IRCAM Grand Metro Station on the Computer Music Subway by Stanley Jungleib [OUTLINE main text OVERVIEW COMPOSITION SYNTHESIS SMALL SYSTEMS sidebars 4X SYNTHESIZER BOULEZ WESSEL appendices BIBLIOGRAPHY/DISCOGRAPHY EUROPEAN COMPUTER MUSIC CALENDAR, FALL 1988] Paris: showcase of Faure and Saint-Saens, Franck and Widor, Ravel, Debussy, and Satie; of Stravinsky's riot-provoking 1913 premier of 'The Rite of Spring,' and of the influential 'Les Six' and 'La Jeune France' groups. From here, Messiaen and Leibowitz led their influential post-war reaction against impressionism and German serialism. The art of assembling sound collages by manipulating tape fragments, 'musique concrete' was invented here in the late 40s by Pierre Schaeffer. His work inspired today's digital sampling. And leading contemporary composers such as Stockhausen, Berio, Xenakis and Risset have all done important work in Paris. Against such a backdrop, under the irremovable scrutiny of history, the very existence of a Parisian computer music center sponsored by the French Ministry of Cultural Affairs arouses the highest expectations. Its creator, Pierre Boulez, has emerged as not just a world-class composer and conductor, but as a dominant figure in contemporary classical music as well. (See interview beginning on p.?.) Not content to merely conquer the symphonic repertoire with every major orchestra, Boulez has throughout his career championed new music. And in creating the Institut de Recherche et Coordination Acoustique/Musique (IRCAM), Boulez has accepted the challenge of bringing classical music -- ultimately, all music -- to terms with a future which inevitably includes computers. IRCAM is not just a computer music lab -- it is a socio-aesthetic vortex of the kind which can only exist in Europe. As a French thinktank, IRCAM reflects the interests of a dynamic culture accustomed to balancing its daily life between grand tradition and vital future. IRCAM is not duplicable anywhere in the world, because nowhere in the world is like Paris. Artists are sustained by its milieu of ceaseless renewal, symbolized in the living architecture, where new structures co-exist alongside 12th- or 17th-century fixtures. And as it grapples with interfacing computer music to society, IRCAM's challenge is to place computer music power into the hands of the contemporary composer with as much grace and determination as how the new glass pyramid has been slipped into the Baroque courtyard of the Louvre. Or, for that matter, as the Pompidou Center for Contemporary Art has been slipped into the renovated Les Halles market district. IRCAM is both physically and fiscally a part of this center, having been conceived by President Pompidou's request to Boulez, in 1970. The main building is renown for being built "inside-out." Exposed ducts, plumbing, conduit, and iron beams allow it to better serve its role of providing flexible, usable human space. In the role of aesthetic or ethical symbol, the center's unromantic surfaces actually serve to focus on, rather than detract from, the humanity bubbling up from every tile: the fashionable, the common, the charitable, the degenerate, and the deranged that populate nearby Stravinsky Plaza. Frolicing in an olympic pool-sized fountain, motorized caricatures dance a slapstick homage to Igor himself. Alongside, stands an 18th-century schoolhouse which contains IRCAM's administrative headquarters and Yamaha-equipped MIDI lab. In its shadow, a singing organ grinder cranks fan-folded pages through his mechanical sequencer. A few stories below his feet lies the most extensive computer music research facility in the world. ((photo fountain/entrance)) Below? Yes, the entire research and performance complex is underground. The entrance looks like a Metro station (the Parisian subway). Inside, no sunlight. No windows. No street noise. Check your sense of time with the guards. Completed in 1977, IRCAM's subterranean world under the plaza can be described as early "Voyage to the Bottom of the Sea," continuing "the Pompidou's" theme of exposed ducts and pipes, catwalks, bulkheads, and sealed doors. The design contrast with Stanford's CCRMA (see Keyboard, Dec 87) couldn't be sharper. In addition to the obligatory complement of offices, labs, computers, recording equipment, and studios, there is an anechoic chamber for listening studies. IRCAM's main performance arena is the notable Espace de Projection (ESPRO), which can be fully tuned to suit the complete spectrum of music. The ceiling can be located several stories or several feet up. The surfaces of ESPRO are arrays of triangular columns which can be rotated so that they either reflect sound from their flat sides, or absorb sound by forming V- shaped anechoic traps. Any acoustic configuration, from completely dead space to Notre Dame emulation, can be preset. (We are talking REAL programmable reverb.) Seating is temporary and flexible. ((photo ESPRO)) IRCAM's original charter emphasized strong support of composition, and this has not changed. Early research addressed problems in synthesis, psychoacoustics, acoustics, and the physiology and psychology of performance, and was originally organized into departments headed by Jean Claude Risset (computer), Vinko Globokar (instruments and voice), Luciano Berio (electronics), Gerald Bennet (multi-disciplinary), and Michel Decoust (pedagogy). But by 1980, most of these originals had slipped away, and Boulez reorganized the institute towards computers, primarily because this was the area in which the most interesting results were being obtained (most significantly, by Giuseppe di Guigno's synthesizer design team -- see sidebar.) Today over 90 people actively work at IRCAM under the general direction of David Wessel (also, see sidebar). This includes engineering and research staff, resident composers, visiting composers, graduate and undergraduate students, support and administrative staff. COMPOSITION: UNION OF WORLD AND IDEA To understand why this multi-million dollar installation exists, let's first consider the general status of computer music. The effect of computer music technology on popular music is enormous and still only a fraction of what it promises to be. But the effect of computer music composition on popular music has been almost negligible, and this shows little sign of changing. One reason for computer music's slow growth is that many pieces require the typical listener to adopt different modes of listening than those to which they are accustomed. Computer pieces are often written according to a formula or process (which is how serialism enters the picture), and the composer is often asking you to appreciate that process rather than simply sing along and tap your toes. Algorithmic music strives to point to a supposedly purer realm of dynamic streams and forms. Sadly, but understandably, in doing so it often leaves many of us mere mortals behind. Few listeners can entirely forego rhythms and harmonies which after 50 centuries of human listening have come to signify musical communication. From this perspective, cerebrally-entombed composers tied up in elaborate private algorithms and sterile etudes are missing the cultural boat, if not just plain copping-out. Why is computer music so hated and feared? IRCAM's Jean-Baptiste Barriere threw responsibility for the situation squarely back on the shoulders of computer composers when he criticized most music performed at the 1986 International Computer Music Conference (ICMC): "The most interesting music today using the computer is, so to speak, inaudible in the noise of the conference. It is somewhere else, and it does not claim to be computer music; it desires simply to be music." After calling attention to the arrogance and narrow-mindedness of most computer composers, Barriere says that, "It is only in the confrontation with other forms, and not folded in upon itself, that computer music can come into the world and disappear into music itself." (Computer Music Journal 11/2, p. 36). 1987 ICMC witnessed continued controversy as to the merit of most of the compositions, as well as research reports. This growing self-criticism within the computer music community is a very healthy sign. Now, how do these issues affect the world's largest computer music facility? If Barriere is right, then drastic changes are needed before computer music can leave the laboratory and stroll down to the local bistro to mingle with the citizens. To begin with, composers need to be able to easily juxtapose traditional instruments and computers through powerful, integrated compositional systems. Next, our current ability to play any sound by computer far outstrips our ability to describe to computers the sounds we want and when we want them. The path from the composers mind to the listener's ear is cluttered with primitive arbitrary, limiting, and time-consuming rules and rituals. Between the musical imagination and the computer's DAC stands the formidable problem of compositional and notational language. Most of the computer power now within any computer music center will be available in the personal studio in three to five years. Given this fact, how can we ensure that musicians can use it to make computer music "disappear into music itself," and thereby close the composition gap? In leisure-rich societies, the demand for these tools is so strong, that fortunes will be created by those who can efficiently package and distribute them. This is why an aware government sponsors such work: aesthetics, ethics, and politics all argue a strong case for better computer tools for composers. And this is IRCAM's mission. IRCAM designs musical tools, but it also trains musical tool designers. It starts by populating its labs with a variety of composers. Certainly you need acousticians and psychoacousticians, engineers, and programmers. But especially you need to have lots of idiosyncratic composers around, who feel secure enough to demand impossible feats of the resources, who suggest far- fetched examples which soon become common practice, and who are hyper-critical enough to serve as vicious alpha/beta testers. The IRCAM environment includes the realization that if you want to design compositional tools, then you need to spend a lot of time allowing composers to exemplify their processes and talk about what composition is. For example, on one hand you have composers who claim to hear the final finished version in their heads in a flash, and who are just trying to steer the computer toward that goal. On the other hand, you have those who don't mind considering the contributions of the computer instrument, improvisation or discoveries made along the way. Some composers may concentrate on traditional instruments and use IRCAM's synthesis resources sparingly. Others, of course, may program complete pieces using real- time synthesis or processing. Some composers begin from a traditional note basis, others think in terms of sonic events. 'Vive le difference!' If in the final analysis composition turns out to really involve any number of strategies (as of course it does), then IRCAM necessarily learns how to select or adjust systems for different types of musicians. SYNTHESIS: THE ROOT OF COMPOSITION Barriere has also pointed out that "synthesis is already composition" (Computer Music Journal 11/2, p. 39), which is to say that the quality of the synthesis affects the quality of the composition. The timbral powers of machines like IRCAM's 1000-oscillator 4X synthesizers (see sidebar 1) allow the composer to go far beyond lesser systems in the sheer mass of sound resources which can be brought to bear upon each instant of time. Its flexibility, strength and character allow the 4X to contribute a completely new and forceful voice to the symphonic repertoire. And "voice" is precisely the term we intend here. The human voice is a great general-purpose synthesizer. (If you don't believe me, ask your local Tibetan monk.) Of course language is literally quite musical: specific, sequenced pitch and timbral changes over time. Thus there ought to be a connection between voice synthesis and instrumental synthesis. And, it in fact turns out that if you can convincingly synthesize sung or spoken voice, then you have created a system inherently capable of accurately synthesizing orchestral instruments, as well as new sounds. Since 1979, Xavier Rodet, Yves Potard, George Bennet, and Barriere have been analyzing and building a computer model of the voice, and creating the CHANT composition language to control it. Since this synthesizer is based on a physical model of the voice, it posesses virtually the same general and complex powers, and has in fact produced convincing versions of most instruments, not to mention its own class of unique timbres. Compositional experience with CHANT synthesis led to the development of the FORMES interactive artificial intelligence composing language. The FORMES project shows one way that "expert" systems will be appearing in compositional tools. As a LISP program, FORMES allows the composer to build a "knowledge base" of sound objects and of rules expressing the relationships between these objects, in a form that can be easily (graphically) manipulated. The "patches" can be freely composed, and allow the composer to achieve completely new depths in the control of real-time synthesis. The CHANT/FORMES environment has been the seat of numerous prize-winning computer compositions. Today, it is the starting point for most IRCAM composers. Recently the team has begun to compile a dictionary of CHANT/FORMES algorithms for studying the essential characteristics of the different instrument families. This, in preparation for allowing the composer to more easily control transitions between these different instrument classes. They also continue to refine their resonance models and demonstrate impressive, detailed vocal synthesis. Many other synthesis methods are being refined at IRCAM, some of which combine traditional approaches (such as additive, or FM), while others are completely new (like the Karplus-Strong algorithm). Each of these techniques have different sonic strengths and weaknesses, and more importantly for commercial consideration, different hardware or processing speed requirements, and therefore widely differing costs. On the performance end of synthesis is the question of how performers on either traditional instruments or non- traditional controllers, or conductors, can by physical gesture control real-time sound processors or MIDI processors. Besides for expression, gesture can be used to select different algorithmic patterns of machine response or accompaniment. Among numerous experiments along these lines, one of the most successful is by Larry Beauregard. It uses physical sensors as well as acoustic analysis. The hall-effect switches mounted on each key of an otherwise standard flute give the processor a good physical basis to start from. From this point on, performer's pitch and envelope nuances control synthesis and processing parameters on the 4X in real-time. Barry Vercoe's "Synapse" (1984) surprised and unnerved its audience in allowing the computer to process the flute according to its own agenda. Boulez's "Repons" also uses the 4X to re-process some instruments. SMALL SYSTEMS MIDI appeared in late 1982, but IRCAM didn't budget seriously for small systems until late 1985. Now, small systems research is Wessel's pet department. October 11 - 13 1986, IRCAM Conference on Personal Computer Music Systems. MIDI-LISP team includes Boynton, Lavoie, Orlarey, Rueda, Wessel, Duthen, Potard, and Rodet. Recipe for Personal Music Artificial Intelligence (AI): Start with Le LISP for Macintosh, by ACT Informatique, add close consultation with IRCAM, and you get MIDILisp. A programming toolkit optimized for MIDI, and ideal for learning AI-programming for music. ExperTelligence in the USA (3122A Eton Ave., Berkeley 94705 654-3902 ?verify) Watch for summer workshops that will be offering MIDILisp courses. LeLisp for the Macintosh also made it possible for FORMES to run on it. Led to new PreFORMES. Puckett. (In lightspeed C, no?) Continues interest in interactive - graphic reconfiguration of sequencer possibly available ? ((illus PreFORMES screen dump of example configuration)) Yamaha research studio under the supervision of David Bristow. Here is a facility that has one of basically everything Yamaha makes. Bristow is Yamaha's eyes, ears, and we may venture, mouth at IRCAM. Working on the edge of where computer music meets the MIDI world, he alpha-tests new boxes as well as gets feedback on Yamaha products from interested IRCAM staff. ((photo Bristow in action)) Adrian Freed at IRCAM developed MacMix, "for browsing, viewing, mixing, and editing sounds. was with Droid Works. Now with IMS.No, make that Waveframe (verify) At this point, discussion of IRCAM's commercial influence becomes more diffuse. An overwhelming number of people that work in or around computer music have spent or will spend time here. Around both CCRMA and IRCAM, numerous companies have spun off. Most no longer exist, but constantly reform into new organizations that usually get just one chance to strike the right chord among a substantial number of musicians, who also happen to be serious enough to be investing thousands of dollars in their studios. If you are reading this magazine now, then chances are good that you'll soon be making buying decisions based largely on conversations occurring in these halls today. SIDEBAR 1 IRCAM'S 4X SYNTHESIZER/PROCESSOR ((no photo: see photo of Wessel with 4X in Wessel sidebar)) Unlike CCRMA's time-shared "Samson Box", IRCAM boasts a number of separate synthesizers which are controlled by individual Sun Unix workstations. Usually, the synthesizer is a 4X. In 1974, IRCAM's electronics director Luciano Berio asked Giuseppe di Guino to build IRCAM's first synthesizer. Di Guino was a high-energy physicist in Rome, who had successfully built a computer-controlled, eight-oscillator, eight-filter hybrid synth in his spare time. Berio asked for a system with 1024 oscillators -- an outlandish request for the time. But diGuino accepted the challenge. In 1975, the IRCAM staff visited Stanford's CCRMA. Di Guino learned the MUSIC 5 system there, and was "furious" to have produced only a few seconds of sound after a week's work. Immediately he began scheming for ways to compute MUSIC 5 with "lots" of oscillators in real time. The result was the 4A, a 256- oscillator wavetable synth built in 1976 (one year before CCRMA's Sambox). This system had no filtering and no envelopes, but excellent audio quality (due to some 24-bit audio paths). It sounded tremendous when producing large clusters of slowly-changing harmonics. In only two years, with four 4As, di Guigno had delivered the promised 1000 oscillators. But it was limited to additive techniques. There was now a demand for other methods, particularly FM, which could not be implemented on the 4A. So, in 1976, in collaboration with Bell Lab's Max Mathews, who served as a scientific adviser to IRCAM, they also designed the 4B. This board had 64 oscillators, but included a filter and amplifier envelope for each oscillator. A flexible interconnection matrix allowed implementing a variety of synthesis techniques. Then came the 4C, which allowed even freer interconnection between digital functions (algorithms). By now IRCAM was sustaining an increasing body of composers. They asked for reverberation and sampling. They asked for much more of everything. Di Guino went back to work expressly to conceive a system capable of serving any compositional technique or intention. The resulting 4X is a general purpose sound analysis and synthesis system that can process up to 64 analog input or output channels, in real time. The 4X design rests on a digital function module that can be instantaneously reconfigured via software. In other words, through the course of the piece, the 4X may be constantly redefining itself and allocating its resources in different ways. For example, you can set up a 1000-oscillator cluster, then within a few milliseconds, switch over to a reprocessor employing 500 filters. Your choices include real-time additive and subtractive synthesis, nonlinear distortion (waveshaping), frequency modulation, ring modulation, filtering (up to 450 simultaneous second-order filters in real time), signal analysis, frequency transposition (harmonizing), reverb, multiple time delays, and pretty much anything you can think of. And yes, it has MIDI. Also unlike the custom Sambox, 4Xs are actually a commercial product. They are used as sound sources in commercial flight simulators. However, this means that the hardware is made to military standards, so they cost $20,000 to $160,000, depending on the number of oscillator cards. Some 4Xs are used at other European computer music centers, which allows for some portability. And the firm Techniques Numeriques Avancees (TNA) offers the 4X as part of its Musical Workstation. (?verify) But before you head to the bank, you may want to know that di Guigno has a 5A in the works. It promises the power of the 4X with both a size and cost of -- can you guess? -- 1/10 that of the original. Asked about his 14 years of experience designing instruments for world-class computer composers, di Guigno said: "I hope that someday the composers will give me a specification of the machine they want. For years they say, 'What we want is this...' but they don't know. They never know from one day to the next. How can they know? Their job is to create complexity. What I say is, 'I do the best I can do,'. If you give them 1000-oscillators, they want 2000. If you give them a 500-kBaud parameter update, they want 1-meg Baud. You know, Stockhausen was here four years ago. He asked for all of the 4Xs in the place for his piece. Eight thousand oscillators! And there was one moment when we had a slight problem with one of the boards, so that he only had 7,000 oscillators -- and he was furious. There is never enough! But this is exactly the crisis in music today: Mozart's 'Don Giovanni' opens with four or five lines and it sounds immense. Today, they can bring in twenty four tracks and it still sounds like nothing." [Note: G.D. recently left IRCAM for Bon Tempi-- verify] SIDEBAR 2 PIERRE BOULEZ AND THE FORMATION OF IRCAM ((photo Boulez)) Born in 1925, Boulez's early education emphasized math and technology, but by 1944 he was studying composition at the Paris Conservatory with Olivier Messiaen and Rene Leibowitz. He then completed several works, including two piano sonatas (1946, 1948) and 'Structures 1a' (1952) for two pianos. 'Structures' was notable for its systematic application of a numerical row, which organized the piece along strictly serial lines. Having established his composing voice, Boulez turned his energies outward and in 1952 founded the Parisian concert series for modern music, the 'Domaine Musical', which served as a rallying point and meeting ground for Europe's most active composers. Throughout the 50s and 60s, Boulez extended his conducting activities, and consolidated his personal style and reputation. In this regard Karlheinz Stockhausen vividly described Boulez in 1968: "Boulez is a composer for whom the quest for technical perfection is absolute, and this technique serves him as a basis for the formation of an unalterable personal style ... Towards his goal, Boulez has not ceased to constantly lay hold on new methods which serve to individualize his work ... Harnessed to this quest there is a constant endeavor to come to terms with tradition ... Boulez has a very definite idea that he must continue the French tradition, and he also has a clear idea of how he must continue it ... Boulez has a synthesizing mind; he studies everything, absorbs everything into his system; and his urge towards synthesis can even successfully embrace the unification of phenomena so apparently opposed as those of the Viennese School and of French Impressionism. It is not surprising that he absorbs elements of Asiatic music; this too is well founded in the French musical tradition." (Karl Worner, 'Stockhausen: Life and Work,' 1973) With its high degree of order and planning, the serial tradition Boulez inherited and developed leads logically to programmed composition. This, as well as any "urge towards synthesis" has encouraged Boulez to embrace the computer, both as a compositional tool and as a symphonic instrument. [SOME INTERVIEW QUESTIONS] ?How do we get from concert music to computer music? What specifically happened (between 1960 and 1972) to introduce Boulez to computers? ?How did he use them in his pre-IRCAM compositions (as synthesizers and/or as composing aids)? ?What lead Boulez to think about IRCAM as a project? Boulez began sketching out the idea for IRCAM in 1972, meeting with the movers and shakers on the Paris scene, through 1973. ?What were the original and main goals (charter) of IRCAM? ?Which goals or projects generated the most enthusiasm among people Boulez approached for assistance? ?Risset has said (Computer Music Journal 9/1 p16) "IRCAM seemed a very unlikely venture, especially in France, but Boulez was in a position to demand its existence." Please elaborate on the factors which made the center "unlikely," and how Boulez prevailed. ?How did Boulez talk the French Government into sponsoring the center? (If that is indeed what happened?) IRCAM was originally organized into departments, with Jean Claude Risset (computer), Vinko Globokar (instruments and voice), Luciano Berio (electronics), Gerald Bennet (diagonal), and Michel Decoust (pedagogy). ?What is the diagonal department? (psychoacoustics?) Max Mathews served as a scientific advisor. 75 IRCAM consulted with CCRMA 1976 IRCAM generated its first computer sounds -- what machine? 1977 moved into present building 1981-1985 Boulez's 'Repons' requires an army of players, technology, technicians and assistants. Piece for three groups: 24 instrumentalists, six soloists, tapes and live sound processing by the 4X. [See separate notes on Repons article by Boulez] ?Today, what is Boulez's prime directive for IRCAM? What is his favorite area of research? What does the French government expect from IRCAM (aesthetic decisions? forum? commercial success?, income?) SIDEBAR 3 DAVID WESSEL IRCAM CZAR [Apparently wessel has resigned from IRCAM, to head Berkeley's new computer music center -- verify . ergo, delete this sidebar and work just some mention into body] ((Photo Wessel)) David L. Wessel b. 1942 began from experimental psychology at Michigan State University, extensive research on timbre perception scaled in various ways. Organized first ICMC 1976 Wessel joined the diagonal dept -so-called because it interconnected other depts. ?Assumed current post when --title Responsibilities main research main compositions In Wessel's 'Contacts Turbulents' "a keyboard is used to control a bank of sequencers that record, transform, and replay phrases produced by another performer in real time." [-- not much room for appendices --] BIBLIOGRAPHY Stanley Haynes IRCAM reports? (haven't seen them) Boulez, Pierre and Gerzso, Andrew "Computers in Music" Scientific American Volume 258 Number 4 April 88. DISCOGRAPHY IRCAM-un portrait. Research sound examples and composition extracts. 70,45 FF need discography Boulez computer works Wessel works anyone else? For current offerings and ordering information: IRCAM, 31, rue Saint-Merri F75004, Paris, France COMPUTER MUSIC CALENDAR, FALL 88 Aug 24-31 "Computer Music for Composition" workshop, Berlin. Contact: Folkmar Hein, Electronics Studio, Technical University of Berlin, FB 1, Sekr. H 51, Strasse des 17. Juni 135, D-1000 Berlin 12, West Germany. Sept 7 - 17. Advanced Macintosh Programming Workshop: Paris, France and Utrecht, Holland. Contact: Bruno Spoerri, Swiss Center for Computer Music, 4, rue St-Laurent, CH-1207 Geneve, Switzerland. Sept 12 - 15: Paris Computer Music Open House Contact; Cornelia Colyer, CEMAMu, GRM, CIAMI, .3 avenue de la Republique, 92131 Issy-les Moulineaux, France, or David Wessel, IRCAM, 31 rue Saint-Merri, 75004 Paris, France Sept 15 - 16. First International Workshop on Artificial Intelligence and Music, St. Augustin West Germany. Contact Cristoph Lischka, GMD F3, Postfach 1240, D5205 St. Augustin 1, West Germany. Sept. 17 - 20 PreICMC, (see next) Sept 20 - 25. 14th International Computer Music Conference (ICMC), Cologne, West Germany. Contact: GIMIK, P.O. Box 600 323, D-5000 Koln 60, West Germany Sept. 24 - Oct 1: First International Symposium on Electronic Art, The Netherlands. Contact FISEA/SCCA, P. O. Box 23330, 3001 KJ Rotterdam, Holland, The Netherlands. // Stanley Jungleib edited the MIDI 1.0 specification and wrote the first articles about MIDI. He works on new products for Sequential/Yamaha. Previous Keyboard editorial and ccrma .../// [extra background] Notes on Repons, from: Boulez, Pierre and Gerzso, Andrew "Computers in Music" Scientific American Volume 258 Number 4 April 88. "The computer synthesizes sounds the composer's mind hears." Right away there are problems. It sounds so much better to say that you hear the sounds first and then use the computer to realize them. But just as often, if not more so, the computer and empirical exposure create the framework which shape imagination. We have these people who claim to hear the entire piece in their head in an instance -- they have heard that this happened to Mozart, so naturally they want to emulate that model. But I think that reality shows there must be give and take between the media and the masseur. Think about it. When you finish a piece, how often is it exactly what you first set out to do? In addition to IRCAM director, PB is president of Ensemble InterContemporain and professor at College de France. Studied at Paris Conservatory .. Yes .. yes .. we know 1946-56 director of Renaud-Barrault theater company. 60-63 professor of musical analysis and comp has been principal guest conductor of Cleveland Orchestra, principal conductor of BBC Orchestra, music director of New York Philharmonic CHALLENGES OF COMPUTER COMPOSITION composer working with technician to articulate the integration of computer and traditional sounds Attempting to use the computer forces composers to try to codify and explain their processes to technicians that build the tools. challenge of providing coherence. Convincing ideas must be readily translatable into both computer and acoustic. The ideas must be resilient enough to be passed back and forth between traditional instruments and computer. distinction between modelling synthesis/resynthesis, and real-time processing of acoustic instruments tape media lacked timing suppleness and bored the audience now computers can do 16 - 40 K sample rate real-time processing advantage of real-time transformation; 1. you can write for traditional instruments and simultaneously create material for the new process 2. study contrasts between the two by "creating close and [or] distant relations 3. Since the transformations are in real-time, they can be spontaneous and expressive. (and imperfect!) REPONS for conductor, orchestra six soloists: cimbalom, xylophone/glockenspiel, vibraphone, harp, piano, piano/DX-7 4X, Matrix-32, six speakers the 4X compactness allows it to be used in concert halls (doesn't require a huge mainframe) manufactured 1984 by SOGITEC ?? What is SOGITEC` Philippe? sounds like a government agency? Let me guess: SOciety Gaul Information and TEChnology? eight processing boards, each of which can be independently programed for any combination of processing methods for example, for additive technique up to 129 waveforms or, up to 128 filters , which can be used for processing. Four seconds of sample memory per board digital functions are arranged as modules the patches which control the modules, and program the interconnections between them, can be loaded (from hard disk) in half a second. So the 4X can be reconfigured any number of times during performance. a language designed by Gerzso , implemented by Patrick Potacsek and Emmanuel Favreau, controls the synthesizer configuration in addition to the configuration program, the 4X runs an event scheduler (Miller Puckette, Michel Fingerhut, and Robert Rowe) which tells the reconfigurations when to occur Matrix-32 is a 16 x 16 programmable patch bay. It controls routing and levels of soloists to the 4X and from 4X to speaker. Built by Michel Starkier and Didier Roncin. Example, in one moment route a soloist to a speaker, in another, route different soloists to different processing modules and send them to different speakers. M-32 can be re-programed in 1/10 Sec. FEATURES OF THE PIECE 1981 commissioned by Southwest German Radio Repons: medieval choral form; call and response. Dialogue between soloists and ensemble, soloist and soloists, between non-processed and processed passages. pitch, rhythm, dynamics, timbre are dimensions upon which the dialogue can occur` thanks to the powers of the real-time processing. Pitch displacement is transposition, rhythm displacement (time) is delay. (Dynamic and timbre displacement we know about.) All of the solo instruments have similar, percussive envelopes. All of the envelope decays are pitch dependent as well. The louder the soloists sound, the faster they move on a specific path between six speakers The peak point of the amplitude envelope is detected and used as a control over the clock rate of a multi-tapped delay line. By arranging a loop of these processors "a sound can be made to circulate repeatedly from speaker to speaker at a speed determined by its amplitude." The computer emulates the "multiplication and proliferation effect" of the medieval call and response, by processing a single note or phrase into chord stacks and arpeggios. ?? How much of this can be done via MIDI today. For example, a programmable keyboard arpeggiator (like the Prophet VS) and a programmable delay line can get you a long way. Soloists enter with broken chords played as arpeggios. spatial modulation takes attention from the center ensemble (orchestra) outward to the soloists and speakers. spatial speed is governed by an envelope follower the decay rate is different for each instrument and for each pitch on each instrument, therefore the spatialization rates change; the sounds die down and slow down at different rates. Also, a decrease in volume level tends to raise the sense of immobility. The envelope follower modulates the clock rate on a multi-tap delay line. Therefore, the louder, the faster the rate. Suppose you arpeggiate CEG3. You might get CEG4, CEG5, GBD5 as responses. The pattern of three notes gets repeated, but also the pattern of three determines that there are three groups of the original arpeggiation. The conductor cues the soloists' playing of notes into the 4X. Each arpeggio can be sampled into 14 registers and played back with different pitches and delays. All harmonic material derives from five chords heard in the first bar. There are a lot of tricks played with chord inversions, and transpositions of transpositions. This arpeggiation / transposition patch is only the first one and is one of fifty patches used 30 seconds Comments on limitations of sampling: frequency shifting the harmonics of a timbre by a fixed amount changes their ratios, hence the timbre is altered. [My example: Fund Res Should be 100 + 25 Hz = 125 200 225 250 300 325 375 400 425 500 Here is a simple and clear demonstration of why simpler sounds -- those with fewer harmonics -- transpose more successfully. -- S] Need more powerful systems, to shift the frequency of each partial. (IRCAM working on this.) Complaint against the trend toward specialized devices, which are difficult to coordinate at the level of integration required to pull off something like this. (He might very well have a point here. -- SJ) Closing box offers demo cassette of excerpts of several pieces, including Repons for $8.50 to Department Diffusion IRCAM [end] *** please confirm receipt *** WS>