Fast, Realistic Lighting and Material Design using Nonlinear Cut Approximation

We present an interactive rendering system for realistic lighting and material design under complex illumination with arbitrary BRDFs. Our system smoothly integrates accurate all-frequency relighting of shadows and reflections with dynamic per-pixel shading effects such as bump mapping and spatially varying BRDFs. This combination of capabilities is typically missing in current systems. We build upon a clustered piecewise constant representation called cuts to accurately approximate both the illumination and precomputed visibility as nonlinear sparse vectors. Our key contribution is an efficient algorithm for merging multiple cuts in a single linear traversal. We use this algorithm to simultaneously interpolate visibility cuts at each pixel, and to compute the triple product integral of the illumination, interpolated visibility, and dynamic BRDF samples. The theoretical error bound of this approach can be proved using statistical interpretations of cuts. Our algorithm extends naturally to computation with many cuts; moreover, it maps easily to modern GPUs, resulting in a significant performance speedup over existing methods based on nonlinear wavelet approximations. Finally, we present a two-pass, data-driven approach that exploits pilot visibility samples to optimize the construction of the light tree, leading to more efficient cuts and reduced datasets.