Archives - 1977
Archives Main | 1977
Electronic Music
October/November 1977
ROBERT A. MOOG
INTRODUCTION:
What is electronic music? Is it a brand new kind of music, or is it rooted firmly in our musical traditions? Must it use only sounds that are electronically generated, or are electronically processed "natural" sounds also permissable? Must electronic music be consigned to tape's magnetic domains, or may the performer be allowed to play directly to an audience? And how about the sounds themselves? May traditional timbres be simulated? Must electronic music sound "electronic"?
The above questions are sufficient to convulse the argumentative among us into paroxysms of noisy, futile debate. Electronic music has existed as a potential artistic resource for a century. As a subject of widespread experimentation since the end of the World War II, and as a major factor in the music business for the past decade. Throughout the development of electronic music, there have been uncountable attempts to categorize and restrict its musical domain, yet today it is nigh impossible to identify an area of music that has not been enriched through the use of electronics. Synthesizer recordings have shed new, clarifying light on the substance of traditional music, while opening our ears to new timbres and textures. The vast expanse between pure sound effects and pure music has been explored with electronic instruments in opera, film, TV, and records. Students from kindergarten through graduate school are gaining new insight into the acoustical and physiological aspects of music as well as the aesthetic aspects. And, of course, composers and performers are producing music that would be impossible with only traditional acoustic instruments. In fact, with each succeeding year, it becomes more and more difficult to objectively consider electronic music as a distinct, bounded entity.
My own definition of electronic music is therefore very broad: Electronic Music is music which is made with electronic equipment. This particular definition simply serves to point out that electronic music is a medium of musical expression which, as it happens, is miscible in any proportion with the traditional pallette of acoustic musical instruments. Although it has spawned an awareness of tone color as a compositional element, the electronic music medium has been employed to interpret the song and dance of vernacular music and the tonal counterpoint of baroque as successfully as it has been used to develop and realize the most aggressively avant-garde composition.
In focussing on the musical resources rather than the technological features of the electronic music medium, we identify four distinct areas:
1) Electronic circuits vibrate (produce audio signals in ways that are fundamentally different from those of acoustic instrument components. This is because electronic circuits are at most always discrete systems that are described by ordinary differential equations, whereas acoustic components are distributed systems, and are therefore described by partial differential equations.
2) It is often useful to separate the sections of an electronic audio signal producing system so as to allow the musician to assemble his is own combination of elements, and to adjust each element according to his particular
needs. In other words, with electronics it is practical to supply a musician with a "kit of parts" out of which he assembles his own instrument. This capability does not exist with acoustic instruments.
3) The means for electronically generating and modifying audio material may be kept separate and independent from the control interfaces which the musician "plays" or through which lie otherwise issues commands, in order to produce the audio signal changes that constitute music. This affords musical instrument designers the freedom to optimize some portions of an instrument system for best audio signal production, while optimizing other portions for the most efficient communication of the musician's commands. Except for very few cases such as the pipe organ, acoustic musical instruments have vibrating elements which also serve as control elements. An example is the violin string which vibrates while the musician' s finger is actually on it.
4) Electronic technology includes the capability of storing sequences of sound events and, more recently, the capability of storing sequences of command (control) signals that define sound changes. This single capability is perhaps the most profound contribution that technology has ever made to music. Not only has it freed the musicians from the necessity of making his music in real time (live), but it has enabled him to physically manipulate and arrange sounds by handling a recording medium such as magnetic tape.
In examining specific uses of the electronic music medium, the most interesting question to ask is: "To what extent are those features of the medium that differentiate it from acoustic music being utilized? It is simply not necessary or desirable to argue the purists position that only certain types of music are "right" for realization through electronic music, that there are fundamental musical limitations to the medium, that live music is better than recorded music, or that the musician does not have the right to "cherry pick" as he strolls through the orchards of technology in search of musical resources.
Neither is it worthwhile to even consider the question, "Is electronic music less natural than music played on acoustic instruments'?" No music is natural! Music is produced only after people invest strenuous and extended effort to gain intimate control over vibrating systems such as vocal cords, a violin, or a synthesizer. Furthermore, all musical instruments, except for the human voice, are highly controlled by technological artifices. They are differentiated not by their degree of "naturalness," but by the technological periods in which they were developed. The string instruments were perfected when woodworking was a flowering technology. Piano designers utilized the processes of a fully industrialized society. And the electronic music medium is being developed now, a time when electronic technology is dominant and the golden age of manufacturing appears to be yielding to what people are calling "the postindustrial era." What people really mean when they accuse electronic music of sound being unnatural, mechanical, or, worst of all inhuman, is that it sounds just plain different from what they are used to, or that the musician, in exploring the new medium, has not yet succeeded in gaining the control necessary to properly realize his music.
This discussion of electronic music deals primarily with the relationships between the functions of electronics devices, and the types of resources that musicians have found useful. It will steer clear of making subjective evaluations or discussing aesthetic aspects of specific pieces of music. First, however, it will sketch a brief history of electronic musical instruments to illustrate that precious little is new under the sun.
ELECTRONIC MUSICAL INSTRUMENTS PRIOR TO 1945
Much of the information in this section is taken from Dr. Thomas L. Rhea's Ph.D. thesis [I]. Through the years, the technologists of the electronic music medium have tended to be frustrated musicians and chronic putterers. Thaddeus Cahill, developer of the Telharmonium, was also endowed with the ability to think big. Although not strictly electronic (it predated the invention of the vacuum tube by about a decade), the Teleharmonium embodied many basic principles that have been used in the electronic music medium: the generation of pitched tones from alternating electricity, the addition of harmonics to determine tone color (additive synthesis), and a touch sensitive keyboard to shape the sound and control their strengths. Cahill generated his signal with a bank of over
a hundred alternators, each capable of producing as much as 15 killowatts of power. The tones were combined in giant mixing transformers, and then sent out over leased telephone wires to subscribers who heard the music on telephone receivers fitted with horns. The musicians performed at the touch-sensitive keyboard console, while listening to them playing over "monitor speakers" of the same type that subscribers had. Cahill's Telharmnonium was built in Holyoke, Massachusetts and shipped to New York in 1906 in some thirty railroad box cars. This first polyphonic , touch sensitive music synthesizer remained in service in New York for only a few years, until the forces of electronic technology (the invention of radio) and crass economics conspired to end telharmnonic music as a viable commercial venture. The basic idea was resurrected again in the 1930s in an instrument that was somewhat more of a commercial success: the Hammond Organ. Instead of monstrous generators, the Hammond Organ tiny tone wheels to generate individual pitches and employed electronic amplifications to boost the tone wheels to drive loudspeaker. Designed and tooled by a group of master machinists (Hammond manufactured electric clock s before it got into the music business), the Hammond organ offered musicians a few potent musical resources plus that rarest attribute of all: reliability. Hammond remained the largest manufacturer of electronic organs for decades, a tribute to the enduring value of a well designed musical resource.
The Theremin was one the very first wholly electronic musical instruments. Developed during the 1920s by a Russian physicist and amateur musician, Lev Sergeivitch Terman, the Theremin is played without being touched. The performer moves his hands in the space surrounding the instrument to vary pitch and loudness of the tone. Terman licensed RCA to make and sell his instruments in the United States. RCA built Theremins on the same production line as their first superhet receivers, and attempted to sell them through the same dealer network of radio stores. From this venture, RCA learned that introducing a new musical instrument into the consumer marketplace requires a combination of delicacy and tenacity. A merchandising approach involving classical promotion and distribution is sure death. This lesson is being relearned even today by musical instrument manufacturers, large and small. After a brief time in the musical instrument market place during which only a few hundred Theremins were sold, RCA discontinued production and sold the rights back to Terman, who continued developing the instrument as an experimental venture.
In Europe, the same pattern emerged with different instruments and inventors. Friedrich Trautwein developed the Trautonium, an electronic musical instrument controlled by a fingerboard-like rheostat. The rights were sold to none other than Telefunken, which, it is reported, made a grand total of fifty instruments before bowing out. Maurice Martenot, a French instrument designer, developed the Ondes Martenot, a monophonic (single tone) electronic musical instrument whose control means included a six-octave keyboard, continuously variable pitch band, and touch sensitive articulator bar. Unlike Terman or Trautwein, Martenot not only kept production of his instrument under his own control, but he also established a school for developing and teaching performancen techniques. Original compositions for Martenot are routinely performed in Europe, even today.
All of these instruments, Theremin, Trautonium and Martenot pointed the wav to experimentation with new control devices for electronic musical instruments. The Solovox, and similar instruments that were less successfully marketed, were some direct precursors of the whole class of instruments called synthesizers. The Solovox contained a single tone generator: and a variety of waveshaping frequency dividing, filtering and envelop shaping circuits. Each circuit could be switched in by the musician. In other words, the musician synthesized, or assembled out of aural components, the sound quality he desired. Manufactred by Hammond, the Solovox was designed to be mounted on a piano right in front of the keyboard. By depressing a specified combination of switches, the player could set up an orchestral-like timbre, which he could then play monophonically with one hand while he played the piano with the other.
The use of electronic technology to assemble, store, and manipulate sound sequences may be traced back to the Coupleaux-Givelet synthesizer which was introduced in 1929 at the Paris exhibition. The Coupleaux-Givelet system used punched paper tape to pneumatically activate control that determined the parameters of four independent voices. Programmed mixing and level control was accomplished by varying the coupling between two coils, by moving one of them with a bellows. The Hanert Electrical Orchestra, developed by John Hanert around 1945, enabled the musician to draw a "score" (program) for a piece of electronic music on a continuous roll of paper. A carriage containing and elaborate photocell array travelled down the paper, "reading" the score and "conducting" the ensemble of sound producing circuitry. After World War II, the RCA Electronic Sound Synthesizer was developed under the direction of Dr. Harry F. Olson. Programmed by a paper roll punched with a set of binary codes, the RCA Synthesizer produced four independent lines of sound. Each line was programmed with a degree of detail that allowed the sound parameters to be redefined as often as thirty times a second.
ENTER THE TAPE RECORDER
The development of the Theremin, Trautonium, and the many dither performance instruments early in the electronic music medium followed in the tradition of acoustic musical instruments. Musicians mastered these instruments simply by developing suitable manual dexterity and applied these skills to real-time performances of a limited (or even unique) set of tone colors and response modes. The instruments of Coupleaux-Givelet, Harnet, and RCA, on the other hand, sought to bypass the performing musician by going directly from composer's score to finished music. The availability of the tape recorder immediately after World War II added a new dimension to electronic music. With tape recording, any sounds, electronic or acoustic, brief on extended, ordinary noises or pure pitches, could be manipulated, assembled, fragmented, combined and stored. The work of experimental musicians in Cologne, Paris, and New York focussed on developing technical procedures and aesthetic rules for tape composition. Music using electronically generated sounds came to be known as just plain "electronic music," music using recorded " natural" sounds came to be known as "musique concrete," and the facilities that were established to support tape composition came to be called "classical studios."
A typical classical studio consists of three main sections: sound generation and modifying, mixing and patching, and recording. A basic feature of the classical studio that differentiates it from both the early electronic performance instruments and the early programmed synthesizers is the separation of the studio components from one another. In a classical studio, all signal paths are established with patch cords. The musician organizes his studio to provide the desired function or operation. Tape composers working in classical studios soon learned that there is a conceptual continuity between physically organizing sounds on tape on the one hand, and assembling a network of electronic components to produce or modify sounds, on the other hand. It became possible to regard the medium itself, as well as the pallette of sounds of which it is capable, as a carrier of musical content. Music need not be defined only by fixed patterns that imply underlying harmonic structures. With patch cords and enough things to patch together, music can also imply elegantly-proportioned networks of signal flow. Thus, in addition to greatly, increasing timbral resources available to musicians, the classical studios added a whole new dimension to musical communication.
THE VOLTAGE CONTROLLED SYNTHESIZER
As first conceived, the voltage-controlled synthesizer is not a uniquely defined piece of equipment, but rather an approach to systems design in which a) each system component performs a single signal generating, modifying, or control function, B) a large number of signal path networks is possible, and c) the principle of voltage control is utilized to expand the means of sound control. The combination of these properties results in an open-ended, generalized musical instrument system that can easily be used (and indeed has been used ) to produce virtually any type of music. Each of these features will now be described in detail, and their musical value will be discussed.
As in the classical studio, the signal generating, modifying, and controlling of a voltage-controlled synthesizer are performed by single function components which are connected into a network by the musician. Audio signals originate in oscillators that generate regularly repeating waveforms, in noise sources that produce random waveforms, or in devices external to the synthesizer. Regularly repeating waveforms give rise to pitched tones. The shape of the wave itself determines its harmonic content and therefore its beginning tone color. The repetition rate of the waveform determines their perceived pitch of the resultant tone. Random waveforms give rise to pitchless, hissing-type sounds.
Audio signal modifier components consist of three main groups: filters, amplifiers, and mixers. Filters attenuate some portions of the signal's spectrum (set of frequencies) while allowing other portions to pass. Amplifiers (or attenuators), merely change the amplitude of the signal, while mixers combine signals in a predetermined proportion.
These are the basic audio signal generating and modifying processes of electronic music. In a classical studio, these processes are static or quasi-static. That is, the functional components are set up to produce a sound or group of sounds. During the actual production of the audio signals, the operating parameters of the components remain constant, or at most are slowly varied by manual control. In a voltage-controlled synthesizer, the main operating parameters of the modules (functional components) may also be varied by applying control signals. These control signals are entirely separate from the audio signal path. They do not correspond directly to sound vibrations, but to changes in the properties of the produced sound. As an example, consider voltage-controlled oscillator, the output of which is an audio pitched tone. Now, a second oscillator is brought into the picture. Assume the second oscillator produces a 5 Hz sine wave. If the second oscillator signal were mixed with the output of the first, no sound change would be heard. If, on the other hand, the second oscillator signal were fed to the frequency control input of the first oscillator, then the frequency of the first oscillator would rise and fall with each cycle of the second oscillator's output. The first oscillator's tone would have periodic pitch variation or vibrato. Using synthesizer terminology, we would call the first oscillator the tone or audio oscillator, and the second oscillator, the control or modulating oscillator. It is important to note that these two oscillators are not distinguished because they are being used in fundamentally different ways. A wide range voltage controlled oscillator may be used for either audio or control signal generation. Furthermore, a control oscillator can itself be the receptor of a second control signal. A hierarchy of control signals thus enables extremely complex sound textures to be generated by a single audio source.
QUANTITATIVE ASPECTS OF VOLTAGE CONTROL
Acoustic parameters are generally proportional to the exponentials of musical (subjective) values. In particular, musical pitch intervals are frequency ratios. Stated another way, frequency is exponentially related to pitch. Likewise, audio signal amplitude is exponentially related to subjective loudness.
Voltage-controlled oscillators (VCOs) whose frequencies are exponentially related to the magnitudes of their applied control voltages are ideal for electronic music applications. Since the introduction of the early voltage-controlled synthesizers of Buchla and Moog in 1964, this type of oscillator has become a standard of the electronic music medium. The usual control signal increase of one volt doubles the frequency, or raises the pitch by one octave. The equally tempered chromatic scale is generated simply by increasing the frequency control voltage in 1/12 volt steps; pitch transposition is accomplished by adding multiples of 1/12 volt to the control signal that produces the pitch pattern. Furthermore, timing relationships are just as simply manipulated when the voltage controlled oscillator is operated in the subsonic region. If an oscillator is programmed to produce a control signal applied to it will double the speed of the pattern, a 7/12 volt increase will give very close to a 3:2 speed increase, and a 5/12 volt increase will produce a nearly 4:3 speed increase.
It is desirable to design voltage-controlled filters (VCFs) so that the relationship between the filters' characteristic frequencies and the voltages applied to their control inputs are the same as the exponential oscillator frequency output/control voltage input relationships described above. If, for example, the center frequency of a voltage-controlled bandpass filter increases on octave for each one-volt increase in control input, then that filter will impart the same pitch patterns to "noisy" (randomly pitched) audio material as an oscillator would produce when controlled by a given control signal sequence. When voltage controlled oscillator and voltage-controlled filter are driven by the same control sequence, then the filter tracks the oscillator, thus establishing a harmonic structure that is independent of pitch.
Voltage-controlled amplifiers with exponential gain/control voltage input relationships are also useful for music production. In a typical VCA, a one
-volt increase in control signal input quadruples the amplifier's gain, or increases it by 12 dB. A negative-going control voltage ramp causes the amplifier's gain to decay exponentially, thus imparting a percussive amplitude envelope to a constant-amplitude audio signal being fed through the VCA.A positive going control ramp results in an exponential increase in gain. Adding a fixed voltage to a given VCA control signal pattern shifts the loudness level without affecting the shape of the amplitude/time contour: a negative going ramp to which one volt gas been added imparts a percussion envelope to an audio tone that is simply 12dB louder than that imparted by the the ramp itself.
Accurate, wide range voltage control was first made practical by the availability of silicon diodes and transistors. In fact, the volt-ampere characteristic of a typical forward-biased silicon junction is accurately exponential over five decades of current. The first exponentially voltage-controlled synthesizer circuits, developed in 1964, used conventional silicon transistors and diodes to compute the exponential control characteristic. Those early VCOs were typically accurate to 1% over a 10:1 frequency range, and the exponential scale factor was proportional to the absolute temperature of the circuit components. Th most accurate currently-available VCOs use temperature-regulated silicon junctions and precision integrated circuits to achieve an accuracy of 1% over a 1000"1 frequency range, with a scale factor that is independent of normal variation in ambient temperature. For a musician, this improvement is of crucial importance. Tuning for tonal music was always difficult with the early VCOs, while the newest circuitry is suitable for microtonal as well was diatonic music.
The synthesizer system
Most present-say voltage controlled modular synthesizers incorporate the same basic components: wide-range multiwaveform VCOs for audio pitch and periodic control signal generation, wide range VCAs for imparting amplitude/time contours to audio as well as control signals, and one or more type of VCFs for imparting dynamic spectrum variations to audio material. The most popular type of VCF is the 24 dB/octave lowpass/resonant filter; the sound variations it produces have become identified with "pop" electronic music more than any of the single device. Other audio signal generating and modifying modules include fixed frequency filter arrays, reverb units, mixers, and noise generators. Control signal generators include envelope (transient) generators, audio signal envelope extractors, and sequential pattern generators. Manual controllers (performer interfaces) include keyboards, touch sensitive plates, fingerboard-like ribbon controllers, and pitch-to-voltage converters. Most modular systems utilize conventional patch cords to establish signal paths; other systems use pin or switch matrices. In all cases, the synthesizer system is completely general. Any logical interconnection network (as well as a large collection of illogical but interesting patches) contains the potential for making music. The musician is free to think of music making either as the construction of a process, or as the production of a series of well-defined sounds
The commercial aspect of electronic music
The listening public first became aware of the electronic music medium subliminally, through radio and TV commercials. Eric Siday, Raymond Scott, and other pioneers of the sound logo, explored electronic sounds in widely-heard commercials during the 1950s and 1960s, well before electronics infiltrated pop music through the rock and roll idiom. The early commercial musicians were generally aware of the activity in tape composition. They were the first to use modular synthesizers for commercial music. One of these musicians, Walter Carlos, studied at the Columbia-princeton Electronic Music Center before setting up his own multitrack tape studio in a corner of his living room. Working part time in his studio over the 2-year period from 1966 to 1968, Carlos developed his technique and, in collaboration with Benjamin Folkman, produced the music for Switched-on Bach, the largest selling classical LP album of all time.
The success of Switched-on Bach and the trend toward heavy use of electronics in pop music spurred a furious activity in electronic music record production during the late 1960s. A few of these were successful. In response to requests by musicians for small synthesizers for live performance, several manufacturers introduced relatively small-scale prepatched instruments during the early 1970s.English keyboardist Keith Emerson pioneered the use of synthesizers in rock music by including a programmed modular synthesizer in his act. At the present time, music production techniques that are tracable to the classical electronic music studio are routinely employed by musicians in studios and on stage. Several successful performance groups play on nothing but synthesizers and related electronic hardware. Musical experiments of the 50s and 60s are now earning points toward cliché status. And synthesizers are the fastest-growing segment of musical instrument industry.
The digital computer
The use of large digital computes to synthesize audio signals predates the introduction of the voltage-controlled synthesizer by nearly a decade. In terms of hardware operation, digital computers make audio signals in entirely different ways than analog synthesizers. Waveforms are computed one point at a time. Each point is a high accuracy number that specifies the magnitude of a narrow time slice of the desired waveform. Instruction from the musician are fed to the computer through punched cards or other conventional data entry means. The computer generally takes longer than the actual duration of the sounds to compute all the waveforms.
In terms of software, however, sound synthesis programs such as Music IV and Music C are closely related to the operation of a voltage-controlled synthesizer. These programs present the musician with a collection of functional elements that correspond closely to synthesizer modules. The musician assembles the elements into "instruments," and the instruments into an "orchestra," by conventional computer programming rather than by patching and knob-turning. In effect, the digital computer models an analog synthesizer, which in turn models a simple vibrating system.
Direct digital synthesis provides an extremely powerful experimental resource. However, the operation is non interactive. In practice, a musician spends several hours to program a short segment of music, then waits several hours or days more until the computation and conversion are complete.
In order to have a real time interactive audio signal generating system with the advantages of digital data processing. Max Mathews and his colleagues at Bell Labs designed and built the Groove system. This system uses voltage-controlled analog modules for the audio signal chains, and a small digital computer to generate and process the slow moving control and timing signals in real time. The musician has a variety of control interfaces at his disposal: keyboards, joystick, rotary knobs, and a CRT readout that graphically displays as many as fourteen control signal contours. A disc file stores the programs for complete pieces of music. A system of this type goes far beyond the composing and editing capabilities of conventional voltage-controlled synthesizers. It even has the potential for bypassing the multitrack recorder in the composition process.
The state of the synthesizer art
The digital synthesizer recently designed and constructed by Hal Alles of Bell Labs is representative of the current state of the art in electronic music hardware. Alles' instrument includes a) a digital audio signal synthesizer, b) a variety of manual controllers, c) a small controlling computer, d) a data entry terminal, and e) two floppy disc drives for data storage. The digital audio synthesizer in effect contains 64 high accuracy oscillators, 32 filters, 32 amplifiers, 256 envelope generators, and similar numbers of supporting function. The manual controllers include two five-octave touch sensitive keyboards, 72 slide levers, and four 3-axis joysticks. The data entry terminal contains a full alphanumeric keyboard and CRT display. And the on-board computer is capable of coordinating the control of the entire instrument in real time. The instrument is self-contained (except for power source) and occupies some 12 cubic feet of space. It is capable of producing and storing music of a level of complexity approaching that of a small orchestra.
What about the future?
Developments in the electronic music medium come about when musical needs and emerging technology are combined with a touch of imagination and foresight. Imagination often involves the utilization of ideas that were previously discarded. In future decades, the influence of personal computing will be the major technical force shaping electronic music. Already digitally-programmable synthesizer components are being sold to computer hobbyists. Focus will shift from sound production to sound control. Even in present day commercial synthesizers, the audio circuitry is eclipsed in size and complexity by the control circuitry. As the complexity of control means increases even further, the information content of the sounds themselves, as well as the music into which they are assembled, will also increase.
I do not believe that brain wave control will become a significant factor in electronic music in the next decade or two, despite the appeal that the idea has for many musicians. Human hands happen to be extremely efficient manipulators. A virtuoso pianist transmits as much as several hundred bits per second of a pitch, timing, and dynamics information to his instrument. Electronic instrument control interfaces designed to fully exploit the capabilities of a musician's hands will undoubtedly supplant the simple keyboards and controls of today's synthesizers. An example is an experimental touch-sensitive keyboard currently being developed at the Indiana University School of Music. Each key on this keyboard is touch-sensitive in three dimensions. Accusations that the requirement of lengthy practice on such a keyboard enslaves the musician are countered by assertions that musical content increases in proportion to the amount of control that a musician exercises.
Finally, the exotic, technological mystique surrounding the electronic music medium will wither away as the whole electronic involvement continues to infiltrate our daily lives. As more and more musicians explore electronic music with open minds and ears, our preoccupation with the electronic music medium itself will give way to an increased awareness of the musicians and their music.
Bibliography
[1] Thomas L Rhea, "The Evolution of Electronic Musical Instruments in the United States" PhD Dissertation George Peabody College, Nashville, Tennessee, (1972).
[2] Herbert A. Deutsch, Synthesis, Alfred Publishing Co., New York (1976)
[3] Elliott Schwartz, Electronic Music, A Listeners GUIDE, Praeger Publishers, New York (1973).
[4] Max V. Mathews, The Technology of Computer Music, MIT Press, Cambridge, Mass. (1969).
[5] Hubert S. Howe, Jr.Electronic Music Synthesis, W.W. Norton, New York (1975).
