Motherboard

Imagine musical trees made of vinyl, year-rings functioning as grooves of audio information. Record labels would go out into the black woods and harvest tall black stalks of acetate, hauling these trees back—CD-R leaves and vines of magnetic tape dragging in the dirt—to shave off cross sections and punch holes in the middle. The first shaving is a test pressing. It’s given a listen and judged. Some are beautiful, and it’s obvious at first listen that nature has grown important, lasting music. They’re then sent to a mill, where the trees are sliced to specifications, packaged, and shipped to record stores.

OK, now flip the situation; instead of changing the trees, change the record player. This is what Bartholomäus Traubeck did with his “Years” project: he made a record player that plays cross-sections of trees, analysing their year rings and reproducing/interpreting the information as gorgeous piano music. The idea is simple and poetic; the process, not so much. I talked to Bart earlier this week.

Do different kinds of trees make different music?

Up to a certain extent they do. Since a lot is predefined by the rules of my [computer] program and the technical setup, there is only space for the interpretation of my machine by the tree’s structure. It is more like a machine to output generative music on the base of a tree’s year rings than an actual sonification of the matter itself as titled by many blogs I saw my work in.

But in terms of piano-music: yes every tree produces a different composition. Sometimes it’s more obvious , sometime it’s not. The effect can be examined best by comparing different types of trees, for example fir tree sounds a lot more minimalistic and has a very abstract rhythm compared to an ash tree which is more full sounding and rhythmic and loud.

Can you explain the process a bit more? What’s the analytic mechanism, and how is it mapped to a musical scale?

OK, unfortunately this is not that easy. I should’ve built a simpler machine I guess. The tonearm is equipped with a modified [PlayStation] Eye camera which streams an almost microscopical closeup image of the wooden disc to the computer. There the image is analyzed in a node basing programming platform called vvvv. I examine the image for (obviously) year-rings and if one is detected, it is analyzed for its thickness, darkness and growth factor.

These parameters determine the rhythm, the tone strength, tone length as well as the pitch and also of how many single tones an ‘event’ is composed of. Since there is no groove to drag a needle all the way to the center, there is a small stepper motor mounted under the tonearm. It is driven by an Arduino microcontroller which is also in charge of handling all the other ‘physical’ inputs like switching the thing on and off.

As the tonearm makes its way to the center, the year-rings usually grow closer together and the signal is more differenciated.The signal itself, in order to produce an output is, after being analyzed and reshaped, mapped onto a scale of piano keys. This scale is defined by the overall appeareance of the wood. It can happen that some wood plays in a wholetone scale, and another plays in a regular C-minor or C-major. These tones are then sent to an ordinary digital piano software via MIDI.

Here’s also the excerpt from my website:

A tree’s year rings are analysed for their strength, thickness and rate of growth. This data serves as basis for a generative process that outputs piano music. It is mapped to a scale which is again defined by the overall appearance of the wood (ranging from dark to light and from strong texture to light texture). The foundation for the music is certainly found in the defined ruleset of programming and hardware setup, but the data acquired from every tree interprets this ruleset very differently.

How much of the music do you consider to be in effect your own composition?

The word composition in relation to me might be a little irritating because the composition and the interpretation is the only part of the music that comes almost entirely from the wood’s texture.

I rather define the ruleset and its boundaries in which a possible composition can be acquired from the wood. My subjective choices in this are plenty, like the choice of using a piano, the choice of how to get the data, and in which resolution to process it. But the actual pitches loudnesses, rhythms, and so on are mostly defined by a) the wood itself, and b) the technical setup and its own native rules of processing data.

Apart from that it’s hard to say, because its a question of how to look at it. It’s like if you would build a special piano with, say 24 predefined keys all in C-minor and a big cage over the keyboard and drop a couple of ferrets in that cage, whose music would it be, the ferrets’ or your’s since you thought the whole process up and built it? But anyways: yes, there is a lot of my very subjective decisions involved in the process to generate this sound.

What do you personally feel is being revealed by this process?

That there are many ways to experience a database. And that everything is a database in some way. I think you can not gain much insight in ‘the life of a tree’ by listening to this, apart from maybe a very romantic or poetic perspective which I have my problems with personally, because those notions are very pathos-laden. The fact that data can be experienced by looking at its overall structure rather than its content is fascinating to me. Wood was the perfect material to demonstrate this, because the structure of the data is what actually makes the material in this case, and that you can see it with your bare eye.

What would this thing do with an actual LP?

Not much I guess, there would be no signal because the machine can’t isolate the small grooves of a record. The microscope doesn’t go that far in since year-rings are much bigger. So it will only see a black plane which isn’t enough data to output sound for my machine. You could try to put a large pancake on it though (or something else that is round and has a finely drawn texture and lots of dark/bright contrasts on it’s outside). That would maybe result in some rather chaotic sound. Basically everything visual with dark/bright contrasts can be read, but the computer is programmed and the setup is optimized to to read year-rings

Can you pick another material you’d like to translate like this? And what instrument would you like to map it to?

Satellite-coordinates and Waterphone (ocean harp, it makes for the eerie sounds in many old horror movies). Best instrument ever:

But seriously I think for my next work I want to use something more direct with less layers of algorhitms and programmed abstraction.

Reach this writer at michaelb@motherboard.tv.