Massive

SIGGRAPH report

SIGGRAPH 2002 was held in San Antonio, Texas, last week. The numbers were far smaller than previous years and there didn't seem to be a huge amount of innovation in the trade show. Once again, photo-realism seemed to be the goal of most products on display. What you do with all this was less talked about but there were some good exceptions.

SIGGRAPH in Texas wasn't all about meat but it did feature heavily in advertising. *gulp*

Highlights of the week:

The course on interactive multimodel media systems design given by the Human-Computer Interaction group at the University of British Columbia. They talk about vision is constructed and how cognitive science is unravelling how we perceive things. Wait till Marketing get a load of this.

The Making of Spiderman session. OK I'm a sucker but I've seen that ending sequence almost fifty times I reckon now and I still think it's cool. Cool is not a word I usually feel comfortable about using around SIGGRAPH but the artistry that went into this high end production was fascinating to watch.

The U.S. army blew our minds with their beaming pride over America's Army. Educating the world as to what it's like to be in the army. Perform well in 32 man online squad combat and the recruiters will love you. You might just get yerself a placement. The course on this project, and scalable MMPGs in general, by the Moves Institute was illuminating to say the least.

Disney showed off their Digital Human Project and had a plastic looking actor alongside his identical digital double. Of course, they then had the real actor (we assume he was real) come in and meet his double, about twenty years younger.

A refreshingly not technical course on Imagery and Symbology was startlingly upgraded to a Cinematrix interactive cinema experiment one afternoon. You held a paddle with red and green panels on either side and saw yourself as a dot onscreen along with all the other audience dots. Flying a plane, mexican waves, the creative possibilities were endless.

Procedural modelling and animation is clearly the future of visualisation and was on everyone's lips. Procedural cities, sculptures, landscapes. SideEffects announced a free version of their procedural software Houdini while Ken Perlin demonstrated the online possibilities in his "Web as Procedural Sketchbook" course.

Interactive fiction was represented by the Facade project - a highly mathematical approach to plot and character via AI routes. The demo was more intellectual than emotional but compelling nonetheless. The system uses AI to recognise story arcs and sequence dramatic beats through a natural language interface. The team, including the man behind Catz and Dogz, will also be releasing ABL - a language for reactive planning interfaces and AI structures for crafting personality. One to look out for...

Will Wright, creator of The Sims, spoke at length about "possibility space" as the landscape of gaming, the idea of 'story recognition' (where the game identifies narrative elements driven by the user) and how successful games provide participants with a way to map their own experience onto the game itself.

A panel on the future of gaming was fascinating with a few speakers referring to an ecology of media in passing, though none thought to expand on their views. Game designers spoke of the need to start with the player experience, rather than with a notion of story or character sketch. Will Wright's "big three" for the future of gaming were METRICS, PLAYER CONTENT, and CONNECTIVITY. The notion was raised that The Matrix is purely a prequel to The Matrix MMPG. The notion of operating "under the cultural radar" was liberating.

Other mentions of note, PingID - a public domain identity management system, TextArc - a visualisation system for text, and the Virtual Terrain Project - free tools for procedural scene construction.

SPIDERMAN

Sometimes cool is not so cool, especially when it feels like you're at a rendering orgy for its own sake. As far as other entries to the special effects arena this year, this was one of the big ones and it was heartening to hear how the guys (perhaps a few women but they didn't get to speak) threw themselves into upping the ante with visual effects and managed to inject a sense of humour into their presentation of it. The "not allowed to tape" 2hr presentation focused on the building of a virtual New York from photographs, painstakeningly stitched together. You've really got to wonder what kind of detail they put into films these days for the benefit of the cast and crew. Not unlike, easter eggs in software, the artists showed considerable pride in showing how that window Spiderman zooms past for, ooh maybe half a second, looks into someone's living room with a Kenny Rodgers style framed photograph of the director. Oh, and they had to reverse every street sign and marking on the left side of the street because, after all, they filmed this in a one-way street in L.A. not Manhattan. The final sequence was shot post-Sept 11 with hired N.Y. cabs in downtown LA. The shot of Spiderman swinging low between two cars and then back up into the sky was achieved using a camera flying down a cable slung between two cabs.

Interesting stuff also with the thought processes going into creating the first spider-sense sequence in the school corridor. They used young mime artists in the scene to hold poses, as in a freeze frame, while the camera tracked around the space, overlaid with CG elements.

THE SIMS

Will Wright was the speaker I kept hearing at the conference. He expressed little enthusiasm for non-linear storytelling, describing a linear story as 'a ride on rails'. Will produced an incredible wealth of statistics about The Sims usage showing the success of a meta-game, where players expand out to other avenues uncontrolled by the game designer. He positioned film and game at opposite axis. He spoke of an appeal stretching from empathy (to a film) to agency (in a game) without seeking to muddy the waters.

He used The Truman Show as an analogy of how games can provide freedom of movement and a story, in hindsight.

He offered a blueprint for dynamic story representation:
STORY - INPUT - VISION.

He divided presentation management into
1) Parsing
2) Presentation
3) Influence
4) Replay

OTHER REPORTS

Why take my word for it? Also see Hyperfiction.org, the movie zone, and SIGGRAPH itself (including a mention of my little flic - HfC).

Posted by .M. at August 3, 2002 10:54 AM