[FFmpeg-devel] About guess_correct_pts / AVFrame.best_effort_timestamp

Ronald S. Bultje rsbultje
Thu Feb 17 18:27:30 CET 2011


I originally wanted to respond more elaborately, but I'm not longer
motivated to do so.

The timestamps in FFmpeg are a mess. I'll some day look into fixing them.

On Wed, Feb 16, 2011 at 3:51 PM, Nicolas George
<nicolas.george at normalesup.org> wrote:
> * lavf and lavc can be used together, of course, but they also can be used
> ?separately: lavc can be used to decode packets coming from a non-lavc, and
> ?lavf can be used to feed non-lavc decoders.
> I hope I have summarized most of the aspects of the problem.
> Unfortunately, I think that the last point makes your (Ronald) premise
> impossible. Imagine a codec with B-frames in AVI: the format has only (more
> or less) DTS. Using a parser, lavf can derive PTS. But we also want lavc to
> work with a dumb parser that does not have a parser, and thus provides only
> DTS.

I disagree here. If other frameworks, SDKs or wrapper-style thingies
want to interoperate with FFmpeg, they have to fullfill our API
requirements. We can simply require that PTS and DTS be set, and
provide the API for doing so in lavf, and have it be applied
automatically when calling av_read_frame().

But like I said, my motivation for this subject is gone. The patch is
rejected, I've explained why, and that's all I'll say about it.


More information about the ffmpeg-devel mailing list