[MPlayer-DOCS] Improved vi_qfactor and (tcplx|scplx|p)_mask

Guillaume POIRIER guillaume.poirier at ifsic.univ-rennes1.fr
Sun Sep 26 15:39:54 CEST 2004


Hi,
Le dim 26/09/2004 à 15:12, Diego Biurrun a écrit :

> >  .B tcplx_mask=<0.0\-1.0>
> >  temporal complexity masking (default: 0.0 (disabled))
> > +Imagine a scene with a bird flying across the whole scene, tcplx_mask
> > +will decrease the bird's macroblocks quantizers (thus decreasing their
> 
> macroblock

I think it needs to be plural since the bird is made out of several
macroblocks. I have problems with that area of English grammar, so I
couldn't say for sure. Maybe The Wanderer should have a look at it...


> > +Imagine a scene with grass (which usually has a great spatial complexity),
> > +a blue sky and a house, scplx_mask will raise grass' macroblocks quantizers
> 
> raise the macroblock of the grass

Same here, grass is probably made out of several macroblocks


> > +quality of P-blocks even if only DC is changing, which will probably not
> 
> P-blocks or P-frames?
> 
> > +(default: 0.0 (disabled)).
> > +p_mask=1.0 doubles the bits allocated to each intra block.
> 
> Same question here as in the paragraph above: Are we talking about
> blocks or frames?

My understanding of masking functions is that they would on a
block-by-block basis, so for consistency reasons, it only reads about
blocks. 
I believe that P-blocks are included in P-frames, so talking about blocs
is probably more specific and precise than talking about frames.
Does that make sense?

The attached patch includes all your suggestions (even those about
macroblocks, but leaves untouched the issue about (blocks|frames).
Please tell me if you agree with it.

Regards,
Guillaume

-------------- next part --------------
A non-text attachment was scrubbed...
Name: man_masking_options_and_vi_qfactor.patch
Type: text/x-patch
Size: 2350 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/mplayer-docs/attachments/20040926/e7afca08/attachment.bin>


More information about the MPlayer-DOCS mailing list