[Radiance-general] New member
Gregory J. Ward
gregoryjward at gmail.com
Thu May 18 18:32:19 CEST 2006
Actually, this is less true than it used to be. Even for highly
reflective, enclosed scenes, Radiance 3.7R2 with Russian-roulette
enabled (-lr -20 or so) and no irradiance caching (-aa 0), you can
generate many bounces to get a converged result.
Of course, you are better off with a zonal cavity approximation at
that point!
-Greg
P.S. You should seriously consider picking up a copy of "Rendering
with Radiance" from BookSurge.com <http://www.booksurge.com/
product.php3?bookID=IMPR02587-00002>. It will make your life easier,
and I'm not just saying that because it makes me money. (Actually, I
don't think I've gotten any money from this yet -- where's Chas?)
> From: John Mardaljevic <jm at dmu.ac.uk>
> Date: May 18, 2006 9:14:40 AM PDT
>
> Chris,
>
> What sort of "lighting chamber" do you have in mind? Radiance is
> optimised for plausible architectural scenes and, for those, it's
> proven to be capable of high accuracy. However, it might not
> deliver comparably reliable results (at least, not without some
> intervention) if the scene is somewhat 'exotic'.
>
> -John
More information about the Radiance-general
mailing list