[giacob]

glass ncrown, absorbance 0, abbe 200, ior 1,5,
they let pass the same amount of light as their frame..
that is not phisically correct!!!!

[giacob]

do you happen to have any comparison images to show? so we all could see the problem

here u are the reference image

[Fernando Tella]

I've managed to get caustics behind a glass but with an emitter placed very close to it; anyway it seems too difficult for dielectrics to let the light through. It looks like they are much more opaque to light in this version.
Is this a bug or what?

[giacob]

i precise that caustics r on

[Fernando Tella]

With all caustics on.

What I see is:

-in the first samples dielectrics produce as much shadows as a diffuse object;
-after some samples dielectrics start to let through some dots of light caused by caustics (very slowly) which clear up the shadow (I only see clearly the effect with an emitter very close to the object);


why is so hard for a thin box of dielectric (0,003m) to let the light through? It should be almost like direct light because, in reality, it does not distort the light ray so much.

In my tests I had a sphere of dielectric too. It let the light through faster than the thin box.

[Fernando Tella]

Something is strange in caustics. In this image you can see the caustics caused by an emitter. I've placed some thin boxes (0,003m), the same box rotated +45º (more or less) and -45º, two bent boxes ")" and "(" and two spheres (one of them is booleaned by a cylindre) plus some diffuse objects. Caustics go through some angles only.
The dielectric material is "Glass - Crown of Borosilicate BSC1"
Absorbance 0,11
uv roughness 0,00
Abbe 56,0
Nd 1,511
Color white

Test done with 3dsmax plugin and beta1.2

sample level 17

[Fernando Tella]

I guess this post should be in "Bug reports"