[FFmpeg-devel] [PATCH] avformat/mpeg: Adjust vid probe threshold to correct mis-detection

[wm4]

On Tue, 15 Nov 2016 20:06:42 +0100
Michael Niedermayer <michael at niedermayer.cc> wrote:

> Fixes: _ij.mp3
> 
> Signed-off-by: Michael Niedermayer <michael at niedermayer.cc>
> ---
>  libavformat/mpeg.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/libavformat/mpeg.c b/libavformat/mpeg.c
> index 1c441c5..423877e 100644
> --- a/libavformat/mpeg.c
> +++ b/libavformat/mpeg.c
> @@ -112,7 +112,7 @@ static int mpegps_probe(AVProbeData *p)
>                            : AVPROBE_SCORE_EXTENSION / 2; // 1 more than .mpg
>      if ((!!vid ^ !!audio) && (audio > 4 || vid > 1) && !sys &&
>          !pspack && p->buf_size > 2048 && vid + audio > invalid) /* PES stream */
> -        return (audio > 12 || vid > 3 + 2 * invalid) ? AVPROBE_SCORE_EXTENSION + 2
> +        return (audio > 12 || vid > 6 + 2 * invalid) ? AVPROBE_SCORE_EXTENSION + 2
>                                                       : AVPROBE_SCORE_EXTENSION / 2;
>  
>      // 02-Penguin.flac has sys:0 priv1:0 pspack:0 vid:0 audio:1

Works for me (I reported this, so same test sample). But I'm not
convinced whether it's the right solution. Not opposed to applying it
either since it at least makes the bogus heuristic return better
results.

I also have a hard time believing that detecting straight mp3 or
mpeg-ps is so complex.