When rendering the wallstucco scene the renderer never proceeds beyond
parsing & bounding (beta 18, sse2, on a dual core system). The renderer
just continues using up processortime (according to task manager) and the
rendertime in the statusbar keeps increasing. The STOP button doesn't work
anymore but for the rest the program responds normally to user interaction.
Closing the program works, but gives - after a while - a Visual C++ runtime
error "application requested runtime to terminate in an unusual way", after
which processor time consumption stops.

After some detectivework the problem appears to originate from the
pigment_map in the Wall declaration, and more precisely due to the crackle
pattern used.

The following minimal scene gives a similar problem:

camera { location 5*y look_at 0 }

light_source { <10,10,0> color rgb 1 }

sphere {
  0, 1
  pigment {
   crackle
    colour_map {
      [0 color rgb 1]
      [0.5 color rgb x]
    }
  }
}

Only now the render produces a complete image on screen which is not saved,
but the render time keeps on increasing forever and processortime keeps
being used without change. Stopping behaviour as above. When any other
pattern is used instead of crackle the scene renders as expected.

Jeroen.

Post a reply to this message

jva nous apporta ses lumieres en ce 23-12-2006 06:15:
> When rendering the wallstucco scene the renderer never proceeds beyond
> parsing & bounding (beta 18, sse2, on a dual core system). The renderer
> just continues using up processortime (according to task manager) and the
> rendertime in the statusbar keeps increasing. The STOP button doesn't work
> anymore but for the rest the program responds normally to user interaction.
> Closing the program works, but gives - after a while - a Visual C++ runtime
> error "application requested runtime to terminate in an unusual way", after
> which processor time consumption stops.
> 
> After some detectivework the problem appears to originate from the
> pigment_map in the Wall declaration, and more precisely due to the crackle
> pattern used.
> 
> The following minimal scene gives a similar problem:
> 
> camera { location 5*y look_at 0 }
> 
> light_source { <10,10,0> color rgb 1 }
> 
> sphere {
>   0, 1
>   pigment {
>    crackle
>     colour_map {
>       [0 color rgb 1]
>       [0.5 color rgb x]
>     }
>   }
> }
> 
> Only now the render produces a complete image on screen which is not saved,
> but the render time keeps on increasing forever and processortime keeps
> being used without change. Stopping behaviour as above. When any other
> pattern is used instead of crackle the scene renders as expected.
> 
> Jeroen.
> 
> 
> 
Hit me with the cells pattern in the foloowing scene from the short code contest:
//Title: Pure Fun 225 bytes
#macro C(c,d) //pigment and  color map  macro
  crackle color_map{[0 rgb.3*c][.5 rgb 1-d]}
#end

plane{z,4
  pigment{cells
    pigment_map{[0 C(z,x)][.3 C(y,x)][.8 C(x,y)][1 C(y,z)]}
    turbulence<15,50,20>// very big turbulence
    scale 4
    warp{repeat-x flip x}}// acts as a mirrow
  finish{ambient 1}//chang ambient to -1 to see whats reflected
  translate<9,7,0>} //moves the whole plane to
                    //adjust whats refelected
                    //the z value has to be >-4



-- 
Alain
-------------------------------------------------
Save the whales. Collect the whole set.

Post a reply to this message

Alain <ele### [at] netscapenet> wrote:
> Hit me with the cells pattern in the foloowing scene from the short code contest:
> //Title: Pure Fun 225 bytes
> #macro C(c,d) //pigment and  color map  macro
>   crackle color_map{[0 rgb.3*c][.5 rgb 1-d]}
> #end
>
> plane{z,4
>   pigment{cells
>     pigment_map{[0 C(z,x)][.3 C(y,x)][.8 C(x,y)][1 C(y,z)]}
>     turbulence<15,50,20>// very big turbulence
>     scale 4
>     warp{repeat-x flip x}}// acts as a mirrow
>   finish{ambient 1}//chang ambient to -1 to see whats reflected
>   translate<9,7,0>} //moves the whole plane to
>                     //adjust whats refelected
>                     //the z value has to be >-4
>
>
>
> --
> Alain
> -------------------------------------------------
> Save the whales. Collect the whole set.

With me, same thing happens here as with the short scene. Rendering
proceeds, but doesn't stop after image is finished. Image doesn't get saved
(a 'state' file is produced though in the output folder).

Jeroen.

Post a reply to this message

jva wrote:
> When rendering the wallstucco scene the renderer never proceeds beyond
> parsing & bounding (beta 18, sse2, on a dual core system). The renderer
> just continues using up processortime (according to task manager) and the
> rendertime in the statusbar keeps increasing. The STOP button doesn't work
> anymore but for the rest the program responds normally to user interaction.
> Closing the program works, but gives - after a while - a Visual C++ runtime
> error "application requested runtime to terminate in an unusual way", after
> which processor time consumption stops.
> 
> After some detectivework the problem appears to originate from the
> pigment_map in the Wall declaration, and more precisely due to the crackle
> pattern used.

This has been fixed. Thanks to all who reported it.

-- Chris

Post a reply to this message