 |
|
|
Design Challenge 5There are many innovative software systems today for producing music. However, there is a frustration with these systems in that we often see no physical embodiment of the system. Many of us have been to "laptop gigs" where we see several musicians behind their laptops coding away and we hear the resulting performance. But there is a frustration here, because we cannot equate the artistry or the innovation of performers say behind their laptop in the same way as watching a concert pianist (say) at work. In this design challenge we are specifically interested in Live Music Algorithms. These are autonomous, digital musical improvisers. See www.livealgorithms.org for more details. An increasingly important part of this research agenda focuses on embodiment. Although at a formal level a LA need only deal with sound, there is no doubt that human performers also rely on visual cues. The theatrical elements of any performance are also vital for the audience, and audience involvement feeds directly back to the performers. Any performer will go through periods of assimilation, collaboration, and innovation, and visual cues help us (fellow performers and other listeners) enormously to understand what is going on "under the hood". This helps sustain our credibility in the performer. These issues are already important in human improvisation but become crucial in the inert world of laptop performance. Basically, how can we develop a "sense of presence" with a machine? Possibly, quite simple cues might do the job. Suppose a machine has two modes innovation and imitation. In order to imitate, the machine must attend to a source, and this could happen by a rotation of a microphone towards the collaborator.Or some visualisation indicating that the algorithm is "leaning towards" another player. Similarly, during moments of innovation and leadership, the machine might "back away" from the group, We need to infer that the algorithm is concentrating deeply, just as we might with a human, for example by noticing a furrowed brow. Can this be animated? These tricks might enable us to believe that the machine is more than automatic, is autonomous even. But it could go even further - audience and performer involvement could be parameterised, a kind of "inverse embodiment". This data would then influence our machine, for example controlling the rate of switching between modes of behaviour. There are several live algorithms that we could use, including my own experiments with virtual swarms. Please see www.timblackwell.com for more information. |
|
|||||||||||||
|
|||||||||||||||||
| copyright | privacy | credits | |||||||||||||||||
|
|||||||||||||||||
|
|
|
|
|
|
|
|
 |
|||||||
