Piano mechanics

You are a well-known particle and astrophysicist who works primarily in cutting-edge research. Where does your interest in playing piano come from?

Before I decided on physics, my goal was to become a concert pianist. My children play the piano, and we have a grand piano in the middle of our apartment. This particular instrument needed some adjusting, and that’s how I met Laurent Bessières, the piano tuner of the Paris Philharmonic. We talked about piano mechanics, and he asked me as a physicist if there was a way to give the stroke more force.

Why does the stroke need to be more powerful?

When Mozart and Beethoven performed their piano concertos, their audience comprised a couple of hundred people at the most. Today, we have concert halls that can accommodate up to 2,500 visitors. The orchestras have also become bigger and louder. Yet the mechanics of the piano have practically remained unchanged for more than a hundred years. Especially for fast passages that are very loud or quiet, the boundaries of physics are exhausted. The pianist is simply unable to vary the tone beyond these boundaries.

What defines these boundaries?

Even with the most extreme training, there is a natural limit to the force that fingers can exert. Secondly, the mechanics of the piano are subject to physics. The most important elements in this regard are the stroke path of the key, secondly the lever ratio with which this path is transmitted to the movement of the hammer, and thirdly the weight of the hammer. The stroke has indeed been extended over time, from about seven millimetres to one centimetre. That is the maximum in order for the pianist to still be able to play fast passages. The lever ratio of five is mostly invariable for design reasons. And with a weight of ten to twelve grams, the hammer has also reached its limit. More weight would encumber the playing inertia.

To what extent can the limits be expanded?

The idea is to completely separate the generation of sound from the stroke and to introduce a mechanical power source instead. This is what I proposed to Laurent Bessières, and this was the direction we started looking into. Finally, I found the LM 1247 linear motor from FAULHABER online after sorting trough innumerous unsuitable models. This drive was very powerful and had just the right dimensions. Its width precisely matches that of a piano key! A couple of millimetres more or less would also be feasible, but details like these sure give you a sense that this is the perfect match.

What came next?

In recent years, we were able to start a research project at Centre National de la Recherche Scientifique. As head of research at the CNRS, I have access to excellent specialists in electronics and mechanics who can contribute significantly to making our idea a reality. And, by the way, they surely would have told me right away where to find the right motor, if only I had asked. We were also fortunate enough to get the support of piano maker Stephen Paulello, who is well-known in his field for his extraordinary innovations and who helps us in many regards.

How does it work from a technical standpoint?

We fasten an acceleration sensor to the hammer handle, which moves due to the pressure applied to the piano key. It records the movement of the stroke very precisely in terms of force and speed. Its signal is transmitted by the electronic control to a linear motor. The motor translates this into a hammer movement with very high precision. The hammer is situated on the axis of the motor, in the same location below the string, just like in traditional mechanics. I would like to point out that we are not endeavouring to design a self-playing mechanical piano - that already exists. Those produce sound, but not real music. Instead, we aim to support the pianist and facilitate new possibilities. The artist has full control over the sound. We call this an assisted piano.

What are the benefits of this assistance?

The external source of energy allows us to change a few things. The weight of the hammer can be easily increased to fifty grams or more. A heavier hammer produces more volume, and it can trigger a wider range of overtones due to the greater force. We can dynamically change the lever ratio between stroke and hammer path, for example by introducing a fourth pedal, thus increasing or decreasing the force of the “lever arm”, i.e. the motor. This gives the pianist entirely new possibilities if he or she wants to play extremely fast passages very quietly or very loudly. We can adapt the keys’ path to the artist’s preferences. A piano tuner can replace all hammers in about ten minutes, adapting them to the pianist's needs. With conventional mechanics, such a reconfiguration would entail many hours of work.

Isn’t there a delay between the stroke and the sound due to the interposition of sensor, control and motor?

Actually, it’s the other way around. Conventional mechanics are subject to considerable inertia. When playing softly, for example, the hammer moves about 0.5 metres per second, while that number is tenfold with fortissimo. This factor of ten is not an issue for the pianist. The keystroke alone takes about ten milliseconds when playing very quickly, and accordingly longer when playing slowly. The reaction time of the motor will be clearly below ten milliseconds, probably even under one millisecond. Our electronics also react in the micro-second range. FAULHABER measured the longest interval in our system, which is the delay that occurs between the controller outputting current and the motor standby current. This is in the range of just a few hundred microseconds, so clearly below a millisecond. I am certain that the pianist will not perceive a delay.

Where are you now at with the research project?

Last year we started with a monochord, which is just a single string, in order to check the feasibility of the technology. Since then we know for certain that we have found the right acceleration sensor and motor. FAULHABER has developed an accelerationoriented control that suits the instrument better than the traditional speed controller. We are now in the process of working out the details of the assisted mechanics. By June, we aim to cover an octave, which is 12 notes. Once we get to that point, we will fit the entire grand piano with the assisted mechanics. Hopefully, we can have the first concert next year.

How important is the motor to the project?

Without it we wouldn’t have been able to even start. Until a few years ago there was no motor with the performance characteristics that we required. The motor has to be very powerful despite the small dimensions, accelerate extremely quickly, and it must be controllable with high precision. And notwithstanding all of that, it must operate completely silently. It was only the development in neodymium permanentmagnets in recent years that has made such a motor even possible, and of course also the development work by FAULHABER.