Mechanical release and propagation study · unapproved
A sharper release.
Not another melody.
The old pick coupling withdraws over 1.5 ms, suppressing upper harmonics in probes. B uses 0.15 ms withdrawal, with a phase-accurate solver that clears the timestep check the old solver failed.
Same complete 248-note performance and backing. This changes release and numerical propagation—not just EQ, but also not a completed realistic instrument. Existing recordings stay unchanged.
With the band
Guitar only
What changed—and what did not
The six strings and reciprocal bridge keep state through every pick and connected arrival. Both real integration stages contribute force work and losses. The old sharper prototype failed refinement; this is a different, validated backend, not that failure renamed.
Same neck-position aperture, electrical circuit and amp/cabinet/EQ/room/echo. One constant gain matches input RMS, another guitar playback loudness. No per-note fixes, samples, new score or whole-band dynamics.
Release duration is an assumption, not a measured pick. Hand reaction, unilateral contact, switching-energy terms, stiffness/polarization and spatial high-band convergence remain unfinished. Bare-instrument work does not prove complete hand energy or good tone.
It still feeds both cabinet paths one mono physical source. The exact stereo riff benchmark remains separate. No claim of a completed guitar or alto.
Actual six-string backend · Structured solver · Gesture and temporal checks · Independent trajectory/work oracle · Dense linear oracles · Whole-piece pitch/levels/preservation · Method and limits