Nowadays we hear less and less of Computer Music, often confused with electronic music. But precisely, what is the difference?
The latter refers to the exclusive use of synthetic sounds, produced by electric oscillators, which generate sound effects, such as white noise. On the contrary the computer music, is a musical style based on the use of electronic processors. In the last period the way of composing, performing and reproducing music has totally changed. Therefore in this article we will analyze specifically, the history of Computer Music, with its major exponents, and the stages that have contributed to its change up to now.
WHAT IS COMPUTER MUSIC
Let's start by saying that computer music or computer music is a branch of electroacoustic music (contemporary genre, which uses sound recording and sound synthesis techniques), which uses computer programming. In a broader sense, any musical production that uses a computer medium can be defined as computer music. It was born around the 1963 with the publication in the famous American magazine “Science” of the article “The digital computer as a musical instrument” by Max Mathews. Music production is exclusively based on the use of computers, audio devices, audio / MIDI sequencers, plug-ins VST, AU, RTAS, software samplers and sample libraries. The style had as its major representative Iannis Xenakis, who deepened it at the turn of the fifties and sixties.
BUT PRECISELY, WHAT ARE ITS ORIGINS?
In 1945 the only machines capable of performing computational processes were too large and operated according to electromechanical principles. An important point was the equivalence demonstrated in 1938 by the Japanese Akira Nakajima, between the electrical switching circuits (on/off) and the boolean logic. Originally these computers were not "programmable", but the first computer was Z3 built by Konrad zuse in Berlin. The era of software programmed electronic computers began with the UNIVAC I (1951) and theIBM 701 (1952), systems based on vacuum tubes but already equivalent to general purpose computers. The IBM 701 is an example of the design architecture developed by the mathematician John von Neumann of the Institute for Advanced Study of Princeton University. In the 1954, Bell Labs built the transistor computer TRADITION, without the use of vacuum tubes.
THE BELL LABORATORIES
The decisive developments in the field of digital audio, or rather the numerical representation of sound and musical informatics, took place at Bell Labs in the second half of the 50s, a place that became the world-class scientific and technological research center. In these years it was thought to apply the methods then in use, extended to the coding and study of telephone communication signals. Clear reproduction of speech on a telephone requires a "sample rate" of several kHz, sufficient to cover the first two or three formants of the human speech spectrum. In the 1957, Guttman and Matthews programmed the IBM 704 tube mainframe to produce a digital signal corresponding to the notes of a short melody composed by Guttman (Silver Scale), for a total of 17 seconds of music. This was the first time that an audible signal was synthesized by an electronic computer connected to a digital-to-analog signal converter, applying Shannon's laws of information theory.
The silver scale
The program prepared in 1957 was called MUSIC I (Music One): it produced a triangular wave signal, the frequency of which could be determined. In 1958 it was the turn of MUSIC II (Music two), which ran on the IBM 709 and allowed the simultaneous synthesis of four sounds. In 1960, Mathews developed MUSIC III (Music three) on the "next generation" IBM 7090 computer, which served as a programming system following a set of standards that could be exchanged in a single numeric variable, such as:
1. instructions describing the sound synthesis algorithms
2. instructions that calculate and store wave functions and control functions
3. instructions for remembering the synthesis algorithm based on a list of times
To expand his software, Mathews felt the need to involve real musicians: in 1961, James Tenney began working at Bell Labs. The Canadians made what should be considered the first truly computer-generated music within months, and that was Analog #1 (Noise Study).
Analog # 1 (Noise Study)
In 1963 and 1964, MUSIC IV ended, introducing sufficient functionality not only for sound synthesis, but also for the development of compositional structures, with the possibility of automatically organizing musically significant data. It was the progenitor of a real "family" of sound synthesis and processing programming languages. Later there was MUSIC IV-BF running on the IBM 7094 in 1967, which was converted to analog and recorded on tape at Bell Labs.
On the other hand, MUSIC V it was originally written in Fortran (General Electric GE 645 mainframes). It has established itself as the definitive standard for digital sound synthesis in different times. In 1968, Jean-Claude Risset produced a series of synthetic examples based on MUSIC V and published them as "Introduction to Computer Synthesized Sound".
MUSIC V's opcodes and unit generator enabled the creation of software oscillators, filters, math operators, amplitude and frequency controls, and noise generators. In fact, it was an example of the development of other audio programming systems, such as CMUSIC extension, developed in C by Richard Moore in the Computer Audio Research Laboratory in 1980. Over time it has remained available on numerous computing platforms still in operation on microcomputers, such as the MUSIC V version of the LMA-CNRS of Marseilles and the CSC of the University of Padua.
Let us observe how the beginning of digital synthesis and the first attempts at digital recording occurred very early: between 1967 and 1969, the researchers of the Tokyo NHK (Japanese public service broadcasting and television) built a stereo digital audio recorder, which used videotape as a backup. Furthermore, in 1970 Thomas Stockham, a professor at the University of Utah, created a digital recorder also capable of editing the present signal.
In a later stage, the digital signal was then converted into an analog signal, where a digital-to-analog converter device was available (DAC). Thus the signal could finally reach the tape recorder or speakers to turn into sound.
To describe the first example of computer composition, it is useful to clarify the meaning of some general categories. From the very beginning there have been two directions of research:
• “Algorithmic composition”: IT formalization of the musical structure of a repertoire or of a new composition for instrumental or vocal execution;
• “Sound synthesis and digital processing”: computer creation and variation of sound signals, structured in musical forms.
ALGORITHMIC COMPOSITION
The invention of formal programs, the "algorithms" that determine the structure of a piece of music, has been a major focus of computer music since its inception. There are two general approaches: one closer to popular music theory and the other more related to musical experimentation.
During 1955, before Max Mathews began his research at Bell Labs, Hiller and Isaacson had begun writing a mainframe computer called ILLIAC (IIllinois Aautomatic Cocomputer). They modified the musical structure with statistical models, using random number generators, selection rules and random methods. In 1956, numerous musical compositions appeared, presented at concerts, primarily the first compositions actually obtained from a computer, or the first form of "algorithmic music".
Illiac suite
An algorithmic approach to composition was that of the celebrated cellist Pietro Grossi (1917-2002). His first approach to computer resources dates back to 1961, at the Department of Mathematics of the University of Florence, when he wrote a series of FORTRAN programs (combination rules) on a IBM 1620 mini computer. Based on data obtained from a computer, Grossi directed seven oscillators whose sounds were assembled at the S2FM studio in Florence. He then concentrated on the electronic transcription of pages of the European instrumental repertoire from 1967, when he worked at Olivetti in Pregnana Milanese using a computer General Electric GE115. The algorithmic compositional models of authors such as Iannis Xenakis e Gottfried Michael Koenig or the French Pierre Barbaud e Michael Philippot, were based on rules rather than specific pre-existing musical models. These composers already had pre-IT experience and were used to giving themselves precise, more or less rigid, generative procedures. Xenakis was already using “hand-calculated” probabilistic processes in orchestral works in the late 50s and early 60s. A relevant example will be Analogique A composed in 1959, for nine string instruments. In 1962, however, the composer used an IBM 7090 in Paris and ran programs he wrote in a Fortran language, based on the rules used for composition. A similar point of view embodies the proposal of Pierre Barbaud (1911-1990), who in turn used automatic processes to create instrumental music and founded the GMAP extension (Ggroup of muse aThe algorithmic of Paris) in 1959.
NUMERICAL SYNTHESIS OF SOUND
In the 60s, however, Tenney and Risset began working in the Bell Labs. In 1961, to get Analog # 1, Tenney determined a unique process of sound synthesis involving the modulation of the amplitude of random numbers controlled by a generator; the results were then edited to tape. Jean-Claude Risset developed an interest in a wider range of sound synthesis techniques, and was able to precisely define the structural spectral dynamics of sound. For Risset, the possibility of "composing the spectrum" and articulating the partials over time, creating a continuous compositional space between the harmonic musical structure and the sound spectrum, was decisive. Computer studies on the timbre of musical instruments were important for him: the aim was not imitation, but the understanding of the psychoacoustic criteria which underlie the human perception of timbre.
He composed two major works at Bell Laboratories: Mutationif Computer suite for Little Boy. The former was more convincing due to its simplicity; the second, on the other hand, was more articulated but limited.
He later had important collaborations with Paris: in the mid-70s he was involved in the creation of Pierre Boulez's IRCAM in Paris, together with Luciano Berio and Max Mathews) and in this context he produced some of his most important works, such as Inharmoniques and Songes, both created with MUSIC V. The latter, developed at IRCAM, was equipped with the possibility of acquiring digital recordings. In fact, in Songes, the study of timbre composition goes in two opposite directions: on the one hand, short instrumental recordings form a "melody of timbres"; on the other there are synthetic sounds that evoke bells with naturalism.
Inharmonic
Dreams
DIFFERENT DIRECTIONS OF COMPUTER MUSIC ALGORITHMIC COMPOSITION INTEGRATION AND SOUND SYNTHESIS
In the realization of Quartet in paris, the work of the American James Randall, computers were used both for encode continuous musical structures for both synthesize the sounds themselvesusing fixed waveform synthesis techniques. The connection was simple: timbre selection was controlled by serial processes that informed the structure of musical pitches.
In other authors, the two layers seem to intertwine even more, almost to the point of confusion, since sound and music are actually formed by the same operating standards. In Analog #1 for example, the collection of single sound events, and the macro-structures in which they are placed, are governed by the same rules, in different time scales. While for Risset and Chowning the computer elaborated new sonorities, for Brün sound was treated as the result of a combinatorial process applied to a sound synthesis process designed according to independent logics with respect to known acoustic or perceptive models.
With Koenig, in the years '70, the numerical synthesis was created through procedures SSP (Sound Ssynthesis Program) in a non-standard form. These digital synthesis techniques led to a very significant development, in fact they eliminated the distinction between the description of the sound, i.e. the synthesis algorithm, and the description of the musical structure, i.e. the algorithmic composition.
The last example of numerical synthesis was that of Iannis Xenakis. In 1972 the Greek-French composer founded the CEMAMu (center for mathematics, automation and music studies) with other university researchers. Xenakis applied to sound synthesis the same criteria that he used in "stochastic music" over the years; furthermore he saw synthesis as a microscopic combination governed by the laws of statistical change, with various degrees of order and disorder and he called his method of numerical synthesis of sounds "dynamic random synthesis".
Fifteen years later he returned to work for Gendy3 (1991), which in a sense represents his approach to information technology. To do this, in fact, he used a probabilistic procedure to synthesize the sound and to articulate each of the 11 parts of the composition.
Xenakis had a development in the computer field: in 1977 he completed the system Upic (interactive polyagogic unit of CEMAMu), which made it possible to transform straight and curved lines into sounds through a minicomputer composed of digital oscillators and graphic data: the trace could then be understood by the computer as slipped, envelopes, algorithms etc.
AMERICAN MUSIC COMPOSERS
The name of John Chowning it is often connected to the technique of digital sound synthesis by frequency modulation, had as tutor Leland Smith, who had introduced him to computer science and digital audio, and who in 1964 had sent him to study with Max Mathews. In 1967, he struggled with sparrow 10 (the software written by him) Chowning investigated the typical "vibratoOf certain singing styles: by mistake he happened to give the computer a vibrato frequency value greater than necessary (usually it is about 5/6 Hz), with rather surprising results; through an engineer friend, George Gucker, he understood that the phenomenon was similar to "frequency modulation"(FM). Chowning decided to deepen and learned to control the process, obtaining complex dynamic spectrum sounds with very limited computational effort (it was enough two sine oscillators and some control variable): it was a much less expensive procedure than additive synthesis and other techniques used in the experimentation and industry of electronic musical instruments of the time. It was this that motivated the further research on FM digital synthesis, and that explains why the first digital synthesizers produced on an industrial scale used this particular technique: it was possible to obtain timbrically complex sounds with few computing resources, even in real time. Chowning was the first to develop on the computer spatialization algorithms e dynamic sound distribution on multiple speakers, and implemented the first forms of digital simulation of reverberation phenomena. These research interests are all well present in Chowning's computer music works, at least starting with Turenas (synthetic sounds on quadraphonic magnetic tape, 1972). All the sounds of Turenas (anagram of the word "natures") were obtained with frequency modulation and were managed according to a dynamic spatial distribution on four speakers, with simulation of movement sources in space along trajectories drawn during the composition phase.
Dashow collaborated with MUSIC IV – BF in Whispers out of time (synthetic sound on tape, 1975-1976) e Side Effects (clarinet and synthetic sound on tape, 1976), considered the first Italian computer-generated music. Among the first hybrid systems developed in the late 60s (in which computers were often used as controls for real-time analog synthetic devices), we remember GROOVE (“Real-Time Operation Generation on Voltage Control Devices”) designed by Max Mathews and Richard Moore at bell laboratories.
The first and most complex real-time synthetic numerical computing system was designed by the physicist Giuseppe Di June (1937). The 4A processor, built by Di Giugno at the University of Naples in 1975, was essentially a set of digital oscillators in hardware that had to be controlled by an external computer. The device, capable of activating up to 256 oscillators simultaneously, aroused the interest of Luciano Berio, who was involved in the foundation of IRCAM in Paris. L'IRCAM (Institut de Recherche et Corder Acoustique /Musique) officially opened in 1977 after going through an organizational process.
It was conceived and strongly desired by the composer and conductor Pierre Boulez (1925-2016). Boulez saw it as a public institution intended for musical research (with a different meaning from the concept of musical experimentation). Disappointed by his first exposure to electronics in the 50s, Boulez had long since matured into a project with a highly skilled, interdisciplinary team that firmly combined artistic and technical-scientific skills. It was in this spirit that he established IRCAM, of which he was director until 1992. In the mid-70s, Di Giugno joined IRCAM at the invitation of Boulez and Berio. At the end of the '70s, Di Giugno designed a series of increasingly performing hardware systems. The study of American computer music of the 60s allows us to know the evolutionary path that music has had thanks to the contribution of new electronic instruments. The purpose of this article is precisely to understand the change in the way of composing and performing today, compared to the past. Furthermore, it highlights how significant the dialogue between technology and music is when the former provides the musical language with tools for the discovery of expressive horizons in line with new sensitivities.
Bibliography
Circuits of time. A historical-critical path in electroacoustic and computer music creativity Author: prof. Augustine of Scipio
Article made by the students of High School of Music GM Galanti of Campobasso as part of the PCTO ROUTE.
Cristofano Giulia
Petrillo Valeria
Roman Maria Michela
