TransQuant/MBCoding (was: Re: Re[2]: [XviD-devel] [BUG] BFrame encoder doesn't force intra for 1st frame)

Christoph Lampert xvid-devel@xvid.org
Thu, 8 Aug 2002 15:18:09 +0200 (CEST)


> > So I need a hyperfast routine, where:
> > * given coefficients are quantized and dequantized
> > * the SAD between quant and dequant calculated
> 
> > * OR, if that is faster: The routine to be called is of course a
> >   "bitcounter", measuring how many bits are needed to encode this block.
> >   As far as I know this is based on a lookup table using the quantizer
> >   values, and a run-level-method to encode 0s.

> Dont you want both then? One indicates the distortion, the other the
> rate. 

Yes. Sorry, if that was a misunderstanding: I need both, distortion and
rate. Both can be calculated using small modification of existing
function and I asked for a new and faster method to calculate distortion
and I wanted to keep "old" method of calculating the rate, because it's
not too slow anyway. The new "distorsion" routine was supposed to call the
old "rate" routine. 

However I noticed (when writing it down) that it might be much better to 
combine both into one, because _then_ quantized coefficient don't have to
be saved to mem. So making "distorsion" fast is easier, if old "rate" is
replaced, too.

gruel


-- 
Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi-     
Beringstr. 6, Raum 14 Tel. (0228) 73-2948 | nell erstellt und bedarf
Sprechstunden: keine, aber meistens da    | keiner Unterschrift. AZ 27B-6