[FFmpeg-devel] Designing a subtitles infrastructure (was: DVBSUBS and TXTSUBS)

Nicolas George nicolas.george at normalesup.org
Thu May 31 16:24:26 CEST 2012

Le septidi 7 prairial, an CCXX, Wolfram Gloger a écrit :
> Please check libavformat/utils.c:ff_interleave_packet_per_dts().
> We already have a threshold (delta_dts_max < 20 seconds) for
> noninterleaved streams.
> Why isn't this working in your case?  Is it because that mechanism is
> per input context, and you have two of them?

I do not know why it does not have effect in this particular case, but if it
did, it would result in a broken file with all video first and then all
audio, that would not be much better, because it does not affect the
scheduling algorithm inside ffmpeg.

I am more and more convinced we need "nothing new" dummy subtitles packets
to transform the sparse streams into full ones.


  Nicolas George
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20120531/69c4f402/attachment.asc>

More information about the ffmpeg-devel mailing list