On Tue, 2 Apr 2002 14:35:24 +0100, "Andrew Cocker" <big### [at] mariner9fsnetcouk>
wrote:
> Well, I'd *definitely* like to see the inclusion of HDRI, and subsurface
> scattering if possible. Dunno about the latter, but I think the former is
> possible. Also, am I correct in thinking that there is room for improvement
> still in the radiosity solving section of POV-Ray.. I think I read it
> somewhere on these newsgroups.. although maybe that will have to wait until
> POV-Ray 4.

Yeah to subsurface scattering! You can use media, but if you specifically want to use
a thin, single
layer mesh that has no 'interior', media won't work. Not to mention it being
significantly more
complicated to get right.

Another item would be some sort of adaptive smooth mesh. In general getting any mesh
to look good
require either a rediculous number of triangles or the scene being done at sufficient
distance from the
model to 'hide' the flat surfaces. This and the lack of an easy non-media way to do
subsurface
scattering are the main reasons why models noever look real, unless you are using ones
that are very
high quality (like those produced by companies like Dreamworks). So I have wondered
for a while if
some way could be found to create additional subdivisions and smoothing with triangles
whos surfaces
are nearly parallel to the camera direction and are thus most visible. A more complex
model will always
be better, but being very bad at mesh model building and nearly as bad at post work
needed to clean
up the artifacts. Besides with animation you don't generally have the luxery of
checking every frame of
a long animation. A method that could take a realtively small slice of time to
optimize the parts of a mesh
that contain visible problems would be worth it, especially since it is far less
likely to miss something than
the human doing the post work. ;)

Anyway, there are bound to be new things to keep adding.

Post a reply to this message