[libav-devel] [PATCH] matroskadec: don't warn about unknown spherical medata when none is present

James Almer jamrial at gmail.com
Thu Nov 2 21:32:49 CET 2017


On 11/2/2017 5:12 PM, Sean McGovern wrote:
> Hi James,
> 
> On Nov 2, 2017 10:03, "James Almer" <jamrial at gmail.com> wrote:
> 
> track->video.projection.type is 0 by default, and is the value set by the
> demuxer for files without the element.
> 
> Signed-off-by: James Almer <jamrial at gmail.com>
> ---
>  libavformat/matroskadec.c | 3 ---
>  1 file changed, 3 deletions(-)
> 
> diff --git a/libavformat/matroskadec.c b/libavformat/matroskadec.c
> index c6e1a190a8..5ed03bb642 100644
> --- a/libavformat/matroskadec.c
> +++ b/libavformat/matroskadec.c
> @@ -1659,9 +1659,6 @@ static int mkv_parse_video_projection(AVStream *st,
> const MatroskaTrack *track)
>          }
>          break;
>      default:
> -        av_log(NULL, AV_LOG_WARNING,
> -               "Unknown spherical metadata type %"PRIu64"\n",
> -               track->video.projection.type);
>          return 0;
>      }
> 
> --
> 2.14.2
> 
> _______________________________________________
> libav-devel mailing list
> libav-devel at libav.org
> https://lists.libav.org/mailman/listinfo/libav-devel
> 
> 
> Errrrr... I'm not sure this is a better than what I had (with which I agree
> on your review point). Isn't this log message potentially useful for
> corrupted streams?

That's a good reason to add a "do nothing" case for
MATROSKA_VIDEO_PROJECTION_TYPE_RECTANGULAR (aka, none), which is the
default value that the demuxer will fill for every single mkv file it
parses, and keep keep the warning for default:.

> 
> Also please note that the sample from BZ #1055 currently registers as
> projection type 15 which maps to the _NB. Pretty strange for a sample
> hailing from 2008. I have a feeling the real bug is elsewhere...

Not projection type, stereomode type 15. And in that case then the file
is corrupt, and it should perhaps be handled in ff_mkv_stereo3d_conv()
or similar.

> 
> -- Sean McGovern
> _______________________________________________
> libav-devel mailing list
> libav-devel at libav.org
> https://lists.libav.org/mailman/listinfo/libav-devel
> 



More information about the libav-devel mailing list