Retrieving Material Properties of Implicit Surface Acceleration Structures - Colorizer: A library for retrieving Material Properties of Implicit Sur faces generated from Triangle Scenes

Abstract

Distance field traversal offers a hardware-independent path for real-time ray–surface intersection, avoiding the specialized hardware that triangle-based ray tracing requires. Libraries such as AMD FidelityFX Brixelizer efficiently resolve these intersections, but provide no mechanism for retrieving the surface properties — albedo, emission, roughness — that illumination algorithms require. This thesis presents Colorizer, a C++/DirectX 12 static library that addresses this gap by maintaining a world-space surface property cache. The cache organizes the scene into a camera-centred circular grid of chunks, each decomposed into a voxel grid whose six cardinal faces form local inverse cubemaps. Surface properties are baked into two-dimensional texture atlases via a scan-conversion rasterization pass, using an atomic depth sort to resolve per-voxel fragment priority. A spatial hash maps voxel face tuples to atlas tiles, and three layered error corrections — shell expansion, weighted bilinear reconstruction, and cardinal face fallback — absorb the imprecision inherent to distance field hit points. The cache is updated amortized across frames, targeting real-time budgets without requiring full per-frame coverage. The system is evaluated in the Amazon Lumberyard Bistro scene on an AMD Radeon RX 7800 XT. Library overhead is measured at 0.12 ms per frame, and surface cache sampling is competitive with barycentric reconstruction against full resolution textures. Bake jobs average 0.25 ms with a P95 of approximately 0.6 ms at an 8 m chunk size. The cache produces plausible albedo and emission backing for diffuse and high-roughness global illumination, with the error correction stack empirically absorbing predicted hash collisions at fill rates up to 37.9%. The primary remaining artifact is intra-voxel leakage, addressable through a separate sparse emissive cache structure proposed in the discussion.