-
Advanced Global Illumination Using Photon Mapping
Photon mapping provides a practical way of efficiently simulating global illumination including interreflections, caustics, color bleeding, participating media and subsurface scattering in scenes with complicated geometry and advanced material models. This class will provide the insight necessary to efficiently implement and use photon mapping to simulate global illumination in complex scenes. The presentation will briefly cover the fundamentals of photon mapping including efficient techniques and datastructures for managing large numbers of rays and photons. In addition, we will describe how to integrate the information from the photon maps in shading algorithms to render global illumination.
-
Bidirectional Path Tracing and Multiple Importance Sampling (2 Minute Papers)
With a classical unidirectional path tracer, we'll have some scenes where it is difficult to connect to the light source, and therefore many of our computed samples will be wasted. What if we would start not only one light path from the camera, but one also from the light source, and connect the two together? It turns out that we get a much more robust technique that can render a variety of "packed" scenes with lots of occlusions with ease.
-
Doug Crockford - Top Down Operator Precedence (aka Pratt Parsing)
In this Youtube video, the "top down operator precedence" method of parsing is discussed. In a nutshell, TDOP, also known as "Pratt parsing" is sort of like a combination of precedence climbing parsers and recursive descent parsers. Notably TDOP does not use a grammar, which makes it somewhat unique among other parser techniques. A working example by the author can be found here: https://github.com/douglascrockford/TDOP
-
Matrix row-column sampling for the many-light problem
Computing global illumination using the many-lights approach is expensive, this algorithm presents a method ("exploration") for reducing the amount of lights to be sampled.
-
Metropolis Light Transport (2 Minute Papers)
Metropolis Light Transport is a powerful technique that can outperform the convergence speed of Bidirectional Path Tracing on most difficult scenes (what makes a scene difficult is a story on its own). It promises optimal importance sampling "along multiple steps" in the stationary distribution of the Markov chain.
-
Path Tracing and Next Event Estimation (2 Minute Papers)
A description of Path Tracing, a ray tracing solution to the rendering equation, and one of its traditional optimizations, Next Event Estimation.
-
Photon Mapping (2 Minute Papers)
Describes the Photon Mapping solution to light transport.
-
Practical Global Illumination with Irradiance Caching
Lighting usually only changes smoothly over a surface, Irradiance Caching is a classic approach to reduce the computational cost by only computing lighting at points where it changes quickly.
-
Radiosity
Lecture describing how to solve the radiosity equations used to compuse diffuse global illumination.
-
Ray-Traced Irradiance Fields (Real-Time Light Probe Rendering)
[...] how to combine glossy reflection ray tracing with diffuse ray-traced irradiance field probes. This creates a complete dynamic global illumination solution that scales across all DXR GPUs and minimizes light leaking without requiring manual per-probe or lightmap work by artists.
-
Stochastic Lightcuts
A solution to the global illumination problem by sampling from many lights. Stochastic Lightcuts is an improvement on the earlier Lightcuts algorithm. Presentation at HPG 2019.
-
Stochastic Progressive Photon Mapping (2 Minute Papers)
Describes improvements to Photon Mapping, called Progressive Photon Mapping, Stochastic Photon Mapping, and their combination.