Posts Tagged ‘Participatory Chinatown’
Participatory Chinatown
May 6th, 2010
I recently finished work on a slightly different project than I’m usually used to working on. It is a multi-player game called Participatory Chinatown. The game’s function is to facilitate community meetings in Boston’s Chinatown neighborhood by allowing community members to interact in a virtual environment that simulates the actual neighborhood. The game was conceived by Assistant Professor Eric Gordon at Emerson College. I was contracted by the Asian Community Development Corporation and worked closely with the game developers at Muzzylane Software and the Metropolitan Area Planning Council.
My part in the project was to take a collection of rigid character models, rig them for animation and then to animate several animation cycles for use in the game. Going from a film environment to a game environment was quite a learning experience as many of the tools I am used to using do not translate very well to the specific and procedural nature of games. For instance, I was unable to use complex and realistic rigging tools that are used in film since the game engine used Collada formatted models, which only supports a select variety of deformers. To work around this, I had to get quite creative in the way I rigged the model.
Generally, characters are rigged in such a way that animators actually control meta objects that then control the joints that actually affect the model’s vertices. In other words, these controls act as the puppet strings. However, for this particular application, the animation data had to be contained within the actual joints for use in the game. Not only was this a technical necessity, but also an efficiency issue since every part of the game had to be optimized for size and computational footprint. Unfortunately, this also made it rather difficult to animate. I could have chosen to struggle with the simple, bare-bones rig I had created, however the though of having to revise my work on a rig with no complex controls bothered me quite a bit, so instead, I made two rigs. One with no controls, and another with controls that animators are used to seeing. Using a series of scripts and constraints, I was able to take my complex rig and bake the animation curves into the game-ready version of the character which could be successfully integrated into the game by the software developers.
40 characters later, the game is complete and is now actively being used in community meetings! One of which I was able to experience last night. It was great fun to see the game in action with real end-users!