First of all, this isn't an attempt to debunk Samuel Benge's method. I just
wanted to see for myself a direct comparison to media SSS. I got out an old
(and sparse of triangles) Venus de Milo statue mesh I had gotten long ago
(POV-Ray Objects Collection?) and rendered it both ways. Unfortunately, the
object pigment way was very slow (12 min.) but media went six times faster
(2 min). I figure there's still good reasons to use a non-media method
sometimes, mainly to avoid multiple media artifacts; or just for other
effects, all of which I couldn't begin to guess right now.

Media settings were very simple, being nothing more than absorption < 0.3,
0.6, 0.9 > scattering { 4, 0.3 } without any other items (everything
default).

There's an obvious color difference between these two images. I guess I
should have changed the averaged pigment of rgb 1 in the 'surface' SSS one
to better match the media. Object SSS top, media SSS bottom.

Bob H.

Post a reply to this message

Attachments:
Download 'sss_comparitive-test1.jpg' (18 KB)

Preview of image 'sss_comparitive-test1.jpg'