A · existing controller
The preserved post-gate short take, copied byte for byte.
One fret hand.
The old controller keeps settled fret contacts on different strings independently. This version prepares only the next needed lower stop, before picking when the whole hand fits, and releases unused supports after the note is quiet. A few phrases move to another string; the written music stays the same.
Giant Steps · same four-second phrase · matched guitar loudness
B · rolling supports
Current stop, next lower stop when needed, then a quiet release.
A playing-controller experiment, not a claim of convincing electric guitar. No extra melody, samples, bends or exact Scofield imitation. Ruth and all earlier recordings remain unchanged.
What changes—and what doesn’t
The same 20 written notes retain their onsets, velocities, gates, 14 picks and six connected arrivals. Five notes change string. Both string choice and physical contact scheduling change; this is not a one-parameter comparison.
The coupled strings, bridge, neck pickup, circuit and car-song amplifier stay fixed. Amplifier input calibration stays fixed; a constant post-amp gain matches guitar loudness. Bass and drums use the exact same source-clock crop.
The full-score plan uses a provisional 95 mm longitudinal contact-span target, not measured anatomy. The mechanical render checks every integration frame for support counts and span. Finger identity, hand/arm dynamics, unilateral fret collision and full-song audio validation remain unfinished. Numerical tests are not listening approval.
The original pitch detector follows a strong overtone at one pull-off. Its octave result remains in the report; a separately tested broad periodicity estimator checks the actual fundamental. That is a measurement correction, not a sound improvement or an audio edit.
Comparison measurements · Mechanical controller tests · Whole-score schedule checks · Controller source · Study notes