Calculated Music - A Letter From the Founder
This project started in 1988 at the Sibelius Academy in Finland. I had studied composition at the Sibelius Academy with professor Enojuhani Rautavaara since the Autumn of 1986. During my studies with him I created, among other compositions, StringQ for string quartet and Reflex for orchestra, where I used algorithmic methods derived from serialism and set theory.
In 1988 I continued my studies with Paavo Heininen at the Sibelius Academy, concentrating on algorithmic composition. I examined and experimented with various mathematical and technical methods which have been used in computer programs for musical compositions, e.g. probability and random processes, fractional noise, Markov chains, neural networks, chaos theory, and methods formulated by researchers and composers such as Max Matthew, Milton Babbit, and Iannis Xenakis.
While a lot of music has been produced by way of these methods, I found them problematic for the following reasons; they tend to be derived from mathematics and science, and applying them directly to musical composition may result in music that is more appealing scientifically or theoretically than musically interesting. These methods can yield enormous amount of data but do not suggest how to convert the data into music. As these methods are not constructed to tackle musical problems they are unsatisfactory when it comes to tasks such as evaluating colour or tension in harmony, melodic resemblance, or musical progression.
For these reasons I began the construction of CALMUS (Calculated Music), a computer program for musical composition where I attempted to create functions that more directly deal with musical problems, both compositional and technical. The program has been constructed and developed through compositional exercises and actual compositions. Each piece has brought up new technical and musical problems that have pushed forward the development of the program.
Founder & Composer