Re: [FFmpeg-devel] [PATCH 3/3] avfilter/yadif_common: fix timestamps with very small timebases

[Michael Niedermayer <michael@niedermayer.cc>]

[-- Attachment #1.1: Type: text/plain, Size: 3033 bytes --]

On Sun, Jan 28, 2024 at 04:01:36AM +0100, Marton Balint wrote:
> Yadif filter assumed that the output timebase is always half of the input
> timebase. This is not true if halving the input time base is not representable
> as an AVRational causing the output timestamps to be invalidly scaled in such a
> case.
> 
> So let's use av_reduce instead of av_mul_q when calculating the output time
> base and if the conversion is inexact then let's fall back to the original
> timebase which probably makes more parctical sense than using x/INT_MAX.
> 
> Fixes invalidly scaled pts_time values in this command line:
> ffmpeg -f lavfi -i testsrc -vf settb=tb=1/2000000000,yadif,showinfo -f null none
> 
> Signed-off-by: Marton Balint <cus@passwd.hu>
> ---
>  libavfilter/yadif.h        |  2 ++
>  libavfilter/yadif_common.c | 20 ++++++++++++++------
>  2 files changed, 16 insertions(+), 6 deletions(-)
> 
> diff --git a/libavfilter/yadif.h b/libavfilter/yadif.h
> index 2c4fed62d2..c144568242 100644
> --- a/libavfilter/yadif.h
> +++ b/libavfilter/yadif.h
> @@ -86,6 +86,8 @@ typedef struct YADIFContext {
>       * the first field.
>       */
>      int current_field;  ///< YADIFCurrentField
> +
> +    int pts_divisor;
>  } YADIFContext;
>  
>  void ff_yadif_init_x86(YADIFContext *yadif);
> diff --git a/libavfilter/yadif_common.c b/libavfilter/yadif_common.c
> index 933372529e..90a5cffc2d 100644
> --- a/libavfilter/yadif_common.c
> +++ b/libavfilter/yadif_common.c
> @@ -61,7 +61,7 @@ FF_ENABLE_DEPRECATION_WARNINGS
>          int64_t next_pts = yadif->next->pts;
>  
>          if (next_pts != AV_NOPTS_VALUE && cur_pts != AV_NOPTS_VALUE) {
> -            yadif->out->pts = cur_pts + next_pts;
> +            yadif->out->pts = (cur_pts + next_pts) / yadif->pts_divisor;
>          } else {
>              yadif->out->pts = AV_NOPTS_VALUE;
>          }
> @@ -150,8 +150,8 @@ int ff_yadif_filter_frame(AVFilterLink *link, AVFrame *frame)
>          ff_ccfifo_inject(&yadif->cc_fifo, yadif->out);
>          av_frame_free(&yadif->prev);
>          if (yadif->out->pts != AV_NOPTS_VALUE)
> -            yadif->out->pts *= 2;
> -        yadif->out->duration *= 2;
> +            yadif->out->pts *= 2 / yadif->pts_divisor;
> +        yadif->out->duration *= 2 / yadif->pts_divisor;
>          return ff_filter_frame(ctx->outputs[0], yadif->out);
>      }
>  
> @@ -168,9 +168,9 @@ FF_ENABLE_DEPRECATION_WARNINGS
>      yadif->out->flags &= ~AV_FRAME_FLAG_INTERLACED;
>  
>      if (yadif->out->pts != AV_NOPTS_VALUE)
> -        yadif->out->pts *= 2;
> +        yadif->out->pts *= 2 / yadif->pts_divisor;
>      if (!(yadif->mode & 1))
> -        yadif->out->duration *= 2;
> +        yadif->out->duration *= 2 / yadif->pts_divisor;

you can use >> instead of division for all above

otherwise should be ok

thx


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

Avoid a single point of failure, be that a person or equipment.

[-- 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".