From: James Almer <jamrial@gmail.com> To: ffmpeg-devel@ffmpeg.org Subject: Re: [FFmpeg-devel] [PATCH] lavc/av1*: fix exporting framerate Date: Sun, 14 May 2023 16:50:06 -0300 Message-ID: <cac40f46-9059-06ed-d84f-45521db0f1f0@gmail.com> (raw) In-Reply-To: <20230514193900.10254-1-anton@khirnov.net> On 5/14/2023 4:39 PM, Anton Khirnov wrote: > * take num_ticks_per_picture_minus_1 into account, since that is a part > of the framerate computation > * stop exporting num_ticks_per_picture_minus_1 into > AVCodecContext.ticks_per_frame, as that field is used for other > purposes (in conjunction with repeat_pict, which is not used at all by > av1) > --- > libavcodec/Makefile | 2 +- > libavcodec/av1_parse.c | 14 ++++++++++++++ > libavcodec/av1_parse.h | 3 +++ > libavcodec/av1_parser.c | 13 +++++++++---- > libavcodec/av1dec.c | 16 ++++++++-------- > 5 files changed, 35 insertions(+), 13 deletions(-) > > diff --git a/libavcodec/Makefile b/libavcodec/Makefile > index 3cf4444b7e..9587e56493 100644 > --- a/libavcodec/Makefile > +++ b/libavcodec/Makefile > @@ -1143,7 +1143,7 @@ OBJS-$(CONFIG_AC3_PARSER) += aac_ac3_parser.o ac3tab.o \ > ac3_channel_layout_tab.o > OBJS-$(CONFIG_ADX_PARSER) += adx_parser.o > OBJS-$(CONFIG_AMR_PARSER) += amr_parser.o > -OBJS-$(CONFIG_AV1_PARSER) += av1_parser.o > +OBJS-$(CONFIG_AV1_PARSER) += av1_parser.o av1_parse.o > OBJS-$(CONFIG_AVS2_PARSER) += avs2.o avs2_parser.o > OBJS-$(CONFIG_AVS3_PARSER) += avs3_parser.o > OBJS-$(CONFIG_BMP_PARSER) += bmp_parser.o > diff --git a/libavcodec/av1_parse.c b/libavcodec/av1_parse.c > index 59ea0bc6e7..26b70845f2 100644 > --- a/libavcodec/av1_parse.c > +++ b/libavcodec/av1_parse.c > @@ -108,3 +108,17 @@ void ff_av1_packet_uninit(AV1Packet *pkt) > av_freep(&pkt->obus); > pkt->obus_allocated = pkt->obus_allocated_size = 0; > } > + > +AVRational ff_av1_framerate(int ticks_per_frame, int units_per_tick, > + int time_scale) > +{ > + AVRational fr = (AVRational){ 0, 1 }; > + > + if (ticks_per_frame && units_per_tick && time_scale && > + ticks_per_frame < INT_MAX / units_per_tick) { av_reduce takes int64_t as input, so you could just cast units_per_tick. > + av_reduce(&fr.den, &fr.num, units_per_tick * ticks_per_frame, > + time_scale, INT_MAX); > + } > + > + return fr; > +} > diff --git a/libavcodec/av1_parse.h b/libavcodec/av1_parse.h > index f4a5d2830e..11f9c77de4 100644 > --- a/libavcodec/av1_parse.h > +++ b/libavcodec/av1_parse.h > @@ -181,4 +181,7 @@ static inline int get_obu_bit_length(const uint8_t *buf, int size, int type) > return size; > } > > +AVRational ff_av1_framerate(int ticks_per_frame, int units_per_tick, > + int time_scale); > + > #endif /* AVCODEC_AV1_PARSE_H */ > diff --git a/libavcodec/av1_parser.c b/libavcodec/av1_parser.c > index 14dae92fe9..0d81c29e63 100644 > --- a/libavcodec/av1_parser.c > +++ b/libavcodec/av1_parser.c > @@ -21,6 +21,8 @@ > */ > > #include "libavutil/avassert.h" > + > +#include "av1_parse.h" > #include "cbs.h" > #include "cbs_av1.h" > #include "parser.h" > @@ -162,10 +164,13 @@ static int av1_parser_parse(AVCodecParserContext *ctx, > avctx->color_trc = (enum AVColorTransferCharacteristic) color->transfer_characteristics; > avctx->color_range = color->color_range ? AVCOL_RANGE_JPEG : AVCOL_RANGE_MPEG; > > - if (seq->timing_info_present_flag) { > - const AV1RawTimingInfo *timing = &seq->timing_info; > - av_reduce(&avctx->framerate.den, &avctx->framerate.num, > - timing->num_units_in_display_tick, timing->time_scale, INT_MAX); > + if (seq->timing_info_present_flag && > + seq->timing_info.num_ticks_per_picture_minus_1 < INT_MAX - 1 && > + seq->timing_info.num_units_in_display_tick < INT_MAX && > + seq->timing_info.time_scale < INT_MAX) { You could pass all uint32_t values as they are and check the return value of av_reduce(). If it's 0 (non exact result) then don't set avctx->framerate (or set it to { 0, 1 }, whatever is cleaner). There could be samples with bitstream values > INT_MAX where the reduced AVRational is still within INT_MAX for num and den. > + avctx->framerate = ff_av1_framerate(seq->timing_info.num_ticks_per_picture_minus_1 + 1, > + seq->timing_info.num_units_in_display_tick, > + seq->timing_info.time_scale); > } > > end: > diff --git a/libavcodec/av1dec.c b/libavcodec/av1dec.c > index c90c9c1a69..dae1d8e8da 100644 > --- a/libavcodec/av1dec.c > +++ b/libavcodec/av1dec.c > @@ -26,6 +26,7 @@ > #include "libavutil/pixdesc.h" > #include "libavutil/opt.h" > #include "avcodec.h" > +#include "av1_parse.h" > #include "av1dec.h" > #include "atsc_a53.h" > #include "bytestream.h" > @@ -709,14 +710,13 @@ static int set_context_with_sequence(AVCodecContext *avctx, > } > avctx->sample_aspect_ratio = (AVRational) { 1, 1 }; > > - if (seq->timing_info.num_units_in_display_tick && > - seq->timing_info.time_scale) { > - av_reduce(&avctx->framerate.den, &avctx->framerate.num, > - seq->timing_info.num_units_in_display_tick, > - seq->timing_info.time_scale, > - INT_MAX); > - if (seq->timing_info.equal_picture_interval) > - avctx->ticks_per_frame = seq->timing_info.num_ticks_per_picture_minus_1 + 1; > + if (seq->timing_info_present_flag && > + seq->timing_info.num_ticks_per_picture_minus_1 < INT_MAX - 1 && > + seq->timing_info.num_units_in_display_tick < INT_MAX && > + seq->timing_info.time_scale < INT_MAX) { > + avctx->framerate = ff_av1_framerate(seq->timing_info.num_ticks_per_picture_minus_1 + 1, > + seq->timing_info.num_units_in_display_tick, > + seq->timing_info.time_scale); > } > > return 0; _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
next prev parent reply other threads:[~2023-05-14 19:50 UTC|newest] Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top 2023-05-07 13:32 [FFmpeg-devel] [PATCH 01/13] lavu/frame: extend AVFrame.repeat_pict documentation Anton Khirnov 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 02/13] fftools/ffmpeg: fix computing video frame duration from repeat_pict Anton Khirnov 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 03/13] lavc/codec_desc: add a property for codecs that support field coding Anton Khirnov 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 04/13] lavf: use AV_CODEC_PROP_FIELDS where appropriate Anton Khirnov 2023-05-08 14:12 ` Michael Niedermayer 2023-05-09 8:37 ` [FFmpeg-devel] [PATCH v2 " Anton Khirnov 2023-05-08 14:15 ` [FFmpeg-devel] [PATCH " Michael Niedermayer 2023-05-09 8:44 ` Anton Khirnov 2023-05-15 18:59 ` Michael Niedermayer 2023-05-15 20:44 ` Anton Khirnov 2023-05-16 17:41 ` Michael Niedermayer 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 05/13] lavc/av1*: fix exporting framerate Anton Khirnov 2023-05-10 11:52 ` James Almer 2023-05-14 19:39 ` [FFmpeg-devel] [PATCH] " Anton Khirnov 2023-05-14 19:50 ` James Almer [this message] 2023-05-15 8:22 ` Anton Khirnov 2023-05-15 11:41 ` James Almer 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 06/13] lavc/libdav1d: " Anton Khirnov 2023-05-15 8:22 ` [FFmpeg-devel] [PATCH] " Anton Khirnov 2023-05-15 11:47 ` James Almer 2023-05-15 12:22 ` Anton Khirnov 2023-05-15 12:41 ` James Almer 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 07/13] lavc/ratecontrol: use AVCodecContext.framerate when available Anton Khirnov 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 08/13] lavc/msmpeg4enc: " Anton Khirnov 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 09/13] libaomenc: " Anton Khirnov 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 10/13] lavc/libkvazaar, libopenh264enc: drop redundant checks Anton Khirnov 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 11/13] lavc/libvpxenc: send frame durations to the encoder Anton Khirnov 2023-05-09 1:18 ` James Zern 2023-05-09 9:09 ` [FFmpeg-devel] [PATCH v2 " Anton Khirnov 2023-05-09 18:17 ` James Zern 2023-05-10 6:34 ` Anton Khirnov 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 12/13] lavc: deprecate AVCodecContext.ticks_per_frame Anton Khirnov 2023-05-07 13:32 ` [FFmpeg-devel] [PATCH 13/13] fftools/ffmpeg: stop using deprecated ticks_per_frame Anton Khirnov 2023-05-07 16:59 ` [FFmpeg-devel] [PATCH 01/13] lavu/frame: extend AVFrame.repeat_pict documentation Kieran Kunhya 2023-05-07 18:01 ` Anton Khirnov
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=cac40f46-9059-06ed-d84f-45521db0f1f0@gmail.com \ --to=jamrial@gmail.com \ --cc=ffmpeg-devel@ffmpeg.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: link
Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel This inbox may be cloned and mirrored by anyone: git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git # If you have public-inbox 1.1+ installed, you may # initialize and index your mirror using the following commands: public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \ ffmpegdev@gitmailbox.com public-inbox-index ffmpegdev Example config snippet for mirrors. AGPL code for this site: git clone https://public-inbox.org/public-inbox.git