Advances in Real-Time Rendering in 3D Graphics and Games (two-parter; second part here): This (somewhat awkwardly-named course) has been my favorite thing at SIGGRAPH for the past three years. Each year it presents all-new material. Previous courses have seen the debut of important rendering techniques algorithms like SSAO, signed distance-field vector texturing, and wrinkle mapping, as well as details on the rendering secrets behind games like Halo 3, Crysis, Starcraft 2, Half-Life 2, Team Fortress 2, and LittleBigPlanet. Not much is known about this year’s content, except that it will include details on Crytek’s new global illumination algorithm; but this is one course I know I’m going to!
Beyond Programmable Shading (another two-parter): GPGPU was promoted by GPU manufacturers in an attempt to find non-graphics uses for their products, and then turned full circle as people realized that drawing pretty pictures is really the best way to use massive general-purpose computing power. Between CUDA, OpenCL, and Larrabee, this has been a pretty hot topic. This is the second year that this course has been presented; last year had information on all the major APIs, and some case studies including a presentation on id’s Voxel Octree research. A subsequent SIGGRAPH Asia presentation added some new material, such as a presentation on real-time implementations of Renderman‘s REYES algorithm. This year, presenters include people from NVIDIA, AMD, Intel and Stanford; I expect this course to add significant new material, given the rapid development of the field.
Efficient Substitutes for Subdivision Surfaces: Tessellation is another hot topic; Direct3D 11 is mostly designed around this and Compute Shaders (the topic of the previous course). There has been a lot of work on mapping subdivision surfaces and other types of high-order surface representations to D3D11 hardware. Including presenters from ILM, Valve, and NVIDIA, this course promises to be a great overview of the state of the art.
Color Imaging: Color is one of the most fundamental topics in rendering. This course is presented by some of the leading researchers on color and HDR imaging, and should be well worth attending.
Advanced Material Appearance Modeling: Previous versions of this course were the basis for an excellent book on material appearance modeling. This is a great overview of an important rendering topic, and well worth attending if you haven’t seen it in previous years.
Visual Algorithms in Post-Production: It is well-known that you can find the latest academic rendering research at SIGGRAPH, but there is always a lot of material from the trenches of film effects and animation production as well. A surprisingly large percentage of this is relevant for real-time and game graphics. This course has presenters from film production as well as graphics academia describing ways in which academic research is used for film post-production. I think our community can learn a lot from the film production folks; this course is high on my list.
The Digital Emily Project: Photoreal Facial Modeling and Animation: Last year, Digital Emily was one of the most impressive technology demonstrations; it was the result of a collaboration between Paul Debevec‘s group at USC and Image Metrics, a leading facial capture company. In this course, presenters from both ICT and Image Metrics describe how this was done, as well as more recent research.
Real-Time Global Illumination for Dynamic Scenes: Real-time global illumination is another active topic of research. This course is presented by the researchers who have done some of the best (and most practical-looking) work in this area. It will be interesting to compare the techniques from this course with Crytek’s technique (presented in the “Advances in Real-Time Rendering” course).
Tags: SIGGRAPH 2009