Frequency Histograms For Predictor Variables Ninety Nine 20-Second Songs for Human Body & Sampled Kitchen Items Originally released as LMM 2.6, part of Learning Music Monthly. August 2009
LMM issue 2.6 is a collection of 99 audio tracks, each twenty seconds long, made of only sounds from my body (singing, clapping, etc.) and manipulated samples of objects I found in our kitchen (metal pots, knifes, cups, spray bottles, cumin seed, a small plastic cooler, tupperware, crystal wine glass, etc). These tracks are meant to be played together in random order (on shuffle).
When I started thinking about this project, I was trying to consider what untapped benefits may be available from presenting songs in a digital format (and more specifically on a computer). I hear and read a lot of people talking about the degradation of sound quality caused by mp3 compression. I myself would much rather hear something on vinyl than from a computer. But surely there are formatting possibilities offered by the computer– just like vinyl provided two separate yet conjoined canvases (an obvious example is Joe Jackson’s Night and Day)– that are only starting to be tapped. Many artists (like Brian Eno, 77 Million Paintings) are utilizing computerized randomness to create pieces that will sound different nearly every time they are played. Most of us already have the tools to do this with our own music library (using the iTunes “shuffle” function). To play songs from a single album in random order used to be relatively difficult, having to move the arm of the record player or fast forward or rewind the tape to the right spot. Even the shuffle option offered on most CD players is less than ideal, as it takes at least a moment for the laser mechanism to adjust it’s position for the next track. With the right technology, this album should be shuffled with NO gap between tracks, creating one 33-minute piece. The number of possible randomized combinations is huge (approximately 9.332622e+155), which means that every time you play this piece, it will be essentially unique.
Where is all this going? Frequency Histograms for Predictor Variables has 99 tracks simply because that’s how many we can fit on an audio CD (and I did pick a specific track order for the CD edition). As more artists realize their freedom from the restraints of linearity and duration (thanks to new technology), I think there will be many albums or even single songs (broken into smaller parts) intended for shuffle playback or for user reordering. This approaches the ‘Composer’ stage of music economy as predicted by French theorist Jacque Attali in his seminal work Noise: The Political Economy of Music. I hope every listener feels a little more empowered by having their own version of this piece every time they listen.
-John Wood
• all songs written and performed by John Wood • released under a Creative Commons BY-NS-SA license • artwork by Matt Lipps :: mattlipps.com
market
