This is really impressive. Did you try to do a candle ? Does your 
technique work with a very close light ?

My guess for your technique is a two pass render, first pass with a 
black and white spherical pigment depending on the light position that 
you apply to all objects with subsurface scattering, and then a second 
pass where you use the first one to evaluate the color. While explaining 
my guess I have no really precise idea on the how you did it though.  :-)

Am I any close ??   8-p

JC

Samuel Benge wrote:
> I think I have found a new way to simulate subsurface scattering(SSS). I 
> have attached an image of an object with a yellowish diffuse and reddish 
> SSS.
> 
> Can you tell how I did it? I did not use media or photons. I did not use 
> double_illuminate.
> 
> Four stars to the person who figures it out.
> 
> 
> ------------------------------------------------------------------------
>

Post a reply to this message

Samuel Benge wrote:
> I think I have found a new way to simulate subsurface scattering(SSS). I 
> have attached an image of an object with a yellowish diffuse and reddish 
> SSS.
> 
> Can you tell how I did it? I did not use media or photons. I did not use 
> double_illuminate.
> 
> Four stars to the person who figures it out.

Tricky, but i have an idea.  Could you render it with the light source 
in the middle between the two shapes?

Christoph

-- 
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 25 Oct. 2003 _____./\/^>_*_<^\/\.______

Post a reply to this message

JC (Exether) wrote:

> This is really impressive. Did you try to do a candle ?


Thanks JC. Nope, haven't tried a candle yet, it was getting quite late 
last night.

 > Does your
 > technique work with a very close light ?

I don't think it will, but only because my math is sloppy. I'm hoping 
when I release the code that somebody with better coding skills will 
develop it further.....


> My guess for your technique is a two pass render, first pass with a 
> black and white spherical pigment depending on the light position that 
> you apply to all objects with subsurface scattering, and then a second 
> pass where you use the first one to evaluate the color. While explaining 
> my guess I have no really precise idea on the how you did it though.  :-)
> 
> Am I any close ??   8-p


You are very close. A black/white pigment is involved, and a preliminary 
pass to establish subsurface scattering effects is used.


-- 
Samuel Benge

stb### [at] hotmailcom
See my website@: http://www.goldrush.com/~abenge/Top/index.html

Post a reply to this message

Christoph Hormann wrote:

 > Tricky, but i have an idea.  Could you render it with the light source
 > in the middle between the two shapes?


Nope, not until the math is fixed. Then I should be able to render it 
with very close light_sources.


 


-- 
Samuel Benge

stb### [at] hotmailcom
See my website@: http://www.goldrush.com/~abenge/Top/index.html

Post a reply to this message

Hi Samuel,

my guess is a mlpov render with use of shadow_pigment, use of the object on
itself, a nonzero ambient value for nonshadowed and ambient zero for
shadowed parts. For the smooth transition you perhaps used an area light to
define the shadow pigment and a point light for the scene.

Norbert







Samuel Benge wrote:
>I think I have found a new way to simulate subsurface scattering(SSS). I
>have attached an image of an object with a yellowish diffuse and reddish
>SSS.
>
>Can you tell how I did it? I did not use media or photons. I did not use
>double_illuminate.
>
>Four stars to the person who figures it out.
>
>--
>Samuel Benge
>
>stb### [at] hotmailcom
>See my website@: http://www.goldrush.com/~abenge/Top/index.html
>

Post a reply to this message

Samuel Benge wrote:
> Christoph Hormann wrote:
> 
>  > Tricky, but i have an idea.  Could you render it with the light source
>  > in the middle between the two shapes?
> 
> Nope, not until the math is fixed. Then I should be able to render it 
> with very close light_sources.

Well, i thought the slope pattern might be involved.  In this case it 
would only work well for a distant light source, otherwise you would 
need the aoi pattern instead.

Looking closer at the image reveals that there are indeeed problems, 
strong scattering occurs at places where it should not occur.

I have two other things in mind:

1) you use saved radiosity data from a different but related scene 
setup.  This is far fetched of course but it might be an intersting idea 
for such effects

2) you use trace() to place blob components on the surface and use the 
blob with the object pattern (not sure how you could achieve the smooth 
transit with this method though)

Christoph

-- 
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 25 Oct. 2003 _____./\/^>_*_<^\/\.______

Post a reply to this message

Christoph Hormann wrote:

 >
 > 2) you use trace() to place blob components on the surface and use the
 > blob with the object pattern (not sure how you could achieve the smooth
 > transit with this method though)
 >
 > Christoph

* * * * !
Congratulations, you get four stars :) While you didn't explain the 
whole technique, you explained the guts of my setup.

I used trace to place blob components randomly onto the object. The blob 
object is made into a pigment using the object pattern . The blobby 
pigment is then jittered about 100 times inside an average pattern's 
color_map.

I have made a few enhancements and bug fixes, and I think it's about 
ready to be released. In order to allow the use of multiple 
light_sources I might have to make it into a macro which you place into 
an average pattern's texture_map.

It will be ready very soon.

-- 
Samuel Benge

stb### [at] hotmailcom
See my website@: http://www.goldrush.com/~abenge/Top/index.html

Post a reply to this message