Another piece of work from my time at CEMA; this one a paper, co-authored with Mark Guglielmetti and Troy Innocent. This paper started with some discussions about models in generative systems, and a feeling that certain kinds of models, or rather certain ontologies - formally defined networks of entities and relations - play an important role in defining the generative outcomes of formal systems. Troy and Mark are also very much into gaming (more than me anyway, my peak gaming experience occurred about twenty years ago and involved an Amiga 1000); as we talked it seemed that these generative ontological structures might also be at work in some of the more interesting games and game art projects around. Mark made me sit down and play Portal (below). Then we started discussing social software...
So in this paper we consider both philosophical and computational senses of "ontology", and propose that computational ontologies (or data models) actually implement philosophical ontologies (notions of what "is"). What's more these ontologies become dynamic, interactive processes; and that's when things get interesting. We focus on "strange ontologies": where default, common-sense or conventional ontological structures are tweaked or hacked, or where emergent phenomena pop out from apparently straightforward structures of being and relation. We draw on examples from social software, gaming (including Portal and Warcraft), art games or game art (including Julian Oliver's Second Person Shooter), new media / generative art (Guglielmetti's own Laboratories of Thought (below) and Jonathan McCabe's Origami Butterfly Method).
The paper has been submitted to an upcoming issue of ACM Computers in Entertainment; for now, grab the pdf and cite it via the permalink for this post. We're seeking feedback on this too - let us know your thoughts.
Thursday, April 03, 2008
Strange Ontologies
Tuesday, May 22, 2007
Fijuu, 3D Synaesthesia and the Loungeroom
Game/art notable Julian Oliver (aka delire) has been using game engines for audiovisual performance since way back. In 2001 I saw him play a Quake mod that had been rigged with audio samples and proximity triggers to create an immersive first-person performance tool; a digital hardcore jumping castle (I think the system was related to the later q3apd). In conversation at the same event, he argued for the potential of this approach. I saw the Quake mod as an ingenious sample trigger interface - a kind of 3D drum machine - but Oliver was looking ahead to realtime manipulation and deformation of geometry and sound. In retrospect he was evoking a form of synaesthetic media, where spatial and sonic attributes are fused and cross-mapped, so that the form is the sound. Gesture is significant here too - in performance practice gesture is at the interface of space, motion and sound. Oliver was imagining dynamic form as an articulation of sonic gesture, but also the prospect of folding back 3D form into sound; procedural texture-mapped geometry as a sonic provocation. What does this sound like?
This conversation came back to me vividly when I ran into fijuu2, a project by Oliver and Steven Pickles. Fijuu comes close to realising what Oliver imagined in 2001: a plastic, gestural, realtime audiovisual 3d environment. Forms twist, shatter and rotate, hovering inside cylindrical arcs of a gesture sequencer. Sound and form transform in unison, evoking a third, more abstract thing, the map or pattern that links them. Global filters influence sound and image, making another (logical) map between pixel shaders and audio effects. It's great to see lush, gaming-grade 3d graphics diverted towards a more abstract aesthetics of play.
fijuu is also interesting as a case study in art/game crossover and the free software ecology. It runs on Linx, using the Ogre graphics engine; all free and good, but installation looks fairly daunting. Apparently it will one day be available as a Linux live CD, bypassing the installation process and maybe approaching a game-console level of runnability. This is the exciting bit, for me. It would be great if artists could join the current battle for the "converged" screen (MediaCentre, PS3, FrontRow, TiVo, etc). One way would be for artists to somehow gain a toehold in mass market console gaming - but that doesn't seem likely, despite the odd promising exception. Another is for art to occupy interstitial attention, with screensavers like The Endless Forest and E-volved Cultures. Fijuu signals a third possibility, for art to temporarily occupy ubiquitous PC hardware as if it were a console. Back to the demoscene, and forward, to generative AV 3D interactive art in your loungeroom.
