 24bpp, little-endian [NOT in AVI] +P210 Semi-planar YUV 4:2:2, 20bpp, 6 unused LSB per 16bit value, little-endian [NOT in AVI] +P216 Semi-planar YUV 4:2:2, 32bpp, little-endian [NOT in AVI] + Since many exotic planar YUVA pixel formats are not contemplated by the AVI/QuickTime FourCC lists, the following scheme is adopted for representing them.

1 0

[PATCH] Add st_sd_displaymatrix for storing the stream display matrix
by Matthias Neugebauer 13 Dec '20

13 Dec '20

See FFmpeg patch: https://patchwork.ffmpeg.org/project/ffmpeg/patch/F7848CEE-46C1-4CAF-9191-A… --- docs/nut.txt | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/docs/nut.txt b/docs/nut.txt index 3641c01..ea49eba 100644 --- a/docs/nut.txt +++ b/docs/nut.txt @@ -1,5 +1,5 @@ ================================== -NUT Open Container Format 20201104 +NUT Open Container Format 20201120 ================================== @@ -1045,6 +1045,12 @@ info packet types framerates in the stream). Note, this value is just a guess! For example, if the time base is 1/90000 and all frames have either approximately 3600 or 1800 timer ticks, then r_frame_rate will be 50/1. + "st_sd_displaymatrix" + Display matrix of the stream. + This stores the 3x3 display matrix of the stream as nine unsigned + integers seperated by colons. The display matrix is a side data of a + video stream and necessary for displaying the stream correctly, e.g. + rotating it. Note: Nonstandard fields should be prefixed by "X-". value -- 2.29.2

2 1

[nut]: r695 - docs/nut.txt
by michael 06 Nov '20

06 Nov '20

Author: michael Date: Fri Nov 6 23:22:24 2020 New Revision: 695 Log: Add r_frame_rate, which has been stored without X- prefix for a long time The description is taken from the API documentation of libavformat Modified: docs/nut.txt Modified: docs/nut.txt ============================================================================== --- docs/nut.txt Sun Feb 2 12:47:56 2020 (r694) +++ docs/nut.txt Fri Nov 6 23:22:24 2020 (r695) @@ -1,5 +1,5 @@ ================================== -NUT Open Container Format 20200130 +NUT Open Container Format 20201104 ================================== @@ -1038,6 +1038,13 @@ info packet types instead? is there some other unique way to identify fonts? This stream uses the specified font, if it is unavailable presentation might be less pretty. + "r_frame_rate" + Real base framerate of the stream. + This is the lowest framerate with which all timestamps can be + represented accurately (it is the least common multiple of all + framerates in the stream). Note, this value is just a guess! + For example, if the time base is 1/90000 and all frames have either + approximately 3600 or 1800 timer ticks, then r_frame_rate will be 50/1. Note: Nonstandard fields should be prefixed by "X-". value

1 0

[PATCH 1/7] [RFC] Revert "avcodec/adpcm_swf: support decoding multiple fixed-sized blocks at once"
by Michael Niedermayer 06 Nov '20

06 Nov '20

The reverted code split at block_align boundaries, but there was already code which splits at a hardcoded 4096 sample boundary. reverting this seemed like the easiest fix but this is a RFC in case another solution is preferred Fixes: out of array write Fixes: 26821/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_ADPCM_SWF_fuzzer-5764465137811456 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg This reverts commit e9dd73d30d09043446ac6dd7b8ad31e557873852. --- libavcodec/adpcm.c | 15 ++------------- 1 file changed, 2 insertions(+), 13 deletions(-) diff --git a/libavcodec/adpcm.c b/libavcodec/adpcm.c index 701b125c47..d018c1f91b 100644 --- a/libavcodec/adpcm.c +++ b/libavcodec/adpcm.c @@ -880,7 +880,7 @@ static int get_nb_samples(AVCodecContext *avctx, GetByteContext *gb, } case AV_CODEC_ID_ADPCM_SWF: { - int buf_bits = (avctx->block_align ? avctx->block_align : buf_size) * 8 - 2; + int buf_bits = buf_size * 8 - 2; int nbits = (bytestream2_get_byte(gb) >> 6) + 2; int block_hdr_size = 22 * ch; int block_size = block_hdr_size + nbits * ch * 4095; @@ -889,9 +889,6 @@ static int get_nb_samples(AVCodecContext *avctx, GetByteContext *gb, nb_samples = nblocks * 4096; if (bits_left >= block_hdr_size) nb_samples += 1 + (bits_left - block_hdr_size) / (nbits * ch); - - if (avctx->block_align) - nb_samples *= buf_size / avctx->block_align; break; } case AV_CODEC_ID_ADPCM_THP: @@ -1770,17 +1767,9 @@ static int adpcm_decode_frame(AVCodecContext *avctx, void *data, } break; case AV_CODEC_ID_ADPCM_SWF: - { - const int nb_blocks = avctx->block_align ? avpkt->size / avctx->block_align : 1; - const int block_size = avctx->block_align ? avctx->block_align : avpkt->size; - - for (int block = 0; block < nb_blocks; block++) { - adpcm_swf_decode(avctx, buf + block * block_size, block_size, samples); - samples += nb_samples / nb_blocks; - } + adpcm_swf_decode(avctx, buf, buf_size, samples); bytestream2_seek(&gb, 0, SEEK_END); break; - } case AV_CODEC_ID_ADPCM_YAMAHA: for (n = nb_samples >> (1 - st); n > 0; n--) { int v = bytestream2_get_byteu(&gb); -- 2.17.1

1 7

[PATCH] Add r_frame_rate, which has been stored without X- prefix for a long time
by Michael Niedermayer 04 Nov '20

04 Nov '20

The description is taken from the API documentation of libavformat --- docs/nut.txt | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/docs/nut.txt b/docs/nut.txt index 2cccb51..3641c01 100644 --- a/docs/nut.txt +++ b/docs/nut.txt @@ -1,5 +1,5 @@ ================================== -NUT Open Container Format 20200130 +NUT Open Container Format 20201104 ================================== @@ -1038,6 +1038,13 @@ info packet types instead? is there some other unique way to identify fonts? This stream uses the specified font, if it is unavailable presentation might be less pretty. + "r_frame_rate" + Real base framerate of the stream. + This is the lowest framerate with which all timestamps can be + represented accurately (it is the least common multiple of all + framerates in the stream). Note, this value is just a guess! + For example, if the time base is 1/90000 and all frames have either + approximately 3600 or 1800 timer ticks, then r_frame_rate will be 50/1. Note: Nonstandard fields should be prefixed by "X-". value -- 2.17.1

1 0

[nut]: r694 - docs/nut.txt
by michael 02 Feb '20

02 Feb '20

Author: michael Date: Sun Feb 2 12:47:56 2020 New Revision: 694 Log: docs/nut: Mention nut4cc.txt Modified: docs/nut.txt Modified: docs/nut.txt ============================================================================== --- docs/nut.txt Sat Jun 8 17:23:25 2019 (r693) +++ docs/nut.txt Sun Feb 2 12:47:56 2020 (r694) @@ -1,5 +1,5 @@ ================================== -NUT Open Container Format 20131223 +NUT Open Container Format 20200130 ================================== @@ -557,6 +557,7 @@ fourcc (vb) FourCC in AVI then the same FourCC MUST be used here. For raw video and audio, the used sample format (signed/unsigned, 8/16/32bit, big/little endian, rgb/yuv) is indicated by the fourcc. + See nut4cc.txt for a list as well as a place to add new entries to. time_base_num (v) / time_base_denom (v) = time_base the length of a timer tick in seconds, this MUST be equal to the 1/fps

1 0

[PATCH] avutil/log: Add av_log_once() for printing a message just once with a high log level
by Michael Niedermayer 31 Jan '20

31 Jan '20

Compared to ad-hoc if(printed) ... code this allows the user to disable it by adjusting the log level TODO: APIChanges & version bump Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> --- libavutil/log.c | 9 +++++++++ libavutil/log.h | 21 +++++++++++++++++++++ 2 files changed, 30 insertions(+) diff --git a/libavutil/log.c b/libavutil/log.c index 0a7b169bc0..57301d42d6 100644 --- a/libavutil/log.c +++ b/libavutil/log.c @@ -368,6 +368,15 @@ void av_log(void* avcl, int level, const char *fmt, ...) va_end(vl); } +void av_log_once(void* avcl, int initial_level, int subsequent_level, int *state, const char *fmt, ...) +{ + va_list vl; + va_start(vl, fmt); + av_vlog(avcl, *state ? subsequent_level : initial_level, fmt, vl); + va_end(vl); + *state = 1; +} + void av_vlog(void* avcl, int level, const char *fmt, va_list vl) { AVClass* avc = avcl ? *(AVClass **) avcl : NULL; diff --git a/libavutil/log.h b/libavutil/log.h index d9554e609d..9c14188a9c 100644 --- a/libavutil/log.h +++ b/libavutil/log.h @@ -233,6 +233,27 @@ typedef struct AVClass { */ void av_log(void *avcl, int level, const char *fmt, ...) av_printf_format(3, 4); +/** + * Send the specified message to the log once with the initial_level and then with + * the subsequent_level. By default, all logging messages are sent to + * stderr. This behavior can be altered by setting a different logging callback + * function. + * @see av_log + * + * @param avcl A pointer to an arbitrary struct of which the first field is a + * pointer to an AVClass struct or NULL if general log. + * @param initial_level importance level of the message expressed using a @ref + * lavu_log_constants "Logging Constant" for the first occurance. + * @param subsequent_level importance level of the message expressed using a @ref + * lavu_log_constants "Logging Constant" after the first occurance. + * @param fmt The format string (printf-compatible) that specifies how + * subsequent arguments are converted to output. + * @param state a variable to keep trak of if a message has already been printed + * this must be initialized to 0 before the first use. The same state + * must not be accessed by 2 Threads simultaneously. + */ +void av_log_once(void* avcl, int initial_level, int subsequent_level, int *state, const char *fmt, ...) av_printf_format(5, 6); + /** * Send the specified message to the log if the level is less than or equal -- 2.17.1

1 0

[PATCH] docs/nut: Mention nut4cc.txt
by Michael Niedermayer 30 Jan '20

30 Jan '20

--- docs/nut.txt | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/docs/nut.txt b/docs/nut.txt index f88c9b2..2cccb51 100644 --- a/docs/nut.txt +++ b/docs/nut.txt @@ -1,5 +1,5 @@ ================================== -NUT Open Container Format 20131223 +NUT Open Container Format 20200130 ================================== @@ -557,6 +557,7 @@ fourcc (vb) FourCC in AVI then the same FourCC MUST be used here. For raw video and audio, the used sample format (signed/unsigned, 8/16/32bit, big/little endian, rgb/yuv) is indicated by the fourcc. + See nut4cc.txt for a list as well as a place to add new entries to. time_base_num (v) / time_base_denom (v) = time_base the length of a timer tick in seconds, this MUST be equal to the 1/fps -- 2.17.1

2 1

[nut]: r693 - docs/nut4cc.txt
by michael 08 Jun '19

08 Jun '19

Author: michael Date: Sat Jun 8 17:23:25 2019 New Revision: 693 Log: Add CPiA Author is: Stephan Hilb <stephan(a)ecshi.net> Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> Modified: docs/nut4cc.txt Modified: docs/nut4cc.txt ============================================================================== --- docs/nut4cc.txt Thu Mar 30 13:43:23 2017 (r692) +++ docs/nut4cc.txt Sat Jun 8 17:23:25 2019 (r693) @@ -13,6 +13,7 @@ us so we can add them to the list. 3IV1 non compliant MPEG4 generated by old 3ivx ASV1 Asus Video ASV2 Asus Video 2 +CPiA CPiA video format CVID Cinepak CYUV Creative YUV DIVX non compliant MPEG4 generated by old divx

1 0

[PATCH] docs/nut4cc: add CPiA video format
by Stephan Hilb 07 Jun '19

07 Jun '19

--- docs/nut4cc.txt | 1 + 1 file changed, 1 insertion(+) diff --git a/docs/nut4cc.txt b/docs/nut4cc.txt index 7c70099..f411a7d 100644 --- a/docs/nut4cc.txt +++ b/docs/nut4cc.txt @@ -13,6 +13,7 @@ us so we can add them to the list. 3IV1 non compliant MPEG4 generated by old 3ivx ASV1 Asus Video ASV2 Asus Video 2 +CPiA CPiA video format CVID Cinepak CYUV Creative YUV DIVX non compliant MPEG4 generated by old divx -- 2.21.0

1 0