Research Projects

Stochastic Lightcuts


The Conference scene illuminated with 1400 light sources on the ceiling, rendered with direct illumination only using (left) lightcuts and (right) stochastic lightcuts with up to 10 light samples for estimating the illumination from all lights. Notice that the sampling correlation of lightcuts leads to severe virual artifacts.

Abstract

We introduce stochastic lightcuts by combining the lighting approximation of lightcuts with stochastic sampling for efficiently rendering scenes with a large number of light sources. Our stochastic lightcuts method entirely eliminates the sampling correlation of lightcuts and replaces it with noise. To minimize this noise, we present a robust hierarchical sampling strategy, combining the benefits of importance sampling, adaptive sampling, and stratified sampling. Our approach also provides temporally stable results and lifts any restrictions on the light types that can be approximated with lightcuts. We present examples of using stochastic lightcuts with path tracing as well as indirect illumination with virtual lights, achieving more than an order of magnitude faster render times than lightcuts by effectively approximating direct illumination using a small number of light samples, in addition to providing temporal stability. Our comparisons to other stochastic sampling techniques demonstrate that we provide superior sampling quality that matches and improves the excellent convergence rates of the lightcuts approach.

Images


The Crytek Sponza scene illuminated with 1644 light sources, rendered with path tracing using (left) lightcuts and (right) stochastic lightcuts with up to 10 light samples for estimating the illumination from all lights. Notice that the sampling correlation of lightcuts leads to severe virual artifacts.


The Sibenik scene illuminated with one million virtual light sources, rendered using (left) lightcuts and (right) stochastic lightcuts with up to 10 light samples for estimating the illumination from all lights. Notice that the sampling correlation of lightcuts leads to severe virual artifacts.


Comparison of different sampling methods for lighting estimation from many lights, showing that stochastic lightcuts provides the lowest noise. The scene contains 10,000 virtual lights and it is rendered using 100 light samples per pixel.

Videos

Stochastic Lightcuts, High-Performance Graphics 2019 paper video.

Stochastic Lightcuts paper presentation at High-Performance Graphics 2019.

Publications

Cem Yuksel
Stochastic Lightcuts
High-Performance Graphics (HPG 2019), Wolfgang Straßer Best Paper Award
Revised Paper - 15.5MB
Revised Supplemental Document - 15.3MB
Publisher's Page
Youtube Video
BibTeX
* The revised paper has been changed from the published version to include a note regarding the sampling bias of dead branch avoidance.