My english is not very good and this is a little hard to explain, but
I'll try.
  I'm making an image for the IRTC and I have a problem.
  I am using radiosity and the calculations take very long. I have access
to two Dec Alpha 300MHz machines, so what I am doing is calculating one
half of the image in one machine and the other half in the other machine
(of course just for final renderings, not fast test renderings). It takes
more than 8 hours to calculate a half of the image.
  There is a tree in the image. I'll try to ascii-draw the situation here:

                                     \     /
                                      |   |
   Upper half of the image            |   |
                                      |   |
   - - - - - - - - - - - - - - - - - -|- -|- - - - - - - - - - - - - - -
                                      |   |
   Lower half of the image            |   |  <- Tree
                                      |   |
                                     /     \

  The problem is that when I turn the radiosity on, the lower part of the
tree is much more brighter than the upper part. Everything else looks ok.
When I join both parts of the image, the difference is flagrant.
  I have not tried to calculate the entire image in one machine (since it
takes so long), but I suppose that the problem would disappear if so.
  (I am using the switches +er and +sr to calculate just a half of the image).

  Have anyone else noticed something like this? May this be a bug or
something?

-- 
                                                              - Warp. -

Post a reply to this message

Nieminen Mika <war### [at] assaricctutfi> writes:

>   The problem is that when I turn the radiosity on, the lower part of the
> tree is much more brighter than the upper part. Everything else looks ok.
> When I join both parts of the image, the difference is flagrant.

  I once heard that the first draft taken by POV-Ray about radiosity
calculates an average on the global luminosity of the image, then uses it as
a parameter for recalculating the whole thing. So, if your halves do not
have the same average lighting, the parameters taken to calculate each of
them (during the final pass) will differ, inducing the differences you
see.

>   I have not tried to calculate the entire image in one machine (since it
> takes so long), but I suppose that the problem would disappear if so.

  I am afraid you will have to... During the development phase, however, try
lower resolutions or use "light" parameters for the radiosity, to get a
convenient speed.

Happy ray-tracing,
Roland.
-- 

bob### [at] casimirrezelenstfr -- Linux, POV-Ray, LaTeX

Post a reply to this message

In article <6kkin7$hcs$1@oz.aussie.org>, war### [at] assaricctutfi says...
>
>  I'm making an image for the IRTC and I have a problem.
>  I am using radiosity and the calculations take very long. I have access
...
>  There is a tree in the image. I'll try to ascii-draw the situation here:

If there's a tree, I assume it's an exterior scene ("nature" topic).
Generally, radiosity is hardly needed for exterior scenes.
If your image really needs indirect lightning, you could try to fake it with
a large area_light; much quicker...

Images that benefits from radiosity are rare, and must be set to use radiosity
from the start.  Radiosity don't adds quality if the images hasn't been tuned
accordingly...

Cheers,
Fabien.

Post a reply to this message