Richard Causton & Per Bloland 4/4 : The Electromagnetically-Prepared Piano – Going the Extra Mile
Per Bloland and Richard Causton’s artistic research residency, which started at the end of 2022, is now reaching its end, with a final workshop scheduled for next January at IRCAM. In early October 2024, the two colleagues were working together in Studio 5 on their electromagnetically-prepared piano studded with coils and sensors – the perfect opportunity to take stock of their ongoing project, which is constantly evolving and getting more promising.
It should be pointed out straight away that, even though it may not be obvious at first glance, their work has considerably progressed since we last saw them earlier this year! First of all, most of the testing, concerning for instance the longitudinal location of the sensors and the activators along the strings – has been completed. The active control system of the strings’ vibration modes is also functional, the relevant equations having been refined.
“One major breakthrough is that we can now modify with precision the fundamental vibration frequency of a string, which was not an easy task!”, Per Bloland happily explains.
Everything is now working smoothly, or nearly so. Considering this, one might assume that their work is now over – but not quite yet! While the system is operating properly, it requires a specific interface which is not only expensive (almost 10,000€) and cumbersome but also difficult to pilot with a computer. Emulating the operation of the interface on a computer is not simple either because it processes only one sample at a time. To get a better grasp of this issue, we need to understand first how all digital equipment dedicated to processing and broadcasting sound works.
Unlike analogue equipment – which follows the audio signal continuously – digital or computer equipment samples the signal, meaning that it divides it in small parts. The audio signal is represented as a succession of point values, which digital devices are in charge of translating by interpolating an uninterrupted signal between each value. The standard audio sample rate (for compact discs, for instance) is 44100 Hz, which represents 44100 points per second, or one point every 1/44100, or 0.00002267573 second. This standard was chosen because the human ear cannot detect it – not by a long shot!
© Deborah Lopatin
Let us go back now to the electromagnetically-prepared piano, whose interface is processing only one sample at a time. 44100 times per second, the interface receives an audio sample from the sensor, analyses it and calculates the signal required to control the string accordingly, before sending it to the activator. All this is done almost instantaneously: the processing time is shorter than the duration of a sample! Control precision is thus virtually perfect.
The only problem is that computers work in ‘sample packs’. In other words, they need to have compiled a certain number of samples before processing them – the number of samples depends on the quality of the computer and its sound card. Let’s say that this number is relatively low: 256 (or 28 – computers like powers of 2), which makes 256*1/44100 = 0,006 seconds approximately – a delay time that the human ear is very unlikely to perceive. Which is not the case for the active control system! In fact, the active control system needs to be able to target a specific spot on the audio signal in order to be efficient. For instance, to cancel a vibration, the control signal must be exactly in phase opposition to it. But with a latent period of 0,006 second, it gets extremely difficult to adjust the signal accurately. By way of comparison, for an A=440Hz (tuning standard for the musical note of A), the period of the sine wave (the length of a complete cycle) lasts approximately 0,002 second!
Anti-phase signals
Henri Boutin’s mission therefore consisted in writing a code that enables the device to achieve an active control and perfectly calibrate the signal despite the latent period caused by the way computers operate by packets. Spoiler alert: he succeeded in doing so at the term of their last working session. Today, the electromagnetically-prepared piano can be analysed and monitored through a simple interface in the software Max – the dedicated 10.000€ interface is therefore no longer needed. While Henri is coding, Per Bloland et Richard Causton are testing their first patches – with some unexpected results!
Amongst their discoveries, they found out that by varying the stimulus frequency of a string from one end of the spectrum of possibilities to the other, we can hear natural harmonies, overtones and even a multiphonic sound! It is difficult to know exactly where the latter comes from, however. Does it come from a single string that produces both pitches at the same time? Another possibility has to do with the piano itself. Because each pitch can be produced by playing three different strings (to intensify the sound of the instrument or recalibrate registers), if two strings were untuned, even imperceptibly, could they produce two different overtones with the same electromagnetic stimulus? It goes without saying that both composers are thrilled and cannot wait to carry on with their exploration of the many possibilities that this augmented instrument opens up.
© Deborah Lopatin
As for researcher Henri Boutin, he dreams of being able to get rid of the sensor/activator pair. Today, the string’s vibration is measured by an electromagnet located at a good distance from the other magnet which serves as the activator – both needing to be placed far enough from the nodes that produce the natural harmonies of the string. Boutin’s ambition is to be able to use only a single coil both for activating and capturing the string’s signal. To achieve this, he needs to overcome a big challenge: to isolate, within the electrical signal measured at the coil terminals, the response of the coil to the string’s vibration, by subtracting from it the signal sent to activate it – the researcher gets a taste of his own medicine!
As it is often the case in artistic or scientific research, every single breakthrough opens up new perspectives and new project opportunities.
Interview conducted by Jérémie Szpirglas
