A new method for interactive rendering of complex lighting effects combines two algorithms. The first performs accurate ray tracing in heterogeneous refractive media to compute high-frequency phenomena. The second applies lattice-Boltzmann lighting to account for low-frequency multiple-scattering effects. The two algorithms execute in parallel on modern graphics hardware. This article includes a video animation of the authors' real-time algorithm rendering a variety of scenes.

We present a novel technique for physically based rendering of participating media like cloud, smoke, wax, marble, etc. We solve the radiative transfer equation (RTE) for participating media using the Modified Discrete Ordinate Method (MDOM), which computes the final solution as a combination of a direct and an indirect component. We propose a scalable GPU based parallel pipeline, for solving the RTE using the MDOM. This parallel RTE solver is capable of rendering intermediate results such as single scattering approximation. We overcome GPU memory size limitations by using low resolution radiance storage while doing high resolution radiance propagation. Furthermore, we achieve scalability by implementing an efficient volumetric data streaming mechanism. Our results demonstrate the ability of our method to render high quality multiple scattering effects.

This paper presents an accurate method to compute the bidirectional reflectance distribution function (BRDF) due to subsurface scattering inside the material of the objects. This computation requires iterating over the different lighting directions, and solving the integro-differential equation of light transport (scattering and absorption). Solving the light transport equation is expensive, and solving it independently for different directions adds even further to the expense. However most of the computations are very similar between directions. We make use of Green's function of the transport problem to have a better separation between computations that are independent of incident directions from those that are dependent. This allows us to avoid as much repetition in the computations as possible, thus gives us a faster BRDF computation method without any loss of accuracy. We validate our method against a standard light transport solver and use it to compute BRDF for a variety of materials.

No abstract