Applied Anomalies — Nonlinear Acoustics, Hysteresis & Perceptual Edge-Cases

Every enclosed space has a resonant fingerprint — standing frequencies set by dimensions, wall absorption and air temperature. Most audio engineering treats the moment those frequencies feed back as a failure to suppress. The Applied Anomalies studio inverts that premise: feedback is an instrument, hysteresis is a compositional parameter, and the room itself is a performer.

Our bench work pairs directional hypercardioid and shotgun transducers with soft-clipping JFET optical limiters and fractional all-pass phase rotators. As loop gain approaches saturation, the limiter excites odd-order harmonics instead of collapsing into squeal, yielding self-regulating polyphonic drone that shifts with human movement and HVAC drift. PROT-001 (FEEDBACK-RES) and its successors have been calibrated in Studio A’s anechoic chamber, the Salton Vault’s 18.4-second reverberant bunker, and the Mojave desert array where wind and thermal gradients become part of the score.

Research extends to mineral acoustics — 24-note basalt lithophones driven by solenoid actuators — and to boundary-induced phase inversions that allow two feedback voices to coexist in one room without cancellation. The division publishes failure post-mortems when containment fails (see post-mortems), because the edge of stability is where new instruments are discovered.

Bench methodology is deliberately slow: each prototype logs 40–80 calibration notes before a research note is published. Frequency sweeps are captured with laser Doppler vibrometry and compared to predicted room modes calculated from measured dimensions, not nominal blueprints. Discrepancies — a 0.8 Hz low-shift near the east wall of Studio A, a persistent 118 Hz node in the Vault that resists damping — are cataloged as anomalies and become the seed for the next instrument. This is how PROT-017’s tactile floor and PROT-025’s optical scanner both originated: not from a brief, but from a measurement that refused to converge.

Field deployments close the loop between lab and landscape. The Mojave array’s 32-channel ground plane lets us test whether feedback stability holds when the “room” is a basin with no walls, only ground reflection and temperature inversion. Data are not presented as success stories; the 22 anomaly post-mortems include three feedback events that exceeded 118 dB SPL before optical limiting engaged, with containment protocols and decommission notes. Publication is via defensive patent disclosures (PAT-2021-001 series) and the two Transactions monographs that treat feedback as compositional dialogue rather than control loop.