A Body Knows the Pattern

A Body Knows the Pattern is a performance piece mapping movement and muscle tension to different aspects of rhythmic phrase and texture, debuted in London, UK in September 2025. Percussive improvisations and samples are shaped, recombined, and placed in dialogue with one another via the body, interspersed with sculpted layers of noise, drone, and field recordings, ultimately describing a sense of dynamic presence in the lived relationship of the body to the world. The work is part of longer-term research on the translation of embodied musical knowledge to a digital context, and builds on prior composition work with microtiming and expressive variation.  

To create this work, I harnessed a glove-based movement sensor along with muscle sensors placed on both arms, using these signals to directly control multiple aspects of the performance system. I also used lightweight, accessible machine learning tools to train a gesture-recognition model on some of my most common movements, drawing from a detailed performance analysis of my embodied rhythm associations. I combined outputs from the trained model with direct sensor signals to compose a bespoke performance system exploring multiple aspects of my embodied associations with rhythm and texture.

Sonically, I pursued a lo-fi and field recording-influenced aesthetic, reflecting the intimacy of the relationship of the body to its environment, and the ways in which we absorb influences from our surroundings and reflect them back outward in our musical expression. For the sonic materials themselves, I collected samples from materials and surfaces in my living space while also performing percussive improvisations on these, processing the results to create new sounds, patterns, and textures. The construction of this performance system created a prism through which to investigate and speculate on the relationship of the body to its surroundings in a digital context, while providing a platform for harnessing embodied knowledge directly in improvised, rhythm-focused experimental music.

  For the performance documented in this video, I composed several distinct sections, each exploring a different approach to performing rhythm- and texture-based material via wearable sensors, and allowing for improvisation within each theme. Aside from the use of a MIDI controller to transition between sections, all musical output is controlled by sensor input into Max.

(0:00) The piece begins with a demonstration of some of the primary movement-sound interactions employed, including the use of muscle tension to sculpt playback rate and dynamics, and hand movements to trigger percussive clips.
(1:38) Each arm controls the playback of parallel percussion tracks, varying volume and speed. A RAVE model is introduced partway through to reinterpret the body-based improvisations via call-and-response. Additional percussion tracks enter, and arm movements shape the mix of these clips.
(7:02) This section explores a “cascade” strategy, in which a short improvised phrase based on right hand movements triggers a series of responding rhythmic phrases, which can be further sculpted by body movement but not silenced.
(8:23) Multiple layers of quieter percussive improvisations and textures are stretched and shaped by movement and muscle tension via a network of subtle, detailed sensor mappings, creating a shifting, meditative “web.”
(11:18) At first, muscle tension and hand position shape the dynamics of a looped drone file. After 1-2 minutes, I trigger a self-trained machine learning model which changes the mix of a group of percussive textures based on recognized gesture types.
(15:39) This section explores the application of sensor input to noise and beat-driven music, using muscle tension to shape layers of distorted field recordings and the playback rate of the underlying beat.
(20:26) A combination of direct sensor mapping and machine learning shapes the balance of lighter rhythmic textures and drones, allowing for more delicate interactions, approaching silence.