In 2023, sleep slipt away me. Sleepless nights, grey days, fatigue, confusion. I dreamed of a sleep that would hold me as a mother embraces and comforts her child's grief. 

From doctor to doctor, I ended up at SommeilSanté clinic for a sleep study, a Polysomnography (PSG). I spent a night in the lab, hooked up to devices that examine and record the signals of this inner life that escapes me. Would they help me regain the sweet unconsciousness of sleep? Would they help me find the soul adrift in my cringing body? 

In the morning, still groggy, I saw the lab computer's screen and was immediately fascinated by the amount of data collected: all these endless variations of brain waves, breathing, and heartbeats. I asked to have access to the data to make it sing. I was certain it would help me exorcise my nights. 

That’s how I met Dr. Alessandra Coeytaux-Jackson, the clinic director, a neurologist and sleep specialist. She is also passionate about sound, bioacoustics, and electronic soundscapes. The project emerged from our discussions and the exploration of the electrophysiological data from my sleep. I havebeen working on it for over a year now. 

Nowadays I sleep again, but the obsession with reading the polysomnography data with my synthesizers remains. I welcome it as a mother welcomes the joy of her child.

How is the music composed? 

My composition method is based on two principles: staying close to the continuity of biological processes in the human body (which occur continuously, day and night), and creating sound timbres that, through their vibrations and harmonic structure, provide the sensory experience of a sleeping body. 

Since this experience is out of reach, sound is a wonderful away to open the gates of feeling - in the sense of something being felt and recorded in our consciousness and in the sense of being touched emotionally. I call this the "data body", a fluid virtual body created out of abstract digital matter.

The data comes from sensors placed on the body (skull, face, chest, abdomen, legs) that record waves of different amplitudes and frequencies. Like sea waves, they overlap in continuous motion, sometimes very flat (a windless sea), sometimes very steep (a stormy sea). Where the novice sees only disorder, the trained eye of specialists (doctors, researchers) detects patterns, meaningful configurations that are interconnected and provide a reliable image of the functioning of the organs and the body as a whole. 

In its simplest form, these data are numbers, tens of thousands of numbers that follow each other over time, each denoting a slight variation in a brain wave, blood oxygenation, eye movement, etc. 

Yes, that's it. And I find there's a spiraling, hidden harmony when I slow down the “speed” at which the data is typically recorded and decoded. An electroencephalogram (EEG) records the brain's electrical activity at 200 variations per second (a sampling frequency of 200 Hz). I want to make audible the transition from one number to the next, as if watching a slow-motion video of the sea in motion to observe the rolling of water from one wave to another. I then slow down the data reading speed, for example to 0.25 Hz, or one numerical value every 4 seconds. At that speed, it’s as if one could grasp the original impulse of a wave. As if one could connect to its starting point, to its very source.

The tempo is sometimes very slow, sometimes very fast; there are variations that correspond to the actual variations in the data. 

That said, another way of working with continuity is to completely set aside the temporal dimension and analyze the structure of the sea waves themselves—in this case, the EEG waves of sleep—and transpose them to the sound plane. It’s similar to how musicians break down a chord into a fundamental note and harmonics, or how scientists decompose a complex waveform into several sine waves (Fourier transform). 

More simply, it’s like looking at a wave under a microscope and taking a cross-section to see what it’s made of. For example, I apply an FFT (Fast Fourier Transform) to a deep sleep EEG sequence and select the 5 most important frequencies (all delta waves, under 4 Hz), which I then transpose into the audible range (above 20 Hz). This way I get multi-layered sine drones extracted from the original sleep waves. A friend told me that when he heard the first one, he "stopped thinking". It's a subjective experience, of course, but it's nice to know that sound can sometimes create silence.

Another interesting fact is that the superposition of the 5 transposed waves from a deep sleep phase gives me F-sharp as the fundamental note and four other notes from the F-sharp minor scale! The different parts of the project are built around this scale to maintain harmonic coherence. 

<<< TO BE CONTINUED >>>