IRCAM: Institute For Research & Co-ordination in Acoustics & Music

If you've heard of Paris's IRCAM, you probably imagine it's a government‑funded research bunker devoted to a kind of avant‑garde sonic and musical experimentation that has little relevance to the average hi‑tech musician. Over the last few years, though, as Paul Tingen discovers, IRCAM has been coming down to earth, and a new spirit of openness and commercial awareness now shapes its work.

Right in the middle of Paris stands an outrageous piece of modern architecture called the National Centre of Art and Culture, better known as the Centre Georges Pompidou. Everyone who has done the tourist sites in Paris will vividly remember this building — it's that bizarre. Yet few will have realised that right next to it, some of the world's most advanced and serious work in a field that's of primary interest to the reader of this periodical is taking place. For this prestigious location is the residence of IRCAM, or the Institute for Research and Co‑ordination in Acoustics and Music. The institute is legendary in the classical music world for pioneering avant‑garde computer‑generated music, yet strangely it's largely unknown or ignored in the world of popular music. And if you work in the field of popular music and have actually heard of IRCAM (pronounded 'ear‑cam'), the name is likely to evoke images of white‑coated science boffins working with bespectacled and goatee‑bearded classical composers at homebrew computers, creating atonal screeching noises that have little relevance to most of us.

Despite still being hampered by its ivory tower image two decades after it was founded by France's number one avant‑garde classical composer, Pierre Boulez, IRCAM is still unique in the world.

This picture's rather unfair, though; the original concept of IRCAM — to put scientists and artists together in the pursuit of new music — was, and is, truly visionary. Peter Gabriel's Real World studio and record label may have been far hipper in pursuing a similar aim, but they stated their credo almost a decade later ("...developing new ways for artists of all sorts to work together with technologists") and their results are much less radical and revolutionary. Despite still being hampered by its ivory tower image two decades after it was founded by France's number one avant‑garde classical composer, Pierre Boulez, IRCAM is still unique in the world. It was laudable and far‑sighted of the French government to invest big money in something as 'far out' as electro‑acoustic and computer music. But with government funding safely in hand, IRCAM appeared to lock and bolt its doors to the outside world, and set about developing large and unusual music‑producing computers that had little relevance to anyone not working in the field of avant‑garde classical music. Yet in the process they developed technology, such as real‑time Digital Signal Processing (DSP) and physical modelling, that eventually had a far‑reaching impact on the popular music field as well.

Momentum For Change

IRCAM today is becoming more interesting and relevant to anyone working with electronic music than ever before. The institute abandoned the development and construction of its own computers in the early '90s, and after a period of using Next and Silicon Graphics computers, IRCAM personnel now work almost exclusively on the Macintosh platform, dedicating much of the institute's time and resources to developing DSP and graphics software for the Mac. Three years ago, IRCAM began to make much of this software available to outside users via their user‑group, called Forum (see 'IRCAM Forum' box). The software available via this user group clearly benefits from IRCAM's unique 20 years of experience in the field, and includes physical modelling, virtual acoustics simulation and directionality packages. IRCAM's new openness and relevance to the outside world have also been driven by the appointment of a new, dynamic young director, Laurent Bayle, who took over from Pierre Boulez in 1992, and by the dramatic breaking down of boundaries between different musical styles, disciplines and cultures that has been happening everywhere during the last decade.

New Spirit

The Finnish musicologist Risto Nieminen has been IRCAM's artistic director since 1991: part of his brief is overseeing the choice of young composers that come to study at IRCAM, and organising the 80‑odd concerts that IRCAM is involved in every year around the world. In a large, airy and light‑filled office in a building called the Piano Tower, he muses about the dramatic changes that are affecting IRCAM: "There's been a big change here since Laurent Bayle became director, that's true. But it's also true that many of these changes were part of a natural evolution. Twenty years ago it was much more difficult to collaborate and communicate with other art forms and we had to build computers that made sounds from scratch ourselves, because they didn't exist yet. That was a pioneering phase, during which we had to invent and build everything in‑house, and it was natural that we were closed and didn't have a lot of contact with the outside world.

"Then, when we switched to the Next computer in the early '90s and built the IRCAM Workstation, or Black Box, then later switched to the Mac, it was suddenly possible for IRCAM to communicate much more easily with other studios and performance bodies. We didn't need to drive a truck with a huge computer inside any more, to play a piece somewhere, and we didn't need to bring specialist technicians along. It's now much easier for us to do outside collaborations and still maintain our artistic standards. This is also reflected in the Forum, our user group. The fact that much of our software is now available to anyone who has a good Mac means that we don't have, and don't want, control any more over the kind of music that is made with our technologies. IRCAM technology is now used to make a wide variety of music. This also reflects the fact that different music styles have become much more integrated in general. In the '70s, music styles were very segmented, we had one aesthetic here and a composer couldn't do much else. Now composers feel free to mix different styles, whether classical, avant‑garde, rock, jazz or world music."

Computers have taken a lot of energy from people in recent years, so that they have little left for creativity, and we try to reverse that trend...

One of the most far‑reaching changes initiated by Laurent Bayle, a 45‑year old former theatre and festival director and former artistic director at IRCAM, was to set up a department called 'Valorisation', a word that translates as 'to promote', 'to develop' or 'to put to good use'. Valorisation started in 1993, and its aims radically changed the working methods at IRCAM. The information and technology transfer expert Vincent Puig was made director, and he initiated not only Forum, but also numerous institutional contacts with the outside world, which led to many industrial and commercial applications of IRCAM‑developed technology, some of them quite spectacular. Puig remembers the new wind that blew through IRCAM: "Laurent gave many impulses for new ideas and for a new spirit, in the sense that we opened the doors to the public and offered much more information to it. He also pushed to make our education department much bigger and attracted a new director for our R&D department, Hugues Vinet, who was the developer of GRM Tools, a Digidesign plug‑in, and who comes from the development rather than the research side. This made a big change in our working methods too."

Puig acknowledges the existence of IRCAM's ivory tower image, and asserts that "there's a conscious effort to change that, but it's difficult to erase an image that's been around for so long and that's very strong. But we are trying to make everything available that we research and develop here, and have much more contact with the outside world. We have to do that to survive. Forum, for example, was partly set up to get feedback from the users. They get very early versions of our software, and their feedback is essential, not only to improve the software, but also to make better manuals and documentation and provide everything that comes with regular commercial software: training, service, and so on. So there's been a real change in the spirit of our development. In the past it was more a matter of the R&D team coming up with technological developments and the composers waiting to find out what was presented to them and then seeing how they could apply it. There were, of course, discussions between composers and researchers — and occasional battles and confrontations — and some very interesting things came out of this. But now the work is structured in teams of composers, users, students and researchers working together in development groups for specific projects. The researchers are not just doing their own thing any more, but are also answering the needs that are expressed in these groups."

...IRCAM continues to play an important role in the increasingly fast‑moving world of computer and DSP technology...

The youthful director of R&D at IRCAM, Hugues Vinet (34), tells me that these development groups are for subjects such as acoustics, real‑time systems, on‑line studios, analysis/synthesis and DSP, information systems, and so on. Within these, there are two strategies: "One level is where a user has a specific need and we try to adapt software to his or her needs. That's a short‑term development. But there are also longer‑term strategies that deal with designing brand‑new software in response to people's needs, and examining at a higher level all the problems that occur when people spend time with technology — how they lose time converting formats, or with badly‑designed interfaces and so on — and we try to integrate these findings into our software developments. The fact that our software is now available to the outside world also requires a different organisation of the R&D department, because we now have to guarantee a certain level of quality and we have to support the software in the long term."

Another aspect of going public with its technological developments is that IRCAM has to keep a close eye on the movements of the commercial players in the field, such as Opcode, Steinberg, Digidesign, and so on. Presumably IRCAM chooses what software to develop on the basis of what it perceives as its specific strengths. Asked what these are, Vinet stresses first and foremost "the people we have here at IRCAM. We have a dynamic team with many brilliant people who have been here for a long time and who are well‑connected to the scientific community around the world. We know all the top researchers working in this area, and you can count them on the fingers of two hands. The second thing is that we do both research and development. Companies like Opcode and Digidesign are developers who apply the results of the research. They're not interested in everything we do, but take from researchers what they want. So we are the originators. At the same time, many researchers aren't connected to developers, let alone users. We have a close relationship between research and development here, which is a unique situation in the world, and on top of that we also work in close communication with composers and other users."

Expression

Vinet stresses that this close connection between research, development and user still forms the core and raison d'etre of IRCAM: "The chain between research and useable software is not always easy to create. We try to have as much communication as possible, and a number of composers work here every year to help with scientific research and give our development teams feedback. For example, when giving feedback on new software, they may either make a set of musical sketches with it, to evaluate it from a musical point of view, or they make what we call a tutorial, a set of musical examples that demonstrates the whole range, from the easiest way to use the tool to the most complex way, so that other people can try it out and get trained. Composers are very demanding — time‑stretching came out of the needs articulated by avant‑garde composers, for example. Because of these high demands, the tools we develop usually are very complex. But we don't want to force users into having to adapt their music to the parameters of our technology.

Forum, the IRCAM user‑group, was set up in 1993 and marked one of the most far‑reaching changes in policy and direction in the institute's history.

"Computers have taken a lot of energy from people in recent years, so that they have little left for creativity, and we try to reverse that trend and make our tools as transparent and easy to use as possible, so that people can concentrate on music, rather than technology. Developing good graphic interfaces is an important part of this. We have to identify our specialities in the work that we do, and we have decided that they lie mainly in the area of developing software tools for composers to work with, developing interfaces for inputting information into a computer that comes from gestures, movements, playing and so on, and designing musician‑friendly graphic interfaces. These are ways of introducing meaning into computers and keeping a musician's work away from equations and computations and needing to be a technical expert. Technology must be at the service of the composer's or performer's expression, and I think that musicians who have something to say will try to hide technology rather than show it."

The successes which IRCAM is having in achieving these aims do not mean that the institute will soon follow in the footsteps of the big commercial players in the music technology field. Puig, Nieminen and Vinet all stressed the fact that IRCAM works within the constraints of government funding, and artistic considerations take precedence. Vinet: "The industry can do mass production and marketing much better than us. We're simply not set up to do that. We therefore concentrate on high‑level technology, which we try to licence to any of the major companies, who can either adapt our technology to their own needs, or produce it exactly as we made it, as Opcode did with MAX." Puig: "The only place where we ourselves make our technologies directly available to larger amounts of users is via Forum, which is commercial, or with the few bits of hardware that we make."

Fun

To the same question about IRCAM's particular strengths, Vincent Puig points to the wealth of data that the institute has accumulated during 20 years of acoustic testing, and describes how this has given rise to some very peculiar bits of software, available via Forum: "You cannot imagine all the measurements of instruments that we have here, done in our anechoic chamber or in the concert hall. It's incredible, and they, and the know‑how and studies that we've built up during 20 years, are now being applied for physical modelling in software like Modalys. You can build your own virtual instrument with Modalys, and we also use it for industrial applications. For example, a company came to us asking us to design a sound for their future electric cars, and we're working with France Telecom on designing a new acoustic for teleconferences. We've also used our computer experience and acoustic data to create 3D software called the Spatialisateur. It has an interface that's designed for musicians, and with it you can create illusions of distance, brightness, presence, placing and even out‑of‑speaker placing." (see 'IRCAM Forum' box). One groundbreaking commercial application of IRCAM's acoustic and computer know‑how was their fantasy creation of the voice of a castrato for the movie Farinelli: Il Castrato, in 1994, by skilfully blending and colouring the voices of a counter‑tenor and a coloratura soprano. Another notable achievement is the breakthrough of the aforementioned MAX software, as marketed by Opcode, and apparently Oberheim are close to releasing IRCAM‑developed software for the Mac, based on additive synthesis, which allows, for instance, the sound of someone playing the guitar to be remodelled into another sound, with the added option to morph between the two sounds. It works in real time, and a pedal to establish the morph point between the two sounds is part of the system. IRCAM are also looking for a partner with whom they can make their amazing directionality software, which alters the directional characteristics of sounds, commercially available.

All these developments demonstrate that IRCAM continues to play an important role in the increasingly fast‑moving world of computer and DSP technology, and that their ivory tower image is as defunct as the monster computers they made more than a decade to go. Clearly, IRCAM remains as outrageous and revolutionary as it was 20 years ago. Moreover, it appears that a lot of the technology they produce is now not only highly practical, but also fun — and maybe even hip...