I was checking multiple reflections using a test scene of two mirrored
planes, one behind the camera and one in front. Put some spheres between
them, 100% red, green, blue, white and black. The planes are rotated
around the y to offset the reflections as seen from the camera viewpoint
so there are multiple reflections of this one group. A max_trace_level
of 15 used so there would be plenty enough to see.
Aside from the fact that the mirror pigments used (I made renders using
rgb 1, rgb 0.5, rgb 0, red, green, blue, yellow, magenta pigmented
mirrors) contribute to the mirrors colors even at reflection 1, which is
to be expected, an odd thing occurs concerning the black (pigment {rgb
0}) sphere. It only shows up in the first reflection, the mirror behind
it in this case. The other spheres remain as expected unless of course
they blend in with the equivalent colors of the 2 planes. The black
sphere is fine if the mirrors are white or gray (maybe black too but
then they aren't visible anyway) but they disappear if red, green, blue,
yellow or magenta pigment is used. I hadn't tried other shades of color
yet to check further since I just did a routine set of renders.
There seems no doubt this must be wrong. Since if the rgb 0 is showing
in one reflection yet none of the other multiples, and you would expect
it to not show at all if it is a case of "non-contributable color" or
some such.
I can't see how this effect has been overlooked before so it's probably
been encountered before and maybe ignored.
Using the Windows version 3.1 r1

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

Post a reply to this message

Bob Hughes <inv### [at] aolcom> wrote:
: Aside from the fact that the mirror pigments used (I made renders using
: rgb 1, rgb 0.5, rgb 0, red, green, blue, yellow, magenta pigmented
: mirrors) contribute to the mirrors colors even at reflection 1

  This is caused by the diffuse and ambient factors. To get a perfect
mirror you have to set them to 0, ie:

finish { reflection 1 ambient 0 diffuse 0 }

  The vector after the 'reflection' keyword (yes, it is a vector) is used
to specify the color of the mirror. '1' means just '<1,1,1>' ie. white, ie.
it reflects all color components. '<1,1,0>' would mean yellow window, ie.
it only reflects the red and green components but not the blue one.
  Funny things can be achieved by applying values greater than 1 to the
components...

-- 
main(i){char*_="BdsyFBThhHFBThhHFRz]NFTITQF|DJIFHQhhF";while(i=
*_++)for(;i>1;printf("%s",i-70?i&1?"[]":" ":(i=0,"\n")),i/=2);} /*- Warp. -*/

Post a reply to this message

On Tue, 09 Feb 1999 02:02:13 -0600, Bob Hughes <inv### [at] aolcom> wrote:
>There seems no doubt this must be wrong. Since if the rgb 0 is showing
>in one reflection yet none of the other multiples, and you would expect
>it to not show at all if it is a case of "non-contributable color" or
>some such.
>I can't see how this effect has been overlooked before so it's probably
>been encountered before and maybe ignored.
>Using the Windows version 3.1 r1

I assume you've tried switching the pigments around to make sure it's 
a color problem and not just a trick of geometry.  If that is the case,
and it's still broken, can you show us a simple scene that demonstrates
the problem?

Post a reply to this message

Yes, I realize the diffusion factor and all (reflection <vector>), it
was the intent to see how the fading occurs for some basic pigments. I
simply used default values while testing this multiple reflection stuff.
The example remains similar concerning how rgb 0 does not fade object
reflections and rgb 1 does most, despite other finish properties.

Nieminen Mika wrote:
> 
> Bob Hughes <inv### [at] aolcom> wrote:
> : Aside from the fact that the mirror pigments used (I made renders using
> : rgb 1, rgb 0.5, rgb 0, red, green, blue, yellow, magenta pigmented
> : mirrors) contribute to the mirrors colors even at reflection 1
> 
>   This is caused by the diffuse and ambient factors. To get a perfect
> mirror you have to set them to 0, ie:
> 
> finish { reflection 1 ambient 0 diffuse 0 }
> 
>   The vector after the 'reflection' keyword (yes, it is a vector) is used
> to specify the color of the mirror. '1' means just '<1,1,1>' ie. white, ie.
> it reflects all color components. '<1,1,0>' would mean yellow window, ie.
> it only reflects the red and green components but not the blue one.
>   Funny things can be achieved by applying values greater than 1 to the
> components...
> 
> --
> main(i){char*_="BdsyFBThhHFBThhHFRz]NFTITQF|DJIFHQhhF";while(i=
> *_++)for(;i>1;printf("%s",i-70?i&1?"[]":" ":(i=0,"\n")),i/=2);} /*- Warp. -*/

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

Post a reply to this message

Not sure how much color switching (or what kind) you mean but I've got
both a sample image and script I used to make it with. I'll post to the
images group and text files group shortly.
Maybe it'll help explain.

Ron Parker wrote:
> 
> On Tue, 09 Feb 1999 02:02:13 -0600, Bob Hughes <inv### [at] aolcom> wrote:
> >There seems no doubt this must be wrong. Since if the rgb 0 is showing
> >in one reflection yet none of the other multiples, and you would expect
> >it to not show at all if it is a case of "non-contributable color" or
> >some such.
> >I can't see how this effect has been overlooked before so it's probably
> >been encountered before and maybe ignored.
> >Using the Windows version 3.1 r1
> 
> I assume you've tried switching the pigments around to make sure it's
> a color problem and not just a trick of geometry.  If that is the case,
> and it's still broken, can you show us a simple scene that demonstrates
> the problem?

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

Post a reply to this message