[FFmpeg-devel] [PATCH] Generic part of frame multithreading

Uoti Urpala uoti.urpala
Wed Aug 20 21:23:31 CEST 2008

On Wed, 2008-08-20 at 21:00 +0200, Michael Niedermayer wrote:
> > mplayer counts frame delay using avctx->has_b_frames, which works for  
> > h264 and doesn't work with this. ffplay counts delay by adding pts to  
> > frames in get_buffer, which works with this but might not work with  
> > h264, because of the frame num gap code. I thought about making avctx- 
> >  >delay accurate for decoding too, which should solve all this.
> mplayer is violating lavc API ...
> IIRC ive rejected the use of has_b_frames back then when uoti suggested it

You remember things backwards. The preliminary version of -correct-pts
used buffer callbacks to count the number of buffered frames, and it was
you who suggested using has_b_frames instead.

> from your mail i assume that this design was commited anyway.

Yes the current MPlayer code does use has_b_frames to get the number of
buffered frames (actually it is only used as an upper bound since it
doesn't reset after seeks etc). This has also been mentioned in several
mails since, so it shouldn't come as a surprise (will you forget it
again in a few months if I don't change it for multithreading support?).

More information about the ffmpeg-devel mailing list