two minutes retunedfor tape (~42')
UA: IEM Graz, Open CUBE, 2006-01-17
The title of the work refers to an aspect of the compositional method. As a side-effect of the processing of a recording of two minutes length, pitch changes occur.
Except for final mixing, I produced the work within the sound synthesis language SuperCollider – a programming language that, in addition to mathematical and logical operations, includes sound objects that one can combine in a very concise syntax. Talking about algorithms in the context of music composition, we should be aware that this is a generic term that can mean synthesis methods or methods for producing control data (e.g., just for these synthesis methods). Within SuperCollider, both are possible, the definition of sound-producing objects ("synths") and their call, respectively modification by sequences of time and control data.
Here I experimented on double synthesis methods. Using a "primary" synthesizer (a kind of simulated plucked string instrument), I generated a sequence of equidistant events (virtually constant quavers). Subsequently, this recording was processed using granular synthesis. I developed the "primary" synthesizer in parallel, concerning the results of the granular synthesis. At one point, I regarded the recording just as given and continued experimenting only on granular synthesis. Now different sections were produced with similar algorithmic control setups to achieve formal connections.
A great variety of control setups in SuperCollider comes from the possibility to combine continuous and discontinuous controls. Here, for example, the perceived "tempo" mainly depends on the velocity of the granular cloud's moving within the buffer and on the rate of reading it out – constant quavers were just selected to restrict the possibilities of rhythmical arrangement to the granular synthesizer. In these and other parameters of the granular instrument, continuous changes (e.g., oscillations) but also jumps are possible. So the space for experiments and aesthetical decisions strongly depends on the preliminary decisions on data types and their tendency towards continuous or discontinuous control.
With friendly support of GesFEMA and: