Chapter 17 Quiz

Closed book first pass — that's the honest score. Multiple choice, then true/false with justification (the justification carries most of the points), then short answer, then one applied scenario: an over-polished demo on the table and you holding the scalpel. Answers hide under each Verify fold; scoring guide at the end.

Section 1 — Multiple Choice (2 points each)

1. The production polish layer sits between which two passes?

A) Recording and editing B) Arrangement and mixing C) Mixing and mastering D) Sound design and arrangement

Verify **B.** Polish happens after the arrangement is pruned ([Chapter 16](../chapter-16-arrangement/index.md)) and before the mix begins (Part V): it adds support, motion, and event detail to a structure that already works, creating the material the mix will then balance.

2. "Complementary, not identical" means each layer in a committee should:

A) Use the same sample at a different volume B) Contribute a region or property the other layers don't cover C) Be tuned to a different key for richness D) Play a different melody

Verify **B.** Layers are hired for different jobs — different frequency band, envelope segment, or texture. Identical layers sum to louder (the fader could do that) and risk comb-filter trouble; they don't add size. (Different keys and different melodies aren't layering at all — that's harmony and arrangement.)

3. Duplicating a vocal track and playing both copies produces:

A) A thicker, wider vocal B) The chorus effect C) The same sound about 6 dB louder D) Phase cancellation that thins the vocal

Verify **C.** Identical waveforms sum to pure addition — louder, not thicker. There's no disagreement between the copies, so there's no smear; the thickness of real doubling comes entirely from the micro-timing and micro-pitch drift between separate performances.

4. The reason a real double survives mono summing better than a Haas-delayed copy is that:

A) Real doubles are recorded at higher sample rates B) The real double's timing drift moves continuously, so the interference pattern never settles into fixed comb notches C) Real doubles contain no shared frequencies with the lead D) The Haas copy is quieter

Verify **B.** A frozen 10–30 ms offset between identical signals combs the spectrum with stationary cancellations when summed; a performance's drift varies moment to moment, so the interference animates instead of notching — it reads as gentle thickening rather than hollowness.

5. Which configuration is the standard "chorus costume" for a lead vocal?

A) Lead alone, center B) Lead center with a whispered double only C) Two extra takes panned hard left and right around the center lead D) Lead duplicated to both sides with 20 ms of delay on one

Verify **C.** The wide pair: real performances at the edges, lead holding the center column. (D is the Haas fake with its mono bill; B is an intimacy move, not a size move.)

6. As this chapter defines it, automation is:

A) Any plugin that operates without user input B) A recorded or drawn change of a control over time that the project performs on playback C) Quantizing MIDI to the grid D) The DAW's autosave function

Verify **B.** Any fader, knob, send, or switch, captured once and replayed identically — performed with a controller or drawn by hand. Introduced here as an arrangement tool; its mix-depth craft waits for Chapter 27.

7. The boundary between this chapter's automation and Chapter 27's is best stated as:

A) This chapter automates effects; Chapter 27 automates faders B) This chapter automates the song's architecture; Chapter 27 automates the mix's balance C) This chapter draws automation; Chapter 27 performs it D) There is no boundary — it's the same skill

Verify **B.** Filter opens, send swells, FX flips, and close-downs change what the song *is doing* — they'd survive being printed as alternate clips. Vocal rides, the +0.5 dB chorus lift, and breakpoint craft only make sense against a finished balance, so they wait for Chapter 27.

8. A throw is:

A) A reverb that runs on the vocal for the whole song B) A send automated up for one word or hit, so it blooms into space while everything else stays dry C) Deleting a word from the comp D) A delay synced to a dotted eighth

Verify **B.** Eventhood is the definition: the send sits at zero, leaps for one chosen moment, returns. Canonical placement: the last word before a boundary, tail decaying across the new section.

9. The dotted-eighth delay technique works only when the performer:

A) Plays as many notes as possible B) Plays behind the beat C) Leaves gaps for the delay's repeats to answer D) Uses maximum feedback

Verify **C.** The interlocking rhythm lives in the spaces between played notes — The Edge's sparse figures are the reason the delay can act as the second guitarist. Busy playing smears repeats into mush. T2 in action: the part shrinks so the sound grows.

10. The telephone-EQ intro works because:

A) Midrange frequencies are the most pleasant B) The brain measures change against the recent past, so a starved spectrum makes the full-range return feel enormous C) Phone speakers can't reproduce bass anyway D) It reduces CPU load

Verify **B.** [Chapter 4](../../part-01-sound-fundamentals/chapter-04-listening/index.md)'s relative-hearing principle: withholding spectrum resets the listener's reference frame. The trick spends bandwidth now to buy impact later.

11. The ear-candy budget rule of thumb is:

A) One candy moment per bar B) One per 4 bars C) About one per 8 bars, adjusted by genre D) Candy should be continuous to maintain interest

Verify **C.** A direction, not a law — maximalist genres run hotter, folk runs near zero — but the economics are universal: candy's value is surprise, and surprise devalues on contact. When in doubt, halve it.

12. Ear candy should be placed:

A) Simultaneously with the lead's biggest notes, for reinforcement B) In the lead's gaps — phrase endings and turnarounds — as answers C) Only in the intro D) On every downbeat

Verify **B.** Call-and-response: the lead calls, candy answers, they never talk at once. Candy overlapping the lead is competition, and [Chapter 16](../chapter-16-arrangement/index.md) already ruled on competitions.

13. Which transition tool does the chapter call the best one?

A) The riser B) The impact C) The reverse cymbal D) Silence — the bar-before mute

Verify **D.** Full-spectrum, free, and brave: whatever follows silence lands bigger. The chapter's craft addition: choose the *kind* of cut (full mute, everything-but-vocal, or low-end-only) and protect it with hygiene fades.

14. In drum augmentation, the first two technical checks after placing a sample under a live snare are:

A) Reverb and compression B) Transient alignment and the polarity-flip test C) Quantization and normalization D) Panning and EQ

Verify **B.** Align transient-to-transient (by sight or trigger software), then flip polarity and keep the fatter sum — the [Chapter 12](../../part-03-recording/chapter-12-recording-instruments/index.md) test, forever. Tuning and job-splitting follow.

15. The four cheap energy levers inside a section are:

A) Volume, EQ, compression, reverb B) Subdivision, filter, width, register C) Tempo, key, lyrics, melody D) Layers, doubles, candy, risers

Verify **B.** Densify the clock, open the spectrum, expand the image, lift the ceiling — all without adding a single new musical part. The expensive fifth lever (another part) is taxed by the frequency budget and comes last.

Section 2 — True/False with Justification (3 points each: 1 for the call, 2 for the why)

16. True or false: the more doubles you stack on a vocal, the thicker it gets, indefinitely.

Verify **False.** Diminishing returns with compounding costs: each double after the first pair buys less thickness and more blur. The blur test is concrete — words harder to make out, transients gone soft — and the fix is fewer voices, not tighter edits (over-editing manufactures copies, the other failure).

17. True or false: a layer that plays through the entire song thickens the track more than a layer that only plays in choruses.

Verify **False.** The everywhere-layer resets the baseline — it stops being enrichment and becomes the sound, leaving you where you started but muddier and with a heavier frequency-budget tax. Layers are costumes: section-dependence is what makes the thickness register as an event.

18. True or false: automation belongs entirely to mixing, so touching it in Part IV is premature.

Verify **False.** Automation is introduced here precisely because some of its uses are arrangement: filter opens, send swells, FX flips, and close-downs change what the song is doing, not how the balance sits. The mix-depth craft — rides, breakpoints, modes — correctly waits for Chapter 27.

19. True or false: a good ear-candy sound should impress when solo'd.

Verify **False.** The felt-not-heard test runs the other way: good candy sounds trivial solo'd and leaves the section subtly poorer when muted. Candy that demands attention in context has outgrown the category — it's auditioning to be a hook and belongs in [Chapter 16](../chapter-16-arrangement/index.md)'s courtroom.

20. True or false: the strongest single energy move inside a section is often a removal.

Verify **True.** The subtraction trick: drop the kick and sub for the last bars and the next downbeat detonates — energy is controlled disagreement with expectation, and silence is the cheapest disagreement available. A removal also counts as a "change" for the add schedule.

Section 3 — Short Answer (5 points each)

21. State the six-step layering recipe in order, and name which earlier chapter supplied each of steps 2 through 4.

Verify (1) Assign jobs in band vocabulary before browsing; (2) high-pass every layer above the lowest — [Chapter 13](../../part-03-recording/chapter-13-programming-beats/index.md)'s rule generalized; (3) polarity-flip test every pair — [Chapter 12](../../part-03-recording/chapter-12-recording-instruments/index.md); (4) tune pitched layers to the key — [Chapter 13](../../part-03-recording/chapter-13-programming-beats/index.md)'s tune-what-sings; (5) balance in context at matched loudness with the mute-A/B question "bigger, or louder?"; (6) make layers section-dependent — costumes, not skin.

22. Explain why a whispered double works — what does a whisper carry, what does it lack, and what does that combination add to a synth lead or a soft vocal?

Verify A whisper carries almost no pitch — it's nearly all consonants, breath, and noise — so it adds articulation, intimacy, and human texture without thickening or detuning the tone underneath. Tucked well under a synth lead it donates the breath and consonant detail the machine lacks (Jaylen's move two); under a sung vocal it adds skin and closeness. It's felt, not heard, and it costs one quiet take.

23. Your bandmate wants the full riser-silence-impact stack at all seven section boundaries. Give the two-part argument against, using the inventory's economics and Chapter 16's contrast engine.

Verify First, devaluation: transition tools are candy at the structural scale — the ear learns the riser in two exposures, and by the fourth identical stack the fireworks read as a tic, not an event. Second, redundancy: [Chapter 16](../chapter-16-arrangement/index.md) built the sections to contrast, and when adjacent sections genuinely differ, the doorway is already audible — decorating every doorway delays the payoff and signals distrust of the arrangement. Ration the stack for the biggest seam; let bare cuts (with hygiene fades) carry the rest.

24. Define the restraint log's four columns and its three verdicts, and explain the practical difference between cut and park using the chapter's example.

Verify Columns: IDEA → WHAT IT PROMISED → WHAT IT COST → VERDICT. Verdicts: keep (survived the fresh-ears mute test), cut (failed — logged with the reason, because reasons compound into judgment), park (right idea, wrong chapter — kept alive with a destination). Example: Jaylen's pad-ducking instinct was parked, not cut, because the impulse is correct and its tools are [Chapter 28](../../part-06-advanced-mixing/chapter-28-advanced-mix-techniques/index.md)'s sidechain — park preserves the idea; cut retires it.

Section 4 — Applied Scenario (10 points)

25. A friend sends you their "fully produced" demo with the note: "I did everything the tutorials said — layers, candy, transitions, automation. Why does it sound smaller and more annoying than my rough version?" You listen: the chorus pad is six near-identical patches stacked, all full-range, all song long. There's a vocal chop or sweep in literally every bar. The lead vocal is one take, duplicated, with 25 ms of delay on the right copy, panned wide. Every one of the eight section boundaries gets the same riser-impact combo. Nothing else moves — no filter arcs, no swells, no throws. Diagnose at least five distinct problems in priority order, prescribe the concrete fix for each (with the chapter's tests and numbers where they exist), and close with the one-sentence principle you'd send back with the session.

Verify Strong answers hit most of these, roughly in this order. **(1) The pad committee is a crowd** — six near-identical full-range layers is the clone/crowd failure: louder and muddier, not bigger, with the low-mids pile taxing the whole budget. Fix: fire the committee, keep one core, hire complementary specialists only (sub below, air above, each high-passed above the lowest), then mute-A/B at matched loudness — "bigger, or louder?" **(2) The everywhere-layer** — whatever survives must become section-dependent: air layer in choruses only, verses back to the naked core, so the thickness registers as an event. **(3) The fake double** — a duplicated lead with a frozen 25 ms offset is the Haas trick on the most important element in the song: it'll comb hollow on every mono system. Fix: real doubles — re-sing it (wide pair hard-panned for choruses, tight tuck for emphasis), and mono-check; if the fake must stay, drop the delayed side several dB and check mono today. **(4) Candy hyperinflation** — one per bar has devalued surprise to wallpaper. Fix: ration toward one per 8 bars, place candies in the lead's gaps (call-and-response), rotate the kinds, then run the felt-not-heard mute test on each survivor. **(5) Transition monoculture** — the same riser-impact stack at all eight boundaries is structural wallpaper. Fix: keep the full stack for the single biggest seam, replace the strongest-contrast boundaries with bare cuts or silence, vary the rest (downlifter out of a peak, one fill). **(6) No motion** — nothing travels inside sections; it's polish without verbs. Fix: the minimum viable motion pass — one filter open into the biggest peak, one close-down into the deepest valley, one send swell at the main boundary. Closing principle, in any words that mean: *polish is teaspoons — every move faces the mute test, and the cut list is the style.*

Scoring

Section Items Points
1 — Multiple choice 15 × 2 30
2 — True/False + justification 5 × 3 15
3 — Short answer 4 × 5 20
4 — Applied scenario 1 × 10 10
Total 75
Score Reading
68–75 Polished — and you know why. Go run C7's restraint pass and the Project Checkpoint; your track is waiting.
56–67 Solid. Re-check whichever family leaked: the support tools (questions 1–5, 16–17) or the motion and economy tools (6–13, 19–20) — then redo exercise C4 or C6.
42–55 The vocabulary is landing but the dosage logic isn't automatic yet. Re-read "Ear Candy Economics" and "The Polish Pass in Practice" with your session open, redo C2 and C6, retake.
Below 42 Reread with the records playing — this chapter only sticks when you can hear the layers being counted. Exercise B1 is the best second pass; then come back.