1. Organised Sound
I am delighted to contribute to the twentieth anniversary issue of Organised Sound.
As far back as in the 1920s, Edgard Varèse decided to call his music ‘son organisé’ (‘organized sound’Footnote 1) as an answer to those who objected ‘is it music?’. The expression was meaningful: it supported the emphasis that Varèse put on the sound material, the vocabulary of music, at a time when most composers were primarily preoccupied with the rules of assembling notes reflecting the grammar of music. In this age of electricity and computation - with the sound compositions of Varèse, Cage, Grainger, Scelsi, La Monte Young and Chowning; the appearance of musique concrète, electronic music, computer music - this dimension of musical practice grew considerably during the twentieth century, which fully justifies the dedication of a periodical to organised sound.
Before the journal existed, Cambridge University Press asked me about the justification for such an publication, and I answered in a most positive way. I tried to convince the Press that the organisation of sound was a crucial issue in music, a burning subject evolving rapidly with technology: most music heard today comes from loudspeakers, and the digital domain is becoming more and more pervasive. Yet electronic and computer music, despite their increasing importance, tend to be ignored by musicologists: analysts find it difficult to work without making reference to a document as undeniable as an instrumental score; also the realisation of this kind of music often calls for sophisticated technical methods which they have not mastered.
Thus there are not enough texts commenting on electroacoustic music, which hinders its diffusion and its insertion into the musical patrimony. Of course, this view is not new: it was realised long ago, and institutions promoting various forms of organised sound had been putting out their own publications since the 1950s, especially GRM in Paris and the Cologne Electronic Music Studio, as well as IRCAM and CCRMA since 1974. However, each institution was intent on promoting its own approach: this did not favour discussions, even though certain publications were open to different points of view.
I wish to mention early publications which concerned diverse aspects of organised sound. I have listed a few periodicals and collective publications at the end of my article.
The celebrated conductor Hermann Scherchen had a personal interest in the applications of new technologies to the creation of music. From 1955 to 1966, he animated and financed about thirty issues of the review Gravesaner Blätter (Gravesano Review), with significant articles illustrating diverse innovations of experimental music, all published in both German and English.
Between 1970 and 2011, Françoise Barrière and Christian Clozier directed the Groupe de Musique Expérimentale de Bourges – GMEB – which became Institut de Musique Expérimentale de Bourges (IMEB) in 1995. The directors were quite intent on developing their centre as a forum for electroacoustic music, diffusing many new works created in Bourges and elsewhere and organising meetings between composers. GMEB published several issues of the journal Faire. Starting in 1995, IMEB organised yearly meetings of an Academy gathering experienced composers of electroacoustic music coming from diverse countries and illustrating different aesthetics. Each year, a theme concerning electroacoustic music was selected, and each member of the Academy was asked to write a text expressing his or her own views about this theme. The points of view were stated and discussed, then each member could edit his or her initial text. The final texts were then published in both English and in French. Thus nine volumes came out on various problematics of electroacoustic music: aesthetics; composition and diffusion; time, composition and technology; the diversity of aesthetic relations between composing and listening; and the relation between hearing and vision.
I participated in both ventures, and I found them quite rewarding. I thought that a journal devoted to the musical organisation of sound could clearly incite a wider group of composers, performers and researchers to pursue and broaden discussions on significant issues such as the ones mentioned above. This indeed happened with Organised Sound.
Of course, the Computer Music Journal had existed since 1977, an invaluable source of information, and the editors had established high standards. But the new journal envisioned embracing a somewhat wider scope, going beyond the digital domain. Also it would be published in Europe, the scene of early and crucial developments of organised sound.
In the United Kingdom, a number of composers and researchers were active in electronic and computer music. I had followed the important work of Denis Smalley, Trevor Wishart, Simon Emmerson, Jonty Harrison, Stanley Haynes, Andrew Bentley, Martin Atkins, Rajmil Fischman, Richard Dobson, Adrian Moore, and others, especially the late Jonathan Harvey and Richard Orton. I had been impressed by the birth and growth in the 1980s of the Composers’ Desktop Project, an exemplary international network based in the United Kingdom and helping those working with computers on organising sounds. From 1979, the Electroacoustic Music Association of Great Britain (renamed Sonic Arts Network ten years later) was actively coordinating initiatives. In the 1990s I had participated in a meeting on Science and Music organised at the University of Cambridge by James Woodhouse. I was impressed by the quality of the contributions, among which was a remarkable review on the acoustics of the violin by Woodhouse and Donald McIntyre, and also a large-scale study accomplished in the United Kingdom under the direction of John Sloboda, which demonstrated that achieving a professional level in instrumental performance requires long and careful practice, even for very gifted performers. It seemed clear to me that there was an opportunity to start an international journal of music technology in a local context different from that of the Computer Music Journal.
My optimistic arguments may have helped the decisions of the publishers to go ahead with Organised Sound, but I was not quite sure myself that the journal could indeed survive and grow as it did. I want to pay tribute to the talented editors and contributors, and especially to the imaginative actions of Leigh Landy as an editor. Landy’s enthusiasm is contagious, and he organised the issues in a sound way, selecting themes that inspired the community.
Indeed, many issues that concern music technology and contemporary creation have been discussed in Organised Sound since 1995. I am personally not sufficiently aware of recent trends and accomplishments to write a looking back/looking forward article that would review the main trends documented in the journal and guess future topics. My contribution to this anniversary issue will mostly mention personal opinions, concerns and desires for further progress.
2. The Digital Era has Barely Begun
Some musicians and musicologists consider the digital era as a mere technological upgrading of the electronics era, an upgrading that did not bring substantial innovations. I strongly disagree with that view. Certainly electroacoustic sound had been very influential on instrumental music: it had suggested new ways to deal with musical time, and new uses of polyphony to ensure timbral and textural effects rather than harmonic functions. But I contend that the use of the computer for elaborating sound has deeper and more lasting potential.
Max Mathews, who left us in 2011 at the age of 85, in 1957 started a most exciting venture, harnessing the computer for making sound and music. The demands of organising sound had far-reaching and economically rewarding consequences. One had to wait thirty years for the first commercial computer with a high-quality sound output – the NeXT Cube. Digital technology has since then become pervasive in the sonic domain. Fender-Rhodes electric pianos and idiosyncratic analogue sound synthesisers can be emulated today with the Csound program (which developed from the Musicn series).
Sound synthesis as a mean of composition has provoked a paradigm shift. Programming makes it possible to apply compositional processes down to the level of the sound microstructure – to compose the sound itself; it gives the keys to a sensitive constructivism of sound. The art of sound colour shifts from tapestry – a choice of fixed colours – to painting – making one’s own colour. Monet painted impressions; Cézanne wrote ‘when colour is at his richess, form is at its plenitude’. With synthesis, the composers can make the listener attend to time in sound as well as distributing sounds in time. They can long to evoke phantasmatic worlds, as Borges suggests, or to set up special relations to memory and time, as in Proust’s literary work.
In France, Hugues Dufourt coined the term spectral music for a genre he developed together with Gérard Grisey, Tristan Murail and others from the 1970s onwards. Dufourt has acknowledged the debt of spectral music to the early explorations of digital sound synthesis. Looking at the main evolutions as discussed in the issues of Organised Sound, one can see that most recent developments imply elaborate logical and computational resources.
For instance, the field of sonification has considerably expanded in the digital domain. Sonification means representation in terms of sound: it implies structuring sounds by external informations or models unrelated to acoustic vibrations. Digital synthesis permits us to modulate at will any sonic parameter by frequency-limited functions: the mapping – the nature of the correspondence between the non-sonic information and the sound parameters – becomes critical to the auditory result. Sonification has practical applications. It is used for instance to alert stockholders when exceptional events happen at the Stock Exchange. Representations are often useful to unravel complex structures. This is well known for visual representations, but it is also true for auditory displays. Perception is well equipped to detect regularities and patterns. It performs heuristic operations that are often hard to formalise or simulate, and these operations are different for vision and for hearing. In particular hearing is highly sensitive to variations, especially frequency and spectral variations, and it can in certain conditions either fuse components or analyse them and sort simultaneous evolutions. Beyond its ancillary function, sonification can inspire new compositional models and processes of sound control susceptible of novel musical applications.
The computer is not simply another tool or a new instrument. According to François Bayle, the computer is a workshop. It allows us to design and implement an infinite variety of tools, intellectual as well as material. As Jean-Baptiste Barrière insisted, what is important is not the hardware, but the software, the know-how, the intellectual investment.
The modular Musicn software for sound synthesis and its latest version Csound supported a new way of thinking about music. It provided the means for composing sounds beyond the traditional graphic notation of notes and durations, in ways that can unfurl with the help of specific programs. Similar software resources effectively bridge gaps: between acoustic and synthetic sound material; between real-time control and delayed synthesis; between synthesis and processing of sounds; and between music composition, sound production and performance.
The computer allows the implemention and testing of a huge variety of sonic processes, compositional rules and formalisms. Olivier Messiaen reminded us that formalisms can play a revolutionary role by forcing the composer to envision musical proposals which he or she could never have imagined or made explicit. One should not mistake the limits of a given program or application for limitations attributed to the computer: one could go beyond these limits with a new program. It is too much to ask of the composer to build his or her own tools. But the composer should be aware that he or she can choose his or her own constraints. Imagination is no longer bounded by material limits.
As stated by the late Jean-Marie Souriau, physicist of relativity and cosmology, we are still in the infancy of the digital age, and a lot remains to be investigated and worked upon in order to improve our synergy with computers of all kinds. Music has been in the forefront of this exploration and it should go on, for its own sake and for the sake of euphony in social organisation.
I shall now evoke some possibilities for taking advantage of sound and music computing that seem to me worth being pursued.
3. Marrying Electronic Music and Musique Concrète
In my 1984 piece Sud, I tried to marry electronic music and musique concrète – not merely by mixing them, but by performing genuine hybridisations to take advantage of the strength of both processes. Using analysis-synthesis and cross-synthesis, I endeavoured to impart synthetic sounds with energy fluxes extracted from acoustic recordings and to imprint precisely defined pitch scales onto acoustic sounds with no clear pitch or with specific pitch structures.
Granular synthesis favours a merging of synthesis and sound processing – the digital version of electronic music and musique concrète – by both creating new sounds and processing existing sampled sounds. There exists a host of analysis-synthesis programs. Yet it seems to me that the compositional use of these possibilities could be pushed further, to enhance both the variety and richness of the acoustic material and the precision and compositional consistency of digital computations.
4. Developing Musical Transformations from Extended Structural Notations
Traditional musical notation was initially used for mnemonic reasons: but it clearly suggested specific compositional transformations, especially the intervallic inversions and the melodic retrogradations used in counterpoint. Such transformations are obvious on a score but unused in forms of music of oral tradition. These transformations correspond to the group of symmetries in the plane. The digital representations of sound parameters provide elaborate structural representations of the musical material: programming can suggest powerful musical transformations along the continuum between microstructure and macrostructure. Such possibilities relate to the general notion of group transformations, which appeared in mathematics and physics in the nineteenth century and which has played an increasing role ever since, in particular in quantum mechanics. For instance, in the Standard Model of particle physics, Glashow has invoked the special unitary group SU(2) to link the weak interaction (responsible for ß decay) to the electromagnetic interaction. The structural representations also relate to field theories, useful for physics in both the microscopic and the macroscopic domains (e.g. quantum electrodynamics and electromagnetism).
Early examples had been given by John Chowning in Turenas, with its continuous trajectories in both illusory physical space and in timbral space; in Stria, with its unprecedented relation between the inharmonic structure of sounds and the pitch scales; in Phoné, where differentiated modulations of coinciding vocal-like spectral lines make figures emerge from an indistinct sound mass. In my work, Mutations, I staged a gradual transition between discontinuous and continuous pitch scales.
I believe there is lot of potential here that has been only marginally explored. Dealing with transformations can help imagine original directions. The composer could select musical transformations depending on his or her personal appreciation of their significance to the compositional project. In their multidimensional software Esquisses, developed at IRCAM in the late 1970s, David Wessel and Bennett Smith had already afforded possibilities for the composer to accommodate material to his or her own perception: the composer could scale his or her own evaluation of the musical relations within a selected set of sound material.
I shall give a more specific example of compositional use of digital representations, the use of functions.
5. A Plea for a Broader use of Functions
Early on, I was impressed by the power of the modular programs designed by Max Mathews to synthesise musical sounds. Among the features that I thought had great potential, I enjoyed the possibilities of defining and combining functions in an algebraic fashion, and of using them at leisure at different scales for different purposes.
One of the first syntheses I tried with Music4 in 1964 was a simple but effective imitation of bird songs realised with a simple instrument: a sine wave modulated in frequency by a complex graphically defined function of time. That same year, I worked on trying to imitate brass tones with the computer. Varèse approved of this. He insisted on the need to inject life and identity into synthetic sounds. I initially mimicked the envelopes of each analysed brass tones by simplified functions specifying the evolution of the different harmonics of the tone: this worked, but it required a lot of work for each tone. I tried a simpler recipe: louder trumpet sounds have brighter spectra, so I devised a Music4 instrument to implement this feature by rule. I made the amplitude of each component a predefined function of the amplitude of the fundamental (i.e. the first harmonic), choosing steeper functions for higher-order harmonics so as to implement the rule the louder, the brighter (an increase of amplitude should bring a stronger proportion of high-order harmonics, even during the attack). Such pre-determined functions can be selected to induce various specific spectral behaviours: one can thus mimic brassy sounds, but also the sound of bowed strings as characterised by Max Mathews. Our studies suggested that variability thoughout the sound was a key to liveliness, and also that a relationship between different sonic parameters might be a key to timbral identity.
John Chowning developed his FM technique for the synthesis of complex spectra by a change of scale: he extended frequency modulation rates fm from sub-audio – as in a vibrato – to audio, incorporating negative frequencies. If A is the amplitude, fc the carrier frequency and I the modulation index, the resulting wave is given by the formula y(t)=Asin(2πfct+Isin(2πfmt)). In order to predict the resulting spectra, the frequency-modulated sine wave can be expressed in terms of trigonometric functions whose coefficients are Bessel functions Jn of the first kind. The amplitude behaviour of each spectral component is a function of I (including the reflected contribution of the negative frequencies), which results in rich spectral variations.
After Max Mathews wrote the modular synthesis program Music3 (1959), he realised several pieces combining synthesis with ingenious compositional algorithms, taking advantage of functions in various ways. In 1963, starting with two melodies, he exchanged their rhythmical patterns. When John Cage heard the result, he suggested the title Masquerades: one theme masqueraded under the rhythm of the other. In 1966, Larry Rosler and Max Mathews designed a graphic system; for each musical voice, the amplitude and frequency variations were specified by functions drawn graphically with a light pen. This influenced the design of Xenakis’ Upic, a system designed in the 1970s to specify the music graphically by drawing envelopes, frequency evolutions and wave shapes. Using the graphic system, Mathews drew functions to control the gradual transformation of one melody into another, for instance in Mathews and Fujimura’s International Lullaby – moving gradually from a Japanese lullaby to a lullaby by Schubert and back. In particular, Mathews made a clever use of piecewise linear functions to generate intriguing transitions between two rhythmic patterns.
In 1966–67, the change of computer generation was catastrophic. In order to increase computing efficiency, most computer programs had to be written in low-level languages – so-called machine languages or assembly languages – closely linked to the hardware structure of the computer. The new generation of computers had drastically different hardware, so that the previous low-level languages were no longer usable. Thus most programs had to be rewritten from scratch, which needed a lot of tedious and uninspiring work. From then on, easing portability became an important concern for the survival of software. Max Mathews designed the Music5 program with this goal in mind: he wrote it in Fortran, a high-level language implemented on most new computers, and he isolated the sample-computing loops where the computer was spending most of its time from the rest of the program, so that these loops could be easily rewritten in machine language for different computers to save computing time. I myself was very intent on facilitating the use of different functions: I performed a minute change to Music5 by introducing the SET unit generator enabling the user to change in every note statement the function used in an oscillator. I extensively used this possibility of exchanging functions for exploring various syntheses, some of which are described in my 1969 Introductory Catalog of Computer-Synthesized Sounds.
In the real-time hybrid system Groove, implemented in 1969 with F. Richard Moore, Max Mathews had included in his design an algebra of functions that could be defined and combined in various ways. For instance, the successive pitches of a voice, as imposed by the musical score, could be specified by defining a function with successive horizontal segments with heights proportional to the successive frequencies. Then one could modify this step-wise curve by multiplying it by a sinusoidal curve of frequency 6 to 7 Hz with an amplitude controlled by a knob. Thus vibratos could be added with a width controlled at will in real time by turning the knob. The vibrato amplitude could then either be memorised as a fixed value or left time-dependent so as to vary continuously from tone to tone and even within a tone. A session could be devoted to continuously controlling just the vibrato for this voice, then the evolution of the vibrato could be recorded as a function: it could either be preserved for later use, or discarded and replaced by another curve defined in another session, or combined with some other function. Thus performance nuances could be specified at leisure in different sessions, and when they were finalised, the piece could be ‘conducted’ - that is, one could beat the rhythm, controlling the temporal flux of the music as a conductor. Such an algebra of functions was extremely convenient, and it seemed to me worth mentioning: it could be useful in many musical circumstances.
In the 1970s, Charles Dodge and Michel Decoust illustrated novel uses of functions. In his work, Earth’s Magnetic Field, Dodge plotted the magnetic field value as a function of time, then he quantised that curve according to his own diatonic choice and he used the quantised curve to control the frequency variations of synthetic tones. For Interphone, Decoust recorded the soprano Irène Jarsky reading a poem, then he used curves extracted from the recording to control other parameters – for instance the amplitude curve could serve to define a melodic contour. Such processes are early examples of sonification.
In Inharmonique, I used the same curve to modulate both amplitude and frequency, with an added frequency quantisation. I also used a related process: a set of evolving inharmonic tones was defined by specifying functions controlling the non-synchronous envelopes of the frequency components. These envelopes had the same shape but different durations. Then changing the common envelope shape produced intimate changes in the resulting texture while preserving the harmony: a sharp attack followed by a resonant decay made the components fuse into bell-like or gong-like sounds, heard as distinct sound objects; by changing it to a gradual attack and decay, one turned these objects into fluid textures revealing the ‘interior’ of the sound, since the different frequencies reached their maximal amplitude at different times: the components were dispersed in time just as a prism disperses white light into the rainbow of colours.
In several of my works (Moments newtoniens, Songes, L’autre face, Sud, Elementa), I applied an unusual process to make up a complex frequency envelope. I resorted to Fourier synthesis, usually used to make up sound waves by adding harmonics, to draw complex and supple curves which I then used to control the frequency of a sine wave. I could produce curves with strong oscillations due to the Gibbs phenomenon. With appropriate choices of harmonics, I could get complex ondulations including low and high frequencies in various proportions; by choosing the phase relations, I obtained both sharp peaks or, in contrast, moderate amplitude variations. I could combine these undulations with piecewise linear functions to control complex frequency evolutions. The resulting effect often evoked bird songs, which generally include fast and complex frequency modulations of a sine wave. At the end of Songes, I combined such supple frequency curves with Chowning spatialisation techniques. As Roger Reynolds remarked, illusions of spatial motion were reinforced by occasional pseudo-Doppler effects caused by frequency glides of the complex frequency curves.
There may have been substantial developments of the use of functions in the last twenty years which I am not aware of. In any case, I think the use of functions still holds promises for both composition and sound processing.
6. Some Issues Discussed in Organised Sound
Diverse digital interventionsFootnote 2 in music were discussed in Organised Sound with their technical and aesthetic aspects. Several issues of the journal were realised in collaboration with the International Computer Music Association, dealing with interactivity, networked music, collaboration with intermedia, performing with technology (the latter topic carefully commented on by Simon Emmerson).
Denis Smalley’s spectromorphology was duly considered. Advanced controls of sound space and new concert environments were looked at – the Sonic Lab at Belfast Sonic Art Research Centre and the performance space at Karlsruhe ZKM are impressive resources in this respect. I have already mentioned sonification. Articles have been published about physical models and enaction. Enaction is also a way to look at multimedia. The recent field of real-time programming performance (e.g., live coding) and laptop ensembles has not been forgotten. Daniel Teruggi analysed the problem of dealing with rights as the Web plays an increasing role – actually a dual role: a reservoir of free information as well as an economical space to buy and sell objects and contents, but dematerialised contents such as digital music data blur the distinction. The Organised Sound DVDs with both audio and video files are extremely helpful.
I shall take a closer look on some topics of special interest to me.
6.1. Analysis and reconstruction of computer music
As mentioned above, the analysis of electroacoustic works cannot take advantage of a representation playing the same role as does the score in instrumental music. ‘Listening scores’ help to follow a piece, but they are too rough and approximate to allow significant differentiations. So there are few analyses of electroacoustic works, largely ignored by musicology.
Yet, for works realised with computers, certain data can provide valuable and precise information for analysis. But the significant amount of generated data is generally disheartening or even incomprehensible for non-specialists, and they refer to specific hardware and software that soon becomes obsolete, since technology evolves rapidly, and so do machines and programs.
Nonetheless, structured modular synthesis programs require the specification of the characteristics of the desired sounds in a precise and structured form. All these specifications form the so-called ‘scores’: these are in effect ‘integral’ scores, which specify the relations between the sounds, but also the details of the microstructure of each sound. Thus the synthesis recipes are at the same time complete descriptions, including both macro- and micro-composition, and they can allow for the reconstruction of the music. Unfortunately such descriptions remain esoteric for whomever ignores the conventions of the programs and the purpose of their usage. In order for analysts to be able to use them to investigate certain aspects, it is preferable that they themselves be aware of the techniques of computer music. But the most aware are the composers themselves. Composers should endeavour to help potential analysts by documenting the technical data and notations of their works as clearly and precisely as possible, and by revealing their purposes, their protocols and their cues.Footnote 3 Simon Waters entitled an Organised Sound article ‘Making the archive/archiving the making’.
It would not be right, however, to have the composer as the only analyst of his music. Composers may tend to describe their works according to their intentions or ambitions. A commendable solution would be to associate two analyses – one by the composer himself, another one by a musicologist who analysed it independently. This was done for several works in a book published in 1996 under the direction of Wolfgang Gratzer. The indications from the composer – or possibly of his or her assistant – can shed light for the analysts and facilitate their task, in particular concerning the technical aspects of the realisation.
The participation of composers in the documentation of their works can even permit a complete reconstitution of these works, especially if they have been realised with structured synthesis programs such as Musicn, Csound or MaxMSP. Thus, from the computer score and with help from the composer, works by John Chowning have recently been thoroughly reconstituted: Stria by both Kevin Dahan and Olivier Baudouin, Turenas by Laurent Pottier. Such reconstitutions can permit to restore damaged versions. They can also allow real-time performance which was not possible at the time of realisation. Parts of my own work, Inharmonique for soprano and computer-synthesised sounds, realised in 1977 at IRCAM, were analysed by Denis Lorrain in sufficient detail to allow their reconstitution. Antonio de Sousa Dias has helped revive musical processes and works implemented with software that has become obsolete. Thanks to his and Daniel Arfib’s work, I could use MaxMSP to generate in real-time inharmonic sound according to Music5 recipes described in Lorrain’s report or my 1969 Sound Catalog. In Sousa Dias’s team, Jose Luis Ferreira is working on a reconstruction of Inharmonique to offer the option of a live electronic performance. One should also mention Michael Clarke’s ongoing work in this area.
6.2. Performing with technology
Real-time demands make it hard to ensure the sustainability of the music. Even though the 4X processor used for the realisation of the piece is no longer operational, Boulez’s work Répons, emblematic of IRCAM, has been kept alive at the expense of years of work by dedicated specialists. However, Balz Trumpy’s Wellenspiele real-time version, realised on the 4X and premiered in 1978 at Donaueschingen as an significant production based on the real-time potential of IRCAM at the time, can today be heard only as a recording, whereas Jonathan Harvey’s non-real-time Mortuos Plango, realised with Music5 in non-real-time with the assistance of Stanley Haynes, remains as a prominent work of that period.
This ephemeral character of musical works calling for advanced technology – in particular live electronics – is now well known, yet hard to counter. Even software has problems of portability and survival. Many complex pieces with cutting-edge processes fail to function or to survive. The obsolescence of the technology should not make musical works ephemeral. Some composers are wisely defiant of the latest innovations. Max Mathews quoted Lenin: ‘Revolutionaries are very conservative … in the choice of their weapons.’
In the editorial of Organised Sound 11/3 (2006), Andrew May stated that ‘live performance with electroacoustic sounds (interactive or otherwise) has reached the “awkward age” because many works would require documentation, while reliable practices for documenting the performances have not been established yet’. That issue of the journal emphasised the problems for maintaining live electronics and the possible drifts encountered. In the editorial of Organised Sound 16/2 (2011), Simon Waters brings the concept of the ‘performance ecosystem’ as a framework connecting composition, performance, instrument and environment, appropriate to the systemic concerns of practitioners such as Nic Collins and Agostino di Scipio.
6.3. Musical psychoacoustics and sensory aesthetics
One can now build sounds of practically any physical structure. The challenge migrates from hardware to software, from technology to know-how. The difficulty is no longer the construction of the sounds, but their musical choice. One can design one’s own constraints – or the constraints may be closely linked our perception.
The specification of each sound is not made by stipulating the desired effect, but by describing thoroughly its physical parameters: one should be able to anticipate the auditory effect of the corresponding physical structure. There is a permanent need for understanding the relation between cause and effect, between physical structure and sensory percept: this relation is much more complex than it seems. Prescribed relations between certain physical parameters do not always translate into the expected relations between the corresponding perceptual attributes. For instance, I demonstrated that certain sounds can go down in pitch when the frequencies of their components are doubled. Listening to a synthesised sound allows to experience the so-called psychoacoustic relation between the physical structure and the auditory effect – only the latter counts for music. The exploration of digital sound has produced important scientific advances concerning sound and hearing, leading to a better understanding of musical sound. The assembled knowledge is transmittable.
John Chowning accomplished beautiful simulations of the singing voice and metamorphoses between timbres. Through analysis by synthesis, he has revealed the means that permit us to distinguish two unison tones, namely the vibratory coherence between the partials of each instrument. Using this knowledge, he could in his work Phoné make distinct figures emerge at will from a sound mass. Chowning has proposed the expression sensory aesthetics for a new field of musical inquiry relating to the quest for perceived musicality – including naturalness.
During a conference panel on the future of musical instruments at the SF MusicTech Summit in 2010,Footnote 4 Max Mathews said:
When I first got some – I won’t call it music – sounds out of a computer in 1957, they were pretty horrible. Almost all the sequence of samples – the sounds that you produce with a digital process – are either uninteresting, or disagreeable, or downright painful and dangerous. It’s very hard to find beautiful timbres. A violin may be beautiful but it always will sound like a violin. The computer can sound like anything … What we have to learn is what the human brain and ear thinks is beautiful.
7. Parting Remarks
I was about 12 when I was struck by the sonically organised poetry of John Dryden’s Song for St. Cecilia’s Day, written in 1687:
The double double double beat
Of the thundering drum
Cries Hark! the foes come,
Charge, charge, ’tis too late to retreat!
The foes today are pervasive and insidious: conformism, consumerism, audience ratings, applause meters, zapping, uniformisation. Research and creation must resist this flood of massification. Long life to Organised Sound!
Varèse dreamed of the liberation of sound. Indeed, the twentieh century has fairly well liberated sound, to the extent of giving us the vertigo of the white page: the twenty-first century has to invent new modalities. I shall give the last word to Varèse, in his 1964 Entretiens avec Georges Charbonnier: ‘the last word is imagination’.
