I've been looking at the facets pattern for normals and have thought a
little about it.

It has the following problems:

* The version for rounded surfaces work for rounded surfaces only, and give
very little control.

* The version for flat surfaces work both for flat and rounded surfaces, but
the illusion of depth is very bad, I think it looks more like a pigment than
a normal, i.e. as if it is painted on the surface.

* The two different version of facets really have not very much to do with
each other. It's almost as if it is two different workarounds.

* The pattern is not really a pattern as it works for normals only, and with
the rounded version you can't even scale it. (I know, this applies partly to
the slope pattern too, but anyway.)

I have this idea which would in my opinion make facets much more useful:

Facets shouldn't be a pattern but a normal modifier. Only the version for
rounded surfaces should be used. This would make it more logical that it
only works for normals (because it's a normal modifier.) It would also
explain why you can't scale it like most other patterns. However, the real
advantage is this:

You could use it in combination with any real normal pattern, for example
bumps. The facets modifier should take into account the normal pertubation.
This would mean that it would work for flat surfaces too, because they are
not flat but bumpy if you give them a bumpy normal. It would also give you
way better control over the facets, as you could use it together with any
normal pattern.

The syntax would also be easier. Just add "facets Float" to the normal
statement, where the Float would determine the number of facets, like the
factor value do in the current implementation.

So, what do you think? Isn't my idea both much more logical and useful than
the current implementation? :)

Rune
--
3D images and anims, include files, tutorials and more:
Rune's World:    http://rsj.mobilixnet.dk (updated June 26)
POV-Ray Users:   http://rsj.mobilixnet.dk/povrayusers/
POV-Ray Webring: http://webring.povray.co.uk

Post a reply to this message