But with mirrors etc. the number of normals to be stored would become HUGE,
requiring massive amounts of memory. Monte Carlo sampling around the curent
intersection would give nearly the same result.

Margus

Mike wrote:
> 
> Since every intersection has a normal calculation, what if you were to do a
> prepass step and store all the normals for the intersections within the
> image, then when you render estimate the curvature? I can see that there
> might be problems around the edges of the surface, but could this work
> somewhat?
> 
> -Mike

Post a reply to this message