[FFmpeg-devel] MXF D10 regression tests
Wed Mar 18 13:44:03 CET 2009
On Wed, Mar 18, 2009 at 12:10:43PM +0100, Reimar D?ffinger wrote:
> On Wed, Mar 18, 2009 at 05:57:56AM +0100, Michael Niedermayer wrote:
> > On Tue, Mar 17, 2009 at 10:29:19AM +0100, Reimar D?ffinger wrote:
> > > On Tue, Mar 17, 2009 at 10:20:36AM +0100, Reimar D?ffinger wrote:
> > > > It also fixes the MXF regression test, but either I messed up the test
> > > > or -f framecrc still differs (maybe due to the differing strides?)...
> > >
> > > Ah, correction: the patch fixes the intra-only case. There is still some
> > > other bug in the non-intra case.
> > Note, these kind of changes wil need carefull benchmarks as they
> > must not cause a speedloss
> > the linesize/uvlinesize relation is intentional and was intended as
> > optimization reducing the number of variables&reads needed
> Note that this is _not_ documented, the only places that mentions
> restrictions on linesize is the avcodec_decode_video doxy and that only
> talks about line size. The avcodec_encode_video function does not
> mention any restrictions.
It is not supposed to be a restriction for avcodec_encode_video() though
maybe it is by a bug. (i did not test this)
The restriction is just on the internal buffers, which for encoding are
allocated by lavc.
[regression tests for avcodec_encode_video linesizes are welcome]
> And to whoever copy-and-pasted the avcodec_decode_audio2 doxy: It would
> be a good idea to actually read that stuff, the part talking about
> "linesize" does not make any sense there.
feel free to fix it and reword it as you see fit
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Into a blind darkness they enter who follow after the Ignorance,
they as if into a greater darkness enter who devote themselves
to the Knowledge alone. -- Isha Upanishad
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: Digital signature
More information about the ffmpeg-devel