[FFmpeg-devel] FFv1 performance
Wed Apr 21 12:17:17 CEST 2010
On Wed, Apr 21, 2010 at 10:51:44AM +0200, Peter B. wrote:
> Hello again,
> Thank you all for your feedback about using FFv1 as a video codec for
> long-term archiving. Thanks to your positive input, I was able to argue
> pro-FFv1 and we're now evaluating using it directly for capturing.
> Unfortunately, first tests have shown that FFv1 is very CPU-intensive,
> and unable to capture PAL in realtime on a 2,68 GHz quad-core (about
> 12.5 fps).
> Since the current implementation of FFv1 seems to be using only a single
> core, I'd be interested in:
> - Is it possible to change FFv1 to support multithreading?
> - If "yes", how much effort would that be?
Its not trivial but neither overly complex. The multithreading itself should
be easy but we need some changes to how the AC & VLC coder is initialized,
because multithreading requires more independant chunks and thus
initialization becomes more critical in terms of compression performance
> - If "no", could it be possible to "tune" the implementation of handling
> of YUV444P, YUV422P, etc?
> (Interesting observation was, that only YUV420P and RGB24 could be used
> for capturing without dropped or inserted frames)
444 and 422 will always be slower than 420 because for them the encoder must
encode more samples. general optimizations that make all faster should be
possible though its hard to say how much faster it can be made before
actually trying/doing it.
> If it would be possible to use FFv1, I might be able to get proper
> budget in order to finance the implementation of these improvements.
This would be great
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Thouse who are best at talking, realize last or never when they are wrong.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: Digital signature
More information about the ffmpeg-devel