[Radiance-general] Compute direct component from sky only

[John Mardaljevic]

Nick,

The direct sky component should be a snip.  I guess your points are  
in a horizontal plane, so the spawned rays will not see the ground  
glow and you can leave it in the description.  Set the following:

-ab 1    You need the first level of hemispherical sampling so that  
rays are sent to the 'glow' sky.
-aa 0    This switches off interpolation and is likely to give you  
more precise results (generally not advised when ab > 1).
-ad 1024 then 2048, if there's a significant difference, proceed to  
4096 and so on...
-as 256 or thereabouts.
-av 0 0 0
-ar - doesn't matter because aa is 0
No need to bother with the ae or aE parameters for the direct sky  
component.

And, of course, you have rtrace set to predict irradiation.

I would (strongly) advise against using mkillum to predict the direct  
sky component.  You can get arbitrarily high accuracy in your  
predictions with -aa 0 and cranking up -ad until the results have  
converged to whatever degree you think is sufficient.

Shameless plug for CIBSE Daylight Group session of possible interest:

Wednesday 12th. October 2005
Rights to Light:  A short-paper/discussion session to be lead by John  
Mardaljevic (DMU) and Paul Littlefair (BRE) with contributions from  
Andrew Thompson (Wilks Head & Eve), Delva Patman and Alistair Redler  
(Delva Patman), Alex Schaunowski and Ian Absolon (Schatunowski  
Brooks), Kaivin Wong (Malcolm Hollis) and Lance Harris (Anstey Horne).

14.00 for 14.30 at University College London, Torrington Place, London.

(With a line-up of that size, there is good chance that the speakers  
will outnumber the audience.)

-John

-----------------------------------------------
Dr. John Mardaljevic
Senior Research Fellow
Institute of Energy and Sustainable Development
De Montfort University
The Gateway
Leicester
LE1 9BH, UK
+44 (0) 116 257 7972
+44 (0) 116 257 7981 (fax)

jm at dmu.ac.uk
http://www.iesd.dmu.ac.uk/~jm