ABSTRACT

This chapter describes adaptive volumetric shadow maps (AVSM), a new approach for real-time shadows that supports high-quality shadowing from dynamic volumetric media such as hair and smoke. AVSM generates an adaptively sampled representation of the volumetric transmittance in a shadow-map-like data structure, where each texel stores a compact approximation of the transmittance curve along the corresponding light ray. AVSM can capture and combine transmittance data from arbitrary dynamic occluders, including combining soft media like smoke and well-localized denser media such as hair. The main innovation introduced by AVSM is a new, streaming lossy compression algorithm that is capable of building a constant-storage, variable-error representation of visibility for later use in shadow lookups. AVSM achieves its goal of adaptively sampling volumetric transmittance curves with performance high enough for real-time rendering throughout the Valve software scene. AVSM provide an effective, flexible, and robust volumetric shadowing algorithm for real-time applications.