Silicon Valley ACM SIGGRAPH Past Events

CPU clock speeds have stopped getting dramatic clock speed increases. Performance gains are now coming from adding more and more CPU cores. GPUs have been adding shader cores for many years and can already handle many threads. How do you go about programming a large number of cores with a massive number of threads to do video processing, such as H.264 decoding and encoding?

Mike Schmit is a software engineering manager at Advanced Micro Devices, managing the Digital Video Software team and Stream Computing SDK. He's worked on optimizing video encoding, decoding and transcoding since 1995 when he helped develop the first software DVD player for the PC. He's written books on software optimization, started his own software company, taught computer architecture, programming and software engineering project management classes and served on the Board of Directors and as President of the Software Entrepreneur's Forum. He's currently works on parallelizing video encoders and transcoders as well as other audio and video related applications.

Mathematical Art - The Beauty of Numbers
A retrospective on 30 years of personal experiments using basic math and simple software to create images
by Bruce Puckett

Bruce Puckett graduated from University of Washington with a Chemistry degree. Computer graphics was in it's early stages, but he did manage to create some early math art images while there. While in school, Bruce worked for NOAA, turning fur seal data into density maps. He has always had two sources of inspiration: exploring and visualizing mathematics with the help of custom code, and wilderness landscape photography. What do these two passions have in common? They are both about visual geometry, whether natural or constructed. Bruce next worked for Boeing Computers Systems doing programming support for CAD systems that were being used to design the next generation of airplanes. He attended his first Siggraph Conference in Seattle, in 1980, and displayed some plotter drawings in the art show. After Boeing, he was drawn to Silicon Valley, and joined the Fairchild Research and Development Lab, where he supported CAD applications, and created a prototype circuit layout application. Bruce then changed directions, and became a teacher at Verde Valley School, teaching programming, chemistry, and Earth Sciences. He continued his learning, and then teaching, at Foothill College. He has taught numerous programming classes, using various languages over the years, but his favorite subject to teach remains the 3D Modeling and Animation class.

Currently, he pursues his interests in digital visualization as an Independent Academic by writing and posting small papers and code examples on his web site, for all to use. In recent times he has become worried about the increasing divide between the students trying to enter the field, and the overwhelmingly technical nature of what computer graphics has become. He hopes to help counter the divide by showing how people can have fun and learn by combining just a bit of custom code with just a bit of mathematical geometry, using the great diversity of software that is available.

We propose a framework for the full two-way coupling of rigid and deformable bodies which is achieved with both a unified time integration scheme as well as individual two-way coupled algorithms at each point of that scheme. As our algorithm is two-way coupled in every fashion, we do not require ad hoc methods for dealing with stability issues or interleaving parts of the simulation. We maintain the ability to treat the key desirable aspects of rigid bodies (e.g. contact, collision, stacking, and friction) and deformable bodies (e.g. arbitrary constitutive models, thin shells, and self-collisions). In addition, our simulation framework supports more advanced features such as proportional derivative controlled articulation between rigid bodies. This not only allows for the robust simulation of a number of new phenomena, but also directly lends itself to the design of deformable creatures with proportional derivative controlled articulated rigid skeletons that interact in a life-like way with their environment.

Tamar Shinar is a Ph.D. candidate at the Institute for Computational and Mathematical Engineering at Stanford University working with Prof. Ronald Fedkiw. Her research focuses on the development of new computational algorithms for physically based simulation, with applications in computational fluid dynamics, solid mechanics and computer graphics. In particular, she has worked on level set based multiphase fluid simulation, coupled rigid/deformable solid simulation, and coupled solid/fluid simulation. She plans to pursue a postdoctoral fellowship at the Courant Institute at NYU in the fall.

We present an automated approach for high-quality preview of feature-film rendering during lighting design. By leveraging large portions of the existing final-renderer, we are able to cache light-independent data and interactively compute light-dependent data then re-sample onto the screen, including motion-blur and transparency.

Doug Epps is Director or R&D at ImagemoversDigital, a production studio based in Marin County devoted to the performance-capture films of Robert Zemeckis. Doug's first job was as a Digital Input Device operator at Tippett Studio working on Jurassic Park. Doug worked at Tippett Studio for more than 10 years on projects ranging from "Coneheads" to "Starship Troopers" to "The Matrix" sequels. He has worked in the RenderMan group at Pixar as well as at Exluna on the Entropy renderer.

Scott Snibbe presented interactive works that incorporate reactive video projections, large-scale tracking of humans and vehicles, and his recent work Blow Up which amplifies human breath as a large field of wind. He discussed the philosophical divide between language and visceral perception that motivates his creation of interactive media art. Working with technologies at the forefront of contemporary research including computer vision and synthetic touch, Snibbe explores how a minimal intrusion of technology can provide insight into the nature of observer's minds and their sense of self. Works shown ranged from large-scale body-centric physical installations to interactive sculpture and screen- and web-based interactive graphics.

Scott Snibbe creates immersive interactive art that evokes powerful emotional and social engagement from viewers. His works are known for their positive social effects: fostering a sense of interdependence, promoting social interaction among strangers, and increasing viewers’ concentration. His artworks have been installed in over one hundred art museums, performance spaces, science museums and public spaces worldwide since 1995 including the Whitney Museum of American Art (New York); the InterCommunications Center (Tokyo); Ars Electronica (Austria); and the Institute of Contemporary Arts (London), Science Museum (London); the Exploratorium (San Francisco), the Phaeno Science Center (Germany); and the Cité de Science (Paris, France). He has been awarded a variety of international prizes, including the Prix Ars Electronica and a Rockefeller New Media Fellowship. He is the founder of two companies: Snibbe Interactive, Inc., which sells and distributes interactive installations for public spaces; and Sona Research, which engages in educational and cultural research.

Snibbe was born in 1969 in New York City. He holds Bachelor’s degrees in Computer Science and Fine Art, and a Master’s in Computer Science from Brown University. Snibbe studied experimental animation at the Rhode Island School of Design and his films have been widely shown internationally. He has taught media art and experimental film at Brown University, The San Francisco Art Institute, the California Institute of the Arts, The Rhode Island School of Design and UC Berkeley. Snibbe worked at Adobe Systems as a Computer Scientist where he made substantial contributions to the special effects software Adobe After Effects and research projects at Adobe Research. Snibbe held research positions at Interval Research where he performed basic research in haptics, computer vision and interactive cinema. Snibbe’s research is documented in a number of academic papers, and over a dozen patents.

Thursday Feburary 28, 2008
The SIGGRAPH 2007 Electronic Theater
by the Silicon Valley Chapter ACM SIGGRAPH (Special Interest Group in Graphics)

Video presentation of cutting edge computer animation from the 2007 SIGGRAPH conference.

This special screening of the most astounding achievements computer graphics animated shorts from last year was the highlight experience of the 2007 SIGGRAPH computer graphics conference in San Diego. This showing was projected in high-definition and will include all materials shown in the electronic theater at SIGGRAPH in San Diego.

The Silicon Valley & San Francisco ACM SIGGRAPH chapters announce a joint meeting to see first hand the state of the art in 3D Digital Cinema.This discussion as well as an actual demonstation using demo trailers and clips with a full Dolby-3D theater setup will be begin at 7:00 PM on December 14th at the Dolby Presentation Room, 3rd Floor, 100 Potrero Avenue, San Francisco, CA. Attendies are advised to arrive at 6:15-6:30pm for check-in.

We look forward to and are thankfull for this rare and gracious invitation from John Gilbert @ Dolby.

Please refer to the following press release for details on this remarkable developement in Cinema projection.

Dolby 3D Digital Cinema Expands Global Presence Bringing High-Quality 3D Experiences to Theatres Worldwide Exhibitors in Over 12 Countries Deploy Dolby 3D Digital Cinema for Paramount Pictures' Upcoming Beowulf Release.

SAN FRANCISCO, Nov 15, 2007 (BUSINESS WIRE) -- Dolby Laboratories, Inc. (NYSE:DLB) announced today that its Dolby(R) 3D Digital Cinema system will be available in 75 screens in 12 countries worldwide in time for the upcoming release of Paramount Pictures' Beowulf, premiering November 16. By securing deals wit h dozens of exhibitors in Asia, Europe, and the United States, Dolby is revolutionizing theatrical experiences with its high-quality digital 3D solution. Dolby will continue installing additional screens during Beowulf's two-week global opening.

"In a short time frame, the team executed an aggressive deployment plan to install Dolby 3D systems in theatres around the world for Beowulf, as we wanted to fulfill as many requests as possible from our valued customers," said John Iles, Vice President, Cinema, Dolby Laboratories. "With an unwavering commitment to a better 3D experience, we are confident that Dolby 3D will provide an exceptional presentation of Beowulf and other upcoming digital 3D movies."

The Dolby 3D Digital Cinema solution brings high-quality 3D to every seat in a theatre: -- The ability for exhibitors to play back 3D content on a standard white screen provides moviegoers with an even image across the entire screen minus any hot spots or inconsiste nt light reflection. -- Dolby 3D full-spectrum color-filter technology provides amazing color fidelity, delivering clear 3D images with realistic color. -- Dolby's color filter technology also maintains premium picture quality because the filter wheel is inserted into the light path before the image is formed, delivering a stable and sharp picture.

The result is crystal clear images and vivid colors that pop off the screen with an amazing sense of depth.

"We are thrilled with the quality of Dolby 3D Digital Cinema and excited to show Beowulf in Dolby 3D," said Mike Thomson, Vice President Operations and Technology, Malco Theatres. "We will have the ability to play back 3D content on our big screen at the Malco Paradiso using Dolby 3D Digital Cinema. The large screen creates a 3D experience unlike anything we've been able to offer our patrons before."

"Marcus Theatres recently debuted Dolby 3D Digital Cinema and our patrons have been very pleased w ith the presentation quality," said Bruce Olson, President, Marcus Theatres. "Marcus strives to deliver the best moviegoing experience possible and we believe Dolby 3D reinforces that commitment."

"Dolby is a trusted brand for providing technologies that dramatically improve the cinematic experience, as we have seen with Dolby Digital Cinema," said Joost Bert, CEO, Kinepolis Group. "Kinepolis recently debuted Dolby 3D Digital Cinema at our newest cinema complex in Ostend, Belgium, and our patrons were very impressed with the sharp, clear, and bright images that seem to jump off the screen."

To date, exhibitors using Dolby 3D technology have presented Walt Disney Pictures' Meet the Robinsons and Tim Burton's The Nightmare Before Christmas 3D and in addition to Beowulf (Paramount Pictures/ Warner Bros.), are expected to show the upcoming 3D presentations of Fly Me to the Moon (nWave Pictures), Hannah Montana (Disney) and Journey 3D (New Line). For a complete list of Dol by Digital Cinema and Dolby 3D Digital Cinema locations, please visit www.dolby.com/consumer/motion_picture/ddcinemas/.

Projectors capable of doing thousands of frames per second are now possible and with the latest graphics cards we can feed them. The most interesting application of this technology so far will be the focus of this presentation - the 360 Degree Light Field Display shown at Siggraph's Emerging Technologies. The presentation will describe the system implementation and rendering techniques for an autostereoscopic light field display able to present interactive 3D graphics to multiple simultaneous viewers 360 degrees around the display. The display consists of a high-speed video projector, a spinning mirror covered by a holographic diffuser, and FPGA circuitry to decode specially rendered DVI video signals. The display uses a standard programmable graphics card to render over 5,000 images per second of interactive 3D graphics, projecting 360-degree views with 1.25 degree separation up to 20 updates per second. We describe the system's projection geometry and its calibration process, and we present a multiple-center-of-projection rendering technique for creating perspective-correct images from arbitrary viewpoints around the display. Our projection technique allows correct vertical perspective and parallax to be rendered for any height and distance when these parameters are known, and we demonstrate this effect with interactive raster graphics using a tracking system to measure the viewer's height and distance. We further apply our projection technique to the display of photographed light fields with accurate horizontal and vertical parallax. We conclude with a discussion of the display's visual accommodation performance and discuss techniques for displaying color imagery. The presentation will include video of the system and a demonstration of the fast projector although not the entire light field display.

Ian McDowall is a systems designer and CEO of Fakespace Labs, a company he co-founded in the early 1990's with two partners. Ian was part of a small team that developed a stereoscopic graphics computer for the NASA Ames Virtual Environment Workstation project in 1988. Since then Ian has implemented a variety of pioneering wide field of view and immersive displays at Fakespace Labs. Ian managed the development of Fakespace’s software which integrated various displays and IO devices. With a degree in Systems Design from the University of Waterloo, Ian brings together systems involving hardware, software, mechanical design, and optics. He is the inventor on a number of US patents and is the co-chair of the SPIE Engineering Reality of Virtual Reality conference. He has been a visiting scholar at Stanford and regularly collaborates on demonstrations at Siggraph’s Emerging Technologies.

3D Medical Images: A Nascent Market No More And Why it is Time for a Standard
by Michael Aratow

With the advent of increasingly sophisticated imaging modalities in medicine, fascinating 3D views of the human body are possible. Fueled by the aging global population, these images are being used more frequently in everyday medical tasks for diagnosis and to guide therapeutic interventions. In the past, images like this could only be viewed on specialized workstations housed in the radiology suite, but now they are appearing on desktop computers and even laptops, as Moore’s law drives the power of GPU’s ever higher. This shift in accessibility of 3D medical images, coupled with their growing volume, has significant implications for interoperability and is driving a whole new set of applications. A standard for these images is vital to promote patient care and innovation in medical markets.

Michael Aratow is a Board Certified Emergency Physician with over 20 years of clinical experience in emergency medicine. He currently is Vice-Chair and Director of Quality Assurance in the Department of Emergency Medicine and Chief Medical Information Officer of San Mateo Medical Center, an integrated healthcare system consisting of a public hospital and community clinics serving San Mateo County. Dr. Aratow has 4 years of past research experience in various areas including orthopedics, basic physiology (with numerous publications) and virtual reality, the latter two areas being conducted at NASA/Ames Research Center. His passion with 3D lead him to develop a patent in 3D visualization of aviation weather and terrain data, and he now sits on the Board of Directors of the Web3D Consortium and co-chairs its Medical Working Group. He currently is the Principal Investigator in a contract with the Telemedicine and Advanced Technology Research Center, part of the U.S. Army Medical Research and Materiel Command, to assist in development of an open standard for 3D medical images. He enjoys spending his spare time with his wife, 2 daughters and son, assisting in orthopedic surgery and PC gaming.

High-agency interactive story, in which the player can have a real and complex effect on both the inner lives of autonomous characters and the evolution of the plot, is one of the holy grails of interactive art and entertainment. Unfortunately, attempts to create interactive stories have primarily involved design-only solutions using standard technologies such as finite-state-machines and simple story graphs, resulting in experiences that inevitably trade-off agency and story structure. The consistent failure to combine agency and story has even prompted some designers and theorists to conclude that interactivity and story are fundamentally opposed. Façade, a first-person, real-time, one-act interactive drama (available for free download at www.interactivestory.net), is our attempt to constructively explore the design space of high-agency interactive story. In this talk we describe the process of building Façade, a process that combined three simultaneous and related research and design thrusts: designing ways to deconstruct a dramatic narrative into a hierarchy of story and behavior pieces; engineering an AI system that responds to and integrates the player's moment-by-moment interactions to reconstruct a real-time dramatic performance from those pieces; and understanding how to write an engaging, compelling story within this new organizational framework. We provide an overview of the process of bringing our interactive drama to life as a coherent, engaging, high agency experience, including the design and programming of thousands of joint dialog behaviors in the reactive planning language ABL, and their higher-level organization into a collection of story beats sequenced by a drama manager. We describe the iterative development of the architecture, its languages, authorial idioms, and varieties of story content structures, and how these content structures are designed to intermix to offer players a high-degree of responsiveness and narrative agency. We conclude with design and implementation lessons learned as well as describe current and future research and commercial directions.

Michael Mateas' research in AI-based art and entertainment combines science, engineering and design into an integrated practice that pushes the boundaries of the conceivable and possible in games and other interactive art forms. He is currently a faculty member in the Computer Science department at UC Santa Cruz, where he is involved in launching UCSC's game design degree, the first such degree offered in the University of California system. Prior to Santa Cruz, Michael was a faculty member at The Georgia Institute of Technology, where he held a joint appointment in the College of Computing and the School of Literature, Communication and Culture, founded the Experimental Game Lab, and helped create Georgia Tech's game design degree. With Andrew Stern, Michael released Façade, the world's first AI-based interactive drama in July 2005. Façade has received numerous awards, including top honors at the Slamdance independent game festival (co-located with the Sundance film festival). Michael's current research interests include game AI, particularly character and story AI, ambient intelligence supporting non-task-based social experiences, and dynamic game generation. In addition to frequent paper presentations at AI, HCI and digital media conferences, Michael has exhibited artwork internationally, including venues such as SIGGRAPH, the New York Digital Salon, ISEA, the Carnegie Museum, the Beall Center and Te PaPa, the national museum of New Zealand. Michael received his BS in Engineering Physics from the University of the Pacific, his MS in Computer Science (emphasis in Human-Computer Interaction) from Portland State University, and his Ph.D. in Computer Science from Carnegie Mellon University.

In this talk and presentation, Hank will review his career in animation and interactive computer graphics since the mid-1970's, and share observations and anecdotes about the growth of technology in computer graphics, and show video clips of his work at NYIT Computer Graphics Lab and at PDI/Dreamworks.

1) Traditional animation. Working with Stephen Lisberger leading up to his writing and directing TRON.

2) NYIT Computer Graphics Lab. Stories behind the development of the first 2D paint systems, 2D tweening, 3D keyframing, and early flexibly jointed characters, such as Gumby.

3) Early work with interactive multimedia interfaces, CD-ROMs and interactive TV at Time Warner Interactive.

Hank Grebe has been using computer graphics techniques to create art and animation for over 25 years and attended his first SIGGRAPH in 1983. He currently works at Mobile Greetings in Walnut Creek, designing interactive cell phone applications running on Verizon's wireless services.

Hank led a team of digital painters and motion graphics artists on PDI/DreamWorks Animation's feature, SHREK 2. He has provided computer graphics and video technical direction at Time Warner, AT&T, Elektra Records, Merrill Lynch, Intel, and numerous agency clients, video production and post production facilities.

In 1995 Grebe founded Media Spin, a computer graphics consulting business, which he dissolved in 2003. Hank continues to update the web site, mediaspin.com, with blogging and new art projects.

Grounded in traditional painting and cel animation, Hank pioneered computer animation at NYIT's Computer Graphics Lab by rigging and animating one of the first flexibly jointed 3D characters, a 3D Gumby shown at SIGGRAPH's Electronic Theater in 1984 and 85.

Daniel Lingafelter created the article "Attending SIGGRAPH 2007" for the Silicon Valley SIGGRAPH Chapter

Here is a link to the article:
http://silicon-valley.siggraph.org/MeetingNotes/siggraph2007.htm

Ashley Still and Pete Falco of Adobe will give an overview of some of the new features in Photoshop CS3 Extended, including movie paint, 3D, and automatic alignment and blending of multiple images. In addition to demonstrating these new features, they will provide an overview of the Photoshop 3D Plug-in SDK that can be used to extend the current capabilities. There will be ample time for Q&A.

Ashley Still is currently Sr. Product Manager for Adobe Photoshop. Ashley has been on the Photoshop team since 2004 and is focused on new markets and advanced technologies for Photoshop. Prior to joining Adobe, Ashley worked with an Entrepreneur in Residence at Sutter Hill Ventures developing and evaluating business plans and at eCircles.com, one of the first online sites offering photo-sharing and editing. She holds a BA from Yale University and an MBA from Stanford Graduate School of Business.

Pete Falco is currently Sr. Computer Scientist for Adobe Photoshop. Pete has been on the Photoshop team since 2005 and is focused on 3D and technology transfer for Photoshop. Prior to joining Adobe, Pete worked as an engineer on QuickTime VR at Apple, as the Director of Engineering at Live Picture and co-founded Zoomify. He holds a BS and ME from Rensselaer Polytechnic Institute.

We present a method for animating fluid with unstructured tetrahedral meshes that change at each time step. We demonstrate that meshes that conform well to changing boundaries and that focus computation in the visually important parts of the domain can be generated quickly and reliably using existing techniques. We also describe a new approach to two-way coupling of fluid and rigid bodies that, while general, benefits from remeshing. Several examples of resulting fluid animation are presented. Overall, our method provides a flexible environment for creating complex scenes involving fluid animation.

Bryan Klingner is a Ph.D. candidate in computer science at The University of California, Berkeley. He works on topics in computer animation and computational geometry with professors James O'Brien and Jonathan Shewchuk.

Bryan Klingner's web site (26Mb)

Presentation slides (26Mb)

Thursday April 26, 2007
SketchUp 6: Behind the Wheel and Under the Hood
by Brian Brown and Aidan Chopra, of Google

Brian Brown and Aidan Chopra of Google will give a wide-ranging overview of Google SketchUp, a free, cross-platform application for creating and presenting 3D models. In addition to demonstrating SketchUp and its integration with other programs such as Google Earth and the Google 3D Warehouse, Brian and Aidan will provide insight into the development of some of the new features of SketchUp 6; namely, Styles and LayOut. Ample time for Q&A will be left for those with specific questions.

Aidan Chopra is the Product Evangelist for SketchUp at Google. He lives and works in Boulder, Colorado, and his background is in architecture, illustration and graphic design. He has just completed work on Google SketchUp for Dummies, which is being published by Wiley, and which will be available in late June of this year.

Brian Brown is the Tech Lead for LayOut at Google. He works in Boulder, Colorado, and his background is in architectural engineering, lighting and optical design. He was part of the development effort for LayOut (beta) that was released in January of this year.

Even though the DirectX 10 API and hardware have just shipped, graphics technology continues to advance. This talk outlines the next steps, beginning with the improvements in DirectX 10.1 targeting increased flexibility and image quality. We'll cover technologies currently under investigation for future releases including tessellating subdivision surfaces, generating compressed textures, and continuing improvements in CPU/GPU interoperation. This is a great opportunity to see where your most needed new features are on the list!

Chas Boyd is a software architect at Microsoft. Chas joined the Direct3D team in 1995 and has contributed to releases since DirectX 3. Over that time he has worked closely with hardware and software developers to drive the adoption of features like programmable hardware shaders and float pixel processing. He has developed and demonstrated initial hardware-accelerated versions of techniques like hardware soft skinning, and hemispheric lighting with ambient occlusion. He is currently working on the design of future DirectX releases and related components.

Thursday February 15, 2007
Luxology modo 202...and Beyond
by Brad Peebler, President and a Co-Founder of Luxology, LLC creators of modo 3D software

Luxology will present on the 3D content creation tool, "modo", and what makes it unique. With a significant focus on workflow, Luxology has designed a flexible application interface that allows deep user customization without the need for intense scripting or additional programming. Additionally Luxology will discuss how it has improved artist flow with the unique fusion of various practices such as modeling, painting and rendering. Rather than partitioning these technologies into their own regimented areas, Luxology blends them together so that each can be leveraged regardless of the stage in the pipeline.

Brad Peebler is the President and a Co-Founder of Luxology, LLC creators of modo. Peebler has worked in 3D for over 15 years and has held a variety of roles from technical support on up. With an intense passion for content creation and technology Peebler has always enjoyed a close working relationship with engineers and artists alike. As the son of two teachers, education and training is an area of deep involvement for Peebler and has had a significant impact on the way Luxology develops software and its business.

Thursday January 18, 2007
SIGGRAPH 2006 Electronic Theater
by the Silicon Valley Chapter ACM SIGGRAPH (Special Interest Group in Graphics)

Video presentation of cutting edge computer animation from the 2006 SIGGRAPH conference.

This special screening of the most astounding achievements computer graphics animated shorts from last year was the highlight experience of the 2006 SIGGRAPH computer graphics conference in Boston.

Thursday December 7, 2006
NVIDIA GeForce 8800 Architecture: Stream Processing for Graphics and Computing
by Henry Moreton and Ian Buck, NVIDIA

This presentation by an NVIDIA architect gives a technical overview of the "G80" hardware architecture and covers the unified pipeline which enables geometry, vertex and pixel shading. This architecture provides extremely high performance, load-balancing, efficient GPU power utilization, and significant improvement to the GPU architectural efficiency. The scalar stream processing architecture also enables new avenues for programmability and thread computing accessed via CUDA, a standard C language interface to the GPU that is targeted at data intensive processing.

Henry Moreton joined NVIDIA in the fall of 1998 as a member of the architecture group. From 1984 to 1988, he worked at Silicon Graphics. In 1992 he received a Ph.D. from the University of California, Berkeley. He has published in the areas of curve and surface modeling, rendering, texture mapping, video and image compression, and unmanned submarine control. He has multiple patents in the areas of optics, video compression, graphics, system and CPU architecture, and curve & surface modeling and rendering. Current interests include the evolution of graphics programming models and, API design and hardware architecture of highly parallel programmable devices.

Ian Buck completed his Ph.D. at the Stanford Graphics Lab in 2004. His thesis was titled "Stream Computing on Graphics Hardware," researching programming models and computing strategies for using graphics hardware as a general purpose computing platform. His work included developing the "Brook" software toolchain for abstracting the GPU as a general purpose streaming coprocessor. He currently works for NVIDIA as the GPU-Compute software manager.

Thursday November 16, 2006
Why Ray Tracing is Doomed
by Alexander Reshetov, Ph.D.

From a hardware perspective, two trends define the increasing importance of the ray tracing research today: emergence of consumer-level multicore architectures and expanding functionality of modern GPUs. Even though ray tracing was long recognized as an embarrassingly parallel application, this had a little practical use on a single core CPU. GPUs, from the other hand, allowed parallel execution from the very beginning, but their limited programmability made it rather cumbersome and ineffective. The Situation is rather different now.

All this is augmented by the development of the new very effective software algorithms. This could have heralded a long anticipated transition from traditional rasterization pipeline to ray tracing, except for one little thing: essentially all the progress in ray tracing lately was achieved by pursuing the SIMD friendly approaches. Accordingly, the general flexibility, which characterized the first ray tracing algorithms, was lost. At the same time, propped by the persistent progress of the GPUs, applications based on the traditional rasterization pipeline are getting a second breath.

I will try to make a sense out of all of this and will talk about the current state of the art in the real time ray tracing and beyond. I will also describe the existing problems and hardware features, which would make RT applications more effective.

About the speaker...

Alexander Reshetov received his Ph.D. degree from Keldysh Institute for Applied Mathematics (in Russia). He joined Intel Corporation in February 1997 as a senior staff researcher after working for two years at the Super-Conducting Super-Collider Laboratory in Texas, where he was designing control system for the accelerator.

His research interests span 3D graphics algorithms and applications, and physically based simulation. His recent work focuses on efficient ray tracing algorithms and data structures. He is best known for his work on multi-level ray tracing algorithm, presented at Siggraph 2005, which is recognized as the fastest real time ray tracing algorithm in the world.

Monday October 16, 2006
Pixar Animation Studios: The making of 'Cars'
by Pixar Animation Studios

San Francisco ACM SIGGRAPH and Silicon Valley ACM SIGGRAPH are proud to present this special event "Pixar Animation Studios: The making of 'Cars'"

Academy of Art University Theatre
Morgan Auditorium, Post & Mason
491 Post Street
San Francisco, CA. 94102
October 16th, 2006
Reception 6:30 p.m.
Event 7:00 p.m. to 9:00 p.m.
http://www.academyart.edu/map.html

sponsored by,
Academy of Art University - http://www.academyart.edu/
Ballistic Publishing - http://www.ballisticpublishing.com/

Thursday June 22, 2006
Invisible Visual Effects--Compositing in Mission: Impossible III
by Todd Vaziri, Digital Compositor, Industrial Light & Magic

Cinematic visual effects artisans have been creating illusions for film for well over a century. In most cases, the images created for films fall in one of two categories. Visual effects can either support the scope of the film, advancing the other-worldy spectacle of the story, or support the narrative in its rawest form, with its presence as invisible as the production's cinematography, editing, and production design, allowing characters and story to take center stage.

Invisible visual effects have progressed to a staggering degree with the rise of digital compositing. Advanced 2D, 3D and hybrid techniques give filmmakers a much larger canvas on which to paint. It is up to the digital effects artist to reign in spectacle and understand cinematic reality. We will look at the invisible visual effects of J.J. Abrams' "Mission: Impossible III" and illustrate how compositing helped bring the director's vision to life, bringing synthetic, hand-crafted images to the screen with the utmost importance placed on invisibility. We will look at specific shots and discuss the techniques and tools that made it happen. Examples from the earliest days of cinema, to specific case studies of modern effects films will also be discussed.

About the speaker...

Todd Vaziri
Digital Compositor
Industrial Light & Magic

Todd Vaziri joined Industrial Light & Magic in 2001, and has contributed to Star Wars: Episode II "Attack of the Clones," Signs, The Hulk, Pirates of the Caribbean and Van Helsing.

Vaziri has worked to expand the role of compositor at Industrial Light & Magic by going beyond the confines of straightforward compositing. He consistently takes on shots which not only require 2D compositing but matte painting, particle effects and 3D compositing. This approach allows much more flexibility and efficiency to the production.

Vaziri grew up outside of Chicago, and graduated from Northwestern University in 1995 with a Bachelor of Arts degree in Film and now resides in San Francisco with his wife. Prior to joining ILM, Vaziri worked on several films as a lead artist and as compositing supervisor, such as Dr. Dolittle, American Pie, Driven and Hart's War. Vaziri also created the now-retired Visual Effects Headquarters web site which covered the visual effects industry (http://www.vfxhq.com).

ILM CREDITS
Feature Films

2006	MISSION: IMPOSSIBLE III (currently in production) - Sequence Supervisor
2005	WAR OF THE WORLDS - Compositor
2005	STAR WARS: EPISODE III "Revenge of the Sith" - Compositor
2004	SKY CAPTAIN AND THE WORLD OF TOMORROW - Compositor
2004	VAN HELSING - Lead Compositor
2003	PIRATES OF THE CARIBBEAN: THE CURSE OF THE BLACK PEARL - Compositor
2003	HULK - Compositor
2002	SIGNS - Compositor
2002	STAR WARS: EPISODE II "ATTACK OF THE CLONES" - Compositor

Thursday May 18, 2006
A History of PDI: A 25 Year Retrospective
by Richard Chuang, co-founder of PDI

Take a personal journey through 25 years of the CG industry: a personal view of how the industry has changed, from one who has experienced evolution firsthand.

There were no tools or applications to do animation at that time. Aspiring computer animators saw the crude animations presented at the SIGGRAPH Conference Electronic Theater by academic researchers, and dreamed of the potential of using the computer to help convey the lively visual stories that were present only for instants in their creative minds.

In 1981, Richard Chuang, along with Carl Rosendahl and Glenn Entis, started Pacific Data Images in an empty warehouse. They slaved themselves to their desks, surveying all of the computer graphics literature, and writing software to build the tools, the business and the vision. They took on a few companies as clients and produced some stunning graphics for television, and moved later to commercials and visual effects, Eventually, they got the chance to do some animations for story- telling's sake, and started making feature length movies.

In this presentation, we tell the story of a journey from the startup of a pioneering CG animation studio through the production of mega-hit like Shrek. We follow the evolution of the studio from Pacific Data Images to PDI to PDI/DreamWorks Animation. We take you through the turbulent shake-up of the industry, that left only a handful of studios remaining. We identify the critical points in technology and the business that revolutionized the industry.

Not to be solely a history lesson, Richard will also be showing highlights of the 25 years of animation productions that built the PDI studio from the early 80's through today.

About the speaker...

Richard Chuang
co-founder of PDI

A co-founder of PDI over 25 years ago, Richard Chuang helped create the studio's powerful proprietary animation system, which the Academy of Motion Picture Arts and Sciences recognized with a Technical Achievement Award in 1997. PDI/DreamWorks proprietary animation system has been used on countless commercials, live-action features and throughout PDI/DreamWorks' feature films, ANTZ, the Academy Award-winning SHREK, and SHREK 2 for which Chuang served as head of special projects.

Known for his hands-on creative approach, Chuang's expertise is in computer animation and visual effects for both animated and live action films. He served as Visual Effects Supervisor on several DreamWorks film including EVOLUTION, LEGEND OF BAGGER VANCE and FORCES OF NATURE.

Richard also led the team that created digital superheroes in the Warner Bros. blockbuster BATMAN AND ROBIN. These digital heroes were a follow-up to his earlier pioneering digital character work in BATMAN FOREVER. He has 16 live action film credits in the visual effects area.

Richard helped plan and guide the CG production for FATHER OF THE PRIDE, DreamWorks first all computer animated primetime TV show, created for NBC. Currently, he is part of the executive group responsible in planning and forecasting future projects.

Thursday April 20, 2006
Grokking Emerging Multicore Processor Architectures and How to Leverage Them for Future Applications
by Yahya H. Mirza

With the STI CELL architecture, the DARPA HPCS effort and the recently announced NSF Petascale acquisition effort, a new generation of computational capabilities are now starting to emerge. Unfortunately, exactly how these powerful new computational capabilities will be utilized for a commercially viable next generation killer app can be considered an open question. When looking for new opportunities, a useful exercise can be to revisit old ideas in conjunction with emerging new technologies. To facilitate this quest, this presentation will survey emerging multicore architectures such as the STI CELL, the SUN Rock, the Cray Cascade, as well as future X86 architectures and illustrate their commonalities and differences. We will look at the problems these new chips are being designed to solve and then map to these solutions, their architectural and micro-architectural techniques they utilize, such as superscalar, multithreaded, simultaneous multithreading, vector architectures, short-vector SIMD, stream architectures, etc. We will illustrate how these solutions balance the need for high performance computation while maintaining additional design goals such as minimum power and die size. Additionally, I'll illustrate the tradeoffs as to functionality supported in hardware vs. functionality, which is supported by the compiler and the programmer.

When looking to accelerate the algorithms in one's application using a new architecture, its useful to understand both the application's opportunities for parallelism, and the systems support for various kinds of parallelism. Given a target system, and a problem: either the problem's algorithms can be mapped to the target; or alternatively new algorithms can be developed explicitly for that target system. To guide us, I'll discuss a common set of parallel patterns that have been utilized in the past to create high performance applications. We will then discuss how these parallel programming patterns map and are explicitly supported and or enabled by the various architectural and micro-architectural elements of the emerging multicore architectures. Next we will discuss the tools we will need to program and thus explicitly leverage these multiple parallel capabilities. To illustrate the technical issues involved, we will discuss a few parallel programming models that have been developed in the past such as pThreads, MPI, and OpenMP, and how the processor architectures / computing systems at the time influenced the design of these solutions. Our goal is to provide counter-weight to the often knee-jerk reaction of prematurely selecting a particular programming tool to address a particular problem without fully considering the implications involved.

A useful parallel programming model and parallel runtime for these emerging multicore based systems, will require the usage of parallelism in all its forms including: message passing, systolic arrays, data parallelism (in its many forms including: loop parallelism, short-vector SIMD parallelism, long vector parallelism, streams parallelism), asynchronous futures, SPMD parallelism, parallel collections, etc. The objective of this presentation is to provide intuition and holistic insight on what makes emerging multi-core architectures like the CELL, different from conventional processors, and how these differences impact how we can program and utilize scalable systems built from these new architectures to create our next generation "killer apps".

About the speaker...

Yahya H. Mirza
Aurora Borealis Software LLC
15817 NE 90th St. Suite E340
Redmond WA 98052

Yahya Mirza's original background was aeronautical engineering and was initially employed by Battelle Labs in Columbus Ohio. After transitioning into the software industry, Yahya spent three years as a visiting researcher in the UIUC Smalltalk Research Group. Through his company, Aurora Borealis Software LLC, Yahya has worked on system software projects for Microsoft, CatDaddy Games, Source Dynamics and Pixar Animation Studio's RenderMan team. For the last five years, Yahya has been organizing the Language Runtimes workshops, held at the OOPSLA and Supercomputing conferences. Yahya's interest in scalable multicore computing systems is driven by his passion to create a real-time interactive feature film. Yahya is currently working on a new distributed programming model to explicitly leverage large-scale clusters built from emerging multicore processors.

Conventional photographs record the sum of all light rays striking each position on an image plane. This dissertation explores how digital photography can be improved by instead recording the light field inside the camera: not just the position of each light ray, but also the direction that it is traveling. This talk will discuss the design, prototyping and performance of a modified digital camera that records the light field in a single photographic exposure.

By resampling the recorded light rays, it is possible to compute final photographs more flexibly and at higher quality than in a conventional camera. For example, in digital refocusing we compute final photographs that are focused at different depths from a single shot. Theory predicts, and experiments corroborate, that we can reduce the misfocus blur anywhere in the photograph by a factor approximately equal to the directional resolution of the recorded light rays. In another application, we digitally correct lens aberrations by re-sorting aberrated light rays to where they should ideally have converged. This increases the quality of final photographs by raising the contrast.

Here are some photographs taken with the prototype camera:
http://graphics.stanford.edu/papers/lfcamera/refocus/

Ren Ng is close to completing his PhD in the Computer Science department at Stanford. His focus is on digital photography systems, computer graphics and applied mathematics. He received his B.S. in Mathematical and Computational Science from Stanford.

Thursday February 9, 2006
Games for Teaching Medicine
by Parvati Dev, PhD and LeRoy Heinrichs, M.D., Ph.D.

Simulation technology is being used to address all aspects of health care delivery and management. The primary use is in education but the potential for work flow improvement and risk reduction are new areas of consideration. We will examine a range of patient simulations from simple multimedia patient records to patients who respond through speech and video. We will touch on novel sensory modalities such as haptics and stereo and their use in surgical procedure simulators. Finally we will discuss the potential of modeling teams of people interacting with a simulated patient, and the use of gaming technology to implement these virtual worlds.

Parvati Dev, PhD
Director, SUMMIT; Associate Dean, Learning Technologies
Office of Information Resources and Technology
Stanford University School of Medicine

Parvati Dev completed her doctoral degree in Electrical Engineering on computer models of the brain at Stanford University in 1975. She has worked on the research and teaching staff at M.I.T., Boston University, and Stanford, and spent seven years in industry, where she developed products for three-dimensional medical imaging. Since January 1990, she has been Director of SUMMIT, a research lab at Stanford. Her current research is in virtual reality and simulation for medical education.

LeRoy Heinrichs, M.D., Ph.D.
Associate Director, SUMMIT Lab; Emeritus Active, Dept. of GYN/OB
Stanford University School of Medicine

In 1976, Dr. Heinrichs was appointed Professor and Chairman of Gynecology and Obstetrics at Stanford University where he performed and taught laparoscopic surgery, and recognized the potential for teaching this type of surgery with Virtual Reality systems, like those used by the aviation industry in training pilots. Finding no emerging technologies, he began a start-up company that failed, but retired to SUMMIT where he developed an anatomically correct, human 3D model (Lucy v.2.6) as the virtual anatomy in pelvic surgical simulators. Dr. Heinrichs also initiated a project with Immersion, Inc. for developing a hysteroscopy trainer, now a commercially available trainer. At SUMMIT, where he is now Associate Director, he and colleagues are developing anatomy and surgical simulation projects for distribution over the Next Generation Internet (NGI). His designation, along with SUMMIT, in 2002 for the 8th Annual Satava Award by the Medicine Meets Virtual Reality organization was for his leadership in the field of surgical simulation. He is PI on a Wallenberg Foundation planning grant designing a 3D World for training of medical teams for crisis management of trauma. Dr. Heinrichs writes on Surgical Simulation and lectures widely on this topic.

Video presentation of cutting edge computer animation from the 2005 SIGGRAPH conference.

This special screening of the most astounding achievements computer graphics animated shorts from last year was the highlight experience of the 2005 SIGGRAPH computer graphics conference in Los Angeles. This year's presentation will include footage that wasn't presented at SIGGRAPH. The Electronic Theater Program contains 20 pieces covering animation, effects, storytelling, and visualization. The public is invited and welcome. The content is entertaining and educational for a general audience, not just those with a technical or artistic bent.

Samuel Lord Black has been in the software business for about 20 years, primarily in the graphics field in one form or another. He holds Bachelor's Degrees in Electrical Engineering and Computer Engineering from the University of Michigan, and a Masters Degree in Computer Science from the University of North Carolina. He worked for several years in the workstation and desktop computing industry for Apollo, Stellar, and Masscomp as a pioneer in the X Window System, with an emphasis on its applications in real-time computing. Then, ten years ago, he went to work in the video game industry (Papyrus Design Group) and worked on several projects, with game credits including NASCAR Racing, IndyCar Racing, and Road Rash. He spent eight years working on rendering software for Pixar Animation Studios, followed by a stint as the Chair for the SIGGRAPH 2005 Computer Animation Festival. He is currently working as a graphics software engineer for Autodesk, Inc.

Thursday, November 17, 2005
Recent advances in Photoshop technology
by Ashley Manning and Todor Georgiev

Ashley will be giving a concise tour of some of the most exciting tools added to Photoshop in the last few releases. Her presentation will provide a lead-in to a technical discussion of what's going on behind the scenes.

Todor will talk about the science behind some of your favorite Photoshop tools like the Healing Brush. This will include introduction to image processing in the gradient domain, emphasizing perceptually correct image processing based on invariance to change of illumination.

Ashley Manning is a Product Manager for Adobe Photoshop, where she focuses on new markets and technologies for Photoshop. Prior to joining Adobe, Ashley worked with an Entrepreneur in Residence at Sutter Hill Ventures developing and evaluating business plans and at eCircles.com, one of the first online sites offering photo-sharing and editing. Ashley holds an MBA from Stanford Business School and a B.A. from Yale University.

Todor Georgiev has a Master's Degree in Theoretical Physics from Sofia University (1987), and a PhD in Theoretical Physics from Southern Illinois University (1995). He has been working in Adobe Photoshop since 1997, focusing on transfer of mathematical methods from Theoretical Physics to Imaging. He is interested in color, perceptual and Gradient Domain imaging, HDR, Light Fields and Computational Photography. He has published numerous papers and holds more than 10 patents related to image processing.

Thursday October 20, 2005
Mobile Games of the Future: Total Immersion on a Small Screen
by Justin Radeka

Discussion points:

Enabling Technologies
OpenGL 2.0, which is breaking down the boundaries of game design for multiple platforms - from Playstation to mobile
Falanx recently introduced the Mali200, this graphics core is fully programmable.
Mixing Media (video, 3D, audio, 2D, still images)
Types of Games
Visual Quality
Next level of games in mobile gaming
Extended games, wireless worlds
Immersive gaming on a 2" screen

Justin Radeka serves as Vice President of Developer Relations for Falanx. Prior to joining Falanx, he served as CTO of EVGA, a graphics board manufacturer, and as head of the gaming group at Hewlett Packard. As the leader of the Developer Relations group at Falanx, Justin is working with top tier developers and publishers to develop the industry strategies in bringing the latest mobile graphics technologies to tomorrow's games. Radeka holds a BA from the University of California.

Thursday, September 15, 2005
Model. Paint. Render. modo 201
by Dion Burgoyne and Allen Hastings

With the New release of Modo 201 announced at the SIGGRAPH 2005 Internationational Conference, the developing program adds features for painting, and rendering to its already impressive modeling power. No other program is as fast or formidable in Subdivision Surfaces. (Ed Catmull, et al. Siggraph 1989)

Dion Burgoyne will take you through the new features of the Modo 201. Current 3D applications are taking notice of the developments in Modo. You will be surprised and impressed with the speed and power.

Allen Hastings will take you into the details and wonders of the rendering engines he has been developing. Blazing speed and remarkable results are the result of study and experimentation into the methods of rendering. Come and see a wizard at this black art and appreciate what promises to be a benchmark program from one of the last American 3D software purveyors.

"modo 201 is not the first 3D environment to incorporate modeling, painting and rendering," explains Brad Peebler, president and co-founder of Luxology. "However, modo is the first environment built from the ground up to integrate these workflow steps in a way that dramatically improves the ability of 3D designers to focus more on creating and less on process and repetitive steps."

modo is the fastest, most advanced subdivision surface modeling environment available. Specifically designed to help 3D artists working on games, films, TV, print, architecture and Web productions, modo enables artists to accomplish more in less time. While the current version of modo focuses on delivering an elegant and advanced environment for accelerated 3D modeling, modo 201 takes these benefits to unprecedented levels across modeling, painting and rendering.

Allen Hastings: Chief Scientist. Allen Hastings' far-reaching vision is always tempered by mathematical accuracy and flawless execution. With a background in fine art and music, Allen first started writing software to make 3D digital movies at a time when few believed it was possible. His efforts were some of the first off-the-shelf 3D packages available. He created NewTek's LightWave 3D in 1989 and remained the primary force behind its animation and rendering algorithms through 2001. Allen received an Emmy for his work on LightWave 3D in 1993, and in 2001 Animation Magazine named him one of the top 15 most influential people in the animation industry. At Luxology, Allen continues to make significant original contributions to the art and science of 3D.

Dion Burgoyne. Dion Burgoyne came to Luxology to fill the role of Content Artist and Quality Assurance manager. Dion is a film maker, photographer and technical artist and brings to Luxology a solid balance of art and science, just as we like. You will find Dion actively assisting the modo community through forum posts and training videos online as well as through the creation of PERL scripts that add functioinality to modo users through the third party website www.vertexmonkey.com.

Walter Vannini created the article "Attending SIGGRAPH 2005" for the Silicon Valley SIGGRAPH Chapter

Here is a link to the article:
http://silicon-valley.siggraph.org/MeetingNotes/siggraph2005.htm

Thursday June 16, 2005
Advanced Physics for Multi-Processor Scalability
by Tom Lassanske

This talk explores how a multi-threaded physics API can help a developer make the most of the extra power afforded by multi-processor platforms for achieving more realistic worlds--while still maintaining compatibility with current generation consoles, if necessary. The AGEIA PhysX hardware will also be demonstrated in a couple sample applications.

Tom Lassanske received a BS in Mechanical Engineering from the University of Wisconsin and a MS in Computer Science from the University of North Carolina at Chapel Hill--where his focus was on physical simulation and computer graphics. For the last several years, he has worked with the Games Industry writing and integrating collision and dynamics middleware from NDL, Havok and now AGEIA. Formerly, Tom learned all about customer service and real-world statics and dynamics (including some bloody-ragdoll effects!) as the 10-year sole proprietor of a building construction firm."

Wednesday May 25, 2005
Digital Cinema: The History, The Technology, The Deployment
by Bill Mead, Bill Kinder and Walt Husak

Topics

The Historical Evolution of Digital Cinema
Issues of Deployment
Political Issues
Why is Digital Cinema a Complex Problem?
Cinema Fundamentals
Digital Cinema Overview
Standards
Timeline of Digital Cinema at Pixar
Advantages of Digital Projection over Film
Business Considerations
Authoring Issues
Hidden Roadblocks

Speakers

Bill Mead
founder and publisher of DCinemaToday.com

Bill Mead is the founder and publisher of DCinemaToday.com, an on-line publication focused on the emerging digital cinema industry. Bill has plenty of experience in market development for cinema technologies. For two years prior to launching DCinemaToday in 2003, Bill consulted with TI's DLP Cinema group. Bill previously spent six years as VP of Marketing for Sony's cinema sound group (SDDS) and 19 years with Dolby Laboratories in range of cinema-related technical and market development positions. Bill is a member of SMPTE's DC28 digital cinema standards committee and is a past director of the International Theatre Equipment Association (ITEA).

Bill Kinder
Director, Editorial & Post Production
Pixar Animation Studios

Bill Kinder leads Pixar's efforts in Digital Cinema, realizing the first ever digital theatrical release of a digitally produced film, Toy Story 2. He has pioneered digital mastering methods at the studio since the all digital release of A Bug's Life on DVD. He produced the DVD for Finding Nemo. Most recently he oversaw Pixar's largest international digital cinema release to date, The Incredibles.

Walt Husak
Senior Manager, Electronic Media
Dolby Laboratories, Inc.

Walt Husak is Senior Manager, Electronic Media at Dolby Laboratories. He began his television engineering career at the Advanced Television Test Center (ATTC) in 1990 carrying out video objective measurements and RF multipath testing of HDTV systems proposed for the ATSC standard.

Prior to joining Dolby, Walt has worked on issues related to the deployment of HDTV systems in the US and the rest of the world. Walt demonstrated the world's first Digital On-Channel Repeater to extend Digital Television signals into obstructed areas. He developed a mechanism for capturing RF signals for offline analysis and distribution to receiver vendors. Walt spent many years lecturing on topics such as video compression, RF transmission for DTV, and overcoming multipath signals in urban and rural environments.

Walt joined Dolby in 2000 as the first video compression expert for Digital Cinema and has spent the last several years studying and reporting on advanced compression systems for Digital Cinema, Digital Television, and HD DVD. He has managed or executed visual quality tests for DCI, ATSC, Dolby, and MPEG.

Walt is now a member of the New Technologies Team focusing his efforts on JPEG2000 and advanced MPEG codecs for Digital Cinema and Digital Television. Walt provides industry lectures on Digital Cinema systems and image compression. Walt has authored a number of articles and papers for a number of major industry publications.

Walt is currently Vice-Chairman of SMPTE DC28 and the liaison to and from MPEG and JPEG. He served as the Chairman of the MPEG Digital Cinema Group and as secretary of the SMPTE DC28 Projection Group. He also served as a member of the ITU task group on Digital Cinema. Walt is a member of SMPTE, MPEG, JPEG, IEEE, and SPIE.

Thursday April 21, 2005
CELL: A New Platform for Digital Entertainment
by Mark DeLoura and Dominic Mallinson

This presentation from Sony Computer Entertainment gives a technical overview of the first CELL processor, touching on hardware architecture, programming model and software strategy. The CELL project is a joint venture between Sony, Toshiba and IBM to create a new generation microprocessor architecture. After four years of design and development, the first CELL processor was recently unveiled at the ISSCC semiconductor conference.

Thursday March 24, 2005
Integration of Combustion 4 and 3D Studio Max 7
by Hagé van Dijk

Hagé van Dijk is a computer industry veteran and digital media pioneer working in software and hardware development for over 17 years. At Apple Computer he was a member of the original QuickTime Team and contributed to the development of products that include: Apple Midi Manager, QuickTime 1.0, QuickTime Starter Kit, QuickTime 3.0 and Apple DVD player, Sound and MIDI compatibility for the MacOS (v6.04-8.6), Other critically acclaimed products include; Radius VideoVision and VideoVision Studio, Radius Telecast, Radius Studio Array, Digital Origin's EditDV (first DV based NLE application which kicked off the "DV revolution"), and the award-winning Cleaner family of encoding products (Terran/Media100, Discreet). Recently he was compression and visual effects product specialist (AE) at Discreet for cleaner integration and combustion. In this role he continued to develop and evangelize cleaner technology as well as focusing on 2d/3d integration between 3ds Max and combustion, integration with discreet systems (flame, inferno) and use of vector based paint for restoration, wire removal, rotoscoping, and visual effects. He advises and supports clients in the broadcast and computer industries by creating settings, content, lectures, seminars, webcasts and demonstrations covering many aspects of visual effect creation, media publishing and distribution for all formats. Hagé has presented for the National Association of Broadcasters (NAB), Streaming Media, Apple Computer, SF Art Institute, and seminars internationally. As a creative professional, Hagé has contributed to all phases of product design, development, technical marketing, and support.

Thursday February 24, 2005
Graphical WOW: Art, History and Technology of Flaming Pear
by Lloyd Burchill

Enhance your graphics with impact. Easily create planets, water, weathered surfaces, organic textures, illumination and mood, richness and complexity -- all in very simple ways. Unusual software produces dramatic effects useful to photographers and digital illustrators alike.

Lloyd Burchill will demonstrate Flaming Pear's increasingly inexplicable Photoshop plug-ins, which draw on science fiction, obscure photographic techniques, panoramas, and now image fusion. How does this indie graphics developer thrive? By staying resolutely small, being weird, and keeping up with the research.

Along the way he'll outline the story of this rewarding niche and explain the interesting technology under the hood. Amid abundant pretty pictures you'll hear about about the intertwining pressures of tech, aesthetics, and audience, the influence of Siggraph and the importance of dirt. He'll reveal what's next and share useful tricks he's picked up over the years.

February 3, 2005
How well do people have to see each other for visual telecommunications to work well, or would people prefer not be seen precisely?
byJaron Lanier

Visual telecommunication technologies have never met human factors requirements. Worse, the full extent of the requirements is not yet known. We are learning a lot, however. It is now possible to propose lines of demarcation between high-end video conferencing and tele-immersion. Tele-immersion ought to convey certain cues that demonstrably matter to the outcome of communication even if non-specialist participants are not able to articulate the nature of these cues as they are experiencing them. Fundamental physical limits make the design of high performance tele-immersive interfaces difficult, but strategies are emerging that are likely to overcome known barriers. We are close enough to reaching the goal of quality tele-immersion that it's time to ask how tele-immersive fidelity might need to be limited in precise ways to bring out the best in the curious social species tele-immersion is intended to help.

Jaron Lanier is a Visiting Scientist at SGI, and he is also an External Fellow at the International Computer Science Institute at Berkeley

This special screening of the most astounding achievements computer graphics animated shorts from last year was the highlight experience of the 2004 SIGGRAPH computer graphics conference in Los Angeles. This year's presentation will include footage that is not on the DVDs. The Electronic Theater Program contains 20 pieces covering animation, effects, storytelling, and visualization.

Christoph Bregler is an Associate Professor at the Media Research Lab at NYU. His primary research interests are in vision, graphics, and modeling. He is currently focused on the animation of human movements, This includes visual motion capture, human face, speech, and full body motion analysis and synthesis.

December 16, 2004
Full Spectrum Command: A commercial-platform training aid
by Michael Van Lent

In the commercial platform training aids project, the University of Southern California's Institute for Creative Technologies (ICT) seeks to develop training aids that capitalize on the technologies, development practices, and hardware of the commercial games industry. To date the project has produced Full Spectrum Command, a PC-based training aid currently in use at Ft. Benning, and Full Spectrum Warrior, an Xbox-based training aid currently undergoing pedagogical evaluation by the Army Research Institute. A modified, commercial version of Full Spectrum Warrior was also the #1 selling Xbox title in June of 2004. The story behind this unique integration of military training tool and commercial entertainment product includes elements of research in educational design and artificial intelligence as well an interesting exploration of the business model behind commercial game console. In this talk Dr. van Lent will discuss the process that led to the development of Full Spectrum Warrior, the academic research involved, and present a demonstration of the differences between the military training version and the commercial entertainment version.

Dr. Michael van Lent completed his Ph.D. in 2000 under the direction of Dr. John Laird at the University of Michigan. His dissertation, titled Learning Task-Performance Knowledge through Observation, explored how knowledge for real-time intelligent agents, such as simulated tactical air combat pilots, could be acquired from observations of human experts more quickly and cheaply than traditional knowledge acquisition techniques. After a one year post-doc at the University of Michigan, Dr. van Lent joined the University of Southern California's Institute for Creative Technologies (ICT) as a research scientist. Dr. van Lent was the lead researcher for ICT's Commercial Platform Training Aids project which resulted in Full Spectrum Command, a PC-based company command training aid, and Full Spectrum Warrior, an Xbox-based squad leader training aid. Now a project leader, Dr. van Lent's areas of interest include advanced AI techniques for commercial computer games, the use of commercial game technology for training and education, and the integration of commercial game technology with military simulation technology. His current resea rch focuses on explainable artificial intelligence and adaptive opponents for military training and computer games. Dr. van Lent is also active in the growing collaboration between academic researchers and commercial game developers as the Editor-in-Chief of the Journal of Game Development and a frequent contributor to the Game Developer's Conference.Dr. Michael van Lent completed his Ph.D. in 2000 under the direction of Dr. John Laird at the University of Michigan. His dissertation, titled Learning Task-Performance Knowledge through Observation, explored how knowledge for real-time intelligent agents, such as simulated tactical air combat pilots, could be acquired from observations of human experts more quickly and cheaply than traditional knowledge acquisition techniques. After a one year post-doc at the University of Michigan, Dr. van Lent joined the University of Southern California's Institute for Creative Technologies (ICT) as a research scientist. Dr. van Lent was the lead researcher for ICT's Commercial Platform Training Aids project which resulted in Full Spectrum Command, a PC-based company command training aid, and Full Spectrum Warrior, an Xbox-based squad leader training aid. Now a project leader, Dr. van Lent's areas of interest include advanced AI techniques for commercial computer games, the use of commercial game technology for training and education, and the integration of commercial game technology with military simulation technology. His current resea rch focuses on explainable artificial intelligence and adaptive opponents for military training and computer games. Dr. van Lent is also active in the growing collaboration between academic researchers and commercial game developers as the Editor-in-Chief of the Journal of Game Development and a frequent contributor to the Game Developer's Conference.

Point primitives have experienced a major "renaissance" in recent years, and considerable research has been devoted to efficient representation, modeling, processing, and rendering of point-sampled geometry. There are two main reasons for this new interest in points: First, we have witnessed a dramatic increase in the polygonal complexity of computer graphics models. The overhead of managing, processing, and manipulating very large polygonal-mesh connectivity information has led many leading researchers to question the future utility of polygons as the fundamental graphics primitive. Second, modern 3D digital photography and 3D scanning systems acquire both the geometry and the appearance of complex, real-world objects. These techniques generate huge volumes of point samples, which constitute discrete building blocks of 3D object geometry and appearance, much as pixels are the digital elements for images.

Pointshop3D is a system for interactive shape and appearance editing of 3D point-sampled geometry. By generalizing conventional 2D pixel editors, it supports a great variety of different interaction techniques to alter shape and appearance of 3D point models, including cleaning, texturing, sculpting, carving, filtering, and resampling. One key ingredient of the framework is a novel concept for interactive point cloud parameterization allowing for distortion minimal and aliasing-free texture mapping. A second one is a dynamic, adaptive resampling method which builds upon a continuous reconstruction of the model surface and its attributes. These techniques allow to transfer the full functionality of 2D image editing operations to the irregular 3D point setting. Pointshop3D system reads, processes, and writes point-sampled models without intermediate tesselation. It is intended to complement existing low cost 3D scanners and point rendering pipelines for efficient 3D content creation.

Mark Pauly is a postdoctoral scholar with the Guibas lab at Stanford University. His research interests include point-based graphics, geometric modeling, physics-based animation, and shape analysis. He received his PhD in 2003 from the Swiss Federal Institute of Technology, Zurich, Switzerland. He is a contributor to Pointshop3d, an open-source framework that facilitates design of new algorithms for point-based graphics.

Diane E. Shapiro created the article "Job Searching at SIGGRAPH 2004" for the Silicon Valley SIGGRAPH Chapter

Here is a link to the article:
http://silicon-valley.siggraph.org/MeetingNotes/siggraph2004.htm

June 3, 2004
Electric Sheep: animating and evolving artificial life-forms
by Scott Draves

Electric Sheep is a distributed screen-saver that harnesses idle computers into a render farm with the purpose of animating and evolving artificial life-forms. Each clip of animation has a genetic code, and the collective voting of users determines its fitness. In the next version a P2P network distributes the bandwidth of sharing the video and votes.

Scott Draves a.k.a. Spot is a visualist and programmer residing in San Francisco. He is the creator of the Fractal Flame algorithm, the Bomb visual-musical instrument, and the Electric Sheep distributed screen-saver. All of Draves' software artworks are released as open source and distributed for free on the internet. His award-winning work has appeared in Wired Magazine, the Prix Ars Electronica, the O'Reilly Emerging Technology Conference, and at the Sonar festival in Barcelona. In 1997 Spot received a PhD in Computer Science from Carnegie Mellon University for a thesis on metaprogramming for media processing. Today he regularly projects live video for underground parties and at clubs, and he just released SPOTWORKS, a DVD of abstract animation synchronized with music. http://spotworks.com, http://draves.org

May 13, 2004
Matrix Revolutions: Techniques and Methodologies With Large Scale Sentinel 'Swarm' Scenes
by Mike Morasky

This presentation will involve the techniques and methedologies used in the large sentinel swarm scenes in the Matrix Reloaded and Matrix Revolution movies.

Mike Morasky was a lead Technical Director on the "siege" sequence of the Matrix sequels (Reloaded and Revolutions) at ESC Entertainment. He was involved in the design and implementation of the production pipeline for that sequence and specifically in charge of handling the sentinels and their "swarming" system. Before that he worked at Weta on the "Lord of the Rings" trilogy where he was a Lead Technical Director in the Massive department. Mike is currently a CG Supervisor for "Circle-s FX" working on "Catwoman".

Real time 3D graphics applications continue to demand higher performance platforms. One way to increase the performance of your application is to multi-thread to take advantage of multiprocessor systems as well as systems with Intel processors with Hyper-Threading technology. This presentation will discuss threading concepts and how they apply to multiprocessor systems and Hyper-Threading.

Dean Macri has a B.A. in mathematics and computer science with a minor in physics from St. Vincent College, Pennsylvania, and an M.S. in computer science from the University of Pennsylvania. After completing his master's degree in 1992, he spent five years developing highly optimized C, C++, and assembly language routines for a 2D graphics animation company. He joined Intel in January, 1998 to further pursue his interests in 3D computer graphics. Dean is currently a staff technical marketing engineer in the Software and Solutions Group at Intel. He works primarily with game developers to help them optimize their games for present and future processor architectures and take advantage of the processing power available to enable exciting new features.

April 8, 2004
Making and Measuring Effective Virtual Environments
by Fred Brooks

Hosted by the Software Development Forum (SDForum)

Event Co-hosted by
BayCHI, SEM SIG
Silicon Valley ACM SIGGRAPH

Series Co-Hosts
Computer History Museum, CSPA, ACM San Francisco Bay Area Chapter

Fred Brooks is a legendary figure in computing. He led the development of the IBM System 360, wrote "The Mythical Man-Month: Essays on Software Engineering", and founded the Computer Science department at the University of North Carolina. His many awards include the National Medal of Technology, the A.M. Turing award of the ACM, the Bower Award and Prize of the Franklin Institute, and the John von Neumann Medal of the IEEE.

In this talk, Dr. Brooks will discuss his current work in virtual environments. The Effective Virtual Environments project at Chapel Hill is trying to determine which technological factors are crucial, which important, and which are negligible in making virtual environments illusions effective. Says Brooks, "We have studied eight different factors so far, with interesting and sometimes surprising results. I shall briefly describe the experiments and the chief findings."

March 25, 2004
3D Display Technology - from Stereoscopes to Autostereo displays
by Ian Matthew

In this presentation, Ian Matthew will discuss the history of 3D display devices and the technology involved in bringing out-of-the-screen 3D effects to a viewing audience. The presentation will discuss how we see in 3D, how 3D display technology works, what is involved in displaying 3D on computer monitors and the software requirements involved in making this happen.

In the past few years, the Holy Grail has been to achieve 3D viewing experiences without encumbering the viewer with glasses. A number of companies have achieved this to some extent. This presentation will also cover how this has been achieved and will introduce Sharp's new switchable glasses-free 3D LCD display technology. The new Sharp Actius RD3D laptop computer which includes the 3D switchable LCD will be demonstrated at the end of the presentation.

Since 3D display technology also requires collaboration between computer software and content, Sharp announced the formation of a 3D Consortium. The goals of this consortium will be covered and an attempt will be made to look into the crystal ball.

Ian Matthew is 3D Business Development Manager for Sharp Systems of America. He is responsible for marketing activities and developer partner program for the Sharp 3D laptops and displays. In this role he is continually adding new software providers to the Sharp Actius RD3D platform.

Ian recently joined Sharp from StereoGraphics Corporation, a company based in San Rafael, California which produces 3D viewing glasses for the professional market. At StereoGraphics, Ian held a number of positions, including Director of Marketing and Director of Product Management. As Product Manager for the StereoGraphics SynthaGram, autostereo display, Ian played a major role in bringing their first glasses-free display to market.

Prior to joining StereoGraphics, Ian's background was in the Computer Aided Design industry, and has worked for the major companies including Intergraph Corporation, Autodesk, Rebis Industrial Workgroup, and Dassault Systemes. He worked in software marketing at these companies. Ian has a B.Sc.Tech and a MSc. tech. in Chemical Engineering and Fuel Technology, from Sheffield University in the UK.

February 12, 2004
Valentine's Day Special
Friendship, Dating and Fun in the Virtual World of "There" (www.there.com)
by Will Harvey and friends

Science fiction authors have been writing about virtual worlds for twenty years, and entrepreneurs have been chasing that vision for almost as long, looking for the elusive killer app. Could online socializing finally be it? Online socializing has become a mainstream phenomenon, but remains predominantly a medium of typing text. Is text the end of the line, or is online socializing destined to become a medium in which people can communicate and interact as avatars? In the six years of developing There, we attempted to create an online world where communicating as avatars felt natural and fun, like communicating in the real world. We found the challenge to be much more difficult and subtle than we ever imagined.

This presentation chronicles the challenges There had to overcome in its development, the solutions that they found, and the lessons they learned, giving examples from some of the earliest prototypes to the live service that is running today.

Will Harvey - Founder of There
Will is a seasoned entrepreneur with a strong background in computer science, software, and video game development. Will founded There in 1998 out of a small room in his parent's house, where he recruited the technology team and built an end-to-end prototype before raising capital to grow the company and hire the management team.

Before founding There, Will ran the dynamic media products at Adobe Systems, including AfterEffects and Adobe Premiere, the world's leading video editing program. Will came to Adobe when Adobe acquired Will's previous company, Sandcastle, which Will had founded to develop network technology to enable low latency interaction over the internet. Prior to Sandcastle, Will served as Vice President of Engineering at Rocket Science Games in San Francisco, where he led the company's transition from full motion video based games to games focused on interactivity.

Prior to Rocket Science, Will founded and ran several successful game development companies while simultaneously earning his Bachelor's, Master's and Doctorate degrees in computer science from Stanford. Will's doctoral thesis introduced several important search algorithms which are now used commercially in manufacturing scheduling.

Will's game companies produced Platinum and Gold game titles including Zany Golf, Immortal, and Music Construction Set, with combined sales of over a million units. Will has filed 5 patents related to networking, graphics, and automated scheduling. He wrote his first commercial video game at the age of 15.

Jade Raymond, Producer and Developer
Raymond is a seasoned product manager and producer with a knack for developing fun and addictive online entertainment. She has more than five years of experience leading the development of large-scale "triple A" games and persistent worlds. Before joining the marketing team at There as a Product Manager, Jade was the Producer of The Sims Online, where she was directly responsible for all design and implementation of online game features and content for EA's highest revenue generating wholly owned property. As Producer of EA's most anticipated new product, Jade managed and led multiple teams of producers, artists, engineers, designers and testers, overseeing development and ensuring The Sims's Online's place in the media. Previous to The Sims Online, Jade founded the first Research and Development group within Sony Online. Her team was responsible for leveraging Sony IP across multiple platforms and ultimately building Sony Online's most trafficked offerings: the entire suite of Jeopardy games played by over 3,000 simultaneous users on a daily basis. Prior to Sony Online, Jade Raymond developed the first ever massively multi-user 3D shopping experience as part of a special project for Microsoft's advanced research group and participated in the creation of best selling entertainment titles for both IBM and Crayola.

January 15, 2004
The SIGGRAPH 2003 Electronic Theatre
by Silicon Valley ACM SIGGRAPH

Video presentation of cutting edge computer animation from the 2003 SIGGRAPH conference.

November 13, 2003
Extensible 3D Graphics (X3D)
by Tony Parisi and special guests

Join X3D editor and co-author Tony Parisi and special guests for an evening discussing Extensible 3D Graphics (X3D), the new standard for web and broadcast 3D graphics content.

Tony Parisi - "If you have to ask the question..."
President, Media Machines, Inc.
Tony will provide 10-minute overview of X3D, description of the features, state of the practice, and a look to the future.

Tony Parisi is a technology pioneer and accomplished entrepreneur at the forefront of Internet New Media. Tony is co-creator of the Virtual Reality Modeling Language (VRML), the ISO standard for 3D graphics on the World Wide Web, and is widely recognized as an expert in standards, technologies and emerging markets for interactive rich media. In 1995 Tony founded Intervista Software, an early innovator in real-time, networked 3D graphics technology and developed WorldView™, the first real-time VRML viewer for Microsoft Windows. In 1998 Intervista was purchased by PLATINUM technology, inc. and Tony joined the company to lead business affairs for its 3D visualization group.

In 1999 Tony founded Media Machines to provided business planning, business development and technology consulting services to companies in the bay area and worldwide. Tony is spearheading the development of Flux™ , a real-time 3D technology that continues to push the envelope in interactive graphics for the web. Tony is also a lead editor and co-chair of the Extensible 3D (X3D) Specification, the new standard for Web3D graphics being developed by the Web3D Consortium.

Alan Hudson - The Xj3D Browser
President, Yumetech, Inc.
Alan will give a tour of Xj3D, Yumetech's cross-platform X3D browser and application toolkit written in Java.

Alan has been involved in virtual reality systems for the past 7 years. He currently leads the Open Source task group of the Web3D Consortium, and is a co-author of a book on the Java programming language. His previous projects include the development of immersive training environments, an online library publishing and automation system, and software for collaborative manufacturing design.

David Arendash - Unreal->X3D Converter
Multi-Media Development Engineer, The ManyOne Network
David will show how game-quality spaces can be authored in the Unreal level editor and easily converted to web games running in X3D.

David Arendash is a multimedia design engineer, responsible for getting artistic vision into practical implementation. Mr. Arendash began designing digital hardware at age of thirteen, and has been programming computers since age fifteen. Prior to joining ManyOne Networks, Mr. Arendash ran an independent Development Contractor firm, Quantum Leap Computing, where his clients included Xerox, StoryVision, Strategic Mapping, and Ziff-Davis/PC-Labs. Quantum Leap Computing's areas of focus were on multi-player online card games using Flash5 & PHP and on the full development of many 2D, 3D, and game development utilities including modifications, add-ons, and levels for Unreal/Tournament Engines. Previously, Mr. Arendash was responsible for inter-team middle-ware library and tools at Intuit for over four years. Prior to this, he worked in development and engineering capacities for companies including Software Publishing in Mountain View, California, DEST Corporation in Milpitas, California, Allen-Bradley Company, in Highland Heights, Ohio, and Wang Laboratories in Lowell, Massachusetts. David received his B.S. in Computer Engineering from Case Western Reserve University in 1984.

Keith Victor - Vizx3D
President, Virtock Technologies, Inc.
Keith will demonstrate Vizx3D, an authoring tool for X3D and VRML. Keith will be presenting some content created using Vizx3D, and demonstrating how complex features, such as H-Anim Avatars and MultiTexturing, can quickly be authored.

After graduating with a BS and MS in Mechanical Engineering at the University of Wisconsin-Madison, Keith began his career as a project Engineer for Delco Products Division of GM back in 1989. In 1993, Keith left GM to work for SDRC, a provider of CAD software. In 1999, Keith left SDRC to found Virtock Technologies, Inc., where he released Spazz3D, an early entry into the VRML tool market. Keith sold Spazz3D to Eyematic Interfaces, Inc., where he worked as a Software Engineer until 2003. Keith was able to regain the IP for Spazz3D when Eyematic ceased, thus he revived Virtock Technologies, Inc. with the upcoming release of Vizx3D, an authoring tool for X3D and VRML.

David Petchey - Planet9 Projects
Software Architect, Planet 9 Studios
David will highlight X3D projects and products developed by Planet9 studios, including eSCENE, a 3D Command and Control interface to aid in the real time management of terrorist incidents, accidents and training scenarios.

David has 20 years of experience as a programmer and software project leader covering a wide range of disciplines, with particular expertise in multimedia Windows development. David has designed, implemented, debugged, and shipped numerous products primarily for the PC and can take any software project through its full development cycle. David has worked for large corporations such as Microsoft and Mindscape as well as co-founding and consulting to several startups.

October 16, 2003
SIGGRAPH 2003 Technical Papers Review and Discussion
by Silicon Valley ACM SIGGRAPH

This is a discussion-style meeting (as opposed to our normal presentation style), during which we will discuss the technological advancements represented by the technical papers presented at SIGGRAPH 2003. Participation is encouraged. It is highly recommended that you bring your SIGGRAPH Proceedings.

Because of the participatory nature of this meeting, there will be no charge for attending this meeting. However, we will be ordering a pizza and donations will be solicited to cover its cost.

Are you offended by drawings of cows with udders? Seventy years ago censors in the United States decided if you show a cow in an animated cartoon it had better be wearing a dress to cover those teats.

Animation has been censored for a variety of reason including being too risqué, too violent, and for showing racial stereotypes and possibly for having subversive content. Forbidden Animation is a film program that explores images and words that were banned when formal censorship was established in America in 1934. The program, based on Karl Cohen's book "Forbidden Animation: Censored Cartoons and Blacklisted Animators in America", illustrates what was once considered acceptable, but was then outlawed by a document called the Production Code. From 1934 to 1968 it was enforced by men with scissors who took their jobs seriously.

Karl F. Cohen teaches animation history at San Francisco State University and is president of ASIFA-San Francisco, a chapter of the international animation association. Prior to SFSU, he has taught at Toledo University and was the curator at the Toledo Museum of Art. He had produced various commercial and personal film projects. He has been a guest presenter at various film festivals in Europe, Israel, Canada, and the United States.

The recent trend in graphics hardware has been to replace fixed functionality with programmability in areas that have grown exceedingly complex (e.g., vertex processing and fragment processing). The OpenGL Shading Language has been designed to allow application programmers to express the processing that occurs at those programmable points of the OpenGL pipeline. A desire to expose the extended capability of the hardware has resulted in a vast number of extensions being written and an unfortunate consequence of this is to reduce, or even eliminate, the portability of applications, thereby undermining one of the key motivating factors for OpenGL. A natural way of taming this complexity and the proliferation of extensions is to allow parts of the pipeline to be replaced by user programmable stages. This has been done in some recent extensions but the programming is done in assembler, which is a direct expression of today's hardware and not forward looking. Mainstream programmers have progressed from assembler to high-level languages to gain productivity, portability and ease of use. These goals are equally applicable to programming shaders. The goal of this work is a forward looking hardware independent high-level language that is easy to use and powerful enough to stand the test of time and drastically reduce the need for extensions. These desires must be tempered by the need for fast implementations within a generation or two of hardware.

Jon Leech is the technical lead of the OpenGL engineering group at Silicon Graphics. As Secretary of the OpenGL Architecture Review Board, he has led the group since 1997 and edited the OpenGL 1.2 - 1.5 API Specifications. He also participates in related standards groups including the OpenGL ES working group of the Khronos SIG, and a new joint effort with Sun Microsystems to standardize Java bindings to OpenGL.

Prior to joining SGI, Jon did research in interactively steered molecular modelling and highly parallel computer graphics architectures at the University of North Carolina at Chapel Hill. He has a M.S. in Computer Science from the California Institute of Technology.

Diane E. Shapiro created the article "A Foray into Experimentation, Illusion and Surreal Expression at the SIGGRAPH 2003 Art Gallery" for the Silicon Valley SIGGRAPH Chapter

Here is a link to the article:
http://silicon-valley.siggraph.org/MeetingNotes/ArtGallery2003/siggraph2003.htm

Ask 10 CAD managers to define "CAD standards," and you will probably get 10 different answers. Although most standards include some sort of layering scheme, most offices also include guidelines for plotting, file and directory naming, as well as many other areas that affect the quality of CAD output.

Modern CAD systems, such as AutoCADR 2004 software, are extraordinarily flexible and can be tailored to fit just about any workflow. This flexibility comes at a price, however; different users can use vastly differing methods to produce a drawing, with different visual results. As a CAD manager you need to establish standards to ensure that your firm's output is of a consistent high quality, no matter who produced a particular drawing, or when.

The concept of ensuring consistent output from a design or drawing office is not new. Even in the days of the drawing board, responsible firms followed design and drafting guidelines that specified standards, for instance, sheet sizes and scale, text and dimension sizes, and styles. Modern technology has introduced many new ways to produce drawings. Although computers and CAD systems have brought many productivity benefits, they have to be carefully managed to produce the desired results.

For more information, go to the following link:

http://www3.autodesk.com/adsk/files/2704184_AutoCAD2004_CAD_Stds.pdf

Lynn Allen, CADENCE columnist and worldwide Autodesk Technical Evangelist, speaks to more than 15,000 users each year. For the past eight years she has written a monthly column in CADENCE magazine called "Circles and Lines." Lynn started using AutoCADR software with Release 1.4, over 16 years ago, and has taught at the corporate and collegiate level for 13 years. A sought-after public speaker with a unique comedic style, Lynn is always one of the highest rated speakers at Autodesk UniversityR. Her latest writing endeavor is AutoCAD 2002 Inside and Out.

For more information on Lynn, please visit www.autodesk.com/lynnallen

Nonconvex Rigid Bodies with Stacking
by Eran Guendelman

In this talk, we will present our SIGGRAPH 2003 paper "Nonconvex Rigid Bodies with Stacking". We consider the simulation of nonconvex rigid bodies focusing on interactions such as collision, contact, friction (kinetic, static, rolling and spinning) and stacking. We advocate representing the geometry with both a triangulated surface and a signed distance function defined on a grid, and this dual representation is shown to have many advantages. We propose a novel approach to time integration merging it with the collision and contact processing algorithms in a fashion that obviates the need for ad hoc threshold velocities. We show that this approach matches the theoretical solution for blocks sliding and stopping on inclined planes with friction. We also present a new shock propagation algorithm that allows for efficient use of the propagation (as opposed to the simultaneous) method for treating contact. These new techniques are demonstrated on a variety of problems ranging from simple test cases to stacking problems with as many as 1000 nonconvex rigid bodies with friction.

Eran Guendelman is a Computer Science PhD student at Stanford University. Working under Ron Fedkiw, Eran's research has focused on physics-based modeling for computer graphics.

All-Frequency Shadows Using Non-linear Wavelet Lighting Approximation
by Ren Ng

Ren's talk will be a preview of his SIGGRAPH 2003 paper, titled "All-Frequency Shadows Using Non-linear Wavelet Lighting Approximation." This work was done with Ravi Ramamoorthi at Columbia and Pat Hanrahan at Stanford. This paper describes real-time rendering of objects under all-frequency illumination represented by high-resolution environment maps. Current techniques are limited to small area lights, with sharp shadows, or large low-frequency lights, with very soft shadows. Our main contribution is to approximate the environment map in a wavelet basis, keeping only the largest terms (this is known as a non-linear approximation). We obtain further compression by encoding the light transport matrix sparsely but accurately in the same basis. Rendering is performed by multiplying a sparse light vector by a sparse transport matrix, which is very fast. For accurate rendering, using non-linear wavelets is an order of magnitude faster than using linear spherical harmonics, the current best technique.

Ren Ng is a first year PhD Computer Science student at Stanford, working with Pat Hanrahan at the Stanford Computer Graphics Lab. His current research focuses on real-time relighting with detailed natural lighting, and will be the focus of his talk tonight. He has also published papers on virtualizing graphics hardware in support of real-time shading languages, and on scalable interactive volume rendering using clusters of rendering servers. Ren is actively exploring career options in industry research, graphics hardware, movie making and academia.