Photon shadingmodels: interior VS exterior

[Aron Gombas]

Hi All!

I played a bit with shadingmodels in an interior scene. See my test-images at www.demoscene.hu/~aron/phototests.zip

Basically I changed the shadingmodel of the window-glass polygons:
- matte: no photons inside -> seems OK
- glass, water: some photons inside -> seems OK
- transparent: no photons inside -> bug?

My other question: why are the photon-maps this coarse inside, while using the same no of photon for an outside scene I got much more "dense" photonmaps? The only difference, that the interior scenes are closed geometries, only the window polygons have the photon-model shown above.

Aron

[Aron Gombas]

Sorry, a type-o again, the correct URL is: http://www.demoscene.hu/~aron/photontests.zip

[Aron Gombas]

I'm not sure if it's related here, but using occlusion() I got almost completely black ambient occlusion for an inside scene:

www.demoscene.hu/~aron/!Interior-occ_100K.png

The funny is that outside, the eaves looks correct.

[Okan Arikan]

   Hi Aron,

   The "occlusion" function does not execute the surface shaders and does not check the surface opacity either where the rays hit. I can check for opacity if it is what PrMan does.

   Thanks

   Okan

[Okan Arikan]

   Hi Aron,

   While tracing the photons, the surface opacity is not checked. If you want a surface to be transparent to the photons, you need to mark them as

   Attribute "photon" "shadingmodel" "transparent"

   Okan

[Aron Gombas]

> Attribute "photon" "shadingmodel" "transparent"
This is exactly what I did, but it seems photons can go through those polygons...
The photon-maps are all-black, and if I use a single occlusion(), I get a completely black image.

That's why I wrote above:
> - transparent: no photons inside -> bug?

Aron

[Aron Gombas]

Sorry, my previous makes no sense at all, what I wanted to write is:

> This is exactly what I did, but it seems photons can NOT go through those polygons...

Sorry.

Aron