Chapter 23 Exercises

Work these roughly in order: Part A makes sure the five controls and the taxonomy actually stuck, Part B retrains your ears on records that already made every compression decision you're about to face, and Part C — where this chapter actually becomes yours — puts your hands on the two knobs that matter. Exercise C1 is the chapter's centerpiece; if you do nothing else, do C1. Every DAW lab assumes the discipline you've now been taught three chapters running: matched-loudness bypass A/B on every decision — set the makeup gain so bypassed and engaged play at the same level before you're allowed to have an opinion. Time estimates are honest averages. Selected exercises (marked ★) have worked answers in the back of the book (Answers to Selected Exercises). Listening Lab deliverables go in your Chapter 4 listening journal, which gains a new permanent column this chapter: what's the envelope doing?

Part A — Conceptual Understanding

A1. The five sentences. ★ (10 min) From memory, write the one-sentence definition of each of the five main controls — threshold, ratio, attack, release, makeup gain — plus the knee. Then sort the six into two piles: the how much controls and the when controls. Finish with one sentence stating which pile this chapter claims is "the actual lesson," and why.

A2. Transfer-curve forensics. (15 min) Sketch (ASCII or pencil) a transfer curve for a compressor set to threshold -20 dBFS, ratio 4:1, hard knee. Mark where a -30 dBFS signal exits, where a -12 dBFS signal exits, and how much gain reduction the -12 dBFS signal received. Then redraw the curve twice: once with a soft knee, once at ∞:1. Label which of the three drawings is a limiter and write one sentence on what physically distinguishes it.

A3. The punch paradox, explained to a friend. ★ (15 min) A friend says: "Compressors only turn things down, so how can a compressor make my snare punch HARDER? That's backwards." Write the four-to-six-sentence answer using the envelope vocabulary from Chapter 14: name the part of the snare's envelope a slow attack protects, the part it clamps, and why increasing the contrast between them reads to the ear as more punch, not less. Bonus sentence: what happens to the same snare at a 0.1 ms attack, and what does the chapter call that ditch?

A4. Release does the groove's books. (10 min) Describe the two failure directions for release on a drum loop: what you hear when release is too slow (where does the energy go, and which hits pay for it?), and what you hear when it's too fast — including the special low-frequency failure, with the rough physics of why a release faster than a bass wavelength turns into distortion (a 50 Hz cycle lasts about 20 ms; reason from there).

A5. Meter stories. ★ (10 min) The chapter showed three GR-needle stories: the dance, the twitch, and the pin. For each, describe what the needle is doing, what it means, and whether it calls for action. Then state the GR working zones for leveling, punch, and glue from memory, and the proverb about meters and ears.

A6. Taxonomy triage. (15 min) For each scenario, name the ONE job from the purposes taxonomy and sketch the settings it wants (ratio / attack / release / GR zone): (a) a podcast guest who whispers questions and shouts punchlines; (b) a kick drum that's all click and no thump wait — reverse it: a kick that thumps but doesn't snap; (c) eight drum tracks that sound like eight strangers; (d) a producer who wants drums "thicker without losing the hits"; (e) a track where the producer wants the whole beat to visibly inhale and exhale with the kick. For (d) and (e), name the chapter where the full technique lives.

A7. The opto explanation. (10 min) Without looking at case study 1, explain in your own words what physical component gives an optical compressor its program-dependent release, what "program-dependent" means in plain words, and why that behavior happens to flatter the human voice. Two short paragraphs.

A8. The forum reply. (15 min) A post reads: "Pro tip: slap a compressor on every track, 8:1, fastest attack, makeup until it's LOUD. Instant professional sound. Works on everything." Draft a four-sentence reply in this chapter's spirit — no scolding, diagnose then fix. Name the bias the makeup-gain move exploits, the part of every sound the fastest attack destroys, the law that makes the "instant improvement" evaporate, and the one-job-per-instance alternative.

Part B — 🎧 Listening Lab

Reference playlist open, marked listening level from Chapter 8, journal ready. The skill this lab builds: hearing gain in motion — level that moves because a circuit moved it.

B1. The pump safari. ★ (25 min) Cue three records where pumping is the aesthetic, not the accident: Daft Punk's "One More Time," Stardust's "Music Sounds Better with You," and Eric Prydz's "Call on Me" (substitute any French-house or big-room track you love). In each, listen past the kick to the sustained material — pads, vocals, filtered disco loops — and describe the surge cycle: what ducks when the kick lands, how fast it swells back, and whether the swell lands in tempo. Then find one track in your own library where you can hear pumping that you suspect nobody chose — quieter, accidental, seasick. Deliverable: journal lines on all four, plus one sentence on the difference between the chosen and unchosen pumps. (The technique that creates the deliberate version is Chapter 28's; your job today is only to hear it.)

B2. Punch versus squash, the famous case. ★ (25 min) The loudness-war chapter of this story is Chapter 33's, but your ears can start now. Listen to any track from Metallica's Death Magnetic (2008) — the album whose crushed dynamics fans famously protested — and then to a track from the same band's late-'80s/early-'90s catalog at matched listening volume (turn the older one up so they feel equally loud; your phone's volume is the matched-loudness law in street clothes). Listen for: do the drum hits land or click? Does the music get louder when everyone plays harder — or is it already at the ceiling? Where does your attention go after ninety seconds of each? Deliverable: a punch-versus-squash verdict with three pieces of audible evidence, written in this chapter's vocabulary (transients, envelope, crest).

B3. Glue identification. (25 min) Two drum-bus worlds: cue Miles Davis's Kind of Blue (1959 — an era of minimal processing; the kit's dynamics arrive essentially as played) and any current mainstream pop or hip-hop track. Listen only to the drums in each. In the first: how much does the level of the kit move — ghost notes versus accents, brushes versus sticks? In the second: notice how the whole kit seems to move as one object, every element rising and falling together, nothing poking out unchosen. That oneness is what this chapter named glue. Deliverable: two journal entries, plus an honest sentence about what each world gives up — the jazz kit's risk, the pop kit's freedom.

B4. Vocal-leveling detection. ★ (20 min) Cue a modern pop lead vocal — Billie Eilish's "everything i wanted" works beautifully (a near-whisper that never gets lost) — and one live or older jazz/folk vocal recording from your library. For the pop vocal, listen for the rail: does the voice's distance from your ear ever change, even when the delivery moves from murmur to full voice? Count how many words you lose in either recording at low playback volume. Deliverable: journal lines on both, plus the chapter concept that explains why leveling + makeup gain makes whispers more audible, not less.

B5. The radio experiment. (15 min) Find a broadcast-style voice — FM radio, a major-network podcast, a sports commentator over a crowd. Listen for three minutes with one question: when does this voice ever get quieter or closer? Then listen to a voice memo of yourself on your phone. Deliverable: a description of the broadcast voice's behavior in GR-meter terms (what would the needle be doing?), and a guess at the working zone — then check it against the spoken-word playbook row.

B6. Breathing hunt. (20 min) Heavily compressed rock and pop masters often breathe audibly in the cymbal wash and room tone. Cue three loud-era rock tracks (anything mainstream from the 2000s is a good hunting ground) and listen to the space between and behind the hits — the wash, the tails. Can you hear the background swelling back up after each loud event, like an inhale? Deliverable: one track where you clearly caught the breathing, with a timestamped moment, and one track where you couldn't — plus a sentence on which compressor control most likely sets the breath's speed.

B7. The reference dynamics audit. (20 min) Take the three reference tracks from your Chapter 4 playlist and rank them by how much dynamic life survives: do choruses physically lift? Do drum hits poke through the wall? Does anything ever get quiet? Deliverable: your ranking, with one sentence per track — and a note on where your own song should sit on that spectrum, because that decision (not a meter, not a tutorial) is what your compression pass serves. The gap between your mix and your reference is your to-do list; this chapter you're auditing the dynamics column of that list.

Part C — 🎛️ DAW Lab

C1. The attack/release-only drill. ★ (30–40 min — the chapter's centerpiece; do this one even if you do nothing else) Load your Chapter 13 drum loop (or any decent loop). Stock compressor on it. Fix these and never touch them again for the whole drill: ratio 4:1, threshold set for about 6 dB of GR on the hits, makeup matched to bypass. You may now move only attack and release. Four committed settings, bounce each:

1. Fastest attack, fastest release. Listen to the hats and the kick tail. Name what you hear (the chapter gave you the words).
2. Fastest attack, slow release (~500 ms+). Where did the groove's energy go? Which hits pay?
3. Slow attack (~30 ms), bounce-test release. Sweep release until the GR needle bounces in time — home before each hit. This should be the punchiest the loop has ever sounded.
4. Slow attack, absurdly slow release. The needle never gets home. Describe the specific tiredness.

Finish with the A/B that makes it permanent: bounce 3 against bypass at matched loudness, three honest listens. Deliverable: four bounces plus a journal entry with your attack number, your release number, and one sentence: what did the threshold and ratio contribute to any of this? (Trick question — they were frozen. That's the lesson.)

C2. The five-purposes exercise. ★ (45 min) One source, five intents — proof that the job, not the source, picks the settings. Take a single full drum loop and create five duplicate tracks, one compressor each, configured from the taxonomy map: (a) leveling — 3:1, medium attack, auto release, 2–3 dB only on the loudest bar; (b) punch — 4:1, 25 ms attack, bounce-test release, 4–6 dB on hits; (c) glue — 2:1, 30 ms attack, auto release, 1–2 dB swaying; (d) density — 6:1, fast attack, medium release, 8 dB, level matched (hear what the tail does — and what the hits paid); (e) FX — pin the meter, fastest everything, no apologies. Bounce all five plus the dry, shuffle, listen blind a day later, and identify each by ear. Deliverable: six bounces, your blind-test score, and one line per version naming what it would be for.

C3. The bounce test, formalized. (20 min) On your own song's drum bus: pull the threshold down to an exaggerated 6–8 dB GR so the needle is easy to see, then sweep release slow-to-fast while watching the meter against the groove. Find the setting where the needle gets home just in time for each backbeat. Log the number, then compute your song's quarter-note duration in ms (60,000 ÷ BPM) and compare. Then restore the threshold to glue-zone (1–3 dB). Deliverable: your release number, the math, and whether ear and arithmetic agreed.

C4. The drum-bus glue dial-in. ★ (30 min) The Project Checkpoint move, performed deliberately. On your drum bus: 2:1, attack 30 ms, release auto. Lower the threshold slowly while the chorus loops, watching GR: stop at 2–3 dB swaying with the beat. Match makeup gain to bypass. Now the two tests: (1) matched A/B — does the kit sound more like one kit? (2) the squash check — keep lowering the threshold to 6, then 8 dB and listen for the moment cohesion turns into the vise (transients dulling, groove tiring, pumping starting). Back off to the last setting where everything still breathed. Deliverable: your final threshold/GR numbers and a one-sentence description of the moment you heard glue become squash — that boundary, in your ears, is the chapter.

C5. The rail, rebuilt. (30 min) Recreate Theo's hook scene on your own lead vocal (or any sung/spoken take). First, try to fix the dynamics with the fader only — ride it by hand through one verse and chorus, and feel the impossibility. Then: clip-gain any lightning-strike words down by hand (Chapter 21's hierarchy), insert a compressor at 3:1, attack 15 ms, auto release, threshold for 3–4 dB on the loudest phrases, makeup matched. Deliverable: a before/after bounce pair and your answer to: which specific phrases did the compressor catch, and which did it correctly ignore?

C6. Find the wall, then back off. (20 min) The squash-and-back-off arc as a repeatable ritual, on a snare: 4:1, slow attack, sensible release, then deliberately over-compress — threshold down until GR reads 10+ dB and the snare is audibly dead. Now back off in stages, raising the threshold 2 dB at a time, listening at each stage with matched loudness. Somewhere on the way back up, the snare comes alive again. Deliverable: the GR reading where it came alive, and a journal sentence on why finding the wall first made the right setting easier to hear. (This is how working engineers actually dial compressors: through the far ditch and back.)

C7. The limiter proof. (15 min) Insert a limiter (stock is fine) on a duplicate of your drum bus. Set its ceiling at -6 dBFS and watch your meters prove that nothing — no matter the input — crosses the line. Push the input gain up 6 dB and listen to what the "wall" does to the transients as more of the signal slams into it. Then delete it from the bus and write one sentence on where limiters DO belong (two places, per the chapter) and why your 2-bus, mid-mix, is not one of them.

Part D — Synthesis & Critical Thinking

D1. "Dynamics are interest." (25 min) Write two paragraphs. First: defend the claim with three mechanisms from across the book — contrast as arrangement fuel (Chapter 16), programmed micro-dynamics as groove (Chapter 13), and dynamics as depth cues (this chapter's preview of Chapter 24). Second: prosecute the over-compression epidemic honestly — name the four mechanisms that drive it (the makeup-gain bias, the gamified meter, context-free tutorials, fear) and explain the paradox of why crushed mixes sound smaller. End with the one-question ritual the chapter prescribes for every compressor decision.

D2. The compression plan. (30 min) A 12-track session lands on your desk: lead vocal (dynamic singer), double, two backing harmonies, DI bass (played), acoustic guitar (even player), electric guitar (bursty strummer), programmed kick, snare, hats, 808, pad. Write the compression plan: for each track, name the taxonomy job (or "none" — deletions are decisions), the GR zone, and one settings note. Add the bus moves. Defend your two most debatable calls in a sentence each. (Hint from the playbooks: at least three of these tracks should probably get nothing.)

D3. Two patients, one tool. (20 min) Diagnose these symptom reports in writing, naming the failure direction (the rail problem or the vise problem), the responsible control(s), and the 60-second fix: (a) "My vocal is compressed but it still disappears in the verses and barks in the chorus — so I compressed it harder, and now it sounds squeezed AND uneven." (b) "My drums hit so much harder with the compressor on! But every mix I finish feels flat the next morning and I keep blaming the samples." For (a), consider what should have happened before the compressor entirely.

D4. The two schools. (25 min) Some engineers mix into a gentle 2-bus compressor from the first fader move; this book teaches a clean 2-bus through Part V with glue decisions deferred to Chapter 32. Steelman both positions in a paragraph each — what the mix-into-it school gains (hearing balance decisions through the glue they'll ship with), what the clean school gains (honest dynamics, mastering headroom, no baked-in lies) — then state your own policy for your current skill level, with the tradeoff you're accepting named out loud. There are no rules, only consequences; write down which consequences you're choosing.

Part M — Mixed Practice

Interleaving earlier chapters — spaced retrieval is how this sticks.

M1. (Ch 1 + Ch 23) ★ Chapter 1 dissected a kick into click, knock, and boom. Map a compressor's attack control onto that anatomy: roughly which attack times protect the click, which protect the knock, and what a sub-1 ms attack does to all three? Then explain why the boom region punishes fast release specifically, using wavelength math (a 60 Hz wave's cycle lasts about 17 ms).

M2. (Ch 14 + Ch 23) In Chapter 14 you designed envelopes with ADSR; this chapter claims a compressor redraws them. Write the mapping: which compressor control most directly reshapes a sound's attack portion, which reshapes its sustain/release portion, and what's the key difference between designing an envelope on a synth and reshaping one on a recorded sound? (One direction starts from silence and builds; the other starts from a finished envelope and can only subtract — connect that to T2.)

M3. (Ch 21 + Ch 23) Your compressor's threshold is fixed at -20 dBFS, but your tracks hit it at whatever level your gain staging left them. Explain, using Chapter 21's -18 dBFS habit and the clip-gain-first hierarchy: why does sloppy input level make every compressor preset meaningless, and why does a hot input "compress more" without anyone touching the threshold? Then define crest factor and state, in one sentence, what compression does to it.

M4. (Ch 22 + Ch 23) ★ EQ before or after the compressor? Work one concrete case: a vocal with a low-frequency rumble problem and a 300 Hz boxiness problem. Explain why high-passing before the compressor changes how the compressor behaves (the detector reacts to total energy, and low frequencies carry a disproportionate share — rumble the listener barely hears can still pump the gain), and why a presence boost might wait until after. Both orders are legitimate; your answer should name what each order trades.

M5. (Ch 19 + Ch 20 + Ch 23) Demi asks Jaylen for "the compressed stems" of "Static Bloom" so she can practice harmonies against them. Using Chapter 19's stems discipline and this chapter's taxonomy: what processing should reasonably be baked into practice stems, what should be left off and why, and which single Chapter 20 question should Jaylen ask before assuming a compressor is the reason any element isn't sitting? (The disguise quote is the answer's spine.)