Is it as a simple as generating a super-low LOD for a particular environment and all its dynamic objects, running a real-time, 2 - 3 bounce path tracer through just that low-def geometry, then reprojecting the result onto the high-def geometry? Or is it baked to a certain extent, in quadrants whose lighting information is then interpolated for a dynamic result at run-time?
Explain enlighten to me like I'm a child
- Thread starter inlimbo
- Start date
