Project Name: Machine-Check Exception (or MCE)
The experience of my project begins with the presentation. The idea is that the instrument is a “deconstructed desktop computer” in that there is CPU resting on its side with its innards exposed as well as a partially taken-apart keyboard and mouse. There should be wires and small parts scattered about the project to enhance the “broken computer” experience. This presentation could easily be accomplished in the Ellis gallery or any gallery for that matter. An “ideal” location for performance may actually be a junkyard since the parts would blend in more as “junk”.
There will be a stand with several objects (such as drum mallets, brushes, rubber bands, etc.) to use to bang against the insides of the CPU in which a contact microphone will pick up the disturbances. It is my hope that once I start taking a computer apart, I will also find ways to manipulate the insides without external objects; for example, physically pulling out the cd drive or moving the fan around may create unique sounds for the contact microphone to pick up. If the sounds are limited, I will be forced to actually attach objects to the inside of the computer, such as a door stop or some guitar strings, to create some unique vibrations.
The speakers for playback should probably be concealed underneath the table that the computer is resting on so that the individual insteracting with the instrument feels that the sounds are coming from the broken computer itself. The contact microphone will take the information to a concealed laptop running Max/Msp. Every time a new strike occurs, Max will alerted by the amplitude of the signal to cease it’s previous sound and begin playback of the new sound in which the individual strikes will be repeated indefinitely through granular synthesis (which normally loops sound of around 50 milliseconds or less) and then modified by other filters such as low/high-pass filters, ring modulators, amplitude modulation (and other LFO-related filters), and probably as many other effects that I can find. All of the resulting parameters will be controlled by the keyboard and mouse. Each key will be mapped to a specific effect and the X and Y axis of the mouse position will map to the granular synthesis (which effects pitch) and the volume. This will encourage the person interacting with it to try different combinations of different keys and mouse positions to create completely unique sounds. There will have to be a particular key, likely the space bar, that will stop and start playback and another key, such as the escape key, to reset the effects back to the default setting. There would also need to be a combination of key strokes that ends the program which wouldn’t be known to the general person interacting with the machine.
Xenakis wrote multiple pieces of music that incorporated analogue granular synthesis but the one that I find most compelling is “Analogique B” which is a work for tape that helps make up the more complex piece “Analogique A & B”:
Xenakis made such beautiful and interesting sounds using this method and through some kind of complex math and Markov chains he derived the string quartet parts that would, to some effect, mimic the electronic sounds.
Another such piece that Xenakis created using this technique (but that did not also have acoustic instruments) was Concret PH:
This is a big contrast from the other one because of how simple it is and that the whole piece seems to exist as a single amorphous soundscape.
One more from Xenakis:
Although this is technically a music conrete piece, it incoprates the use of granular synthesis.
This video is a great example of the technique I’m trying to emulate (it really starts at 0:23):
Every moment he plays is extended by granular synthesis and although he is just improvising, the result is beautiful. He’s not adjusting parameters or adding other filters though and I think the result of doing that would be wonderful. Plus I’m hoping to use very small, subtle sounds.
An unused computer including the CPU, keyboard, and mouse.
A laptop running Max/MSP.
Possibly other devices to cause sound inside the computer.
I have been doing a lot of preliminary experimentation in Max for Granualar Synthesis with fascinating results. Every sound seems to have an enormous range of manipulative possibilities depending on the length of the sample and the speed of its playback. I’ve watched many videos of projects using contact microphones and the results have generally been quite good.
My project explores several different avenues of live sound manipulation but the unifying element is definitely the granular synthesis. By taking microscopic samples of sound and looping it as well as slowing it down and speeding it up, a brand new waveform is created that is sustained, thus allowing for further interaction. However, the most important element in this project for me is the presentation. The fact that the performer will be performing music on a deconstructed computer and is physically manipulating the inside of a computer to make sound is extremely exciting for me. It definitely puts a different spin on the term “computer music”.