object as transcription

detail shot
detail shot
front (zoomed)
full frontal
"myspace" angle
"myspace" angle
side view
side view
top view
top view

I've been thinking about the different functions of notation, such as transcription or transmission, and the different forms that these things can take, such as schematics or so-called 'concrete' scores (recordings), and trying to put some strain on the meanings to come up with some murkiness. After seeing the beautiful 'ur-scores' of Doug Wadle-- and initially misinterpreting them to be of mixed media rather than paint, due to the fact that I was looking at pictures-- I decided to investigate the possibility of using a 3-dimensional object as a kind of notation. Precedents include (but are not exhausted by) Cage's "Rocks Role (after Riyoanji)", the aforementioned Wadle, Xenakis' work with architecture, and the use of electronic circuits themselves -- not their schematics-- as scores by the likes of Tudor, De Marinis, etc. How can the ambiguity inherent in using an object as notation be overcome and used as a strength?

After making my portfolio recording of 'Sanction of the Victim', a composition for 2 or more networked computers, I was looking at the waveform up close and found this really cool shape that lasts ~0.09 seconds. It turns out, this artifact happens a few times in the recording, and it is the result of all the tones in a pitch constellation flock being struck at once, and the filterbank momentarily blowing up. Because of the flocking algorithm used to determine rate and alignment between the tones in the constellations this only happens once per computer. It occurs at the moment when the system on a particular computer starts up. I grabbed the brief snippet (only 4096 samples long) and performed an analysis of it, breaking it into 12 layers using haar wavelets. I first cut the shapes of these layers out in foamcore, discarding the top 4 layers because they were way too complex to cut out at that scale. As it turned out, the top two layers were still nearly impossible to do in foam with an xacto knife. I sent the vectors, along with some 1/2" thick medium density fiberboard, to AMS for lasercutting. I was dismayed to find that 1/2" MDF proved too difficult for their laser to cut through without starting a fire, and they sent me home with a charred piece of synthetic wood. I came back with 1/4" thick masonite, which they lasercut without incident. The resulting layers I fixed together with woodglue so that the lowest values lined up and the piece could stand upright. All this was painstaking work, but a nicely different pace from directing vocal pieces for humans or programming computers to simulate them. Also, I do enjoy the feeling of carrying around raw materials in Manhattan. For some reason, there is a level of devotion to a composition that is felt when one's arms ache from schlepping as the result of it, which is different from the devotion one feels to a massively complex piece of software. Well... sometimes. 'Sanction of the Victim' is both a massive piece of software and four computers, so my arms and brain felt the devotion to that one. The weight and dimensionality of this first masonite form (there will be three more) certainly work for me as components inspiring devotedness.

I do not know the proper way to realize this score. I have suggestions, though. One could trace or rub the score onto paper, superimposing staves perhaps. It could be a Rainforest object. It could be a Cartridge Music object. It could be that you're a computer science enthusiast and you make a table of the wavelet coefficients (with applied scale) of the values implied by the score and resynthesize the original recording (at a lossy compression, due to the missing top 4 layers).

2 thoughts on “object as transcription

Leave a Reply

Your email address will not be published. Required fields are marked *