The theremin allows for the transformation of hand movements in space into melody. Certain techniques for the transformation of plane drawings into sound patterns resemble the technology used by the theremin. Thirty years ago, R. Saifullin and I invented and built an electro-mechanical device based on these techniques. Its main component was a large, flat rheostat switched into the frequency (w) control circuit of an electronic sound generator (Fig. 1). Vertical displacement of an electric contact with the rheostat produced a change of the pitch of the sound (Lg(w)), while horizontal displacement (whether to the left or right) produced no change in pitch (w = const, here meaning the value of w at a particular position of rheostat contact along the vertical). Simple as it was, this device nevertheless revealed an interesting correlation between aural and visual perception: when we drew curvilinear shapes (circles, spirals, etc.) on the plane, listeners were able to describe the shapes formed by the lines without visually observing the drawing process. Our experimentation with the device contributed to our research in the field of aural-visual synaesthesia [1]. Years later, I have now come to understand how the principles on which this device was based can also be applied to musical practice by means of computer technology [2].
 |
Fig. 1. The original Galeyev/Saifullin device for sound
production by means of drawing, as photographed in a television
studio in 1961. The operator (R. Saifullin, in background) slides the electric contact along the flat rheostat. (The person in the foreground is the television camera man.) |
Collaboration with two colleagues experienced in computer programming (A. Shumilov and D. Piotrovsky) has brought me back to these ideas after a lapse of many years [3, 4]. We have created a software called Computer Brusophon that represents a "sound field" as a ruled drawing field on a computer display monitor, with horizontal lines demarcating equally tempered scale divisions (Fig. 2). The uppermost line in the drawing corresponds to the highest frequency, while the lowest line corresponds to the lowest frequency of the sound. Each octave is divided into 12 parts. The software allows for the user to scan a preset drawing in order to produce a composition (Fig. 2). Alternatively, one can also draw with the mouse; changes in the curves of the lines thus produced are simultaneously transformed into changes in pitch. Thus, one can both transform preset drawings into sound and also create new sound patterns in real time while drawing.
 |
Fig. 2. Screen shot of Computer Brusophon, software for sound generation by means of drawing, 1992. The contours of the graphic pattern displayed on the computer monitor are scanned by a sound-triggering cursor (the white circles on the lower right are traces left by the cusor as it scans over the line). The buttons on the upper and left peripheries of the screen correspond to functions such as switching the sound on and off, activating the mouse, changing the speed at which the preset drawing is scanned, choosing the scale of the drawing and rotating the drawing. Notes corresponding to the lines in the drawing field are designated on keys to the immediate left of the field |
Our experiments with drawing using this program have not yet produced patterns of sounds that could be called musical melodies. With a view toward producing melodies, we have configured sound fields of particular diatonic (flat, sharp) and pentatonic scales. This configuration consists of excluding, by means of programming techniques, any "excess" notes - i.e. those not used in a particular scale - from the 12 tones of an octave. The condensed range of tones is graphically represented as a kind of vertically oriented keyboard appearing to the left of the drawing field (for example, Fig. 2 shows the sound field of the pentatonic scale that consists of the tones produced by the black keys on the traditional piano keyboard). The sound structures obtained by drawing on such configured sound fields come closer to resembling melodies (especially - in the case of the configuration shown in Fig. 2 - the melodies of traditional Tatar music, which feature relatively smooth transitions between notes in a pentatonic scale) [5]. In the case of European music (or, more precisely, its basic diatonic modes) additional intervals must be introduced into the condensed keyboard in accordance with specific rules in order to produce sound patterns resembling melodies. My colleagues A. Sutchkov and M. Belyalov and I are now busy solving this problem, in addition to conducting experiments related to sound-duration control (rhythm), periodical accenting of sound (meter) and development of the software to produce harmonic accompaniment as well as melody. For the time being, we are limiting ourselves to European music of the seventeenth and eighteenth centuries in order to avoid the problem of musical-language formalization, which is not an easy one to solve due to the impossibility of reducing musical languages to simple uniformity. We plan to use the computer to test an old hypothesis of mine regarding possible correlations between the melodic and ornamental structures of the arts of a given culture, in addition to continuing to develop the current software's potential as a means for creating and performing music by drawing. We have been encouraged by the fact that similar experiments are now being carried out independently by other musicians and researchers (in Moscow, for example). This has compelled us to continue our own research and experimentation and to begin to compile an international collection of articles on the subject for publication under the title "Melody-Drawing Transformation."
Acknowledgments
This research is being carried out under the auspices of the Academy of Sciences of Tatarstan and is supported by the Russian Humanitarian Foundation.
References