[Radiance-general] Re: Radiance-general Digest, Vol 38, Issue 14

[Rob Guglielmetti]

Mark de la Fuente wrote:
> Steve,
>  
> Rob is pointing you in the right direction.  But be careful when you 
> wrap the fixture in an illum box as the light output may be quite 
> right.  It's been a while since I've had to do this, but I recall 
> having issues with "flatcorr" and "lboxcorr" in my brightdata 
> definitions (created by IES2RAD).  I'm not sure where these are 
> documented, but I'm sure other functions are used for spheres and 
> rings.  I believe the sphere option is the easiest to work out, but 
> only works if you can wrap your fixture in the illum sphere and the 
> sphere does not bisect geometry.  In my case I've had to work out 
> linear fluorescent pendants which have required a box.
Yeah, like I said, it's far from straightforward.  =8-)  Mark is right, 
the sphere is easiest, but since the illum enclosure must completely 
contain the luminaire geometry, it's not the best choice for a long 
skinny thing like a fluorescent strip.  The sphere ends up so big you 
introduce inaccuracies.  So those cal files lboxcorr and flatcorr take 
the ies2rad .dat output -- which is essentially a point distribution -- 
and wraps it around more apporpriate geometry: a box in the case of 
lboxcorr, (for a linear pendant fixture) and a flat rectangular polygon 
in the case of flatcorr (for say a recessed fixture where the luminous 
aperture is a single plane).

- Rob Guglielmetti