The Radio Baton and Conductor Program, or: Pitch, the Most Important and Least Expressive Part of Music

Abstract

In the late 1950s, the first computer-synthesized music (Mathews 1961) required many minutes to compute a second of sound. Computation time was greater than Performance was not possible. In order to perform, a performer must be able to hear the sound he is making in real time as he is producing it. He must be able to modify the sound depending on what his ear tells him. The performer's interpretation and modifications of the notes specified in the score add much meaning and live quality to music. We call these changes performance nuance. The impossibility of performance and the absence of performance nuance in early computer music was much missed by some of us. Today, computers and digital synthesizers are fast enough to play powerful and complex music in real time. Performance is possible. One of the first performance systems was Groove (Mathews and Moore 1970). Groove was a hybrid system in which a digital computer controlled an analog synthesizer. Although Groove was limited by its synthesizer and by its large, unwieldy control computer, many of the ideas developed in the conductor program (Mathews 1989a) were first demonstrated there. Although expressive performance with synthesizers has progressed rapidly in the domain of popular music, it has not been nearly so successful in other musical areas. Part of the problem may simply be that a cadre of highly trained synthesizer performers does not yet exist, but in addition, the richness of control possibilities in the synthesizer may exceed the capacities of human performers. We now believe that programs to facilitate expressive performance can be written. One class of such programs, of which the conductor program is an example, allows the performer to concentrate all his attention on the expressive aspects of music and relieves him of data processing and motor control tasks which are very demanding with traditional instruments. The conductor program can make the many hours of practice most performers have to put in to learn technique unnecessary, but these new possibilities raise some fundamental and unanswered questions about performance training. Specifically, do performers learn other crucial musical skills and sensitivities while they practice to achieve adequate technique, and if so can we teach these skills in some other way? The conductor program and its associated device, the radio baton, have been described elsewhere (Lawson and Mathews 1977; Mathews 1989b), so in this article we will only briefly review their characteristics. Mainly we wish to consider the possibilities, problems, and questions raised by programs designed to aid performers in achieving expressive performances. Our most interesting insights concern the possibility of dividing traditional music into a predetermined part and an expressive part. This division is the basis for the conductor program. Since computer-aided performance is very new, and performers require years or decades to learn their art, most of the questions we will raise can only be answered by future developments. A great part of this article is speculation. Computer Music Journal, Vol. 15, No. 4, Winter 1991, ? 1991 Massachusetts Institute of Technology.