December 1999
I CAME, I SAW, IRCAM
A Visit to France's Premier Music Technology Think Tank

ILLUSTRATION: Jack Desrocher
by PAUL D. LEHRMAN

Keeping on last month's topic of travel, the great thing about getting out of your own country for a while is that you realize there are billions of people all over the world who live their lives and do their work quite differently from the way you're used to, and, astonishingly enough, they are completely happy this way.

In France, for example, the citoyens are perfectly okay about the terrible mail service. A package sent to the U.S. costs three times as much and takes six times as long as the same package mailed from the U.S. As one (American) wag put it to me, "If God had wanted us to use the French post office, he wouldn't have given us Federal Express."

They also seem to be copacetic about not having any such thing as unlimited local telephone service. Every time you make a call, the meter is running. So for me to maintain the Mix Online Web site from France for three weeks, even though I was dialing a local number, cost something like $60 in phone bills--not counting the ISP charge. They deal complacently with stores being closed for two hours or so in the middle of the day (not to mention the entire month of August), and they don't seem to mind, or even know, that the Academie Français has never figured out how to translate the phrase "customer service" into their language. (Although the fellow at the computer store in Perpignan who lent me a modem when mine blew out did seem to be familiar with the concept.)

On the other hand, they are passionate about supporting their cultural, educational and scientific institutions--something we seem to have lost interest in here in the United States, except when it comes to building new highways and football stadiums. Among the many things they give lots of money or attention to are their language (which, despite my difficulties with it, I still find beautiful and expressive), their museums and historic buildings, their educational system, and a unique and thriving research center in the heart of Paris, the Institut de Récherche et Coordination Acoustique/Music, better known by its acronym, IRCAM.

Unlike American research centers, IRCAM isn't based at a university, although it does have some high-level academic affiliations. It doesn't get gobs of money from corporate sponsors who then get dibs on anything that comes out, although it does make deals with private corporations, and is not at all averse to making money. There are other public research institutes in other countries that do similar work--STEIM, in Holland, for one--but none are as well-funded or as well-organized. IRCAM is indeed unique, something you could only find in France, where socialism intersects with capitalism, and fascination with the future intersects with pride in the past.

IRCAM was founded in the 1970s by the French government to promote research in music technology. From the beginning, it emphasized science and art equally. Its founding director was Pierre Boulez, arguably the most influential French composer of the second half of the 20th century. At the time, he also was the music director of the New York Philharmonic, where he was known for his cold personal style, amazing musical ear and killer interpretation of Stravinsky's "Rite of Spring." (I met Boulez, and his ears, during my conservatory days when he came to my school to conduct some master classes. We were reading through a Stravinsky piece for large wind ensemble, in which I was playing contrabassoon--the big wood and chrome thing hiding behind the oboes that looks like a plumber's bad acid trip. The first note was an impossibly complex chord, at the bottom of which was my low C--30.8 [more or less] Hz. He stopped the group, pointed at me and said, "Pull out your tuning slide one quarter inch.")

It would be hard to imagine a figure farther away from the pop world than Boulez, who was known for his extremely complicated, difficult music. Not surprisingly, the work that came out of IRCAM during Boulez's reign was very high-level stuff, aimed at composers working with the most advanced tools, in the most rarefied of artistic worlds. Few composers were lucky enough to gain access to the center's resources; an invitation to work at IRCAM was among the highest honors a computer musician could receive. But the work wasn't exactly aimed at a large audience. I recall visiting there 15 years ago for an International Computer Music Conference, and after sitting through endless concerts of dense, intricate music, I arrived at what I modestly christened the Lehrman principle of computer composition: Any sound, no matter how beautifully and elaborately constructed, will become boring if you repeat it often enough.

Many of the developments at IRCAM, however, eventually filtered down to more popular genres. The institute was at the forefront of the digital audio workstation concept with the 4X, a mini-computer-based synthesizing and processing machine. They researched and built machines for doing physical modeling of sound long before Yamaha and Korg made the technique a household word: It was at IRCAM that Barry Vercoe of MIT and Miller Puckette invented the first computer system that could follow a human performer, and they developed Max, a semi-graphical computer language for manipulating musical events in real time that is now marketed by Opcode Systems, and is hands down the most popular computer music programming system ever developed.

But the world of music technology research changed in the 1980s, and it became very difficult for IRCAM to maintain its leadership role. Work being done at private companies began to outpace what was going on at places like IRCAM. The availability of powerful desktop computers meant that even small (at the time) companies like Digidesign, Steinberg, Antares (née Jupiter Systems), Arboretum Systems and IVL Technologies--just to name a few--could be at the cutting edge, developing practical DSP tools for musicians. Meanwhile, musical instrument companies like Yamaha, Korg and Kurzweil made retail products whose capabilities would have delighted even the most technologically advanced composers a decade before. So IRCAM reinvented itself.

Since my French itinerary this past summer had me in Paris for a few days, I decided to give the place another visit. The first indication that this was a "new" IRCAM came when I called to set up an appointment and I was put on hold--the music they were playing was delightful. It was a kind of floaty, breathy, circular piece for vibraphone and electronics, which lent itself perfectly to a music-on-hold loop, and actually had me looking forward to the next iteration. I found out later it was by Edward Campion, an IRCAM composer.

The second indication was that IRCAM is now, literally, above ground. Its original facility was built entirely underneath a plaza next to the hideous Georges Pompidou Center. It was a formidable space, but it seemed cramped right from the beginning, and natural light was nonexistent; it felt more like a rabbit warren than a space for creative human beings. In 1990, the new eight-story Piano tower (named not after the musical instrument, but its designer, Renzo Piano) was completed, and in 1996, two four-story buildings next door, one an old school and the other a former municipal bathhouse, were taken over and refurbished. Today the underground space is empty, undergoing serious renovations, while all the work goes on in decidedly more pleasant, if still somewhat cramped, quarters (at least, until they finish the renovations and move some of the labs back downstairs).

In 1992, Boulez stepped down from his directorship and was replaced by Laurent Bayle. The new leadership and the new buildings allowed IRCAM to take off in some important new directions, all of which are designed to make the center's mission more relevant to the world outside. And this mission has been very successful.

One new area is education: IRCAM hosts some 60 students per year at various levels. About 20, chosen from among 400 or more applicants, take a "short course" in computer music, primarily so they can get their skills up to the point where they can do more advanced work on their own. Some of these students are French, but most come from other European countries or the U.S. The remaining students are working toward Ph.D.s in a program that IRCAM has set up with a number of different French universities and academic centers. Some are scientists, doing research in acoustics, sound processing and computer applications, while others are historians or in other areas of the humanities, working in musicology or analysis.

In 1993, IRCAM created its first marketing department, overseen by Vincent Puig, who was recruited from the field of technology transfer. Puig is still marketing director and was my host when I visited. He describes his task as providing a much more public face for IRCAM by creating products that people can use, getting the products out into the world and soliciting feedback. One high-profile project involved using IRCAM's resources to re-create a sound that few living ears have heard: the singing voice of a castrato. For Gerard Corbiau's Golden Globe-winning 1994 film Farinelli, about the life of a renowned 18th-century castrato (and if you don't know what that is, you won't find out in this column), IRCAM analyzed the voices of a countertenor (a male who sings falsetto) and a coloratura soprano, and fused them to create 45 minutes of stunningly authentic-sounding vocal material. The soundtrack CD for the film, Puig told me, has sold more than a million units, and the technology used is included in two programs--Audioscope (the analysis part) and Diphones (the morphing part)--now distributed by IRCAM.

Among other remarkable software IRCAM has made available is Modalys, a physical modeling program that goes far beyond the popular synthesizer makers' implementation of the technology, in that you don't have to start with any real model--you can reinvent physics if you like. Then there is The Spatialisateur (the French never use four syllables when seven will do), a collection of Max patches for simulating and mixing in multiple surround environments, which according to Puig allows you to do a surround mix in one format, and then do direct remixes in other formats. In the released product, the emphasis is equally on the capabilities of the software and its usability. "We need to develop user interfaces that people will be comfortable with," says Puig, sounding suspiciously like a product manager at a commercial company.

Max has continued to be a success story for IRCAM, and in 1993, legendary programmer David Ziccarelli came to Paris for a year to work with Miller Puckette, taking the DSP objects that were created for Max on the NeXT platform and porting them to the Macintosh. Four years later, the Mac was fast enough for Ziccarelli to develop a whole new DSP environment based on the stuff done at IRCAM, which he is now selling through his own company, Cycling '74. The latest mutation of Max coming out of IRCAM is jMax, a Java version of the program that will run on UNIX and Linux machines, with Mac and Windows-NT versions on the way.

Another way that IRCAM is interfacing with the real world is through the Multimedia Library, or Médiathèque, a really impressive online index of musical works, recordings, articles, documentation, videos and Web sites, much of which (the text-based stuff, anyway) can be downloaded from anywhere, and all of which physically resides in their building. There's even a "virtual reality" tour of the library online, using World View's VRML plug-in (which unfortunately I couldn't get to work).

Then there's the IRCAM Forum. This is essentially a worldwide user group open to any individuals or institutions. There are three subgroups dedicated to sound design, computer-aided composition and real-time interaction. For $300 a year for one group (or $600 for all three), you can get access to training, information, discussion groups, and a large software library, both online and on a CD sent to all members. You also get to participate in biannual meetings, which are sometimes held at IRCAM and sometimes not. As you read this, you will have just missed a Forum gathering at Columbia University in New York. Current membership stands at just over 1,000, with Americans making up about 15% of the total.

And finally, there's the Studio Online. About 125,000 sounds of classical instruments played by some of France's top musicians in traditional and modern styles, recorded in 24-bit digital using several different microphone setups, are painstakingly organized and cataloged, loaded onto a 160-Gigabyte RAID array, and available for download. In addition, you can access some impressive sound-transformation tools for tweaking the samples. Unlike commercial online sound effects libraries, there are no door slams or gunshots--just musical sounds in unimaginable abundance and variety.

I could go on with a laundry list of the resources available from this remarkable institution, but you can find out what's there yourself: Take a look at www.ircam.fr, and be ready to spend a while. The site is a little flaky in spots, and the English-language version of the site sometimes lapses into French at inconvenient times, but there's a ton of material to browse through. A quarter-century after its founding, IRCAM is still the place to be if you want to hang out with the best minds in the computer-music and research biz, or if you're interested in the latest cutting-edge sounds and sound-making tools. Thanks to the Forum and the center's other outreach efforts, you can be there, wherever you are.


By the time you read this, Paul D. Lehrman will have just finished producing a concert of the 20th century's loudest piece of concert music--composed, naturally, by someone in France--and will still be getting the cotton out of his ears. Gory details at www.antheil.org.

These materials copyright ©1999, 2001 by Paul D. Lehrman and Intertec Publishing