Capstone Project 2: Producing Another Artist

The collaboration path — the Wren & Hollow road. Capture-heavy, human-heavy. Project 1 proves you can run the pipeline; this one proves you can run it when the song isn't yours, the voice isn't yours, and the person whose record it is can hear everything you do to it.

Overview & Outcomes

Produce another artist's song, start to finish: pre-production conversation, capture, edit, mix, master (or managed mastering), and a delivery package they own forever. A real artist — not you under an alias, not a hypothetical client. Someone with a song that doesn't exist as a record yet, and a reason to care that it will.

This is the brief where the producer's actual job description surfaces. Most of this book taught you to make sound decisions; this project assesses whether you can make them for someone, with their trust on the line and their name on the cover. The engineering is the same physics you've already proven. What's new is everything Chapter 19 warned you about: meetings that decide whether sessions succeed, performances that have to be produced rather than recorded, feedback that has to be metabolized without ego, and paperwork that has to exist before money does.

On top of Project 1's outcomes, you'll demonstrate that you can:

• run pre-production that converts a vision conversation into a capture plan (Chapters 19, 16, 7);
• prepare rooms and signal chains for capture days that can't be repeated cheaply (Chapters 10, 12);
• produce performances — coaching, pacing, knowing when to push and when to stop (Chapter 11);
• run a feedback and revision protocol with a real client without drowning in rounds (Chapter 19);
• execute the splits and roles paperwork that keeps collaborators friends (Chapter 36);
• and deliver stems and masters so complete the artist never needs to call you (Chapters 19, 31).

There's also a career argument for choosing this road, and it's worth saying out loud: producing other people develops your ears faster than producing yourself. Your own work arrives pre-approved by the person who made it; another artist's song forces you to articulate standards you've been applying by feel. Every producer who has coached a vocalist through take after take discovers they suddenly hear their own takes differently. Objectivity, once installed, doesn't uninstall.

One honest note before you accept the gig: humans are the variable. Schedules slip, takes 14 through 20 go nowhere, notes contradict each other. That's not a defect in the project. That is the project.

Prerequisites

Everything Project 1 requires — Chapters 1–33 as the working minimum, 35–37 before release planning — with four chapters promoted from "read" to "command": Chapter 7 (microphones), Chapter 11 (the vocal session), Chapter 12 (instruments), and Chapter 19, the human chapter. Reread Chapter 19 the week you start. It's the one that saves this project.

And you need an artist. A singer-songwriter with a worktape, a rapper with a verse and a hook, a friend's band, a choir director with an arrangement — anyone whose song genuinely needs producing and who genuinely wants it produced. The standard is simple: the song must not already exist as a finished record, and the artist must want the result, not be doing you a favor. (Producing an imaginary client is Project 1 in a costume; it can't develop what this brief assesses.)

Before you accept, have the pre-condition conversation: scope (one song), timeline, how many revision rounds, what they receive (masters plus stems), and the fact that a split sheet gets signed before anything is released. If any part of that conversation scares them off, you've saved both of you a worse conversation later.

The Brief

Produce one song by another artist, in their genre, to release-ready standard, and deliver the following six items.

1. The master. A release-ready stereo WAV of the artist's song — same technical spec as Project 1 — with measured loudness QA numbers: integrated LUFS, true peak ≤ -1.0 dBTP, and the hedged target justification (Chapter 33).
2. The delivery package, to the artist. Masters plus alternates (an instrumental at minimum), and stems printed from zero, all the same length, documented with sample rate, bit depth, tempo, key, and your wet/dry policy — Chapter 19's standard, labeled so they're usable in any studio, by any engineer, in ten years. Sessions rot; stems are the insurance. This package is the product. The artist walks away owning their record.
3. The session, in pro hygiene. Chapter 19 standard, same as Project 1, graded on reopenability.
4. The collaboration documents. Three, all real: - The roles agreement. One page, plain language, written before capture: who does what (the producer/engineer/mixer vocabulary of Chapter 19), the revision-round count, the deliverables list, the timeline, and what happens if the project stalls. - The split sheet. Executed BEFORE release (Chapter 36): every contributor, every percentage, signed and dated. The kitchen-table scene — your kitchen, your table, your collaborators' names spelled right. - The feedback rounds log. For each round: the consolidated, timestamped notes received; what changed in response; what was declined and the reason you gave the artist.
5. The production logbook. Project 1's format — every stage, the ONE key decision, the reasoning — plus a pre-production front section: meeting notes, the agreed reference list, the capture plan, the session schedule.
6. The release plan — or the release. It's the artist's record, so this one is built with them: a four-week plan (Chapters 35, 37) or evidence of the release itself, with their written consent either way.

Milestones

Suggested span: 4–8 weeks. The calendar bends around humans, so schedule slack where the humans are — capture and feedback.

The two-meeting minimum in pre-production is deliberate. The vision meeting decides WHAT; the logistics meeting decides HOW. Collapse them and you get a capture day that runs on guesses — the $2,400 studio weekend that gave Wren & Hollow two songs they hated was exactly this failure, paid for at studio rates. And the revision rounds are capped at the agreed number on purpose: unlimited rounds don't produce better records, they produce abandoned ones. One consolidated round, executed; one more to confirm; sign-off in writing. If the roles agreement promised more rounds, honor it — but you wrote that agreement, so write it wisely.

Schedule the capture days with a fallback: one reserve session you both agree not to plan anything into. If everything lands early, it becomes the ad-libs-and-experiments session — Chapter 11's while-the-mic-is-hot doctrine. If a voice gives out or a string breaks, it's the difference between a calm reschedule and a compromised record.

Constraints & Variations

The genre is the artist's. Full stop. The Chapter 18 audit runs on their references; your taste shows up in execution quality and in the one or two deliberate violations you propose and they approve — not in quietly steering their record toward your lane.

The three reader archetypes. The bedroom producer runs the remote variant — the Jaylen-and-Demi road: capture happens at the artist's end, your capture plan becomes a coaching document, and the Chapter 19 file protocol becomes load-bearing (sample rates declared, dry signals insisted on, versions named like a professional). The band member can produce their own band — the Theo road — which is legitimate and carries the steepest objectivity bill in this book; the outside ear at the mix gate stops being a suggestion. The podcaster can take this brief by producing a musician anyway (the stretch is the point), or run the capture-heavy variant: a narrator or voice artist's showcase reel, produced end to end — though the music path exercises more of the rubric.

Variations. The two-song EP scope, only with instructor sign-off — every milestone multiplies, and the sequencing questions of Chapter 32 join the brief. The remote-only variant, formalized: all direction over calls and notes, which makes the feedback log the richest deliverable in the package. The swap, for cohorts: two producers on this brief produce each other — efficient (everyone gets a record and a client) and instructive in stereo, because you'll receive exactly the kind of feedback round you just gave, and your roles agreement gets tested from both chairs. And hard mode: one mic, one room, one weekend for all capture — the Wren & Hollow studio weekend inverted, except this time the planning is yours, the room is free, and the plan exists before the clock starts.

What "Done" Means

The gate list applies in full — Gates 0 through 6 for release-ready, 7 through 10 for released; printable in Appendix G, commentary in Chapter 39. This road adds three relationship gates of its own:

• R1 — the roles agreement is signed before the first capture session. Not discussed. Signed.
• R2 — the split sheet is signed before release. Before a dollar, before an upload, before the metadata is typed (Chapter 36). Gate 6's metadata check reads against this document, name by name, percentage by percentage.
• R3 — the artist signs off on the final master, in writing. The last feedback round closes with an explicit "approved," not with silence you chose to interpret generously.

R3 deserves one practical note: "approved" arrives more easily when the artist has lived with the final candidate on their own systems, alone, for at least a day. Send it, name the deadline, and ask for the sign-off sentence explicitly. Silence is not approval; enthusiasm on a phone call is not approval; the written sentence is approval — and it protects the artist exactly as much as it protects you.

Done means the artist could release tomorrow without calling you — and would still want to work with you if they did.

Assessment

Same instrument as the other paths: the Capstone Assessment Rubric, six dimensions, 70% gate. Three emphasis notes for this road. Process & Documentation leans hard on the collaboration documents — the feedback log especially, because it's the visible difference between "made some changes" and "ran a protocol." Release Readiness scores splits and rights as core items here, not conditionals. And Identity & Taste is read as production identity in service of the artist: does the finished record sound like them, on purpose, with your fingerprints in the craft rather than over their face?

Common Failure Modes

Each with its chapter home. Most of these fail producers, not files.

1. Skipping pre-production. The first tracking day spent making decisions a meeting should have made; the capture plan improvised at the mic stand (Chapter 19, Chapter 7).
2. The infinite revision spiral. No round count in the roles agreement, so "one more tweak" never ends (Chapter 19's protocol: consolidated, timestamped, counted).
3. Producing your record instead of theirs. The reference list quietly swaps toward your taste; the artist stops recognizing their song (Chapters 4 and 18 — the compass belongs to the artist).
4. Recording instead of producing. Twenty technically clean takes, none alive — take 14 perfect and dead while take 3 wins unheard (Chapter 11). Your job was in the talkback, not the waveform.
5. Phase mess on multi-mic sources. Discovered in the mix, when it's expensive. The flip test and the 3:1 habit live at capture (Chapter 12).
6. The split-sheet-after-release disaster. Percentages negotiated after money exists. This is how collaborators stop being friends (Chapter 36 — before, always).
7. "Stems" that are 34 raw multitracks at clipping levels. The 4 GB inbox sin (Chapter 19). Printed from zero, same length, documented, or it isn't a delivery.
8. Feedback received as an identity attack. The note stings, the producer defends, the session sours. Separate yourself from the artifact and find the problem behind the note (Chapter 19).
9. The artist ghosts, or the scope creeps. Humans are the variable; the roles agreement's timeline and stall clause is the tool you'll be glad you wrote on day one.

The reward this road pays is specific, and nothing else in this book pays it: another artist trusted you with their song, and you handed them back a record that sounds like them — only finished. That's not a class exercise. That's the job, and now you've done it.