Recently in my class with Ali Momeni, we had an exercise surrounding the idea of a networked ensemble of computers. It turns out to be a fairly easy and effective technique. In a couple of hours we had fifteen computers all talking to each other by sending OSC messages over a wireless network. The messages could trigger a sinusoid at a specified volume and frequency, noise at a specified volume and duration, or one of five samples played at a specified volume and rate. The samples were selected and loaded locally, so only the OSC messages were being transmitted over the network.
By the end of the experiment all fifteen computers were producing chaotic mash-ups of randomly pitched sine waves, random white noise, and an eclectic collection of pitch shifted and/or reversed sampled sounds. Unfortunately I didn’t have a portable recorder on hand, but what I did do was capture the audio that was being generated from the messages that were being received on my laptop. Here’s an excerpt of what that sounded like. Now try to imagine that times fifteen. I’m also attaching the patch in case anyone would like have have a look. Just be aware that it was made quickly (I did clean it up a bit before uploading) and there’s a dependency on an external called ali.samplor that can be found here.
network_instrument.maxpat
Network Instrument Example

For the last two weeks I have been working on a performance application that I’m developing in
I have begun to refine the patch I described in the last entry, creating an instrument that allows you to traverse through samples using an x-y controller and a slider. The x axis controls the forward or backward playback frequency of the grain. The y axis controls the width of the grain; either very narrow (minimum of 10ms), to an adjustable percentage of the sample at the widest (15% by default). The slider controls the position of the grain in the sample. Buttons across the top allow you to choose one of five preset samples. Alternatives can be loaded in the patch or set as defaults.