Case Study 2: The Rebuild — Static Bloom's Width Pass, Logged Move by Move

Jaylen Cole is this book's running bedroom producer, and the session below is a teaching narrative, not field data — but every entry in it is the kind of entry you should be writing during your own width pass, and the protocol is real practice you can run tonight. This case study resolves the chapter's opening hook: when we left Jaylen, four playback systems at the repair shop had returned the same verdict on his weekend of WIDE & HUGE, and the problem shipped with the file. The Project Checkpoint gave you the executive version of the rebuild. Here's the logbook.

The rebuild didn't start in the DAW. It started Thursday night with a pen, because Jaylen had figured out something on the bus home from the shop: the wreckage wasn't a settings problem. Every gimmick he'd stacked that weekend — the widener, the imager, the 18 ms hats, the maxed unison — was the same purchase from the same store: width bought with disagreement between the channels, billed in full the moment any single speaker added them together. Fixing it one plugin at a time would be bailing a boat through the hole. The whole pass had to be rebuilt from agreement upward.

So: Saturday morning, eight days after the bench. Coffee, phone on do-not-disturb, and the session notes open to a flag that had been sitting there since Chapter 20 like an unpaid parking ticket: pad thins in mono — deal with it in the width pass. Today was the width pass.

The Map He Was Demolishing

First entry in the log, because you can't measure a rebuild without a before-state. Here's the stereo field Saturday-night-Jaylen had built, drawn the morning after the bench with the honesty of a man documenting his own accident:

   BEFORE — the WIDE & HUGE map (v2.4 + weekend gimmicks)

   L100         L50          C          R50         R100
   ◄═══════════ PAD, unison maxed ═══════════════════►
   hat copy ◄──────────── 18 ms late ────────────► hat copy
   ░░░░░ DRUMS bus through "WIDE & HUGE" preset ░░░░░░
                   kick · 808 · VOX
   ◄══════════ 2-BUS IMAGER, +40%, everything ═══════►

   ...the smear: everything wide, nothing anywhere.
   Mono verdict: pad = whistle, hats = straw, clap = flam, vocal = +3 dB

The before-map. Note what's missing: seats. Other than the alley — which survived by default, not by decision — no element has an address; everything has a blur radius.

The correlation meter, parked on the 2-bus for the autopsy, spent the chorus wandering between 0 and −0.3. Per the chapter's reading guide: camping below zero, structural fault, addresses available on request.

The Strip and the Zero

Ten minutes, ruthless, logged:

• WIDE & HUGE: off the drum bus, then deleted from the session entirely, so future-Jaylen can't "just try it" at 1 a.m.
• The 2-bus imager: off, deleted.
• The 18 ms hat duplicates: deleted, both sides. The original loop survives, center, alone.
• The pad's unison spread: reset — to be re-voiced properly, not just narrowed.
• Every pan knob in the session — all thirty-four tracks: dead center. Mono button: on.

He played the chorus once like that, the whole record in a single column, and wrote down the sentence the chapter promised he'd write: it sounds terrible and honest, like all foundations. Then the part that surprised him: thirty seconds in, the mono column already sounded better than the WIDE & HUGE bounce had sounded on the bench — because everything was at least present. You can build on present.

The Paper Plan

Before touching a knob, the LCR sheet — every element, three seats, a written reason for anything that would later leave them. This is the artifact that ended up taped above the monitor, and it's reproduced here in full because the reasons column is the entire lesson:

Two entries deserve their own sections, because one was a decision and one was a fight.

The Doubles Decision

The chorus needed to bloom, and Jaylen had two ways to buy it.

Option one was sitting in the session already: back in Chapter 11, while the mic was hot and Demi was warmed up, she'd cut two full chorus doubles after the lead takes — twenty extra minutes that had been waiting five chapters for this exact evening. Option two was the budget path the tutorials push: duplicate the lead, run a doubler plugin on the copies, pan the results.

He A/B'd them properly — matched loudness, full mix, busiest chorus. In stereo, honestly, both worked; the doubler version even sounded a touch tidier, because its fake drift was politer than Demi's human timing. Then the fold. The real doubles summed into the lead as natural thickness — the unison effect, a one-woman choir, arguably better in mono than in stereo. The doubler copies summed into a faint, swirly blur around the lead — not destroyed, the ledger's "budget double" row held up — but audibly a product where the real takes were audibly a performance.

Verdict, logged: real doubles, L100/R100, chorus only. The editing followed Chapter 15's taste — phrase-level tightening so the takes fuse, no sample-level flattening, because a double edited into a clone combs like a fake. And the restraint clause, straight from the ledger's bill: chorus only. The verses keep Demi alone in the alley, which is the entire reason the chorus blooms when the walls arrive.

The One Element That Had to Move

The plan's only casualty was the acoustic guitar, and the fight it put up is the most useful forty minutes in the log.

On paper, hard left made sense — the rock instinct says guitars are walls, and Theo's strummed layer was the track's one organic texture. In practice, the LCR skeleton with the guitar at L100 leaned like a loaded shelf: plucks L40, perc L35, and the guitar all stacked on one side, answered only by a hat at R30 and a delay weave. Thirty seconds of eyes-closed listening, per the chapter's audit: name the heavy shoulder. Left. Instantly.

Attempt one — the bookkeeping fix: duplicate the guitar clip to R100 and call the stage balanced. The correlation meter called it first, easing up while the field got no wider, and the chapter's own warning finished the sentence: a bit-identical copy panned both ways is a phantom-center guitar with extra steps. S = 0. Deleted.

Attempt two — and write this down, because it's the honest part: the Haas temptation, eight days after the bench. Nudge the duplicate 15 ms, and the guitar smeared gloriously across the right half of the field. On headphones it was the best the track had ever sounded, and for about ninety seconds Jaylen stood at the exact crossroads where Saturday-night-Jaylen had turned left. The difference this time was a ritual instead of a feeling: mono button. Straw. The same straw, the bench straw. Deleted — and the log entry reads, verbatim: the trick came back wearing the same clothes. It will come back again. Fold everything. Every engineer learns this one twice; the good ones just compress the interval.

Attempt three — the honest fix: the guitar moved. Not a wall (no second take exists; Theo recorded one layer in Chapter 12, and he's gigging this weekend), but a placed seat at R55, opposite the plucks — far enough to answer the left side's weight, close enough to keep real mono presence. Tone did the rest of the balancing: the guitar's dark, woody strum against the plucks' bright attack reads as counterweight, not mirror — equal weight, different furniture, exactly as the chapter ordered. Two lines in the log: accept the small fold tilt — it's texture, not load-bearing, and if Theo ever cuts a second take, revisit the walls before the mix prints (Chapter 31).

The Verdicts

Then the ritual, by the book: DAW fold-down on the double chorus, the two questions out loud, the hardware fold Monday at the shop — two phones and the register's Bluetooth cylinder — and the one-earbud test on the bus. The verdict table, copied from the log:

The re-touch took ten minutes — pan moves are balance moves, Chapter 21 invoiced as promised — and one marker landed on the timeline for Chapter 27: CH: doubles in, pad voices widen, hats 30→45. The bloom is planned; the riding waits for its own chapter.

   AFTER — the v2.5 map (taped above the monitor)

   L100         L50          C          R50         R100
   CLAP-L                  CLAP-C                  CLAP-R
   GTR DBL*              DEMI · snare              GTR DBL*
   (vox dbl L)           kick · 808                (vox dbl R)
        harm L60                          harm R60
            plucks L40            gtr R55
              perc L35       hat R30
   PAD ◄══ upper voices wide · sub+center voices C ══► PAD
                                          *chorus only

The after-map: every element an address, the alley untouched, the edges earned. Correlation meter: living between +0.4 and +0.9, breathing to ~+0.1 when the chorus blooms. Saved as static-bloom_v2.5_width.

Sunday night he A/B'd v2.5 against the WIDE & HUGE bounce at matched loudness and logged the finding that makes the whole chapter click: the honest version is wider. Not "almost as wide" — wider, because the committed edges finally had a welded center to contrast against. Then Demi's text, Monday, after the new bounce reached her in the van between gigs, one earbud in: "it's all there. even on one ear. what did you do??" The reply is the chapter in four words: "gave everything a seat."

What the Rebuild Teaches

Paper before knobs. The pan plan with a reasons column took fifteen minutes and made every subsequent decision either obvious or visibly wrong. The map isn't documentation of the pass — the map is the pass; the knobs are data entry.

The trick comes back. Eight days after four playback systems convicted the Haas trick, it talked its way back into the session wearing the same 15 ms. Discipline isn't learning the lesson; it's the ritual that re-administers the lesson automatically — fold everything, every time, because your headphones will never stop lying about this.

Seats are negotiable; the alley isn't. The guitar moved, the hat eased inward, the harmonies found their spread — and kick, 808, snare, and Demi never moved a millimeter. A width pass that renegotiates the spine isn't a width pass; it's a new mix.

Verdicts in writing. Three elements in the final table say accept — real losses, chosen consciously, logged. Mono-compatible never meant mono-identical; it means nothing structural lost and every loss a decision. The difference between v2.5 and the bench wreckage isn't that v2.5 loses nothing in mono. It's that Jaylen can tell you, line by line, exactly what it loses, and why that's fine.

Run this rebuild yourself: exercise C1 is this logbook turned into a checklist for your own track, and C2 is the doubles-versus-Haas A/B that inoculates you the way the bench inoculated Jaylen — ideally without the bench.