Band-practice 7.18.26, passage 17:40–23:12 (5:32). One transform per mic, reference 8.07.
FINAL_PASS. Brandon's listen: the composite (7 lanes) sounds better than the isolated stems, approved. Audio here is 192 kbps mp3 for the web; the master package is 44.1 kHz / 24-bit.
On the material: this is a guitar-dominant passage. Guitar sits about 20 dB over the kick and 62 dB over the toms, so the drums barely register and the kick A/B was low-stakes. The composite is the deliverable.
Selected bass is mic 7.31 at +2.24 ms. Eight isolated attacks scattered with no repeatable residual, so the physical offset stands. The two aliases are one bass period (13.7 ms) off, for an ear check that +2.24 is right.
pYIN note on the 7.31 bass: voiced 50 percent, no octave jumps, F0 moves 66–203 Hz, but pitch-tracking confidence is moderate. A confirm-by-ear note that the bass follows the line, not a defect.
All five drum lanes are LOW_RELEVANCE here, RMS more than 12 dB below the guitar, so they stay in the frame-locked package but are excluded from required listening and their source choice does not gate the verdict. Kick, snare, toms, hihat are 8.07; cymbals is 7.31. Kick shown only as a sanity listen.
| lane | mic | offset | verdict |
|---|---|---|---|
| bass | 7.31 | +2.24 ms | accepted (ear) |
| guitar | 7.31 | +2.24 ms | accepted (ear) |
| cymbals | 7.31 | +2.24 ms | accepted (ear) |
| kick | 8.07 | 0 ms | PASS |
| snare | 8.07 | 0 ms | PASS |
| toms | 8.07 | 0 ms | PASS |
| hihat | 8.07 | 0 ms | PASS |
All 7 lanes frame-locked at 14,641,200 frames / 44.1 kHz, polarity keep. Kick+bass 40–160 Hz masking ratio ~1.0.
Overall: FINAL_PASS (human review: Brandon, PASS)