Michael Niedermayer <[EMAIL PROTECTED]> added the comment: On Sun, Feb 17, 2008 at 11:38:39AM -0000, wg wrote: > > wg <[EMAIL PROTECTED]> added the comment: > > If you care so much about AVPacket.duration, then why is it _not used_ in > ffmpeg.c to compute AVInputStream.next_pts? > IOW, duration is currently _never_ used to decide whether there is > a discontinuity!
yes, ffmpeg.c is a buggy mess, i just 2 days ago tried to fix a bug and i still havnt been able to commit it as it uncovers tons of other bugs. Which i have to fix first. (the actual bug is ffmpeg looses timestamps in the audio path, but fixing this results in tons of bugs turing up as timestamps (which where unused and wrong) become used > > I don't want to hack a new patch that sets AVInputStream.next_pts based > on AVPacket.duration _if_ that should ever be non-zero (currently not the case > for all the subtitle demuxers) and set it to AV_NOPTS_VALUE otherwise. To do > this just in the subtitle case, now _that_ would look silly. Yes it also would break various things ... What i meant really is a pkt.duration check where the ist->next_pts != AV_NOPTS_VALUE check is. An alternative might be a flag for the container indicating that it doesnt allow discontinuities. This still would be more correct than a check for subtitles. (a example where it might make a difference are dvd subtitles where discontinuities can happen and having the detection disabled might have some nasty effects) [...] ______________________________________________________ FFmpeg issue tracker <[EMAIL PROTECTED]> <https://roundup.mplayerhq.hu/roundup/ffmpeg/issue356> ______________________________________________________
