Hey DIVA team-
i've run into an interesting, and seemingly frustrating issue. I'm studying a vertical courtyard with reflectors at the top, and finding that when i make the reflectors specular (metal) the calc freaks out and i get large regions of 48384lux (point-in-time calc) or 340 (daylight factor). Below are 2 different runs of the same calc with different results. The surfaces at the top are the reflectors - im testing different shapes, this is just the flat one, which i already rotated in 1 degree to try and account for the freakout. When i make the material diffuse i don't have a problem and the results look correct. . it seems that the reverse raytrace can't find a solution? any ideas? i'd like to run daylight autonomy on the interiors but can't yet trust the results...
You have basically created a light-pipe, which Radiance alone is not capable of calculating. To do so, you need to use photon-mapping which is now integrated into DIVA.
For a bit more information look at the Radiance Online forum, for example:
hark! great to know! yes i figured it was something like this. thanks kera!
wait where is it now integrated? 4.0??
Kera is correct; however, usually you would expect less light to come through a light pipe in those situations, not an excess of light! I would suggest to actually make a visualization under a diffuse sky (same type as daylight factor) to try to see what is happening.
DIVA-4 does integrate photon mapping, but right now only for visualizations, not grid-based calculations.
very interesting indeed. i actually tried the uniform sky as well with same results.
in the mean time i'm using iray in 3ds max which does the job for point-in-time. i was hoping to do DA though which may be a bit beyond reality...
thanks for the insights.