This program has probably been discussed before but I haven't
found any information on it. The basic problem is that when I have an
object with both a normal applied and reflection turned on I get
strange, ugly artifacts. Well, I'll include the source so you can see
for yourself. I'm using v3.02 for Win 95.

#include "colors.inc"

#declare theta=2*pi*clock

camera{
 location 3
 look_at 0
}

light_source{
 <5*sin(theta),5,5*cos(theta)>
 color White
 spotlight
 radius 15
 falloff 70
 tightness 50
 point_at 0
}

sphere{
 0,1
 texture{
  pigment{color Yellow}
  finish{
   reflection 1
   specular.8
   roughness .006
  }
  normal{
   bumps .6
   scale .3
  }
 }
}

plane{
 y,-1
 texture{
  pigment{
   checker color Red, color Blue
  }
 }
}

Post a reply to this message

Nick Bray wrote:
> 
>       This program has probably been discussed before but I haven't
> found any information on it. The basic problem is that when I have an
> object with both a normal applied and reflection turned on I get
> strange, ugly artifacts.

Wierd.  It seems to me to be some sort of self-reflection
due to problems with the normal calculation applied to
rays of near-grazing incidence.

It seems to be eliminated by using ambient 0 diffuse 0 --
a perfectly reflecting surface is what you want, right?

Dan

P.S. A more direct way to rotate the light is to put
a "rotate" statement in the light rather than explicit
sin and cos calculations.... it comes down to preference,
I suppose.


-- 
http://www.flash.net/~djconnel/

Post a reply to this message

Nick Bray wrote:

>       This program has probably been discussed before but I haven't
> found any information on it. The basic problem is that when I have an
> object with both a normal applied and reflection turned on I get
> strange, ugly artifacts. Well, I'll include the source so you can see
> for yourself. I'm using v3.02 for Win 95.
>

    I played around with your scene for a little while and I think what you
are seeing is not an artifact, but rather the normal is giving you what your
asking it to do. With the single light source, a high degree of reflection, and
few if any other objects in the scene to be reflected in the surface finish, you

are seeng the true perturbance function of the normal.
    I did notice that changing the bumps statement to a dents statement with the

same value and scale produced a pretty good perturbed surface without the
"artifacts" your seeing.

K.Tyler

Post a reply to this message