IRCAM
( Institut de Recherche et Coordination Acoustique - Musique )

 

 

There is something weird about Ircam. The place, though a bit austere, is quite nice. It's rather difficult to get in and work there. On the whole, it's very prestigious. But from the inside, it's another story. Though it was much worse a few years ago, Ircam is still some kind of a Boulezian Church, dedicated to its founder and guru Pierre Boulez. Here, Science and Music embrace each other under the severe, yet benevolent glance of the Master.   the Temple

 

Joke aside, what happens at Ircam is very surprising. The researchers claim to follow a true scientific line, but they often forget about the very nature of what they are studying. For instance, an Ircam engineer, who was working on a special speaker, was testing it at the very center of an empty, square room. From observations and measurements made there, he deduced general principles about the speaker's behaviour. Later, the team he was working with was making a demo of this speaker in Ircam's conference room : a normal room with seats etc. Naturally, the demo went very wrong and the speaker didn't behave at all as expected - nothing surprising here, except for our Ircam team, the Room Acoustics team, which had obviously forgotten everything about room acoustics !

 

Ircam also sells the "Ircam Forum", a software suite they develop. But the products are actually awkward : not properly finalized, crashing all the time, very bad ergonomy... When you ask them why on earth those programs work so badly, they answer they are researchers, not software developers. But then, why bother developing software ?   Nice original ideas, bad engineering..

 

And the worst part is that most Ircam researchers are deaf. I mean not really deaf, but they have the most disastrous "sound engineer" kind of hearing. When they develop audio processing tools, forget about transparency, quality of timbre, accurate transients, balanced spectrum etc. None of the usual audio quality criteria one usually expects. I once witnessed the stereo transcription of a live piece with complex spatialization issues. They had to use complex, MSP based tools, to translate the actual live spatialization on two speakers. I don't know about the actual accuracy of the transcription, because the whole thing sounded like it was processed by a very bad MP3 encoder, full of artifacts like badly defined, "flapping" transients, and reduced bandwidth. Amazingly enough, it seemed that I was the only one around aware of this

 

A friend of mine once went to an Ircam summer course for composers. Maybe he was unlucky, but the guy who was teaching actually told him very seriously that what is called in french musique concrète was forbidden, and not only forbidden : anyone who was making this king of music was still lingering in the anal phase - the anal phase, as defined by Freud. The teacher theory was that writing music with real sounds ( sons concrets ) was just like playing with your shit. The "real" music was the abstract one, preferably defined with parameters and compositional processus. This is a typical example of the Bach syndrom. A vast majority of the composers who work at Ircam are subject to this syndrom. Now, this kind of point of view naturally comes from Pierre Boulez. Blinded Boulez zealots still roam free at Ircam...   Revere the Master or prepare to die !

 

Now, to be fair. not all people there are this bad. For instance, the production team is full of competent guys, the musical assistants are often highly skilled, and there are composers who are highly skilled and creative, though not really helped by the researchers supposed to help them working on the tools they want to use. But it remains a very weird place, not at all up to its reputation... at least it's my opinion.   Back to main index