There is enough potential in anything to last a lifetime.
This paper will describe the processes utilized to realize a recent composition. The primary focus will be upon the creation of formal elements through score synthesis techniques, although an overall view is necessary in order to explain them, therefore delineation of much of the process is required.
With the proliferation of sound design software and hardware in the early years of the 21st century, there is sometimes a tendency for composers to put emphasis on sound design and neglect or completely ignore considerations of form engendered by a formal score. This is indeed disturbing because these composers are abandoning one of the primary means at their disposal to make musical sense of these new sounds in favor of a quick easy layering or looping of sounds. Although there are exceptions to the contrary, generally sounds do not make the composition, instead it is the composition that makes the sounds viable, as music, through formal elements. Without the use of some form of syntax, which is created by establishing score relationships, there is no meaning and so, no semantic element.
Algorithmic score synthesis is in essence a way for a composer, using indirect methods, to elicit musical combinations that would not be possible through more direct methods. Today, this means that a computer can be programmed to assist in the creation of musical possibilities through, among other methods, the use of mathematical tendency masks. This is by no means a way for the composer to remove himself from the process of composition, nor from the responsibility of making musical decisions. In fact, it places the composer into an ideal position to make these creative choices because it is not necessary to be as absorbed by the more direct tasks that are normally required in non-algorithmic compositional approaches. The use of mathematical tendency masks is a perfect choice for algorithmic composition of music conceived in the digital realm. Since digital sound is by its very nature based on numeric values, these tendency masks integrate into the very foundation of the sound itself.
Csound, a programming language intended for musical composition, is an ideal starting point for algorithmic composition because it allows the composer to work with sound directly on a numeric level. At first this idea may seem foreign to a person who has studied and composed music in more “traditional” settings. However, consider the fact that, for centuries, music has been divided into mathematical quantities; for instance note durations, in relation to time (a quarter note at 60 bpm, in 4/4 time, is one second long) and divisions of frequency, in terms of scalar relationships (concert A vibrates at 440hz, A an octave higher vibrates at 880hz), thus establishing a musical language. A numerically unrestricted language, such as Csound, greatly expands the possibilities for precision control in these as well as countless other areas of musical creation and expression, and provides seemingly limitless new possibilities for numerical relationships, impossible to realize by traditional musical languages and instruments.
Cmask is a stochastic event generator for Csound scores. It provides a versatile array of algorithms that can be used to create tendency masks that in turn can be used to determine various pfieldvalues in these scores. Quasi-random values are generated based on tendency masks that can be stipulated in several ways, perhaps the most interesting of which is by the use of linear and exponential envelopes.
I am extremely interested in designing score synthesis techniques to utilize in my compositional process. Through an association with Otto Laske, a pioneer in the field of algorithmic composition, I have been deeply influenced by Koenig, Brun, Xenakis and Stockhausen, as well as Laske himself, all of whom have spent lifetimes developing, among other things, score synthesis techniques. These techniques need to be recognized for what they offer and therefore extended and passed on to subsequent generations of composers. This is the impetus for the documentation of the process utilized in the creation of Azimuth, a fifty five minute composition consisting of four movements.
To understand the title of the composition, which is, in part, directly related to its formal elements, an explanation of the correlation is required. Webster defines azimuth as the “horizontal direction expressed as the angular distance between the direction of a fixed point (as the observer’s heading) and the direction of the object.” Those who have worked with tape machines will remember that the azimuth has to do with aligning the tape path in relation to the tape heads. The composition of Azimuth was begun with a very specific sub-score, to be explained later, as a starting point, which evolved for the creation of each of the various sections. This evolution occurs in a linear fashion in that the origin of each score can be traced back to the single sub-score that was used to generate the first sample. This method enables the generation of relational yet time varying score data, creating a sense of overall formal structure as well as a general thematic direction.
<![if !supportFootnotes]> <![endif]> Many terms in this paper, such as pfield, assume a basic working knowledge of Csound.
Go To Page: