Cheers - my diy attempts to deal with the problem there didn't quite seem to work, but they again I'm not a developer, so will hope that someone who is can come up with something. I'll raise a report in the morning and see if it gets anywhere.
As a "sticking plaster" I have used the fact that code in ffmpeg.c throws a warning if two frames have identical DTS fields (as a result in this case of having duplicate frames). By picking up on this warning, I can at least mitigate the effects of these duplicates, even though I have not actually avoided their impact. On Sat, Apr 2, 2016 at 9:16 PM, Moritz Barsnick <[email protected]> wrote: > On Sat, Apr 02, 2016 at 16:42:52 +0100, Ben Barker wrote: > > So maybe this would be a better place to deal with the problem...simply > > don't add the packet to the queue if the two sequences are identical > here... > > Nice analysis and thanks for trying. At this point, it would probably > be good if one of the ffmpeg developers had a look at this. A few are > present on the list, but since you found a few "issues"/shortcomings, > you might consider reporting them to the bugtracker, so that they will > get attention. (Actual proposed patches go to the ffmpeg-devel list.) > > Moritz > _______________________________________________ > ffmpeg-user mailing list > [email protected] > http://ffmpeg.org/mailman/listinfo/ffmpeg-user > _______________________________________________ ffmpeg-user mailing list [email protected] http://ffmpeg.org/mailman/listinfo/ffmpeg-user
