Implementation Question

[Crane-YU]

Hi @Kai-46, In your actual implementation of the lobe sharpness in Eq. (8) in the figure below, why do you treat the half angle vector $\mathbf{h}$ the same as lobe axis in L185?

[Exception1984]

I read the paper and compared it to the implementation, leading to the exact the same question as asked by @Crane-YU. What I could already find out by checking the code is that the halfway vector stored in new_half is just the surface normal, since warpBrdfSGLobes is the reflexion of the view vector w.r.t. the normal, so taking warpBrdfSGLobes + viewdirs is again the normal. I really do not understand why this is done like that, is it possible to get more information on that?