In article <41dd8cc6$1@news.povray.org>, jai### [at] ignoranciaorg 
says...
> Rene Bui wrote:
> >>The problem I think with SSS is that you will need a realistic structure
> >>inside (bones and muscles) to get a realistic head.
> > 
> > ??? .. Do you really think that ? Could you tell me more about it ?
> 
>    I mean: a real head is not a perfect solid with uniform density. This 
> is, IMHO, why it always looks like wax/soap when you apply faked SSS to 
> an empty object. The SSS effect should be different on places with bones 
> more close to the skin surface than in others with more flesh or muscles 
> between the skin and the skull, and this only seems possible by 
> modeling, at least roughly, the internal structure.
> 
That is likely the point of some methods which only apply depth rules to 
the surface. The object is still hallow, but since it doesn't depend on 
'real' calculations using media, it doesn't get treated as though it is. 
More of a case of the sampling occasionally blending point on other parts 
of the texture you give with the point you calculating, so it 'seems' to 
be doing what media does. Accurate subsurface scattering with media 
would, much like adage about trying to accurately simulate the universe, 
almost require a simulation of the body that would require building one, 
layer by layer, with nearly every part in it. Could it be done? Possibly. 
It is practical if all you want is a halfway realistic looking model? Not 
by the farthest stretch of the imagination. lol

-- 
void main () {

    call functional_code()
  else
    call crash_windows();
}

Post a reply to this message