[XviD-devel] [In progress] devapi4 -- mpeg matrices

[Michael Militzer]

Hi,

Quoting Edouard Gomez <ed.gomez at free.fr>:

> Ok i'm now statified with the C framework, it's doing better than old
> code, and is thread/instance safe.

Quoting Edouard Gomez <ed.gomez at free.fr>:

> h263 is identical. mpeg will be  identical too once i reintroduce the RD
> optimized bias.

ok, so obviously the new code is _not_ identical after all - so do you have
any clue at all why it's different?

Quoting Edouard Gomez <ed.gomez at free.fr>:

===============================================================================
> Test name: underworld-trailer
> Using cached first pass stats (hash - a1c5ab3ce4420b302dd41ee9886379ea)
>  
> Results:
> xvid:     Min PSNR y : 32,54 dB, u : 40,08 dB, v : 42,27 dB
> xvid: Average PSNR y : 39,19 dB, u : 43,19 dB, v : 45,49 dB
> xvid:     Max PSNR y : 99,99 dB, u : 99,99 dB, v : 99,99 dB
>
===============================================================================
> 
> Old code was doing 39.11dB with same options. Not a big gain, but still
> nice.

Now that the new code behaves different: do you have any prove that it's
really better? You've tested with just _one_ sequence where you reached
a 0.08 dB gain - do you know how it performs on other scenes? Will it be
an overall win or not? Was the filesize identical to the old code btw?
 
bye,
Michael