[Radiance-general] Modelling an artificial sky : help needed

[Greg Ward]

Hi Ljubica,

I just want to add to what Thomas wrote.  Quoting your webpage:

> Materials are basic :
> walls are mirrors (80% reflexion - mirror material)
> ceiling is translucent (60% transmission all diffuse, trans material)
> floor is gray (60% reflexion, plastic material)
> half-sphere is blue plastic

Since your trans surface blocks visibility of the light sources, I  
wouldn't recommend using the mirror type for this at all.  I'd just  
use an 80% reflective metal material (pure specular) and let the  
interreflection calculation handle it.  I don't think an illum is  
suitable, due to the high variance in luminance over the surface of  
each panel.

I'm actually quite surprised at how bad these results look.  I think  
something else must be going on that I don't understand, like leaks  
at the sides of the structure.  Anyway, try it without the mirror  
type and see if that helps.

Your calculation time could probably be improved without too much  
loss in accuracy by using a glow material for the cylindrical light  
sources -- set the distance limit to something like 4 times the size  
of one of your light panel widths.

Finally, I don't think a pure diffuse translucent material can have a  
transmission over 50%.

Best,
-Greg

> From: jeando <jdlenard at deluminaelab.com>
> Date: October 22, 2008 2:04:22 AM PDT
>
> Hi,
>
> We are building an artificial sky for an architecture school where  
> we teach, and before starting we prepare a Radiance model, to  
> simulate the interior of the artificial sky. The idea is to help us  
> to choose the right bulbs, the ceiling, etc.
>
> To simulate the overcast sky (Moon and Spencer) distribution in  
> Radiance we made walls of mirror material and ceiling is filled  
> with fluorescent tubes (very close to one another) and covered with  
> translucent material. Then we tried to render few images of  
> interior, with  several sets of rendering parameters, but final  
> images still aren't good enough with a lot of spots.
>
> Please see our results on http://docs.google.com/Doc? 
> id=dfvbdbg3_13f66n99kx
>
> Does anyone on the list have an experience with similar modelling,  
> where mirror objects are included?
> Why do we get strange spots on the floor even when ambient  
> parameters are really high?
> What combination of parameters may lead us to good image, for  
> acceptable rendering time?
>
> Thanks in advance,
> Ljubica, JeanDo, Marija