tizian/ltc-sheen

Code for "Practical Multiple-Scattering Sheen Using Linearly Transformed Cosines" (SIGGRAPH 2022 Talk) by Tizian Zeltner, Brent Burley, and Matt Jen-Yuan Chiang

ltc-sheen's Issues

A formula problem in sheen model by patry

According to the slides from "Samurai Shading in Ghost of Tsushima"
https://blog.selfshadow.com/publications/s2020-shading-course/patry/s2020_pbs_patry_slides.pdf,
on page 52 , it can be seen that the denominator for the exponent part of e is udotv * vdotn.

While in sheenbaselines.cpp, class SheenPatry :

Float tmp1           = CosTheta(wo + wi),
      tmp2           = tmp1 / (cosThetaI * cosThetaO),
      numerator      = 1.f - std::exp(-density * tmp2);

the denominator here I suppose is equivalent to udotn * vdotn?
Is this a mistake here? If not, what's the reason?

As I am not familiar with pbrt, please forgive and correct me if I make any mistakes.

Thanks!