Chapter 16 Exercises

Arrangement is the cheapest skill in this book to practice — you don't need a new plugin, a better mic, or a treated room. You need a mute button, a pencil, and the willingness to ask every part the rude question: what exactly are you doing here? Work Part A first to load the vocabulary, then do the Listening Lab before the DAW Lab — you can't budget density you've never counted on a real record. And treat C1 as the chapter's actual exam: the closing pages promised you the mute test, and here it is, written out as a workflow you'll run on every song you make from now on.

Difficulty legend

Exercises marked ★ have worked answers in the back of the book (Answers to Selected Exercises).

Part A: Conceptual Understanding

A1. Three crafts, three sentences. ⭐ (5 min) Write one sentence each defining composition, arrangement, and mixing so that the three can't be confused — what question does each one answer about a song? Then explain, in two or three sentences, why this chapter's title claims that where a part sits matters more than how good the part is. Use the Wren & Hollow session as your evidence: seven keepers, one drowning song.

A2. The section interview. ★ ⭐ (10 min) Give a one-phrase job description for each of the six classic sections: intro, verse, pre-chorus, chorus, bridge, outro. Then answer the follow-up: which section's entire job is to make another section's job easier, and what happens to the song when that section over-delivers and steals the payoff for itself?

A3. Energy is not volume. ⭐⭐ (15 min) List four levers that raise a section's perceived energy without anyone touching a fader. For each lever, name one record or one moment in a record where you've heard it work. Then explain why the fader is the weakest energy lever of the bunch — what does the ear do with "louder" after about thirty seconds, and what does it never stop doing with "different"?

A4. The frequency budget ledger. ★ ⭐⭐ (15 min) Two producers harmonize the same verse. Producer one uses three sustaining parts that all live in the low-mids: a synth pad, a strummed acoustic, and a piano part heavy on the left hand. Producer two delivers the same chords as one sustained pad, one palm-muted plucky part, and one high arpeggio up near the air band. Using Chapter 4's six-band vocabulary, explain which verse costs more against the frequency budget and why — then use this example to explain the chapter's claim that the arrangement is the first mix.

A5. Drops are made of dropouts. ⭐⭐ (10 min) Explain the contrast engine in your own words: why does the biggest-feeling moment in a song almost always arrive immediately after the emptiest one? Name two records where you can hear this — one from your genre, one from a genre you don't make — and identify exactly what disappears in the bar or two before the payoff. Finish with the principle underneath it: what does it mean that hearing is relative?

A6. Intro economics. ⭐ (10 min) An intro has two jobs the chapter named. State them. Then make the honest counter-argument: describe one record you love whose long intro is its identity, and name the cost that record accepted by taking that road. (No statistics needed — platforms don't publish your song's fate in advance. Reasons only.)

A7. The 8-bar rule, obeyed and disobeyed. ⭐⭐ (10 min) State the 8-bar rule in one sentence. List six changes that satisfy it without adding a single new part — every one of them should cost nothing but a decision. Then name the failure mode of over-obeying the rule: what does an arrangement feel like when something dramatic changes every eight bars without fail, and which genres deliberately bend the rule toward hypnosis instead?

A8. Both right. ★ ⭐⭐ (10 min) Demi's claim: "the parts are good." Theo's claim: "the song is sick." Using the language of density and the frequency budget, explain how both claims are true at the same time — and then explain why no volume, pan, or EQ move could resolve the contradiction. Where does the fix have to happen, and in which domain (level, spectrum, or time) does that fix operate?

Part B: 🎧 Listening Lab

B1. Draw the curve. ⭐⭐ (20 min) Pick one famous record you know well — Adele's "Rolling in the Deep," Kelly Clarkson's "Since U Been Gone," and The Weeknd's "Blinding Lights" all work, or substitute a structurally clear hit from your own genre. On paper: sections across the bottom, energy from 1–10 up the side. Plot every section by ear, no skipping around — listen front to back. Then mark the single biggest jump and the single biggest drop, and write one sentence each naming the move that creates them. (Hint: the biggest jump is usually paid for by the drop right before it.)

B2. The part census. ★ ⭐⭐ (20 min) Take one song you've heard a hundred times and count the active parts in every section — a table with sections down the side and your count beside each. Group families if you must (a stack of background vocals can count as one). Michael Jackson's "Billie Jean" famously hires its band one player at a time across the intro, and Lorde's "Royals" keeps the whole payroll countable on two hands — both make great first attempts, and case study 1 in this chapter does the full version of this exercise on the latter. Where's your song's maximum? Its minimum? Does the chorus win by headcount, or by some other lever from A3?

B3. The dropout hunt. ⭐⭐⭐ (30–45 min) Find five bar-before-the-payoff dropouts across at least three genres, and log each one: song, where it happens, what mutes, and what the silence buys. Starters that are easy to verify: the full-band stop under the title line of Walk the Moon's "Shut Up and Dance," the "don't believe me, just watch" stops in "Uptown Funk," the way Billie Eilish's "bad guy" empties out around its title before the beat flips — and almost any festival-era EDM track, where the pre-drop silence is practically a genre signature. Your goal is pattern recognition: by the fifth one, you should hear the vacuum coming before it arrives.

B4. Blind section ID. ⭐⭐ (15 min) Shuffle your library and drop in at a random point mid-song — or have a friend do the dropping. Within ten seconds, name the section you've landed in without using the lyrics. You're allowed only function cues: how dense is it, where does the melody sit, what are the drums doing, how wide does it feel. Ten rounds. Log which cue gave it away each time. This is Chapter 4's structured listening aimed at form — the energy pass, weaponized.

B5. Reference deconstruction (your compass). ⭐⭐⭐ (45 min) Choose one reference track in your genre and build its complete arrangement grid by ear: every section, its length in bars (count along — pickups will fight you; the shape matters more than perfection), what enters and exits at every boundary, and which transition tool carries each seam — fill, riser, mute, or cold cut. Keep this grid. It becomes the measuring stick for C2 and C4, and per the chapter's checkpoint, the gap between this grid and your own song's grid is your to-do list.

B6. The whisper test. ⭐ (10 min) Play a professional track at conversation volume or lower — quiet enough that detail dies. Notice what survives: the sections still announce themselves, the lifts still lift. Now play your own work-in-progress at the same level and write down what disappears. Detail that vanishes at a whisper is mix. Structure that vanishes at a whisper was never there.

Part C: 🎛️ DAW Lab

C1. The mute test — the chapter's real exam. ★ ⭐⭐⭐⭐ (90 min–2 hours, plus a next-day listen) Run the full arrangement-by-subtraction pass on your own track. This is the formal workflow; follow it in order, and don't negotiate with step 7.

1. Save As a new session version — name it something like "SongName_MUTE-PASS." You never prune the only copy; this is Chapter 6's non-destructive thinking applied to arrangement.
2. Create the BENCH at the bottom of your session: a muted folder or track group, fader irrelevant, output optional. (Appendix E has your DAW's name for folders and track grouping.)
3. Draw the song map with bar counts on actual paper — intro, verses, pre-choruses, choruses, bridge, outro, whatever your song has. Pen beats pixels here; you'll be writing on it all night.
4. Loop section one. Mute every track. Silence is the starting roster. Nothing has a seat yet — not even the parts you love.
5. Unmute the anchor — usually the lead vocal or topline melody — plus the one accompaniment part that defines the harmony. That's the minimum viable section.
6. Add parts back one at a time. Before each unmute sticks, the part answers two questions out loud: What is your job in this section? and Is another part already doing that job? No nameable job, or a duplicated job — it stays muted. Good has nothing to do with it.
7. When the section sounds complete, stop. Even if good parts remain muted. Especially if good parts remain muted. Complete is the finish line; full is the failure state.
8. Write the cut sheet for that section: keeps, benches, and one line of why per decision. Future-you will not remember tonight's reasoning.
9. Next section, return to step 4. Every section auditions everyone from zero. A part that earned its chorus seat does not inherit the verse — that assumption is how Wren & Hollow ended up seven banjos deep.
10. Bench the unemployed. Anything that never got unmuted moves to the BENCH folder. Bench, don't delete: benched parts audition again for cameo roles when Chapter 17's polish pass arrives.
11. Full playback, low volume, eyes off the screen. Does the energy curve you drew actually happen in the audio? Mark every flat spot on the paper map.
12. Sleep on it. Tomorrow: one revision pass only — fix the flat spots, second-guess at most two benchings — then bounce "ARR-V2" and A/B it against your pre-prune bounce at matched loudness.

Deliverables: the paper map, the cut sheet, the BENCH folder, and both bounces. Genre adaptations from the chapter checkpoint apply: hip-hop, schedule a beat-mute under a key line and drop the sample to half its layers somewhere; band arrangements, decide who lays out per section; electronic, ration yourself to two risers per drop and spend one full beat of true silence at your highest tension point.

C2. Build the grid. ⭐⭐ (20 min) Put your song map into the session: a named, colored marker or region for every section, bar counts in the names ("CHORUS 2 — 8"). Color verses one family, choruses another, so the timeline reads as a map at arm's length. Print or screenshot it next to your B5 reference grid. Two songs, two grids, one gap analysis.

C3. The 8-bar variation drill. ⭐⭐ (30 min) Take any 8-bar loop — yours or a quick sketch — and extend it to 32 bars as four consecutive 8-bar blocks. Each block changes exactly one thing from the block before: something leaves, something swaps registers, the hat pattern turns over, a fill announces the seam. No new parts allowed; decisions only. Bounce it, listen tomorrow, and ask: do the seams sound intentional, and does 32 bars of one loop now hold attention?

C4. Skeleton rebuild by ear. ⭐⭐⭐ (45–60 min) Using your B5 grid, rebuild your reference track's skeleton in your DAW with dummy parts — any pad, any drum kit, blocked-in regions. You're not copying notes; you're copying shape: section lengths, density counts per section, where the dropouts land, where the peak sits. Then play your own melody or vocal idea over the borrowed skeleton. Structure isn't copyrightable taste — it's craft, and stealing it this way is how arrangers have always learned.

C5. Contrast surgery. ★ ⭐⭐ (20 min) Loop your verse and chorus back to back. Rule: you may not touch the chorus. Remove parts from the verse — one at a time, listening across the boundary each time — until the chorus visibly lifts. Count what's left in the verse when it happens. You made the chorus bigger and you only ever muted things. Write that sentence somewhere you'll see it.

C6. The transition bake-off. ⭐⭐ (30 min) Take the last bar before your chorus and build four versions: (a) untouched, (b) a drum fill, (c) a full-band mute for that bar, (d) the riser you designed in Chapter 14. Bounce all four, shuffle them tomorrow so you don't know which is which, and pick by ear. Write one sentence on why the winner won. There's no universally right answer — there's the right answer for this doorway.

C7. Intro on a budget. ⭐⭐⭐ (30–45 min) Build two intros for your track: a 4-bar version that gets to your anchor almost immediately, and your current intro (if your current intro is already four bars or fewer, build a 16-bar version for science). Tomorrow, cold, play both as if you'd never heard the song. Notice the exact second your attention reaches for the skip button. Nobody's watching; honesty is the whole exercise.

C8. The double-chorus build. ⭐⭐⭐ (45 min) Make your final chorus beat chorus one using arrangement moves only — faders are off-limits. Three levers, pick at least two: add exactly one element that's never appeared before, lift something an octave, or remove a core element for the first half and slam it back for the second. Verify the win two ways: by part count, and by the back of your neck.

Part D: Synthesis & Critical Thinking

D1. No knob for this. ⭐⭐⭐ (30 min, writing) Theo spent two nights trying to make seven banjos "clearer" with volume, pans, and EQ, and every move helped one part by hurting another. Write the 150–250 word memo to your future self explaining why density problems can't be solved downstream: what the frequency budget says about seven parts sharing one spectrum, and why the only fix lives in the time domain — when things play, not how they're processed. End the memo with a prediction about what you'll discover in Part V, where Chapter 20 makes this idea a threshold concept.

D2. Skeleton transplant. ⭐⭐⭐ (30–45 min) On paper, redraw your song into a neighboring genre's skeleton: your folk ballad as an EDM build-drop cycle, your trap beat as a tight pop verse-pre-chorus form, your rock song as a hip-hop loop-plus-events arrangement. What survives the transplant? What dies on the table? Write a paragraph on what the breakage teaches you about your home genre's unspoken contract with its listeners — Chapter 18 turns this intuition into a full toolkit.

D3. The skip-button debate. ⭐⭐ (20 min, writing or argument with a friend) Streaming-era pressure says front-load the song; some records refuse and are loved for refusing. Argue both sides honestly: the case for surrendering to the skip window, and the case for the long intro as a filter that selects your real audience. Then rule on your current song specifically — not on principle, on this song. Where's its line, and what would change your mind?

D4. Rules for cutting together. ⭐⭐⭐ (30 min) Design the ground rules for running a mute-test session with a collaborator whose parts are on the chopping block — rules that protect the song and the relationship. Decide at minimum: who holds the mouse, what happens to cut material, how ties get broken, and what gets said out loud before any muting starts. Write them as a numbered list you'd actually hand someone. Then read case study 2 in this chapter and steal anything Wren & Hollow did better.

Part M: Mixed Practice

M1. Budget audit by band. ★ ⭐⭐ (Chapter 4 + this chapter; 20 min) Build a table: your track's parts down the side, Chapter 4's six bands across the top — sub, low, low-mid, mid, high-mid, air. Mark where each part actually lives (its energy, not its highest overtone). Now scan per section and circle the most crowded band. For every crowd, prescribe one arrangement fix — a part leaves the section, moves register, or trades sustain for pluck — before you ever reach for an EQ. Chapter 22 will inherit whatever this audit doesn't solve, and it will thank you for a short list.

M2. Mute by bus. ⭐⭐ (Chapter 6 + this chapter; 20 min) If your routing still matches the template you built in Chapter 6, the mute test has a fast first gear: run C1's section loop muting whole buses before touching single tracks — kill DRUMS entirely, kill SYNTHS entirely, hear what the section misses at the group level, then go fine-grained inside the buses that earned their seat. If your buses don't map onto musical groups anymore, that's the real finding: fix the routing first, then run the test.

M3. Hat language per section. ⭐⭐⭐ (Chapter 13 + this chapter; 30–45 min) Program three different hi-hat patterns for verse, pre-chorus, and chorus, so that rhythmic density climbs with your energy curve: maybe closed eighths in the verse, offbeat opens joining in the pre, sixteenths with Chapter 13's velocity shaping in the chorus. Then mute every other instrument and listen to the hats alone, front to back. If the kit narrates the song's shape by itself, your curve has a spine.

M4. The vanishing section. ⭐⭐ (Chapter 15 + this chapter; 20 min) Pick the section you're least sure about and cut it out entirely with Chapter 15's edit craft — slice at transients, crossfade the seam so nothing clicks. Listen tomorrow without looking at the timeline. If you miss the section, restore it with a clear conscience. If you don't — it was already gone, and the edit told you what the mute button couldn't. Update the song map either way.

M5. Deploy the inventory. ⭐⭐⭐ (Chapters 13 and 14 + this chapter; 30 min) Take your transition inventory — Chapter 13's fills, Chapter 14's riser and impact — and place them on the song map with a hard budget: at most one tool per transition, and at least one transition gets silence instead of any of them. Bounce the result and run one verification pass: no transition tool may land on top of a vocal entrance. The doorway frames the singer; it doesn't compete with her.

Selected answers — the exercises marked ★ — appear in the appendix Answers to Selected Exercises at the back of the book. Do the work before you peek; an answer you didn't sweat for is a fact, not a skill. And if you only have one evening this week, spend it on C1. The mute test is the one exercise in this chapter that pays you back on every song you ever make.