[FFmpeg-devel] h264 speed regression after PAFF

[Michael Niedermayer]

Hi

On Fri, Oct 12, 2007 at 02:06:12PM +0200, Andreas ?man wrote:
> Hi
> 
> After the PAFF code got in h264 is about 3% slower for non-interlaced
> content. See test below.
> 
> The offending diffs are
> svn diff -r10679:10690 (changes in deblock_mb_fast())
> svn diff -r10691:10696 (The "enable" patch)
> 
> The "enable" patch can obviously not be blamed on its own.
> 
> Either way, I find this a little too much. Michael, whats your opinion?

3% is too much yes
the non PAFF code should not be slowed down by a measureable amount ideally

this also shows how important benchmarks are
someone should do similar benchmarks for mpeg1/2/4 begining from r1
iam sure there are plenty of points where speed was lost and which could
be done differently to avoid the speed loss
but ive suggested such benchmarks already many times, we mainly lack a
volunteer to do them, but that is getting off topic first the PAFF slowdown
should be dealt with ...

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

I do not agree with what you have to say, but I'll defend to the death your
right to say it. -- Voltaire
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20071012/5323da9d/attachment.pgp>