Typos in the discrete form of the rendering equation

[anning5]

Dear Filament team,

Recently when I was reading your documentation I found some potential typos in some of the equations in it. https://google.github.io/filament/Filament.html#toc5.3.4.2.2 equation 82 and https://google.github.io/filament/Filament.html#mjx-eqn-iblSampling equation 124 Both of them are missing 2PI in the equations, as the integrand needs to be multiplied by the integral domain(which is 2PI as the hemisphere surface area) over the sample count. So I think it should be 2PI/N rather than 1/N.

Could you confirm if they are typos or why the 2PI is absent in the equations please?

Many thanks Anning

[pixelflinger]

I think you might be correct -- but the 2pi factor must cancel out somewhere. For the LD term it's obvious in eq 85. But I can't find where it goes away for the remaining "DFG" term.

[anning5]

I see, thanks for the confirmation!