Sound Asleep is a collaboration between Milton Mermikides and eminent sleep scientists Professor Debra Skene (University of Surrey), Professor Morten Kringelbach (University of Oxford), Professor Vladyslav Vyazovskiy (University of Oxford), Doctor Renata Riha (University of Edinburgh) and several sleep researchers and video, audio and computer programmers. The project is developing techniques that allow the systematic translation of sleep data into musical compositions, with a number of outputs and events based on this system.The aims of the project are – through the creation of software resources and associated public engagement artworks and events – to reveal the nature of sleep and the hidden lives that we all share. Through the translation of data into sound, such phenomena as the disruption of sleep in the visually impaired, sleep apnoea and the transitions in brainwave activity between sleep states are captured and translated, allowing a musical communication – and aesthetic appreciation – of this information. The project is being disseminated to members of the sleep science, visually impaired, sleep disorder, music and wider communities, allowing us all to experience this otherwise hidden – yet vital – part of our lives.
Sound Asleep has been presented as a keynotes at the British Sleep Society conference at the Sage Gateshead in October 2015, Royal Physiological Society in December 2018, Royal Society of Medicine February 2020, and has been featured on BBC Radio 4 Inside Science and BBC Radio Scotland where the sleep patterns of visually impaired host Ian Hamilton were revealed sonically. In February 2019, Sound Asleep was exhibited at the Design Museum, London. Some selected outputs are presented below, including a short documentary on the project produced by the National Film and Theatre School.
Nocturne I: Deep Sleep employs an entirely systematic translation system to convert a subject’s polysomnographic data (PSG) into a virtual score, with a flute melody dictated by sleep state (descending when falling asleep, trilling on REM and silent when awake), strings and harp by oxygen transfer (SPO2), breathing interruptions (apnoea), periodic leg movements (PLM) and snoring by pizzicato and percussion, and body position by a bassline. This subject’s enviably good night’s sleep produces a correspondingly peaceful musical output.
Collaborators: Dr Renata Riha (University of Edinburgh) Anna Tanczos (video)
Nocturne II: Breathless. Using exactly the same rules as Nocturne I, a subject with severe apnoea, creates a starkly different soundworld. The orchestra reveals coordinations of apnoea and body movement events, and the fitful sleep episodes.
Collaborators: Dr Renata Riha (University of Edinburgh) Anna Tanczos (video)
Nocturne III: Restless The same translation system produces from a subject with restless leg syndrome, constant movement in bassline and percussion, and a frustrated sleep flute melody that never resolves. The harp gliss at the beginning occurs when the subject remembers to attach their SPO2 sensor.
Collaborators: Dr Renata Riha (University of Edinburgh) Anna Tanczos (video)
Inner Sound of Sleep Electroencephalography (EEG) data is hear translated into audio frequencies, acting as pitch-spectrum oscillators and LFOS. The data is presented in real time, with several extracts through the night sampled. A prominent frequency can be heard falling in pitch with deeper sleep states, and rising rapidly during dreaming.
Collaborators: Professor Vladyslav Vyazovskiy (University of Oxford) Anna Tanczos (Video)
49-48 represents the phasing mismatch between the 24-hour clock and the typical circadian body clock of 24 and a half hours. In the absence of regulating melatonin and light cues, the resulting 49-48 polyrhythm produces a slow but persistent phase.
Sleep Map: Phase In the absence of visual cues, this biological and chronological clock 49:48 mismatch causes phasing, with napping occurring during melatonin peaks, and severely disrupted sleep patterns.
Collaborators: Professor Vladyslav Vyazovskiy (University of Oxford) Anna Tanczos (Video)
Transitions Rather than a simple trajectory from wake to sleep, human sleep involves a complex web of interlinking states. Some can be traversed in two directions, other one way only. The resulting map is remarkably similar to harmonic flowcharts found in analytical models of composition. Here the Hidden Markov model (Stevner et al. 2019) is employed as a Neo-Riemannian harmonic map, offering not just the surface translation of one sleep event, but an underlying blueprint to create countless trajectories.
Collaborators: Professor Morten Kringelbach (University of Oxford)
Stevner, A. B. A., et al. “Discovery of Key Whole-Brain Transitions and Dynamics during Human Wakefulness and Non-REM Sleep.” Nature Communications, vol. 10, no. 1, Mar. 2019, p. 1035, doi:10.1038/s41467-019-08934-3.
