On Sun, Nov 6, 2011 at 15:11, Burk <[email protected]> wrote: > Thanks Alex, I have only tested with AVI/Xvid, but eventually will add > MKV/x264. > I see so ffmpeg is doing the counting internally, but when it seeks by file > position it loses the point of reference. > I guess seeking by index will know the target frame number that's why it > works there. > I was hoping that the container or the codec will "waste" 8 bytes to > remember the frame number in cases like that :-) > > Cheers, Thomas
Unfortunately, they do not work this way. If you are skipping forwards, you probably should be skipping frames one by one, until you reach the expected byte offset. Skipping backwards, you can store the frame boundaries for yourself. Good luck, Alex _______________________________________________ Libav-user mailing list [email protected] http://ffmpeg.org/mailman/listinfo/libav-user
