Perchance to Dream

Music hath charms to soothethe savage breast,To soften rocks, or bend a knotted oak By magic numbers and persuasive sound. William Congreve (1670-1729)The

Music hath charms to soothe
the savage breast,
To soften rocks, or bend a knotted oak …
By magic numbers and persuasive sound
— William Congreve (1670-1729)
The Mourning Bride, act 1, scene 1

The power of music is well known. In addition to William Congreve's famous peroration, researchers at the University of Toronto ( have found that music can help those who suffer from insomnia, which is estimated to affect 30 to 40 percent of the general population. Led by Dr. Leonid Kayumov, director of the university's Sleep Research Clinic, the multidisciplinary team involved in this research includes computer programmers, music therapists, composers, neurophysiologists, and psychiatrists.

Most intriguing is their use of computers and synthesizers to create sequences of notes that are customized for each patient, who then listens to the music while trying to sleep. After documenting a patient's subjective complaints and taking objective measurements to determine the exact nature of the problem, an electroencephalograph (EEG) records the electrical activity at several points on the surface of the brain during various states of consciousness, such as alert concentration and deep relaxation. The low-frequency brain waves associated with sleep are called theta (4 to 7 Hz) and high-amplitude delta (0.5 to 2 Hz).

The EEG signals are sent to a computer for Fast Fourier Transform (FFT) analysis, after which a specially designed algorithm generates MIDI note data based on the FFT. This is not a direct conversion from the frequency spectrum of the electrical waveforms from the brain; the music arises from the relationships between the EEG signals, which form a map of the brain's electrical activity in different areas.

The MIDI messages are sent to a synthesizer, and the resulting audio is recorded onto CD-R, which the patient can take home and play according to personalized instructions. Listening to these sounds seems to stimulate the electrical activity associated with the state of consciousness on which they are based; if the sounds are derived from theta and delta waves, those waves are induced in the patient's brain, which helps him or her to sleep.

In a recent double-blind study, Kayumov and his team generated custom “sleep music” for 18 volunteers who had experienced at least two years of anxiety- or stress-related insomnia. The subjects were divided into two groups: the experimental group got CDs of music based on their own theta and delta waves, and the control group got CDs of music based on other people's low-frequency brain waves. After four weeks of use at home, the experimental group showed objective and subjective improvement in the quality of their sleep, while the control group exhibited no objective improvement.

Kayumov's team is also studying the effects of having a human musician compose music while a patient is connected to the EEG machine; this is a variation of biofeedback. The composer plays music while watching an FFT display with the patient in real time. When the composer finds music that consistently affects the FFT in the desired manner (usually increased power in theta and delta activity) for more than two or three minutes, that music is recorded for use by the patient at home. This method was used with one patient in each of the two groups of the aforementioned study, while the rest of the subjects used computer-generated music.

Music-based solutions for insomniacs are clearly much better than drugs, with virtually no side effects or dependency issues. This research is in its infancy, but it holds great promise. Indeed, the tools of electronic music are playing an integral part in helping us to sleep: perchance to dream.