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

skal skal at planet-d.net
Fri Nov 28 10:22:30 CET 2003


	Hi GomGom and all,

On Fri, 2003-11-28 at 10:41, Edouard Gomez wrote:
> skal (skal at planet-d.net) wrote:
> > 	Sure, no pb. But beware:
> > 
> > 	a) i'm not sure it's bug-free (especially the h263 (de)quant)
> 
> Ok i  integrated the C  part of your  code to XviD  and i get  a strange
> result with  mpeg4 quantization  (using your C  code, not  the reference
> one).
> 
> ===============================================================================
> Test name: underworld-trailer
> Using cached first pass stats (hash - a1c5ab3ce4420b302dd41ee9886379ea)
>  
> Results:
> xvid:     Min PSNR y : 30,13 dB, u : 38,08 dB, v : 39,42 dB
> xvid: Average PSNR y : 38,72 dB, u : 42,98 dB, v : 45,27 dB
> xvid:     Max PSNR y : 99,99 dB, u : 99,99 dB, v : 99,99 dB
> ===============================================================================
> 
> I was getting  39.11 dB on average  with the old code. 


	First things first: is dequantization (decoding) giving
	exactly the same result?

> Before  I go into deep debugging,  
> id you  sacrifice some precision  step that we  had in
> XviD ? 

	Partially, since it uses divide-by-multiply
	[with variable fixed-point precision] instead of
	the exact divide. Also: the bias are not the same
	[Q/2 vs.(3*Q+2)/4, for intra]. What's the file size
	difference? 
	But it might be a bug too.


> (if  i remember correctly, you're also the author  of the C quant
> code of xvid).

	nope, i think it's Suxen_Drol

	bye!
Skal



More information about the XviD-devel mailing list