SPPM Testing

[SATtva]

I see. Thanks for the explanation.

[J the Ninja]

I think lowdiscrepancy with 1 pixelsample is logically the same as random/linear, just slower.

When using sppm, those aren't even used. It has it's own internally defined samplers. I just do that because at one point it was causing an issue when you left it as metropolis. But I don't think that's an issue now.

If you have the direct light sampling option enabled, selecting anything other than random makes it crash.

[binarycortex]

I think lowdiscrepancy with 1 pixelsample is logically the same as random/linear, just slower.

When using sppm, those aren't even used. It has it's own internally defined samplers. I just do that because at one point it was causing an issue when you left it as metropolis. But I don't think that's an issue now.

If you have the direct light sampling option enabled, selecting anything other than random makes it crash.

For the win. Well that explains my crashes.

[giuliojiang]

I have a question: does SPPM support dispersion (like with Cauchy fresnel texture)?
If not, will this feature be added in the future releases?

[SATtva]

I have a question: does SPPM support dispersion (like with Cauchy fresnel texture)?

Of course it does. Currently it only lacks support for scattering media.

[B.Y.O.B.]

I have a question: does SPPM support dispersion (like with Cauchy fresnel texture)?

Of course it does. Currently it only lacks support for scattering media.

Have a look: gallery2.php?g2_itemId=17086

[giuliojiang]

Have a look: gallery2.php?g2_itemId=17086

thanks
I have another question.
In the description of the image I read
"Classic cubo render. Test of SPPM in the dev branch. Photons per pass set to 100M, Starting radius 0.75 alpha 0.7 photon depth 1000."
by default in Lux1.0RC it's set to 48. Do I need to higher this value to get correct colored caustics?

[J the Ninja]

Have a look: gallery2.php?g2_itemId=17086

thanks
I have another question.
In the description of the image I read
"Classic cubo render. Test of SPPM in the dev branch. Photons per pass set to 100M, Starting radius 0.75 alpha 0.7 photon depth 1000."
by default in Lux1.0RC it's set to 48. Do I need to higher this value to get correct colored caustics?

No, you shouldn't. 1000 is huge overkill.

[giuliojiang]

I still can't get correct dispersion

Bi-Directional 12 KS/p


SPPM 1h on my pc


I also increased the Cauchy B value to get more color, but I still can't see anything

'store on glossy' is ON. Does it affect the result?

This is my .blend scene file. Have I misconfigured something?
https://docs.google.com/open?id=0BwOqPT ... VFJZUREMFE

[binarycortex]

How big is your gem, in blender units? Please remember that 1BU = 1 meter.