ohmaya opened this issue a year ago · 2 comments
Enough time : )
This does not really affect me.
Thanks for point it out. This is probably due to a bug in the sampling/evaluating of the Beckmann BSDF model; which ultimately results in NaNs. I will look into it later.