Chapter 35 Exercises

This set is different from every set before it, and honestly, that's a relief for your ears: after thirty-four chapters of critical listening at full attention, most of this chapter's work happens at a desk with a spreadsheet open. Don't mistake desk work for optional work — the artifacts you build here (the metadata card, the ISRC ledger, the four-week calendar) are the actual operating documents of your release, and Chapters 37 and 39 assume they exist. Work Part A first to load the concepts, then do the Listening Lab in one sitting — it's lighter than usual and treats your master as a set of deliverables rather than a mix to judge. Part C is the spine of the chapter: C1 and C4 are mandatory if you do nothing else. One standing reminder: every platform mechanic referenced below is "as of this writing" — when an exercise asks you to check a portal or policy, checking the current version is part of the exercise. Exercises marked ★ have worked answers in the back of the book (Answers to Selected Exercises).

Part A: Conceptual Understanding

A1. The gate, explained to a bandmate. ★ (10 min) Your drummer asks why you can't "put the song on Spotify yourself like YouTube." In five sentences or fewer, explain what a distributor is, why the DSPs route supply through distributors instead of taking direct uploads, and what travels in each direction through the gate. Bonus point for including the one thing the distributor handles that most musicians forget exists until their first statement arrives.

A2. The pricing-model audit. ★ (15 min) Three artists: (a) a producer releasing a single every six weeks who earns modest but real streaming income; (b) a band releasing one EP and then touring for two years without releasing; (c) a brand-new artist with zero audience releasing their first two songs. For each, choose flat-fee, percentage, or hybrid, defend the choice in two sentences using the chapter's tradeoff table, and name the "quiet catch" each artist most needs to read in the terms before signing up.

A3. License plate vs. shipping container. (10 min) Define ISRC and UPC in one sentence each, then apply: an artist releases a single, then an EP containing that single plus four new tracks, then a remix of the single. Count the total ISRCs and UPCs across all three releases, and explain what specifically would break — in plays, placements, and payments — if the EP's copy of the single carried a different ISRC from the original.

A4. The credit that isn't a contract. ★ (10 min) Explain the difference between the writer-credit field in an upload form and a split sheet: what each document establishes, who reads each, and what specifically remains undefined if you complete the upload form perfectly but never write the split sheet. End with one sentence on when the split sheet should be written relative to the release (Chapter 36 will hold you to your answer).

A5. Why day-of uploads die. (15 min) A friend finished their master Tuesday and plans to release Friday. List four specific capabilities their release structurally forfeits, mark each as recoverable or unrecoverable after release day, and write the two-sentence text message you'd actually send them tonight. (Tone matters — you were this person once. Chapter 19's feedback rules apply to release advice too.)

A6. The pool, not the price. ★ (15 min) In plain words a non-musician would follow, explain the pro-rata pool model: where the money comes from, how a month's pool gets divided, and why "Spotify pays $X per stream" is a back-calculated average rather than a rate. Then name two structural reasons the same thousand streams can pay different amounts in different months or countries, and one reason the per-stream figure you compute from your own statement won't match anyone else's.

A7. The two pipes. (10 min) Draw the two-pipe money diagram from memory — one stream splitting into the recording side and the composition side — labeling who collects each side for an independent artist. Then answer: which pipe does your distributor handle, which pipe "fails silently" if ignored, and what does failing silently mean in dollars-and-paperwork terms?

A8. Friday, defended and attacked. (10 min) Make the case for releasing on Friday in two sentences (history and platform rhythm), then make the case against it in two sentences for a specific artist profile you describe. Conclude with the chapter's doctrine in one line. This is a T3 drill: there's no correct day, only consequences.

Part B: 🎧 Listening Lab

Lighter than usual, and deliberately so — these are deliverable checks, not mix judgments. You're listening as the quality-control department of your label-of-one. Use your calibrated level from Chapter 8, but these passes are about catching errors, not making art.

B1. The instrumental top-to-tail. ★ (15 min) Play your instrumental master start to finish, eyes closed, at working level. You're hunting for one specific class of ghost: vocal remnants that survived the mute — printed delay throws carrying a word (Chapter 27's automation events live on FX returns, not the vocal track), reverb tails with a vowel in them, a harmony tucked into a riser. Log every timestamp where you hear a human voice or its shadow. If your full pass comes up clean, run the confirmation sweep: solo each FX return in the mix session for the chorus and bridge sections only — ghosts hide loudest where the throws were. If you find any, diagnose where in the routing the ghost lives (insert vs. send — Chapter 6's distinction earning its keep) and re-print. A sync supervisor will do this exact listen someday; beat them to it.

B2. The clean-edit seam check. (15 min) If your track needed a clean edit, listen to each edited moment three ways: in context at full level, soloed at low level, and on earbuds. You're checking Chapter 15 craft under commercial pressure: are the seams crossfaded or clicking? Does the edit preserve the groove (a reversed syllable or dropped word should land on the pocket, not punch a hole in it)? Would a listener who doesn't know the original notice anything? If you have no clean edit, do this exercise on any commercially released clean version in your genre and log how they solved each masked word — mute, reverse, scratch, re-sing — and which solution was least audible.

B3. The sequence sanity pass. (20 min) Play your EP (or your single plus its B-sides) in release order, straight through, doing nothing else. You verified gaps and level steps in Chapter 32; tonight you're listening one last time as a stranger: does track two's entrance feel intentional after track one's tail? Does any transition make you reach for the volume? Note any gap that feels wrong by more than "hmm" — then check whether the issue is the gap length or the level step (they masquerade as each other). This is the last listen before these numbers are frozen into deliverables.

B4. The encode A/B. ★ (15 min) Take any released track you know intimately (your Chapter 4 reference playlist has candidates) and A/B the streaming version against a lossless copy if you have one — matched loudness, as always. Listen for what lossy encoding touches: cymbal sheen and air (Chapter 1's frequency anatomy), reverb-tail texture, stereo width stability. Then play your own master and ask: is anything in my top octave or my widest elements fragile in the same way? You can't control the encoder, but Chapter 32's -1 dBTP ceiling was your insurance against its worst behavior — this listen is why that spec existed.

B5. The sandwich test. (20 min) Build a three-track playlist on a streaming platform with normalization on: reference track, your master (via a private upload or local-file integration if your platform supports it — otherwise simulate in your DAW with masters level-matched per Chapter 33), then a second reference. Listen straight through twice. T4's bluntest question: when your track arrives between two records in its genre, does anything flinch — tonal balance, width, loudness feel, energy? You did this with mixes in Chapter 30; doing it with the final master in a playlist context is the closest rehearsal of release day that exists.

Part C: 🎛️ DAW Lab / Desk Lab

C1. Build the metadata card. ★ (45 min — mandatory) Reproduce the chapter's metadata anatomy card as a document (spreadsheet or plain text) and fill every field for your single: title, version, primary/featured artists (exact strings), all writers in legal names, producer credit, genre per the reference-playlist test, language, explicit flag, ℗/© lines, label/imprint name, intended release date. Leave the ISRC/UPC rows present but blank, marked "assigned at upload." Then the hard part: have one collaborator or trusted ear proofread it against the actual recording — Chapter 19's outside-eyes protocol applied to paperwork. Save it next to your masters folder with a version number. This document is your release.

C2. The alt-versions print session. (60 min) If Chapter 32's checkpoint alts don't exist yet, print them now: open the mastering session, mute the vocal bus (not the FX returns — then listen for B1's ghosts and decide what else must mute), and print the instrumental through the identical chain. Print a clean edit only if the lyric needs one, building the masks with Chapter 15 edit craft before the master chain. Re-verify each alt's true peak (the limiter behaves differently with the vocal energy gone — re-check, don't assume). Name files per your Chapter 19 convention with the version in the filename. Log each version as a row in C3's ledger.

C3. Start the ISRC/UPC ledger. ★ (20 min) Create the spreadsheet this chapter made non-negotiable: columns for title, version, ISRC, UPC (release-level), release date, distributor, and notes. Enter every recording you've ever released anywhere (codes retrievable from your distributor dashboard if applicable) and rows-in-waiting for your current single's versions. Add one documentation row at the top recording your canonical artist-name string and imprint name — character for character. Store it with your session backups under the 3-2-1 rule. Ten years from now this file is your catalog's birth certificate.

C4. The four-week calendar build. ★ (45 min — mandatory) Take a real calendar and build your runway backward from a Friday at least four weeks and a day from today: upload/metadata-final date, pitch submission, profile claiming, pre-save launch, the Week -2 verification pass, the Week -1 staging, release-day checklist execution, and the Week +1 follow-through. Every entry gets a day, not a week. Mark the gates (▼) from the chapter's diagram as non-negotiable and the campaign items as placeholders for Chapter 37 to fill. Then add the engineer's margin: one buffer day before every gate, because distributor review queues, artwork rejections, and collaborator response times are all variables you don't control — a calendar with zero slack is a calendar that fails on first contact. If you're not releasing yet, build the template calendar anyway with relative dates ("R minus 28") — Chapter 39 will instantiate it.

C5. The mock upload walkthrough. (45 min) Create an account with a distributor (most allow account creation free of charge regardless of pricing model — verify before entering payment details) and walk the upload form to the final confirmation screen without submitting, or use a distributor's documented walkthrough if you'd rather not create an account. Screenshot every page. At each field, fill from your C1 card and note any field the card didn't anticipate. Log: where ISRCs get assigned, where featured artists map, where the pitch tools are referenced, what the stated review time is. You're scouting the route before moving day.

C6. The artist-name collision search. (30 min) Search your artist name — exact and close variants — on every major DSP, two social platforms, and a general web search. Log every collision: active artists, dormant pages, near-misses that would split your search results. Verdict: clear, distinguishable-with-care, or change-it-now. If you find a collision, write the three-sentence plan (distinct name? middle initial? mapping vigilance?) before your first upload makes the decision for you. Do the same search for your imprint name from C1.

C7. The artist-profile audit. (40 min) Claim your artist profiles where you have live releases (or bookmark the claiming portals if you don't yet). Then run the T4 storefront audit: pick three artists one rung above you in your lane and grade their pages against yours on bio, photo intentionality at thumbnail size, catalog presentation, and link hygiene. Write your gap list — three concrete fixes — and execute at least one tonight.

C8. The distributor decision memo. (30 min) Write a half-page memo choosing your distributor category (not brand): your honest releases-per-year, your revenue expectation (calibrated by the sidebar — low), your exit-friendliness requirements (ISRC portability), and which add-ons you're deliberately declining until Chapter 36. End with the category and a one-line review date ("revisit after release three"). Appendix I supplies current names when you're ready to act on the memo.

Part D: Synthesis

D1. The waterfall plan. (30 min) You have five mastered tracks. Design the full waterfall: which track leads and why (your A&R call — defend it in two sentences using your Chapter 18 genre audit), the release order and spacing, and the ISRC/UPC ledger for the entire sequence (every code-event across all releases). Annotate where each release inherits the previous one's momentum and what specifically carries it. Close with the single tradeoff you're accepting versus dropping the EP whole.

D2. The label org chart of you. (25 min) Draw the org chart of a traditional label's release team — A&R, production, mastering/deliverables, label copy/metadata, distribution ops, rights admin, marketing — and under each department, write which chapters of this book trained you for that desk and which artifact from your own project proves it (the manifesto, the masters, the metadata card, the calendar...). Note the two desks still unstaffed until Chapters 36 and 37. Keep this — it's the most honest résumé you own.

D3. The post-mortem letter. ★ (20 min) A friend uploaded their single yesterday for release tomorrow, pitched nothing, claimed nothing, and is now reading about pre-saves with a sinking feeling. Write the honest one-page letter: what's genuinely lost (and that it's gone), what's still winnable this week (profiles, links, the warm audience, the Week +1 follower-feed cycle), and the calendar discipline for their next release. The skill being drilled: triage under irreversibility — Chapter 30's ranked-fix-list thinking applied to a release.

D4. The edge-case memo. (20 min) Your track contains an uncleared sample (Chapter 13's legality flag, about to come due in Chapter 36). Your distributor's form asks you to warrant you own all rights, worldwide. Write the memo to yourself: what the warranty checkbox legally claims, why "release it only in territories where it's probably fine" is not the workaround it sounds like, and what has to happen before this track enters the pipe at all. One paragraph, ending with the chapter number where the answer lives.

Part M: Mixed Practice

Interleaving, as always — retrieval from earlier chapters keeps the whole pipeline warm.

M1. (Chapter 21) Your distributor rejects your upload for clipping. Your DAW's 2-bus meter never touched red. Reconcile these facts using the difference between sample peaks and true peaks (Chapter 32's sidebar) and the gain-staging doctrine, then state the print-spec fix in one sentence.

M2. (Chapter 24) On the B1 instrumental listen you found a vocal ghost in a reverb tail. Explain, using send/return architecture, why muting the vocal track didn't silence the vocal's reverb — and name the two clean solutions (mute order vs. re-print) with the tradeoff of each.

M3. (Chapter 13) Your clean edit reverses one word on the backbeat. The edit is inaudible as a seam but the bar now feels wrong. Using groove anatomy, diagnose what the reversal likely broke (which layer of the groove carried the word's transient?) and propose a fix that preserves both the mask and the pocket.

M4. (Chapter 33) Your master measures -9.8 LUFS integrated. A forum tells you to "re-master to -14 for Spotify." Using the normalization threshold from Chapter 33, explain what the platform will actually do with your -9.8 master, what re-mastering to -14 would and wouldn't change, and the one question (about dynamics, not numbers) that actually decides whether your master is right.

M5. (Chapter 16) Your day-one analytics show strong saves but a completion cliff: a visible share of first-time listeners bail in the first twenty seconds. Before you blame the genre tag or the pitch, make the Chapter 16 argument: which arrangement decision governs the skip window, what the energy curve of your intro promises (or fails to promise), and why this is a problem no distribution mechanic in Chapter 35 can fix. Name the intro-economics tradeoff honestly — what does a shorter intro cost the song?

Done? The quiz checks the plumbing; case study 1 hands you the four-week template in narrative form; case study 2 watches three metadata mistakes nearly ship. Your C1 card and C4 calendar come with you to Chapter 36 — the writers you listed are about to sign a split sheet.