Ray tracing animated scenes using coherent grid traversal

Abstract

Figure 1: Several animated models ray traced using our coherent grid traversal: a) A gesturing hand of 16K triangles. b) An animated "Poser" model (78K triangles). c) Animated wind-up toys (11K triangles) walking and jumping incoherently around each other. d) A rigid-body dynamics simulation of marbles (8.8K triangles). e) A complex scene of 174K animated triangles, where a fairy and a dragonfly dance through an animated forest. Scenes are rebuilt from scratch every frame, allowing fully dynamic animation. Including shading, texturing, and hard shadows, as used in the above images, we can render these scenes at 1024 × 1024 pixels with 15.3, 7.8, 10.2, 26.2, and 1.4 frames per second on a dual 3.2 GHz Xeon. Excluding shading, texturing, and shadows, we achieve 34.5, 15.8, 29.3, 57.1, and 3.4 frames per second. AbstractWe present a new approach to interactive ray tracing of moderatesized animated scenes based on traversing frustum-bounded packets of coherent rays through uniform grids. By incrementally computing the overlap of the frustum with a slice of grid cells, we accelerate grid traversal by more than a factor of 10, and achieve ray tracing performance competitive with the fastest known packet-based kd-tree ray tracers. The ability to efficiently rebuild the grid on every frame enables this performance even for fully dynamic scenes that typically challenge interactive ray tracing systems.

Coherent multiresolution isosurface ray tracing

“…Recent works in coherent ray tracing [16], [17], [18] combined highly-optimized coherent traversal with SIMD primitive intersection to deliver up to two orders of magnitude increase in frame rate, allowing interactive ray tracing on a single processor. For faster isosurface ray tracing, the ray-cell intersection test was adapted to a SIMD SSE architecture by Marmitt et al [19].…”

“…Successful coherent systems have been optimized for relatively small dynamic polygonal data [18], [17] in which many rays intersect common primitives. For sufficiently complex scenes, where each ray intersects a different primitive, intersection costs at least as much as it would in the single-ray case.…”

“…COHERENT OCTREE VOLUME RAY TRACING Having constructed a compact octree volume with an embedded min/max acceleration structure, we now turn to the task of building a coherent ray tracing system. In general, we seek to optimize for coherence as aggressively as possible, namely by implementing a vertical SSE packet architecture and a frustum-based octree traversal similar to the coherent grid traversal of Wald et al [18].…”

“…With coherent ray tracing, we are given the choice between depth-first traversal similar to a kd-tree [21] or BVH [17]; or a breadth-first coherent grid traversal (CGT) approach [18]. We choose the latter for several reasons.…”

“…The original CGT paper [18] implemented a two-level hierarchy, with a single layer of macrocells each corresponding to 6 grid cells. For small polygonal data, this was generally sufficient.…”