Gestural Musical Performance with Physiological Sensors, Focusing on the Electromyogram: The Routledge Companion to Embodied Music Interaction

Physiological computing has become widespread, as the democratization of biomedical technologies has facilitated take-up in interactive computing systems. Signals from the human body can be detected by a wide array of sensors and digitized, providing computational systems information on individual identification, body states, and gross and fine limb movement. These signals have rich potential to be exploited for musical interaction—be it recording performer or audience state in ambient interaction, or capturing instrumentalists’ volitional acts in gestural musical interaction. This chapter focuses on the potential of physiological interfaces to capture performer gesture to create embodied interaction with computer music systems. We begin with a brief history of the use of physiological signals in musical performance. We introduce the range of physiological signals, and focus on the electromyogram (EMG), reporting muscle tension. We then present techniques for using the EMG in music, including signal pre-processing and feature extraction, and analysis by machine learning. We discuss challenges of reproducibility and situate the EMG in a multi-modal context with other sensing modalities. We conclude by proposing gesture “power” as one low-level feature that in part represents Laban’s notion of “effort” to demonstrate the potential of the EMG to capture expressive musical gesture.