Great looking images!

It is really cool to see all of the creative ways that people are testing
this patch!!!!

For the reflections one, you probably will want a higher density of photons.
How many photons were used for that one, by the way.  It does look good the
way it is though... looks like an area light is being used.

Which reminds me, I never documented this:
The photon mapping patch currently treats all kinds of lights like point
lights.  This, of course, will eventually be changed.

But this brings up an interesting point... to simulate an area light, just
use a lower density of photons (doesn't always work).

The reflections pic also shows off one 'bug' with the patch.  You can see
some weird effects between the reflections of the F and L (see my ascii
art... the plus marks show where to look in the image for the artifacts,
which appear brighter than they should).

 ******  **
 ******  **
 **      **
 **      **
 ******  **
 ******  **
 **      **
 ** +++  **
 ** +++  ******
 ** +++  ******

In my final documentation, I'll explain why this happens.  But for now,
to fix this problem, you can either a) increase the density of the photons,
or b) use only one search radius:

if it was
  radius 0.1, 2, 0.1
change it to
  radius 0.2, 1

-Nathan

Shinzaho wrote:
> 
> Here are two of my photon mapping experimens one using refractive caustics
> and the other using reflective caustics.  The glass image uses reflective
> caustics as well, but the contribution to the scene is minimal.
> 
>  [Image]
> 
>  [Image]

Post a reply to this message