[MPlayer-dev-eng] [PATCH]Add support for CoreAVC h264 codec

[Alan Nisota]

On 10/3/06, Michael Niedermayer wrote:
> could you post benchmarks with h264 videos with various parameters with
> coreavc and ffh264? that should be VERY usefull to find out which parts are
> faster then ffh264 which could then help to improve ffh264
>
> things to test
> * low resolution where all reference frames+1 _easily_ fit in the L2 cache
> * CABAC / CAVLC
> * high bitrate / low bitrate
> * intra only
> * B frames vs. no B frames
> * loop filter / disabled loop filter
>
> and also check if the output of coreavc is binary identical to ffh264
> if not that would explain why they are faster
>
> also try to compile (i386)/dsputil* with -O2 that will make lavc
> faster depending on the gcc version
I will work on all of this.  It'll take some time to generate the
data, but I will get back to you.
Just for your info, I didn't see much difference in speed between the
beta (0.4) and the latest version (1.1)
I've been benchmarking using
mplayer -benchmark -nosound -vo null
is this sufficient for everything but the frame-accuracy tests?

I will get you all of the above, but here is the result on a 1280x720
stream (just one I had lying around), which shows 20% improvement.
This is on a 2.8GHz P4 (Northwood generation):
CoreAVC
BENCHMARKs: VC:  37.063s VO:   0.046s A:   0.000s Sys:   1.013s =   38.122s
BENCHMARK%: VC: 97.2213% VO:  0.1217% A:  0.0000% Sys:  2.6570% = 100.0000%
ffh264:
BENCHMARKs: VC:  46.495s VO:   0.047s A:   0.000s Sys:   1.001s =   47.542s
BENCHMARK%: VC: 97.7968% VO:  0.0983% A:  0.0000% Sys:  2.1050% = 100.0000%