[Radiance-general] Re: Aiming Failures and Geometry

Greg Ward [email protected]
Tue, 30 Sep 2003 09:05:16 -0700


Hi Marcus,

Illum's, like all light sources in Radiance, are preferred as 
rectangles or (at least) 4 or more sided convex shapes.  Since Radiance 
samples a rectangle with equivalent area to the source shape, triangles 
are a particularly poor choice.  I recommend replacing any windows you 
plan to use as illum's with rectangles for that reason.

-Greg

> From: "Marcus Jacobs" <[email protected]>
> Date: Tue Sep 30, 2003  8:26:53  AM US/Pacific
>
> I have a question concerning an issue that I am encountering with 
> illum objects that were created with mkillum and the type of polygon 
> that is used in conjunction with it. I perform quite a bit of my 
> modeling with 3D Studio VIZ. This is more or less a "lite" version od 
> 3D Studio Max. As many of you may already know, virtually all of the 
> geometry no matter whether it's a sphere, cylinder, plane, etc. is 
> composed of a multitude of triangles. To create my windows, I 
> generally model them as a plane. When I export the scene to RADIANCE 
> (using ConRad), all of the planes are converted to individual 
> triangles. I have no problem using mkillum to create the illum 
> modifiers for these polygons. The problem comes when using rpict. Once 
> rendering of the scene begins, a bunch of "Aiming failure" error 
> messages are generated for all of my window polygons. It doesn't 
> matter what value for -dj I use, this error message still occurs. As 
> an experiment, I decided to create the windows in AutoCAD. Using 
> Desktop RADIANCE's conversion, the windows planes were translated to 
> Radiance as four sided polygons (as opposed to three when modeled in 
> 3D Studio). The polygons were modified as illums with mkillum and when 
> rendering with rpict I had no "Aiming failure" error messages. My 
> question is does anyone <coughGreg> know if Radiance has any 
> preferences for a polygon type when the polygon is modified by an 
> illum primative?
>
> Thanks
> Marcus