Chapter 32 Exercises

This is a judgment chapter disguised as a tools chapter, so the set is weighted toward your ears and your own record. You'll need: your printed mix from the Chapter 31 checkpoint (24-bit, peaks around -6 dBFS, no 2-bus limiter — if it doesn't meet that spec, fix the print before anything here), your three reference tracks from Chapter 4, your DAW's stock EQ, compressor, and limiter (every DAW ships all three — Appendix E has their names), a correlation meter, your listening journal, and the sticky note. Yes, actually write the sticky note. It costs nothing and it referees everything.

Three ground rules govern every exercise below. First: tomorrow, not today. No DAW Lab exercise is valid on the same day you finished the mix — Chapter 31's next-day rule is doubled at mastering, because limiter judgment is the first thing fatigue eats. Second: the matched-loudness law has no exceptions here. Every A/B in this set happens at matched loudness or it doesn't count; at this stage of the pipeline, an unmatched comparison isn't a shortcut, it's a rigged trial. Third: the ceiling is -1.0 dBTP with true-peak detection on, in every limiter exercise, always. We're not negotiating that one; the chapter's sidebar explains why, and Chapter 33 puts a meter on it.

Difficulty runs ⭐ (warm-up) to ⭐⭐⭐ (will make you sweat productively). Time estimates are honest averages. Exercises marked [A] have worked answers in the Answers to Selected Exercises appendix — try first, check after; the struggle is where the learning lives.

Part A: Conceptual Understanding

A1. The order, argued from consequences. ⭐ · 10 min · [A] Without looking back: write the chain's five stages in order, then give the one-sentence reason for each of these three rules: corrective moves before dynamics moves; dynamics before loudness; loudness dead last. Finish with a prediction: a friend runs limiter → EQ → glue, in that order. Name two specific things that will go audibly wrong, and why the order itself — not the settings — caused them.

A2. The quarter-dB world. ⭐ · 5 min · [A] Explain why a 1 dB EQ move at mastering does roughly the work of a 3–4 dB move in a mix. Then state what the 2 dB line means in practice: your gap list says "low-mids need about -4 dB at 250 Hz." What does this chapter tell you to actually do, and which earlier chapter's discipline is it enforcing?

A3. Humility, direction of. ⭐ · 5 min · [A] Stage 0 pulls the references down to the mix — never the mix up to the references. Give both reasons (one is about judgment, one is about headroom), and explain what you're matching by: peaks, or average? Why that one?

A4. The automation eraser. ⭐⭐ · 10 min · [A] Your Chapter 27 pass put ±1–2 dB vocal rides and a +0.5 dB chorus lift into the mix. Now imagine a master glue compressor at 2:1 working 4 dB deep through the whole song. Do the arithmetic on what survives of those moves, then explain the chapter's budget — ≤2 dB of gain reduction, loudest moments only — in terms of the difference between compressing "the moments between dynamics" and compressing the dynamics themselves.

A5. Spend the decibel. ⭐⭐ · 10 min · [A] A collaborator sets their limiter ceiling to 0 dBFS: "My meter shows no clipping, and lower ceiling means quieter master, so why give up a whole decibel?" Write the correction in four moves: the difference between sample peak and true peak; what an inter-sample over is and which Chapter 2 threshold concept explains its existence; the two downstream systems that will reconstruct or re-encode your waveform without your supervision; and the practice that makes the entire problem evaporate.

A6. The multiband interview. ⭐⭐ · 10 min Three masters land on your bench, each with a "multiband moment": (a) the bass blooms out of control, but only in the choruses; (b) the vocal's esses survived into the master and sting; (c) the low-mids swell whenever everything plays at once. For each: name the stage of origin (arrangement, source, edit, mix — Chapter 30's discipline), the cleaner fix, and the chapter that owns it. Then answer honestly: under what circumstances is master-stage multiband actually defensible — and what gentler cousin does the chapter suggest trying first?

A7. Width court procedure. ⭐⭐ · 10 min · [A] Rank the three master-stage width approaches from most to least honest, with one sentence each on why. State the total budget for side-level moves, the law that follows every single width move, and the reason a width mistake costs more at master stage than it ever did at mix stage. Finish with the prior question you're supposed to ask before reaching for any of them.

Part B: 🎧 Listening Lab

Volume-match every comparison by ear before judging it — the whole lab is void otherwise. Fresh ears only; Chapter 4's hygiene rules govern. Where named records are suggested, substitutes from your own library are fine as long as they fill the same role.

B1. Find the distortion onset — on real records. ⭐⭐ · 25 min · [A] Metallica's Death Magnetic is the documented loudness-war case this book keeps honest company with (Chapter 33 tells the full story). Play "The Day That Never Comes" or "My Apocalypse" at moderate level and hunt the chapter's three onset signatures: sanded transients (listen to the cymbal tops), hardened sustain (the guitars' and low end's texture in dense passages), and your own fatigue (note the timestamp where you first want it to stop). Then play something celebrated for breathing — Daft Punk's Random Access Memories is a famously dynamic modern master; try "Giorgio by Moroder" — volume-matched to the Metallica by ear. Log both in symptom vocabulary. Then the honest second half, because this is not a snob exercise: write one sentence on what the dense master bought — there's a reason people push limiters, and pretending density has no upside is its own ear-training failure.

B2. The era ladder. ⭐⭐ · 30 min Four tracks, four eras: Fleetwood Mac's "Dreams" (1977), Michael Jackson's "Billie Jean" (1982), the loudest late-2000s rock or pop master in your library, and Billie Eilish's "bad guy" (2019). Volume-match each pair by ear as you go, and rank the four by density — not level. Describe each in loudness-character words: does the chorus lift and breathe, or press forward in a slab? Do drum hits poke through the wall or sit inside it? Journal where your own genre's current records sit on the ladder you just built — that's the density culture you'll be deciding about, deliberately, in exercise D2.

B3. Master-character matching, scored. ⭐⭐⭐ · 30 min · [A] The reference-mastering workflow as a drill. Pick three candidate references in your lane that disagree about density — hip-hop/R&B readers might run Travis Scott's "SICKO MODE" against Frank Ocean's "Nights" (dense slab versus air and space); pop readers, "bad guy" against the biggest anthemic chart record they know; band readers, a quiet indie master against an arena one. Gain-match all three to each other by ear. Then scorecard each against the others, Chapter 4 band by band — sub, lows, low-mids, mids, high-mids, air — plus one density line per record. Final verdict, one sentence: which record's clothes does your track want to arrive wearing, and is that the answer your Chapter 5 manifesto predicted?

B4. The transient autopsy. ⭐⭐ · 20 min Same band, two eras: any track from Metallica's Master of Puppets or ...And Justice for All, then any track from Death Magnetic, volume-matched. Attention locked on exactly two things: the snare's crack and the cymbals' tops. Describe the difference in Chapter 1's envelope vocabulary — what survived, what got sanded, what the front edge of each hit does. This is the sound of limiter gain reduction you're learning to hear at a distance; you'll meet it again from inside, on your own record, in C2.

B5. The EP level-relationship listen. ⭐⭐ · 45 min · [A] One beloved album, front to back, volume knob untouched — pick one with a famous quiet moment (Nevermind's "Something in the Way" closes the standard sequence about as quietly as records get; or use your own favorite). Log, track to track: louder, softer, or same as the one before? Where does the quietest track sit in the running order? Did your hand ever actually want the knob — and if not, what does that tell you about how deliberately those relative levels were set? Final question, the one this chapter cares about: did the quiet track read as weaker — or as the reason the loud ones land?

B6. The seam reel, pro edition. ⭐⭐ · 25 min Three seams across two albums: one conventionally gapped record from your own shelf, and Abbey Road's side two, where the medley is famously welded into one continuous build. For each seam, log: the gap length by stopwatch; whether the outgoing tail and the incoming opening flow or stumble (key adjacency — you don't need note names, your ear files the complaint either way); and whether the tempo hand-off feels like momentum or a gear grind. Verdict per seam: default, or designed? You're building the checklist you'll run on your own seams in C6.

B7. Your master among them — the playlist test, formalized. ⭐⭐⭐ · ongoing Run this after C2. Your keeper master goes into a playlist with your three references. Shuffle. Walk away — dishes, bus, gym. Every time your track arrives unannounced, log any flinch with an address: band vocabulary for tonal flinches ("air is shy of ref two"), loudness-character words for density flinches ("mine presses where they breathe"). No fixing during the test; collect first. Re-run after each revision pass until the honest log entry is: it came on, and I just listened. That entry is this chapter's finish line.

Part C: 🎛️ DAW Lab

Each lab starts from a fresh session — your printed mix on one track, references on another, routed around your processing so they stay raw. The chain you build across C1–C4 becomes your mastering template; C7 saves it. Tomorrow-not-today applies to all of it.

C1. Stages 0 and 1: the gap list and the corrective pass. ⭐⭐⭐ · 60 min · [A] The chain's foundation, built with checkpoints. (1) Trim the references down to your mix — usually 6–10 dB, judged on the average, until switching mid-section produces no level flinch. (2) Listen ref-against-mix, section against section, and write the gap list: three to six honest lines in band vocabulary. (3) Corrective EQ from the list only: wide Q (0.5–1.0), one to four moves, nothing past 2 dB — and a matched-loudness A/B verdict logged for every move before it's allowed to stay. (4) If the sides carry low-mid fog, try the M/S side high-pass at 100–150 Hz — followed immediately by a mono check with the correlation meter in view. Deliverables: the gap list, a move log (frequency / Q / dB / verdict), and one sentence naming a gap-list line you declined to fix because it's identity, not defect — your Chapter 5 manifesto holds that tiebreaker. If any move begged for more than 2 dB, write down the mix note instead and be glad the system caught it.

C2. The 0.5 dB ladder. ⭐⭐⭐ · 45 min + 15 next day · [A] The chapter's signature drill — if you do one exercise, it's this one. Limiter on (your C4 two-stage chain if you've built it; the stock limiter if not), ceiling -1.0 dBTP, true-peak on. Climb the input gain in 0.5 dB steps, from zero push to past the point of failure — you need to hear the far side to know where the edge was. At every rung: loop the densest bar of your loudest section, your longest bass or 808 tail, and the snare/hat tops; A/B against the previous rung at matched loudness; write one verdict word in the log — better, same-but-louder, or worse plus the symptom in your own words. When the verdicts turn, step back 0.5 dB: that's your keeper. Bounce it — and bounce one deliberately hotter version from 1–1.5 dB further up the ladder, both with neutral filenames. Tomorrow morning: blind A/B, three trials, matched loudness, verdict in the journal — and a note on whether night-you and morning-you agreed. Keep the full rung log. It's your record's loudness fingerprint, and Chapter 33 is about to hand you the meters that put numbers on it.

C3. Stages 2 and 3: glue and crumbs, auditioned honestly. ⭐⭐ · 30 min · [A] Glue first: 1.5–2:1, attack 30 ms or slower, auto release, threshold raised until the GR meter reads zero through the verses and touches 1–2 dB only at the loudest chorus moments, makeup matched. Bypass A/B at matched loudness; keep the winner and write the verdict ("sections hold hands" or "arrived glued — bypassed"). Then character: your gentlest saturator, lowest drive, drive-and-match per Chapter 26, matched A/B — with one ear on the 200–400 Hz neighborhood you may have just cleaned in C1. The keep-it bar is better, not warmer-and-louder-flavored. Finish by writing which stages your chain actually kept. An empty insert is a verdict, not a missing skill — and for many well-mixed tracks, two empty stages is the chain working.

C4. Two gentle versus one slammed. ⭐⭐⭐ · 45 min · [A] Build both paths to the same total push — use your C2 keeper's push plus 1 dB, so both limiters are genuinely working. Path one: a single limiter doing everything. Path two: a leveler (slightly slower release, roughly half the push) feeding a wall (faster, true-peak on, ceiling -1.0 dBTP, the other half). Bounce both, match the files' levels by hand, then blind A/B three trials: hats and snare tops, bass tails, and the fatigue test — which one can you loop longer without wanting out? Journal the verdict and what you actually heard, in symptom words. Optional third path for the brave: a soft clipper as stage A, crumbs only, same court — test it, don't cargo-cult it.

C5. The width court, in session. ⭐⭐ · 25 min Three bounces of your master chain with side-level lifts of +0.5, +1.0, and +2.0 dB. After each move: mono check, correlation meter in view. Then two rounds of A/B at matched loudness: full stereo (where wider will flatter), and mono fold-down (where the bill arrives). Find the setting where headphone-wow starts costing you the fold-down, and write your own budget in your own words — the chapter says ±1 dB total; your ears should now have an opinion about why. Last line of the journal entry: if a lift genuinely helped, write the mix-stage note (pan map? doubles? Chapter 25) that would have made it unnecessary next time.

C6. Sequence a three-song EP. ⭐⭐⭐ · 90 min · [A] Your mastered track plus two more of your own — older bounces are fine; this drills sequencing, not your catalog. (1) Choose the order with an energy argument, written down: what does the opener promise, where does the heaviest track sit, what resolves? (2) Set level relationships: anchor the loudest track, then place the other two against it by ear, in sequence, at matched listening level — and if one is sparse, give it less push, not just less level. (3) Gaps: start every gap at the 2.0-second default, then decide each one on purpose — momentum or breath — and log the seconds. (4) Build the seam reel: last twenty seconds, gap, first twenty seconds, looped, running B6's three checks on each seam. (5) Tops and tails: trim heads close to first sound, fade tails into silence. (6) The verdict listen: front to back, volume untouched. Deliverable: your level map drawn like the chapter's — relative dB, gap seconds, one-word reason per gap.

C7. Translation insurance, master edition. ⭐⭐ · one evening + next morning Chapter 30's protocol with the master-stage questions attached: the car (your master in a playlist with the refs, shuffled — does it sit or lurch?); the phone speaker (does the low end still imply?); earbuds (post-limiter esses — limiting pushed them forward; do they sting?); the quiet listen (does the vocal still lead at whisper level?); the next room (balance truth). Log per system, rank the fix list, and allow yourself exactly one focused revision pass — Chapter 31's discipline holds at the master too. Then the next-day listen, then print the keeper: 24-bit WAV, peaks at or under -1.0 dBTP, named like an adult — plus your alternate versions (instrumental at minimum) through the identical chain while everything's loaded. Save the session as your mastering template. Chapter 38 is going to borrow it for an experiment.

Part D: Synthesis & Critical Thinking

D1. The self-master brief. ⭐⭐ · 20 min Chapter 31's honest list, applied to your actual next release: stakes, objectivity, monitoring truth, budget, timeline. Write the verdict — self-master or hire — in a paragraph that names the strongest argument against your choice, then the tripwire that would flip it: "I'd hire out the moment ___." Honest answers only; both verdicts are legitimate, and knowing your tripwire is the professional part.

D2. Density as identity. ⭐⭐ · 20 min · [A] The chapter calls a folk record mastered to slab density "a translation error in the cultural sense: it arrives wearing the wrong clothes." Unpack that in a paragraph — what is loudness character communicating before anyone judges the song? Then place your own record on the density spectrum deliberately: cite two references, your B2 era-ladder notes, and your manifesto. Name the direction your genre's culture pulls you, and state — in writing — whether you consent.

D3. The vinyl interview. ⭐⭐ · 15 min Imagine your track is getting cut to vinyl next month. Which of the three physical vetoes bites your record first — low end in the stereo sides, hot sibilance and top end toward the inner grooves, or loudness versus playing time? What would you send the cutting engineer (the chapter is specific about which master), and what two questions would you ask them? No lathe required — the exercise is knowing your own record's physics before a machine votes on them.

D4. The last-stage paradox. ⭐⭐⭐ · 20 min "If the limiter is the loudness stage, and loudness is what everyone wants, why does this chapter spend four stages not touching loudness at all?" Answer as a letter to a friend who masters everything with one slammed limiter and a preset. Use three ingredients: what each stage hands the next, the better-or-louder loop as an experimental design, and — because there are no rules, only consequences — one honest sentence about what their method gets right and when it's genuinely enough (a same-day demo for the group chat is not a mastering session).

Part M: Mixed Practice

M1. (Chapter 2) Your master's sample-peak meter reads -0.2 dBFS — "no clipping" — yet the streaming preview crackles on the transients. Using Chapter 2's not-stairsteps threshold concept: explain why the meter missed it, name the measurement that catches it, and give the setting that prevents it ever happening again.

M2. (Chapter 21) A collaborator sends you their mix "for mastering" and it peaks at -0.3 dBFS. Name the Chapter 21 contract that got broken, what your chain literally cannot do until it's fixed, and the right request to send back — plus one sentence on why "I'll just turn it down myself on track one" is only half an answer.

M3. (Chapter 23) From memory, build the two-column table: Chapter 23's drum-bus glue versus this chapter's master glue — ratio, attack, release, gain-reduction target, and what each is gluing. Circle what stayed identical and what halved, then answer: why does the dose shrink when the scale grows?

M4. (Chapters 25 and 28) Your master sounds huge on headphones and hollow on every Bluetooth speaker you try. Name the likely culprit move, the meter that would have flagged it in real time, the law that should have followed the move, and the M/S concept that explains exactly what the mono fold-down ate.

M5. (Chapters 4 and 22) A friend plays you their "improved" master 2 dB louder than the old version and waits for your praise. Name the two stacked mechanisms that rig this demo — one psychoacoustic, one cultural — and the court procedure that un-rigs it. Then predict, kindly, what the matched-loudness comparison will probably reveal about the improvement.

Selected answers and worked examples: see the Answers appendix. The exercises that matter most this chapter are C1, C2, and C6 — the gap list, the ladder, and the sequence. If you only have two evenings, spend them there, then take the quiz. And leave the sticky note up: Chapter 33 is about to tell you what the platforms think of "just louder," and you'll want to have felt the edge yourself before they weigh in.