The Electromagnetically-Prepared Piano 3/4 : Thomas Hélie
Thomas Hélie, Research Director of the "Systèmes et signaux sonores : audio/acoustique, instruments" (S3AM) team, talks to us about how Per Bloland and Richard Causton's research on the electromagnetically prepared piano ties in with the activities and history of the team, established in 2017.
What attracted you to Per Bloland and Richard Causton's project?
Firstly, team members like Joël Bensoam had already been in contact with them. So mutual trust had been established. Secondly, they weren't just starting out: Per had been working on the concept of the electromagnetically prepared piano for some twenty years - notably in England with Andrew McPherson. Per has been involved for 5 years. It's work we already knew about, and found interesting in terms of sound results and, more broadly, applications. However, until then, they hadn't explored active control. That is, they simply sent vibrations into the string, but didn't measure them or adjust the action according to the measurement obtained. Per Bloland and Richard Causton came to us with this very goal in mind.
On the other hand, our team had never before worked specifically on the piano. We have a lot of experience with active control on wind instruments, on soundboards (and therefore, by extension, on the piano, although in this case it was just one of many), on bowed strings (but the actuator acted on the soundboard or near the bridge, not on the string itself). In this project, we work directly on capturing and actuating the string itself.
So we can't use the same transducers as usual: the electromagnetic actuator used here offers us the prospect of a new instrument to explore, on which no active control has ever been attempted - at least not recently.
Thomas Hélie à l'Ircam © Ircam-Centre Pompidou
Does IRCAM have any experience in this particular field?
If we go back far enough in the Institute's history, vibration control laws using electromagnetic actuators were established a long time ago, but with rather old, analog technologies. Today, we have the means to revisit the subject with more recent technologies, and thanks to our greater knowledge of control algorithms.
How does this artistic research residency fit in with the S3AM team's research?
We're not starting from scratch; we're adapting what already exists. For example, for the electromagnetic actuator itself. The behavior of an electromagnetic coil in a loudspeaker is well known. The sources of distortion in the loudspeaker and in the electromagnetic coupling are well studied, even if there's still a lot to learn and model. But the actuator we're interested in isn't a moving coil, as in a loudspeaker. Here, the coil doesn't move, but acts as an electromagnet that animates the string. Modeling this is all the more difficult as the magnetic field produced is neither homogeneous nor linear.
Second point: the excitation itself, and the way it reacts to the electrical signal (which acts as a relay between the computer, and sound design, and sound production): the relationship between the electrical excitation and the magnetic field produced is not linear, and must therefore also be understood in order to model it and control the system and the sound result as effectively as possible.
From the point of view of active control, this type of electromagnetic actuator has already been the subject of extensive research, but this was aimed more at stabilizing the magnetic field, with a view to levitating objects, such as the Maglev or magnetic levitation train. Nothing had been done to control vibrations for musical purposes. However, as is often the case, modeling and control laws are very much linked to our objectives. To produce precise sound effects (such as changing frequencies or modifying a spectrum), control laws have a very different architecture than those designed to levitate a train. In this case, the algorithms we use are similar to those we have developed for the xylophone or drums.In this way, the system at the heart of Per and Richard's research intersects with many of the subjects that concern the team, enabling us to develop our ideas in a number of possible directions.
In terms of equipment, which avenues are being explored?
The actuators can undoubtedly be improved, or even redesigned, but we're not working on that right now. Per and Richard have looked into it, and the question will undoubtedly come up again, but for the moment they're concentrating on the amplifiers that power each electromagnet, to ensure amplitude ranges suited to musical performance in a wide variety of venues. One of the major bottlenecks today is power, especially in the low register, or if you want to operate the strings outside their natural frequencies, and even more so if you want to act during the attack.
What do you expect from this residency for the S3AM team?
From a scientific point of view, we're looking for a proof of concept. We want to show that the device works on piano strings and is reliable.
Then, potentially, extend the application of the results obtained to other instruments. But, here again, a major challenge is the power deployed.
Finally, from an artistic point of view, the sounds obtained are absolutely unheard-of and fascinating. Thanks to the diffusion via the piano's soundboard, they are also very enveloping - nothing to do with a set of loudspeakers. Richard is currently composing a piece for premiere in March 2024, and Per and Richard are each working on new compositions where these incredible sounds can flourish.
By Jérémie Szpirglas
