[Radiance-general] glow&spotlight material

[Greg Ward]

Hi Woogie,

What you are seeing is likely the result of a known anomaly in the 
hemispherical sampling pattern used by the indirect calculation in 
Radiance.  I think Charles Ehrlich was the first to point this out to 
me -- the polar sampling I use on the hemisphere means that the exact 
pole is never sampled, and there is a slight dip in the response to 
indirect surfaces directly overhead as a result.  Since this is how you 
arranged your scene, it is likely that this is the source of the odd 
dip in value towards the center of the surface illuminated by the glow 
source, as the glow only contributes to the indirect portion of 
illuminance.

To fix this problem would complicate the code and only reduce errors in 
some very odd circumstances.  There are many other circumstances that 
cause similar biases, so I never thought it worthwhile to change it.

-Greg