Re: [Brad] Help:Setting up a radiance system--brad

[Rob Guglielmetti]

On Apr 21, 2007, at 2:11 PM, steve michel wrote:

> Thanks Rob for ltview; it did allow me to quickly 'preview' a  
> lighting fixture via radiance and see the difference with brad's  
> output.
>
> I took the liberty of attaching my results.
>
> Thomas: your item #5 in the list of brad procedures worked  
> successfully to export my lighting data to radiance. For Brad  
> export  I positioned my lightsource close to a surface to 'see' its  
> 'signature'.  The obvious hypothesis for a and up/down T5 geometry  
> is a linear shaped illumination on a horizontal surface (one foot  
> from the ceiling) but I was getting a circular shape (see attached  
> file) and also strange, the lighting polygon is casting a shadow on  
> the cube (no other object in the room). It appears as though a  
> wrong circular is positioned above the self-luminous polygon.

One thing that ltview does is automatically set one of the Radiance  
parameters: -ds .15.  THe default for -ds is .25, which is too high  
to get any appreciable direct source subdivisions, which is really  
what you need with an IES file depicting a linear source.  So in  
ltview the distribution looks correct, but in your Brad rendering -ds  
must still be set too high.  Anytime you do a Radiance rendering  
using ies lights (opposed to daylighting-only) you really need to  
crank the -ds up, but it really becomes critical when you have a  
linear source close to a surface.

The shadow cast seems correct; there is a dim blob of light on the  
top of the box (from the uplight component of the fixture) and a  
brighter blob on the bottom half of the fixture (from the downlight  
component).  The box is so close to the fixture that the dead spot  
along the fixture housing is very pronounced.  THis is also  
exaggerated because the -ds is still too high, so your linear fixture  
is behaving like a point source. Turn up -ds and render your  
luminaire in a real scene where the luminaire is in a room but fairly  
far away from the objects in the room and I think you'll see what  
you're looking for.  You're getting there!

- Rob Guglielmetti