Page 1 of 1

[fixed bug] max_extinction_coeff explained

Posted: Thu Oct 15, 2015 11:28 pm
by CTZn
I'm pretty sure this bug was squashed already, unless if it is depending on some local circumstances (not a systematical bug) because my clouds scenes are rendering alright iirc.

So here is a mesh with a scattering shader output aiming at 2,500.0; it renders clear, while setting scattering to a uniform spectrum of 2500.0 will give the expected result.

I also tried crazy high absorbtion with uniform values and the results were not convincing on this human head sized object... until I deactivated scattering; wether shader or not it is bugging absorbtion.

I'm not sure what this is about. Someone else noticed ?

Spotted with Indigo 3.8.16, present in 3.8.26.

Re: scattering shaders clamped & scattering breaking absorbt

Posted: Sat Oct 17, 2015 12:42 am
by OnoSendai
I think you need to set max extinction coeff properly.
It should be set, in this case, to 2500.

Re: scattering shaders clamped & scattering breaking absorbt

Posted: Sat Oct 17, 2015 2:21 am
by FakeShamus
can this be explained a bit more? I never know what to set the max extinction coeff so it's always left default, but how exactly is it used? are there other situations where it should be set differently?

Re: scattering shaders clamped & scattering breaking absorbt

Posted: Sat Oct 17, 2015 5:12 am
by OnoSendai
max_extinction_coeff should be set to the maximum extinction coefficient that the medium will take at any point.
The extinction coefficient is the sum of the absorption and scattering coefficients.

max_extinction_coeff determines how large a step along the ray Indigo can take when it is calculating participating media. The larger max_extinction_coeff, the smaller the steps need to be.

max_extinction_coeff is only used when the medium is heterogeneous - that is when then the scattering or absorption coefficient varies with position. This will be the case when a shader is used for the absorption or scattering coefficients.

Re: scattering shaders clamped & scattering breaking absorbt

Posted: Sat Oct 17, 2015 8:58 am
by CTZn
Ah well we have learned something new today, plus it's not a bug, great !
We can take, and actually need to know such tech details.

Thanks !