Re: [FFmpeg-devel] [PATCH] avformat/demux: in case of no timestamps, calculate dts based on previous dts and duration

From: Michael Niedermayer <michael@niedermayer.cc>
To: FFmpeg development discussions and patches <ffmpeg-devel@ffmpeg.org>
Subject: Re: [FFmpeg-devel] [PATCH] avformat/demux: in case of no timestamps, calculate dts based on previous dts and duration
Date: Fri, 5 May 2023 22:25:34 +0200
Message-ID: <20230505202534.GJ1391451@pb2> (raw)
In-Reply-To: <20230505113123.28404-1-jeebjp@gmail.com>

[-- Attachment #1.1: Type: text/plain, Size: 2816 bytes --]

On Fri, May 05, 2023 at 02:31:23PM +0300, Jan Ekström wrote:
> From: Martin Machalek <martin.machalek@24i.com>
> 
> Adds DTS to packets which would otherwise break tbr calculation
> within libavformat, as well as packet rate calculation in ffmpeg.
> 
> An example of such input would be an MPEG-TS stream where a whole
> GOP is put into a single PES packet.
> 
> Co-authored-by: Jan Ekström <jan.ekstrom@24i.com>
> 
> Signed-off-by: Jan Ekström <jan.ekstrom@24i.com>
> ---
>  libavformat/demux.c | 5 +++++
>  1 file changed, 5 insertions(+)
> 
> diff --git a/libavformat/demux.c b/libavformat/demux.c
> index 06de71cd6e..9a4ab28679 100644
> --- a/libavformat/demux.c
> +++ b/libavformat/demux.c
> @@ -1098,6 +1098,11 @@ static void compute_pkt_fields(AVFormatContext *s, AVStream *st,
>      if (!onein_oneout)
>          // This should happen on the first packet
>          update_initial_timestamps(s, pkt->stream_index, pkt->dts, pkt->pts, pkt);
> +
> +    if (pkt->dts == AV_NOPTS_VALUE && !is_relative(sti->cur_dts) &&
> +        sti->cur_dts != AV_NOPTS_VALUE && pkt->duration > 0)
> +        pkt->dts = sti->cur_dts + pkt->duration;
> +
>      if (pkt->dts > sti->cur_dts)
>          sti->cur_dts = pkt->dts;

This produces invalid timestamps

example:

 [mpegts  read_frame_internal stream=2, pts=310689000, dts=310687200, size=11356, duration=1800, flags=0
 [mpegts  read_frame_internal stream=2, pts=310719600, dts=310689000, size=74100, duration=1800, flags=0
 [mpegts  read_frame_internal stream=2, pts=310721400, dts=310690800, size=36957, duration=1800, flags=0
-[mpegts  read_frame_internal stream=2, pts=310332330, dts=NOPTS, size=49072, duration=3600, flags=0
-[mpegts  read_frame_internal stream=2, pts=310348800, dts=NOPTS, size=16516, duration=3600, flags=0
-[mpegts  read_frame_internal stream=2, pts=310356000, dts=NOPTS, size=16876, duration=3600, flags=0
+[mpegts  read_frame_internal stream=2, pts=310332330, dts=310335930, size=49072, duration=3600, flags=0
+[mpegts  read_frame_internal stream=2, pts=310348800, dts=310339530, size=16516, duration=3600, flags=0
+[mpegts  read_frame_internal stream=2, pts=310356000, dts=310343130, size=16876, duration=3600, flags=0
 [mpegts  read_frame_internal stream=2, pts=310366800, dts=310332330, size=51652, duration=3600, flags=0
 [mpegts  read_frame_internal stream=2, pts=310363200, dts=310348800, size=15588, duration=3600, flags=0
 [mpegts  read_frame_internal stream=2, pts=310370400, dts=310356000, size=21476, duration=3600, flags=0

Its not possible for a decode timestamp to be after its presentation timestamp

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Everything should be made as simple as possible, but not simpler.
-- Albert Einstein

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

[-- Attachment #2: Type: text/plain, Size: 251 bytes --]

_______________________________________________
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".