The SGGX Microflake Distribution

摘要

We introduce the Symmetric GGX (SGGX) distribution to representrnspatially-varying properties of anisotropic microflake participatingrnmedia. Our key theoretical insight is to represent a microflake distributionrnby the projected area of the microflakes. We use the projectedrnarea to parameterize the shape of an ellipsoid, from which wernrecover a distribution of normals. The representation based on thernprojected area allows for robust linear interpolation and prefiltering,rnand thanks to its geometric interpretation, we derive closed formrnexpressions for all operations used in the microflake framework.rnWe also incorporate microflakes with diffuse reflectance in our theoreticalrnframework. This allows us to model the appearance ofrnrough diffuse materials in addition to rough specular materials. Finally,rnwe use the idea of sampling the distribution of visible normalsrnto design a perfect importance sampling technique for our SGGXrnmicroflake phase functions. It is analytic, deterministic, simple tornimplement, and one order of magnitude faster than previous work.