Closing the energy loop
An ordinary pickup sends a vibration outward. In an active loop, part of the signal returns through a physical exciter. This makes it possible to sustain individual vibrational modes. But there is latency and a frequency-dependent phase shift between sensor and exciter.
The point is not to add a loud copy to a quiet string. The coil returns energy to the material, so the hand encounters a vibration that has already changed. String, bridge, pickup, and amplifier form a shared system. Moving the pickup changes which modes it observes; moving the coil changes which modes it excites efficiently. A single gain setting therefore cannot transfer the behavior to another body.
Sustaining is not the same as controlling
Simply increasing the gain makes the loop select the frequency most favorable to self-oscillation. Musical use requires filters, a limiter, and checks with different hand positions. Damping the string changes the system; it cannot be treated as a minor disturbance.
Delay becomes particularly consequential at higher frequencies. As an illustrative calculation, a pure delay of 1 ms produces a phase shift of −180° at 500 Hz: −360° × frequency × delay. This is neither a measurement of NACRE nor a complete stability calculation. The actual loop includes resonances, filters, and nonlinearities; a limiter restrains amplitude but does not by itself establish controllability.
How we tune the loop
With NACRE, we begin tuning the loop using one steel string, one sensor, and one coil. We compare the spectrum and decay time at several excitation levels and check heating. We then transfer the settings to the full string system and check interactions between channels.
Begin with a passive reference: the same plucking gesture with excitation disabled, recorded without reverb or automatic normalization. Then change one parameter at a time and save the settings alongside the recording. Check the adjacent string, light contact, and full damping separately. If the reference recording clips, its spectrum describes both the string and signal-path distortion, making the comparison ambiguous.
What we assess in performance
The transition from pluck to sustain should be repeatable without an abrupt change in level. We also check how predictably the instrument responds to touch, damping, and loss of the sensor signal.
A useful sustain must still allow phrasing. Testing a note’s ending matters as much as testing its duration: can the performer remove energy, leave a brief residue, and begin the next attack? In NACRE this question connects Ivan Salamin’s acoustics, Adelheid Von Fliegen’s measurement practice, and Emanuel Galois’s electronics. The musical criterion is a legible transition between states, rather than the longest trace on a graph.
What the research adds
Yi and colleagues infer contact from the spectrum of an electromagnetically excited string. Their device is a robotic sensor, not a musical sustainer. Our question for NACRE is how touch changes the operating regime of its loop. [1] Yi et al., 2026
Gonzalez de la Maza and colleagues describe stochastic electromagnetic excitation without feedback or pitch stabilization. The comparison separates energy source from control law. Their reference to quantum models does not imply a quantum mechanism in NACRE. [2] Gonzalez de la Maza et al., 2026
spectral changes induced by contact
A practical check
Prepare three short listening excerpts: a passive pluck, sustained excitation, and the transition into damping. Compare them at similar loudness while retaining the original unnormalized files. Note the fundamental, sudden overtone changes, coil noise, and the ability to begin a quiet subsequent note. These transitions reveal where an effect becomes a playable instrument.
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.