Name the action
Press, slide, hold, release: four different intentions. Before choosing a sensor, it helps to describe the musical consequence of each action. Does it change loudness, spectrum, or the structure of the phrase?
Start with a verb and an audible consequence: approach with the palm to darken HALO’s timbre; increase breath to change AER’s electronic reflection; hold an angle to let SEREIN’s flow thin out. Then identify the reverse action and the resting state. A mapping that makes sound easy to build but hard to stop constrains the phrase before the first rehearsal.
Know what is being measured
A capacitive sensor is good at detecting an approaching conductive object, but it is not a universal force sensor. A key-position sensor shows travel, not load. A useful musical mapping can be arbitrary, provided the performer understands the convention.
In SOMATIC, distinguish raw measurement, gesture estimate, and musical assignment. For a range sensor these might be distance, normalized approach, and filter frequency. Each layer needs clear units. Recalibrating the hand’s range need not change the artistic response curve, and replacing the synthesizer should not require redefining what the sensor measures.
Leave room for error
Even a smooth hand movement contains small variations; humidity and contact with the instrument also change. A dead zone and moderate smoothing help stabilize control. Excessive filtering slows the response and blurs articulation.
Separate activation and release thresholds slightly to introduce hysteresis and reduce chatter near a boundary. This differs from heavily smoothing every movement. SEREIN’s slow tilt and GRAIN’s brief attacks need different temporal responses. Before concealing an unstable sensor behind a long filter, check its mounting, power supply, optical lighting conditions, and the repeatability of the raw signal.
Test it with a simple phrase
Before a complex piece, try a study with three actions: change a parameter smoothly, hold it, and return quickly. If the performer cannot repeat the result, the problem may lie in the control map rather than their skill.
Performers should be able to describe the response in familiar terms: “the sound starts late,” “the level is hard to hold,” or “the tail fills the pause.” These observations help refine the mapping. Nina Vetrova tests movements in short studies; Emanuel Galois compares the experience with sensor readings. The working journal keeps settings and the reasons for changes, including unsuccessful versions.
What the research adds
Shier and colleagues study gesture mappings co-designed with a professional percussionist. Technology can extend expression while steering a performer’s aesthetic. We take this as a reason to revisit musical intention throughout development, rather than only in a final test. [1] Shier et al., 2026
Heyen and colleagues explore practice visualizations across different skills and goals. Our application is selective: an engineer may need the raw sensor trace, while a performer may benefit more from a clear indication of holding and release. [2] Heyen et al., 2026
they also exert aesthetic influence
A practical check
Choose one parameter and three target levels. Play transitions with visual guidance, then without it, and repeat after a short break. Test release and signal loss separately. Record missed targets, fatigue, and unintended triggers. The comparison should reveal which musical task the mapping makes easier, not merely how smooth its trace looks.
Research references
Links point to the specified arXiv versions; publication status is available on each paper’s page. The studies concern their authors’ systems. Connections to Sagittarius A Studio instruments are our engineering and musical interpretation.