ABX wrote:
> Somehow I feel it incorrect. Transition from dark to light on the same
> areas seems unnatural but it can be illusion and/or result of blur. 

  Yes, I feel also the same.  It's a side-effect of the trick used, but I 
will try to minimize it. Blur doesn't has nothing to do with it, and 
perhaps helps hidding the fact.

> Before showing what trick you have used (I assume it gives very fast 
> rendering) please send real radiosity version for comparision.

  That's the problem, and my motivation to do the trick: I can't get that 
kind of lighting only with pure rad and one light_source outside. Surely 
it's possible, but I've not the patience needed to find and test the high 
quality settings... I'm very lazy, do you remember? :)

  The original image on the first post uses an extra area_light placed 
carefully on the window hole, to simulate the external rad coming in, with 
"Radiosity_Normal". The image called "faked_rad_normal.jpg" shows the same 
setup with only the outside light (almost no visible rad effect). The image 
"faked_rad_no-aa.jpg" uses also the same tricky extra light, but without 
the area_light statement (to show better how it is placed). 

 I've tried this trick on my "alchlab" image done with Megapov, but seems 
that it works better with 3.5. I will try to add an spotlight statement 
automated to follow the sun position, to give it a more directional look.

-- 
Jaime Vives Piqueres

La Persistencia de la Ignorancia
http://www.ignorancia.org

Post a reply to this message

Attachments:
Download 'faked_rad_no-aa.jpg' (9 KB) Download 'faked_rad_normal.jpg' (3 KB)

Preview of image 'faked_rad_no-aa.jpg'

Preview of image 'faked_rad_normal.jpg'