The computer generated imagery you see in movies today is created by simulating the way light interacts with a virtual environment. On the one hand, these simulations require sophisticated mathematical models and numerical algorithms to maintain accuracy and speed. On the other hand, the users of these simulations are typically artists, and not physicists, so these technicalities need to be abstracted away to provide a usable artistic tool. This unique blend of constraints exposes interesting research questions at the intersection of applied math, computational physics, and art.

In this talk I’ll summarize some recent work we have been doing in my group at Dartmouth’s Visual Computer Lab on accurate, efficient, and controllable physical simulation of light transport. I’ll discuss how these projects have taken us on a tour of seemingly unrelated areas ranging from nuclear engineering to atmospheric sciences, and lead us to mathematical concepts like Fredholm integral equations, fractals, and stochastic processes.