|
|
|
| The Sound Panel with it's Computer Control during a Test Run | |
NOTE: The Sound Painting was vandalized at Trondheim Kunstmuseum in 2001. No economic compensation has been received from the institution.
Experiment with 64 loudspeakers mounted in an 8x8 matrix covering a squared panel measuring 2 by 2 meters. The panel was not covered with any canvas nor did it present any image but the speakers themselves. By simply moving the sound across the panel, the experiment was to see if the overall sound presented would be perceived as a 2-dimentional quality rather than a full 3-dimentional spatial quality. The sound source used for this experiment was the high frequency noise generated by the sound controlling computer itself. The noise was picked up using a VHF receiver. The audible sound, i.e. the low frequency oscillation, was almost like the humming of bees. The quite hypnotizing Doppler Effect was clearly recognized by the rapid movement of sound being distributed over the panel at a rate of about 10 milliseconds for each speaker. The self-referent artwork included both the VHF receiver and a homebuilt computer.
An interesting comment from the audience was that the sound moved too fast to clearly comprehend any moving pattern. I would compare these comments with similar comments from people looking at abstract art and expecting a clear resemblance with the known reality.
Next step could be a 2-dimentional audio image superimposed over a painted canvas with the intention to see if it's possible to harmonize sound and image without any illusion of depth in either media. It could also be an attempt to extend the expression of painting itself and to experiment with how we comprehend paintings when they interact with another media.
Another improvement could be to furnish each loudspeaker with it's own 5 Watt amplifier and thereby contol an Audio Line Signal (1 volt) instead of the output signal being distributed to each speaker directly. The current method leaves all silent loudspeakers without an energized coil, which means that each time a loudspeaker receives a signal, it first has to energize itself, and then do it's work pushing air. All this within an extremely short timeframe. Using 64 Output Amplifiers within the panel would give me the option to sound more than one speaker at the time without having to worry about load characteristics in the same way as now, and even improve the dynamics of the sound beyond any current possibilities
|
|
| sound panel computer control | |
|
|
| detail from the backside of the panel | |
I made and tested the software that allow me to compose any pattern of sound movement across the panel, i.e. addressing patterns within an 8 by 8 matrix of 64 loudspeakers. This pattern is written to a file and saved on disk. The program uses this datafile and addresses the appropriate loudspeaker accordingly. Any sound source can be used as input to the panel through a standard audio amplifier system.
I used the Velleman Computer Interface Board K8000 hooked up to a parallel port on a homebuilt computer. This card provides an easy way to get started with computer interfacing. The kit includes routines in the C, Turbo Pascal, and QuickBasic programming languages, and comes with adequate documentation.
Jens Laland