On Wed, Aug 17, 2011 at 06:25:19PM +0200, Anton Khirnov wrote: > > On Wed, 17 Aug 2011 18:01:58 +0200, Kostya <[email protected]> wrote: > > On Wed, Aug 17, 2011 at 05:04:50PM +0200, Anton Khirnov wrote: > > > Keep a per-OutputFile instance of it, thus making -t work with multiple > > > output files. > > > --- > > > avconv.c | 22 ++-- > > > tests/ref/fate/feeble-dxa | 1 - > > > tests/ref/fate/lmlm4-demux | 332 > > > -------------------------------------------- > > > tests/ref/fate/real-rv40 | 3 + > > > 4 files changed, 16 insertions(+), 342 deletions(-) > > > > looks OK (especially if you have found out and can explain the reason why > > those fate test changed). > > The big change in lmlm4 is a bug in current behavior -- it produces a > 15s long video stream (and correct 3s long audio) even though -t 3 is > specified. After my patch the video is correctly 3 seconds long. > > The other two....no idea. Some magic inside output_packet() most > probably. I can try to investigate further if you feel it's important.
No, that's enough for me. _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
