Christoph Hormann wrote:
> 
> The first one is testing one of Jaime's new HDR light probe, the other 
> the environment from the MegaPOV sample scenes.  The table is using 
> blurred reflection (with the average texture trick) which works very 
> well but is quite slow (1 hour at this size).

   Lovely blurred reflections... with a bit of banding on the brighter 
parts, but nice anyhow. It's a pain it can be made faster...

> The scene contains different objects for radiosity (without reflection, 
> refraction using no_reflection and no_image flags) and normal appearence 
> (no_radiosity flag) so i don't need a two pass render.  Furthermore i 
> tested some new improvements of the radiosity function.

   Very interesting tips... really worth trying this weekend, I think. 
:) The rad seems very good, and you don't seem to get the artifacts on 
the reflections that I get when using two passes (being the first 
without reflective textures).

--
Jaime

Post a reply to this message