hello,

in your opinion, why both sides of the cube are so darker than the top 
face ?

one plane, one sky_sphere and the cube.
no light.
use of standard "Radiosity_Final" parameter.
all object are white
ambient is set to zero in de default settings



thanks for your help.

-- 
klp

Post a reply to this message

Attachments:
Download 'cube_radiosity_final.jpg' (47 KB)

Preview of image 'cube_radiosity_final.jpg'

> in your opinion, why both sides of the cube are so darker than the top
> face ?
>
> one plane, one sky_sphere and the cube.
> no light.
> use of standard "Radiosity_Final" parameter.
> all object are white
> ambient is set to zero in de default settings

The light is coming from the sky sphere, and you have blocked the lower half of
that sphere with your plane.  Therefore, the sides are only exposed to roughly
half of the amount of light that the top gets.

Post a reply to this message

"Kirk Andrews" <kir### [at] tektonartcom> wrote:
> The light is coming from the sky sphere, and you have blocked the lower half of
> that sphere with your plane.  Therefore, the sides are only exposed to roughly
> half of the amount of light that the top gets.

I just thought the same as Kirk. But how did you came to this image? Mondrian in
black and white as a stock photo? A very nice minimalistic scene IMO.

Best regards,
Michael

Post a reply to this message

kurtz le pirate <kur### [at] yahoofr> wrote:
> hello,
>
> in your opinion, why both sides of the cube are so darker than the top
> face ?

Kirk is right: the ground plane is blocking light from the sky_sphere's lower
half.

Nice image. How did you produce the cube object, BTW? With an algorithm, or
hand-placed boxes?

Sam

Post a reply to this message

kurtz le pirate <kur### [at] yahoofr> wrote:
> hello,
>
> in your opinion, why both sides of the cube are so darker than the top
> face ?
>
> one plane, one sky_sphere and the cube.
> no light.
> use of standard "Radiosity_Final" parameter.
> all object are white
> ambient is set to zero in de default settings

My first thought was the same as the others--that the ground plane is blocking
half the light.  But then I thought, if the ground plane is white, isn't it
reflecting the same light as the sky?

It turns out, if the diffuse on the cube and plane is set to 1.0 /and/ the
radiosity recursion_limit is set high enough (5 in my trials), then all sides of
the cube are the same shade.

Post a reply to this message

kurtz le pirate <kur### [at] yahoofr> wrote:
> use of standard "Radiosity_Final" parameter.

I just looked at Radiosity_Final, and it sets recursion_limit to 1.  If I
understand radiosity sufficiently, this means that the cube is not getting the
effect of the light reflected off the ground plane.  (At the first bounce,
radiosity is calculated only from non-radiosity illumination, of which there is
none in your scene other than the sky_sphere.)  This is why the ground plane
ends up blocking half the light, despite being white itself.

I never use recursion_limit 1 for a final render.  2 is usually sufficient, but
in your case, you'll need to go higher.

Post a reply to this message

hi,


it seems that the problem comes from the plane that block the light.


with the comment of 'Cousin Ricky', i'm testing changing some values
in /Radiosity_Final/ settings :
 * count 800 -> 50 //  just for speed in tests
 * error_bound 0.20 ->  0.05 
 * recursion_limit 1 -> 2
 * brightness 1.0 -> 0.80

these parameters gives this new image which is much better.
i will continue my tests by adjusting other param/values.



about image itself, just an idea like that, i have not thought of 
Mondrian. this image as hand-placed boxes, but, I had already made an 
algorithm for generating Mondrian pictures.

you can see it on my page <http://louisbel.free.fr/scenes/scene021.shtml>
(in french)



thanks a lot for your help.

Post a reply to this message

Attachments:
Download 'cube_radiosity_modif.jpg' (44 KB)

Preview of image 'cube_radiosity_modif.jpg'