Lightning and Sound
by Rael Dornfest
Scene: A five minute elevator ride. Players: You, your idea, and a captive audience. The O'Reilly P2P Conference adapted the popular
Lightning Talks from
YAPC (Yet Another Perl Conference). A sort of "Idea Karaoke," each speaker is on the clock from setup to closing.
I had the pleasure of moderating the Technical Lightning Talks which included pitches by Applied MetaComputing, OpenCola, Roku, PeerMetrics, NeuroGrid, Porivo, OverX and iMesh. Presentations were followed by both technical and business/VC analysis by
R.V. Guha of Alpiri and Barry Taylor of
Warburg Pincus. While the overall usefulness of the session was questionable, the audience were overwhelmingly in favour of a repeat performance at the upcoming fall
P2P conference in DC.
The runaway favourite was a fellow named Steven Hazel and his creation, soundmosaic. Soundmosaic borrows from the the photo mosaic, those posters of Vader, Einstein, et al constructed from countless tiny photographs. Soundmosaic applies this principle to sound, building an approximation of an audio sample out of bits and pieces of other samples. Steven's demonstration was a classic clip of The Beatles introducing themselves ("I'm John and I play the guitar...") featuring samples of John Coltrane's performances. While still quite choppy, Steven believes with enough computing power (read: distributed computing), these sound mosaics will, at the very least, make wonderful audio effects in his music.
I sat down with Steven for a little while after the session to ask the obvious: Why? Steven is foremost a musician with an interest in sound effects beyond what is currently possible with a small circuit (eg. flange, wa-wa pedal). But he's also playing with the line between fair-use and copyright violation. At what point does a recording replaced bit by bit with samples of other sounds cease to be the original recording? As both a programmer and musician (not an uncommon combination, mind you), Steven finds himself on both sides of the current music industry / Napster furor. With the viability of contemporary intellectual property concepts in question, Steven and folks like him are looking forward, experimenting, and testing the limits.
Steven's attendance at the conference is as a
developer. As both content-producer and programmer, "Freenet really raises issues close to home." Yet he's not using Freenet in the distributed architecture behind soundmosaic since Freenet's focus is anonymity which is just not needed for the project at hand; given this, he'll be using something faster and more reliable. This isn't the first time I've heard that Freenet's particular slant can indeed be a hindrance in its use in P2P projects.
Asked why, then, he's so actively involved in the Freenet project: "Every individual feels that they are the person keeping Freenet from sucking!"
I (and everyone else at the session) look forward to hearing more from Steven at the next conference.