|Vol. 42 Issue 2 Reviews||Reviews > Publications >|
Örjan Sandred: The Musical Fundamentals of Computer Assisted Composition
Reviewed by Ross Feller
The Musical Fundamentals of Computer Assisted Composition is a slim, practical book of value to composers, especially student composers, who might wish to formalize parts of their compositional process, using readily available computer concepts and technologies.
The book is essentially organized into two large sections. In the first part resides the bulk of the text: eight miniature chapters that introduce and present various ways to use the computer to assist the composition process. The second part contains images of the patches used to generate all the examples from the first part of the book. The patch examples were written using IRCAM’s OpenMusic, PatchWork GL, as well as additional examples using Lisp.
In the Introduction, Sandred lays out a strong argument for the longstanding importance of inventing musical systems to deal with compositional structures and using computers to help achieve compositional goals. He states: “the computer can give an instant feedback on how a structure fits with a musical concept, either by producing a music score or by giving aural feedback.” He goes on to say that composers can then adapt their music to better fit their intentions because of this immediate feedback.
Chapter one begins by mapping out a basic territory involving musical representation. Sandred demonstrates standard ways to numerically represent pitch and rhythm, largely relying on the MIDI protocol for illustration. The chapter ends with chronometric and rhythm tree examples, sufficiently demonstrating general principles for representing durations, metric stresses, and rhythms.
In the second chapter Sandred outlines some basic operations on pitches and durations. Beginning with simple transposition he goes on to extrapolate to multiplication and division of frequency spectra, inharmonicity, frequency modulation, and distorted spectra, illustrating these concepts with overtone and undertone examples from Gérard Grisey’s composition Modulation. Sandred also illustrates the concept of applying mathematical operations to fractional rhythm with examples from piano pieces by Olivier Messiaen.
Chapter three introduces the novice composer to modular arithmetic in music, through concepts such as pitch classes, intervals, attack points, and rhythmic patterns. In order to illustrate how one might go about using modular arithmetic for rhythm, Sandred provides a mini-tutorial for how Mario Lanza creates rhythmic patterns in his music.
In a mere four pages, chapter four covers how to model scales or other pitch collections. The reader is shown how to go about creating chords or other pitch collections by using pointers or index number locations. An excerpt from a work by Toru Takemitsu is shown to demonstrate how one might go about using pointers to select pitches in an actual composition.
Chapter five covers the use of iterative processes in arithmetic and geometric series. In each case the user begins with two seed values: an initial value and either the step size or a fixed ratio that determines subsequent values. Sandred provides examples of the Fibonacci series and the Golden Section, which is the go-to procedure for many composers’ efforts to create formal balance. The chapter continues with some useful, basic, information about iterative processes and chaotic orbits. Several charts are provided that show how small tweaks in seed values can produce non-linear, chaotic oscillations. Finally, at the end of the chapter, Per Nørgård’s intriguing Infinity Series is explained and shown how it might be used to generate a pitch sequence.
The next two chapters outline, respectively, global and local composition strategies. Chapter six briefly covers the use of profiles and stochastic methods to control musical parameters. Profiles are commonly used to generate and process melodic shapes. Iannis Xenakis’ piano composition Mists is used to illustrate how musical gestures might be generated from random walks that randomly change direction. Sandred goes on to form profiles from the extracted pitch collections, and how one might go about altering these profiles through linear interpolation. He also demonstrates how to combine separate profiles in order to produce a hybrid melodic sequence, similar to how one might utilize the isorhythmic technique but solely applied to pitch content. To illustrate stochastic processes Sandred explains how Xenakis applied a Poisson distribution to group orchestral colors in his orchestral composition Achorripsis. Markov chains are explained via a transition table that Maurice Sendak used in his children’s book Where the Wild Things Are, and Richard Pinkerton’s 1956 pioneering software program “The Banal Tunemaker” which does what you might expect given its title.
Chapter seven is devoted to the local control of pitches and rhythms using rule-based procedures and logical operators. Sandred shows how, by locking certain pitches to certain durations one can create “a logic that can clearly be perceived” even given an initial pitch collection that was selected randomly. As a way to illustrate some of the concepts in this chapter Lejaren Hiller’s Illiac Suite is used as an example of a simple rule-based composition. Sandred also outlines drawbacks of using the trial and error method when computing the musical score, and possible solutions to this problem in the form of constraint solvers that intelligently employ search algorithms. Additional examples are supplied from the work of Witold Lutoslawski, Jacopo Baboni Schilingi, as well as an example from the author’s own composition. Aesthetically, the examples in this chapter seem too pared down to be of much use to a novice composer, but might be useful as conceptual starting points.
Chapter eight String Rewriting Systems focuses on methods for rewriting given sequences according to a set of production rules. Sandred discusses formal grammar approaches and Lindenmayer system grammars.
Finally, there is an Appendix that contains images, at times drastically scaled-down in size, of the patches that produced all of the examples in the text. Although visually problematic including this Appendix was a good idea. Readers will be able to see exactly how the patches look and function.
After reading through the entire book one gets the sense that many of the algorithms or formal procedures, although largely conventional, are from the author’s personalized toolkit. This raises a question about the extensibility of the examples, which at times reflect the author’s idiosyncratic procedures. It is worth noting that the author is primarily a composer and teacher of composition.
There is also the question of the book’s overreliance on standard forms of notation, namely pitch and rhythm. Occasionally the author makes mention of other factors such as timbre or texture but the bulk of the book is devoted to manipulating pitches and rhythms. Given the vast processing capacities of the computer, this traditional approach seems rather passé. There are numerous composers who use OpenMusic or PWGL for much more than pitch and rhythm manipulation.
The book’s formatting includes different typefaces for the primary text (a serif font) and the example text (a sans-serif font slightly larger in size). This works well, especially helpful when attempting to locate an explication from a specific example. The Appendix images are less effective. Many are simply too small to be readable. For example, the many examples that include large chords with a lot of pitches look more like a blot of black ink rather than a distinct chordal structure. To make them readable some of the images would have to be double their current size and spill over onto additional pages. Another solution would have been to increase the overall size of the book, which seems too small at roughly half a standard A4 size page.
Another problem with the Appendix section is that it assumes a high level of familiarity with either OpenMusic or PWGL, two programs that are used largely in Europe but are under utilized in the U.S.A. The required knowledge of OpenMusic or PWGL contrasts markedly with the lack of specialist knowledge in the rest of the text. In the Introduction Sandred states: “There are many books on Computer Assisted Composition that explain advanced concepts and techniques. Most of the available literature assumes however that the reader has a basic knowledge of the fundamentals.” Sandred’s book attempts to fill this gap by providing information available to novices. Unfortunately, in order to understand the Appendix examples, one is expected to already have attained a high level of understanding of two specialist programs. Perhaps it would have made more sense to translate the examples into Max, since it has become the lingua franca of computer assisted software programs. Sandred does provide a link to the patches found in the book, and suggests that both OpenMusic and PWGL are readily available to would-be users. The problem is that after reading a basic outline of various computer assisted compositional devices the novice composer, or composer unfamiliar with OpenMusic, will not be prepared for the learning curve required in order to achieve advanced levels of knowledge with this program. Thus, The Musical Fundamentals of Computer Assisted Composition, as a whole, can be viewed as an appendix to OpenMusic or PWGL manuals.