...at the price of one.
Render this simple, dull scene. Can you notice those nice blue patches everywhere?
It is a real puzzle which shows only in certain conditions.
Turn off one of the flags at will and the problem will disappear.
Sorry if this is a known bug (or if it's not a bug at all of course :-).

#version 3.5;

global_settings {
  radiosity { recursion_limit 2 }
  ambient_light 0
}

box {-1,1
  scale 12
  pigment {rgb 1}
}


light_source {
  <0,2,0>
  rgb 0.7
}

#declare IOR = true;
#declare FRESNEL = true;
#declare CONSERVE_ENERGY = true;
#declare LAYERED_TEX = true;
#declare INTERSECTING_SPHERES = true;


#declare glass_mat =
material {
  #if (LAYERED_TEX)
  texture {pigment {rgbf 1}}
  #end
texture {
  pigment {
    rgbf 1
  }
  finish {
    reflection {0.2,1
      #if (FRESNEL) fresnel on #end}
    #if (CONSERVE_ENERGY) conserve_energy #end
  } }
  interior {
    #if (IOR) ior 1.5 #end
  }
}

union {
  #if (INTERSECTING_SPHERES)
sphere {
  -x/2,1}
sphere {
  x/2, 1}
#else
sphere {
  -x, 1}
sphere {
  x, 1}
#end
material {
  glass_mat}
}


camera {location <0,8,-8> look_at 0 angle 80}

//----end of code

--
Jonathan.

Post a reply to this message

No one confirms? I must be mad then.
Well, nevermind... you will never come across this one

--
Jonathan.

Post a reply to this message

I had passed this message so quickly before I meant to go back to it and
copy/paste into POV and it got skipped.

The blue coloration is a real curiosity alright. Found that gray_threshold 1
in radiosity clears it up, so maybe there's a glitch in that. I really have
no idea though.

Post a reply to this message

On Tue, 3 Sep 2002 17:19:23 -0500, hughes b <omn### [at] charternet> wrote:
> I had passed this message so quickly before I meant to go back to it and
> copy/paste into POV and it got skipped.
> 
> The blue coloration is a real curiosity alright. Found that gray_threshold 1
> in radiosity clears it up, so maybe there's a glitch in that. I really have
> no idea though.

I use binary compiled with ICC and I noticed this when using the
lightsys stuffs:

 #debug concat(" lct: ", vstr(3, lct, " ", 0, 3), "\n")
 #local lct2=vnormalize((lct/REF_WHITE))*3;
 #debug concat(" lct2 #1: ", vstr(3, lct2, " ", 0, 3), "\n")
 #local lct2=lct2+(COLOR_FILTER.gray)-COLOR_FILTER;
                  ^^^^^^^^^^^^^^^^^^^
 #debug concat(" lct2 #2: ", vstr(3, lct2, " ", 0, 3), "\n")
 #local Light_Color=(lct2*(lm/MAX_LUMENS)*lm);
 #debug concat(" Light_Color: ", vstr(3, Light_Color, " ", 0, 3), "\n")

now when I include it, I get:

lct: 1.000 0.459 0.060
lct2 #1: 2.626 1.435 0.212
lct2 #2: 2.399 1.208 -0.015
                     ^^^^^^
Light_Color: 4317.879 2174.049 -26.481
lct: 1.000 0.538 0.077
lct2 #1: 2.516 1.613 0.260
lct2 #2: 2.289 1.386 0.033
Light_Color: 4120.097 2494.310 60.007
...

that negative value for blue causes evil blue areas into the picture
(lightsys.pov).

now, when I change

#local lct2=lct2+(COLOR_FILTER.gray)-COLOR_FILTER;
to
#local lct2=lct2+ COLOR_FILTER.gray -COLOR_FILTER;

I get the same as with the official povray 3.5:

lct: 1.000 0.459 0.060
lct2 #1: 2.626 1.435 0.212
lct2 #2: 2.626 1.435 0.212
Light_Color: 4726.759 2582.928 382.399
lct: 1.000 0.538 0.077
lct2 #1: 2.516 1.613 0.260
lct2 #2: 2.516 1.613 0.260
Light_Color: 4528.977 2903.190 468.887

what's up?
is the parser code buggy?
does ICC generate buggy code for the parser code?

-- 
Safari - y7p### [at] sneakemailcomgovinvalid
"Talk is cheap. Show me the code." - Linus Torvalds

Post a reply to this message

Me 2.

Txemi Jendrix
tji### [at] txemijendrixcom
http://www.txemijendrix.com

Post a reply to this message

On Tue, 3 Sep 2002 23:57:39 +0200, "JRG" <jrg### [at] hotmailcom> wrote:

>No one confirms? I must be mad then.
>Well, nevermind... you will never come across this one

I can confirm it, but I don't have a clues as to why this happens.

/Erkki

Post a reply to this message