Skip navigation

“7 Systems of Control” was to be a modified or adapted opera. After researching the history of opera, one important concept became evident, namely the idea of the “Leitmotif.” Since the “opera” was a collaboration among different groups, an overarching umbrella structure needed to be constructed. The “Leitmotif” would fill that void. Rather than choosing a traditional musical motive, an algorithmic procedure was chosen instead. The algorithmic procedure was an Artificial Neural Network, in a sense a very basic artificial intelligence system based on biological neural networks of the brain. ANN had the capability of learning and therefore recognizing various performance patterns amongst the collaborations. The learned information was then juxtaposed with unknown variables and new unpredictable outcomes were achieved. However, even though outcomes of evaluated information would be different every time, basic shapes would remain the same. ANN was interpreting incoming data via sound analysis, number analysis and OSC data. The interpretative aspects of ANN were programmed via C extensions in PureData and where output as sonic sound scapes that emerged, participated and interfered throughout the performance of the “opera.” The first module involved capturing data from the participating audience. ANN collected data via OSC and tried to discover patterns within the audience participation, which was masked by seemingly static noise à la Xennakis’ ConcretePH, dispursed over 12 audio channels. Out of the collected data during the dispersed noise phase emerged the “Overture.” During the second module, a series of read texts were evaluated and interpreted by performers. ANN was to create a commenting Greek style chorus from previously gathered information that was a 12 voice, 12 channel motet reminiscent of early tape music utilizing voices by Karheinz Stockhausen. The third module consisted of an interactive aria or duet between a visual sculpture and a female singer. ANN evaluated the data and created a processed or granulized  reaction of the singers previously recorded samples, which created another sort of motet for 12-voices, spatialized over 12 channels that served as a transition to the fourth module. Module four had a dance sequence that dealt with words of control projected onto a screen, which was interactively accentuated by dancers. ANN collected data via OSC and functioned as an aggregator for punctuated gestures. The fifth module contained a piano duet between ANN and a piano player. ANN began with a rhythmic pattern animating the piano player to approach the piano. ANN had the task of recognizing patterns within the piano players idiosyncratic improvisation. The module culminated in a battle between man vs. machine, which was on one hand won by the machine because the piano player was not able to replicated or respond anymore to the machine, but on the contrary was won by the human since the calculations ended destroying the system. During the sixth module ANN received a well-deserved break (the neural network had to be restarted since it culminated in its destruction during the fifth module). During the finale or seventh module, ANN would be interweaved into a humanly performed rant and would create a sound scape from all learned data and recognized patterns throughout the evening. However, ANN was not the only creator of sound, but certainly was crucial collaborator for the accumulated Apex.

Pdf files:

ANN_GrainEngine_1

ANN_GrainEngine_2

ANN_OSC_motion_capture

Advertisements
%d bloggers like this: