It's really only surface scattering, rather than subsurface scattering, but
here's the montecarlo approach.  In summary,

1) Only lighting is HDR image.
2) Single object with one uniform texture
3) Interior is absorbing media
4) Surface has diffuse reflection and diffuse refraction
5) Surface has transmit (1-eps) and diffuse 1/eps
   ( This maintains diffuse=1 since (1/eps) * (1-(1-eps)) = 1 )

This works well for thin features, but won't work well for smooth transitions
like a sharp light on a sphere.  That is, unless a second sphere is placed
below the surface so that some subsurface scattering actually occurs.

Another random MCPov tip: render large.  Scaling down decreases noise, but you
can't ever scale up.  In other words, it would be larger, but my computer went
to sleep overnight.

 - Ricky

Post a reply to this message

Attachments:
Download 'sss.jpg' (39 KB)

Preview of image 'sss.jpg'

"triple_r" <nomail@nomail> wrote:
> It's really only surface scattering, rather than subsurface scattering, but
> here's the montecarlo approach.  In summary,
>
> 1) Only lighting is HDR image.
> 2) Single object with one uniform texture
> 3) Interior is absorbing media
> 4) Surface has diffuse reflection and diffuse refraction
> 5) Surface has transmit (1-eps) and diffuse 1/eps
>    ( This maintains diffuse=1 since (1/eps) * (1-(1-eps)) = 1 )
>
> This works well for thin features, but won't work well for smooth transitions
> like a sharp light on a sphere.  That is, unless a second sphere is placed
> below the surface so that some subsurface scattering actually occurs.
>
> Another random MCPov tip: render large.  Scaling down decreases noise, but you
> can't ever scale up.  In other words, it would be larger, but my computer went
> to sleep overnight.
>
>  - Ricky

The image is very dark on my monitor.
But it is very promising!
Cool! ;-)

Post a reply to this message