from BRDF model to Renderman Illuminance

[parsival]

Hello,

I have a question about the porting of a BRDF model to Renderman Illuminance construct.

For example, when I have a physically based model of a BRDF and I want to develop a BRDF shader under sl language, I see pag. 52 of Siggraph 2001 Renderman course that Matt Pharr multiplies the BRDF by pi and N_dot_L.

Is it a right rule that I have to multpiply by PI and N_dot_L my BRDFs every time I want to port a BRDF model to Renderman Illuminance construct?

Thank you

Parsival

[Mario Marengo]

Is it a right rule that I have to multpiply by PI and N_dot_L my BRDFs every time I want to port a BRDF model to Renderman Illuminance construct?

14400[/snapback]

Hi Parsival,

Long story short: no; you don't need to multiply by PI.

The PI term results from solving the integral over the hemisphere, but then the surface albedo (ratio of reflected light to incident light) needs to be expressed as "albedo per PI", canceling out the PI term. If you look closely, you'll notice that he calls brdf_diffuse() with albedo/PI, where albedo is kd*Cs... and so "bye bye PI"

Regarding the dot product N.L, the answer is normally "yes; you need to multiply by that". That particular bit represents "the amount of incident light per unit area" (the "per unit area" thing is important). Irradiance varies in relation to the angle between the surface and the light -- specifically, it varies according to the cosine of the angle between the two, which is what that dot product gives you. Just keep in mind that for that term to "work as advertised", it is assumed that both N and L are normalized; so the actual term is dot(normalize(L),normalize(N)).

Hope that helps,

Cheers!