[FFmpeg-devel] Patch: CrystalHD decoder support

Michael Niedermayer michaelni
Fri Jan 7 19:58:11 CET 2011

On Fri, Jan 07, 2011 at 05:13:21PM +0000, John Cox wrote:
> > [snip]
> >
> >there are 2 timestamps
> >1. pts, they represent the time at which the binary packet should be displayed
> >   once it is decodec (this goes through reordered_opaque)
> >2. dts, they represent the time at which the output that happens to come out
> >   of the decoder should be displayed.
> So this isn't the H.222 definition of DTS (Decode Time Stamp)?  H.222 DTS is
> the time at while a coded frame should be fed into an instantaneous decoder
> (i.e. a decoder that takes no time to process a frame but may then delay it for
> reordering or other purposes).
> Just seeking clarification.

dts of packet X is the time at which packet X is feed into an instantaneous
decoder. The decoder can during this instantaneos decoding at time X output
some bitmap, and this bitmap then should be displayed at time X. The input
bitstream and output bitmap can be unrelated (aka reordered).

H264 maybe allows to decouple dts from pts so that no pts has to match any dts
but that just means that dts becomes useless and can be ignored in favor of pts
for sake of playback so its not something ive looked into deeply as its a matter
of 1 line fix if we run into such file. Until then i assume the mpeg2 definition
holds for h264 too

Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

I have often repented speaking, but never of holding my tongue.
-- Xenocrates
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20110107/5bf18bdc/attachment.pgp>

More information about the ffmpeg-devel mailing list