Vol. 29 Issue 2 Reviews
Workshop in Algorithmic Computer Music 2004

Music Center, University of California, Santa Cruz, USA; 25 June-9 July, 2004.

Andrew Marshall
Santa Cruz, California, USA

In an intensive, three week period, the Workshop in Algorithmic Computer Music (WACM) 2004 at the University of California, Santa Cruz (UCSC), delivered instruction on the essentials of algorithmic music composition. Under the direction of David Cope, lectures on theory, software, hardware, and performance were stacked back to back, covering an immense variety of information. Mr. Cope and his dedicated staff presented compelling demonstrations, which made the information immediately relevant. Participants formulated and constructed individual projects utilizing some of the various techniques covered. The workshop culminated in a full day of performances, and each of the 12 participants was sent home with a CD-ROM containing each project as well as the software used. I found the workshop to be an intellectually transformative experience. I was particularly impressed by the enthusiasm and diversity of backgrounds of the participants, which made for an environment of rich intellectual stimulus and provocative dialogue.

The workshop began with an intensive tutorial in Macintosh Common Lisp. The participants were led to view musical notes as lists of five open-ended parameters: ontime, pitch, duration, channel, and dynamic. Within days we were all using list manipulations to alter, simplify, complicate, and compose musical arrangements. Supplementing this focused tutorial were lectures on the fundamentals of algorithmic music composition: its history, philosophies, goals, and obstacles. We discussed the different hierarchies of choice in composition and viewed algorithmic composition as a subtractive process (i.e., the human composer sculpting away from the computer’s output some essential form). We looked at algorithmic analysis and composition as tools for overcoming a composer’s creative block. We delved into the origins of Experiments in Musical Intelligence, Mr. Cope’s own software composition project built from recombinative algorithms. We also examined some less than obvious examples of pre-computer algorithms in music.

Lectures and demonstrations filled the typical eight-hour weekday, while evenings allowed supervised lab time for the development of individual projects. Weekends featured guest presenters who covered a variety of material, mainly applications that had succeeded in commercial or performance arenas.

Alongside Mr. Cope, three staff members lectured and provided tireless hours of one-on-one instruction and guidance. Peter Elsea, Director of the Electronic Music Studios at UCSC, lectured on fuzzy logic, Max, and a variety of other means for creating different flavors of algorithmic music. Mr. Elsea has a lifetime of experience in experimental electronic music and was able to share with us a delightful selection of interactive musical toys, made using potentiometers on circuit boards connected via USB to laptop software. Additionally, Paul Nauert lectured extensively on set theory and Markov chains, and instructed us on the usage of OpenMusic, a sort of visual approach to computer algorithmic composition developed at the Institut de Recherche et Coordination Acoustique/Musique (IRCAM). Both Mr. Nauert and graduate student Soren Goodman lectured on their own works respectively, and played for us their compositions. As well, both were on hand to facilitate supervised lab hours from 5 p.m. to midnight each day.

Additional topics
Other topics covered during the workshop included sonification and visualization of data for both scientific and aesthetic purposes. We looked at nature as a source of rough algorithm, as in Olivier Messiaen’s birdsong imitations, and also found that cellular automata and Turing Machines provided us with rich systems to sonify. This study included a detailed look into Stephen Wolfram’s cellular automata research. Neural nets and genetic algorithms as ways of building musical intelligence rounded out this comprehensive lecture series.

Participants and projects
The general experiences of participants ranged through the fields of music, computer programming, and mathematics, to varying degrees. I was impressed with the geographical diversity, as some six countries were represented among the 12 participants. The projects were equally diverse. Some focused on analytic algorithms, such as one on classical analysis and one on the music of The Beatles. Another used algorithms to pinpoint a performer’s traits in jazz improvisation, as a step toward developing a virtual jazz accompanist. Plenty of projects focused on the task of pure algorithmic composition, using a variety of aforementioned techniques. I collaborated with one other student to create a duet, with one part algorithmically generated by computer for a live human cello performer, the other part hand-written for a MIDI-controlled Clavinova. All in all, the spirit was one of excitement, collaboration, and vigor.

WACM 2005
The opportunity to study under some of the world’s pioneers in the field of algorithmic music is a rarity. For those of you seriously (or tentatively) interested in algorithmic computer music, I strongly recommend this event. The tools and experience packed into these brief three weeks are enough to keep one going for years. If you decide to attend WACM 2005, plan on staying a few extra days to explore Santa Cruz’s beautiful beaches and redwood forests, for the workshop will keep you too busy to do so while in session!

Further information on WACM can be found at the workshop Web site: summer.ucsc.edu/wacm/.