[Radiance-general] solarbr

[Greg Ward]

Hi Axel,

It's probably fair to say that gensky can't be trusted to produce 
accurate absolute luminances for the sun OR the sky at ANY altitude.  
Skylight is simply too variable to be pinned down, even on average, 
without considering the local climate.  In the case of low altitude 
sunlight in particular, there are a lot of highly variable atmospheric 
effects at play, which is why we don't see many reruns of the same 
sunset.  In the code for gensky, I was going by some general ballpark 
values taken from Wyszecki and Stiles (Color Science), and it's a 
really crude estimate.

For accurate values, you'll have to come up with your own "sources" and 
set the -B and -R options accordingly.

-Greg

> From: "Axel Jacobs" <a.jacobs at londonmet.ac.uk>
> Date: March 29, 2004 6:56:19 AM PST
>
> Dear all,
>
> if I read gensky.c (line 267) correctly, then the solar brightness is
> constant for sin(sun altitude) below 0.16 which is an altitude of about
> 9.2deg.
> Does that mean to say that RADIANCE skies can't be trusted for (very) 
> low
> angle sun?
> Is there a scientific, computational, or philosophical reason behind 
> this?
>
> Cheers
>
> Axel