Just looked again. Yep, right about the inverse, I could have sworn I had it 
working within texture{}, turns out I had failed at putting it elsewhere 
(finish{} etc.) and thought I succeeded at putting it there in texture. I 
mistakenly read the braces of course.
I've looked at reflecting planes more and found a "wrong look" to the thing, 
partly mirrored and partly not at all, about like your test images but it has 
to do with filter or transmit and if they're used or not.
I was trying ior and it makes a lot of difference in the look, however, it 
comes down to the fact that reflection fades with increasing transparency. A 
known fact of the sort that causes 'finish' to fade away until none exists if 
filter or transmit is 1, otherwise no "invisible" container objects could be 
made since default diffuse, etc. would prevent them from being possible. 
Except that reflection does not allow 100% disappearence in case of 100% 
transparency, and refraction makes it all look further intertwined.
All in all I think the reflection_exponent is something to look into. You can 
set it to zero and get a transparency-based reflection it seems unless 
refraction is on, instead of the other way around (whatever).

I'm going to get your test file and try it too, btw.

So anyway, to review:

 filter [0...1] and/or transmit [0...1]
 reflection [0..1] and/or reflection_exponent [0..1]
 ior [1...?]
 inverse (or not)

These all seem to be doing a strange thing together; like I even need to say 
that.

Message <3664BA3F.DC68E210@Kopp.com>, Nathan Kopp  typed...
>
>The inverse keyword does seem to make a difference, but it doesn't fix the
>problem.  I assume that you mean 'inverse' in the object (it is a syntax 
error
>if you put it inside the finish or inside the texture).  Also, on a plane,
>inverse has the same effect as switching the direction of the normal vector.
>
>Go grab my test scene (in the povray.text.scene-files group) and render it. 
>Then add 'inverse' to the plane.  The problem is still there.  At least it 
was
>on my machine. Actually, 'inverse' has a really weird effect that I didn't
>expect... maybe I'll look into that...
>
>Thanks for telling me about the 'inverse' keyword, though.  Even if it had 
made
>my bugfix null & void, it would be good to know!  :-)
>
>-Nathan Kopp

-- 
 omniVERSE: beyond the universe
  http://members.aol.com/inversez/POVring.html
=Bob

Post a reply to this message