Re: [Yafray-devel] Skin shader..

[Alfredo de Greef]

--- Andera <address@hidden> wrote:
> > HDRI is no magic GI solution
>
> Yeah, but it is the case yhat I tried all my
> abilities to
> get BMRT2.6 accept an HDRI on Linux...and it worked
> so bad..

Surprisingly I never tried BMRT with HDRI, but it
should be possible to do, I have seen some results,
although not really very good looking ones. It has to
be converted to tiff format (possibly LogLuv?), but
there are several shaders available for it, one of
them on the rendermania site, but you probably know
about that one. There are also some created by Larry
Gritz himself. Probably the best lighting method (and
fastest) would be the 'indirect' method.
On Linux you could try using Radiance, the 'original'
HDRI renderer used for 'Rendering with Natural Light'
& 'Fiat Lux', but it is a bit hard to use.

> I started to work because a good looking skin shader
> is what MakeHuman needs...
> ..and I am happy to do this for MH.
> But, really..I started this project without knowing
> if
> I could get to goal.....
> So, I am coding also to learn more..on yafray.
> 
> I don't want that you leave the shader because I am
> working
> on it....

Well, it's nice to have some contributions by others
as well, unless you have problems with it, I'll just
move on to other things.

> I got a first result now...but it has to be improved
> a lot.
> I took as first model skin.sl from Matt Pharr
> (Exluna).

I used the same of course, but I'm not sure if it
really looks all that good.

> It is a simple one....my ambitious final target is
> to implement
> a more sophistcated and general "translucent"
> shader...like in Jensen paper..
> .."A Practical Model for Subsurface Light
> Transport"...but at the moment
> it seems so far...

Well, that is really beyond me, like I said, I'm not a
programmer, and I definitely don't have the necessary
math skills to understand even a single word from that
paper. I understand the basic principle, but that is
where it stops for me.
btw, in case you don't know about it, Jensen also
published a paper with a faster bssrdf model:
http://graphics.stanford.edu/~henrik/papers/fast_bssrdf/

I have even seen some real time demos of it.

> What the guys from MH needs is a working shader, not
> a "mine" shader..
> 
> Well, I don't know, let's see what comes out from
> the actual skinshader...
> ...and discuss on it...

I look forward to it. Good luck with coding!

Alfredo

__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com