"Ron Parker" <ron### [at] povrayorg> wrote...
> > - Is a form of ray coherence used to speed up calculation?
>
> Some patterns do this.  Uppermost in my mind is the crackle pattern, which
> caches the centers for the most-recently-visited part of its perturbed
> lattice in case the next ray happens to hit the same part of the lattice.
> This is much faster than recalculating the centers for each ray.

Also, shadow calculations utilize a at least one cache to keep a
recently-used object around for quick intersection tests.

-Nathan

Post a reply to this message