notes: Take110 first headphone listen (d1 tame, d8 piercing, d3 banger, d4 low)

armada/tide-table/performance_notes.md

@@ -857,3 +857,22 @@ video (`REC-20260928-215506-screen.mp4`; the master sits at +3.80 s in it).
   source: USB stayed up the whole time, no stuck or doubled notes. Not confirmed by opening it.
 - Next: hard-press the key 30-40 times, then dismount the key and clean the contact if it sticks.
   Gig workaround: play the note an octave away or transpose in the patch.
+
+## 2026-11-01 — Take110 (noir_soleil demo), first listen to master v0 on headphones
+- **Overall: "great track".** Cut at v0 0:38 ("less than 5 s from pure pad to first d3") and an
+  ending that fades out or dissolves into reverb were both asked for and rendered
+  (`Take110_master_v0_cut_fade` / `_cut_verb`).
+- **Bass is low (d4), asked "is d4 lonely?"**. Measured: d4 is 150-1k (56% 150-500 Hz, 41% 500-2k),
+  almost nothing under 150 Hz; the sub belongs to d1, d10 and d8.
+- **v0 ~5:12: the pad stops abruptly, then d3 resumes** (PLN had muted d3 and dropped the m4 mute
+  as "done"). Cut before it.
+- **d3 (open hats): "banger whenever present".** Stem sits at -36 dBFS RMS yet reads loud: 55% of
+  its energy is above 5 kHz where only d8 competes.
+- **d8 (jungle breaks): "sometimes too piercing"**, no timecode given. Measured candidate: take
+  210-250 s (v0 ~3:29-4:09), where d8's 2-5 kHz band is 7-10 dB hotter than at 150-190 s while its
+  total level moves 2 dB; d8 is 85-100% of the whole mix's 2-5 kHz there, hump at 3.15 kHz.
+  Hypothesis, not confirmed: the slice/`n` variant changed (^56 sets `n 90`).
+- **d1 (kick, sample `jazz`): "super tame, needs sound design".** Measured: body ~80 Hz, almost
+  nothing under 60 Hz, no beater click (1-4 kHz ~37 dB under the body), -20 dB at 119 ms; 163 of 222
+  hits land at the identical peak (+4.4 dBFS in the float stem): level is not the problem, the
+  character is. Others do not mask it (d1 owns 74-94% of 30-300 Hz power in the 120-150 ms after a hit).