Capstone Assessment Rubric

One rubric for all three capstone projects, because the differences between the roads are logistics and the standards are physics. Six dimensions, four levels each, one weighting table, one gate.

How to Use This Rubric

Self-assessment first. Before any instructor, mentor, or peer sees the work, score it yourself — in writing, with evidence cited for each dimension. Then get the outside pass (run it on Chapter 19's feedback protocol: specific, timestamped, about the goal). The gap between your scores and theirs isn't embarrassment; it's calibration data, the same skill as level-matching your A/Bs in Chapter 22 and trusting meters over mood in Chapter 33. Producers who self-score accurately mix better, because assessment is just diagnostic listening pointed at your own work.

Score what's evidenced. The deliverables are the evidence: the master, the stems or package, the session, the logbook, the QA numbers, the documents. An assessor can't credit reasoning that isn't written down — an undocumented good decision scores as luck, because from the outside that's what it is. This is why the logbook is a deliverable and not a suggestion.

Match the level that mostly fits. Read each dimension's four descriptors and pick the one that describes most of the work. Ties break downward: claims aren't evidence, and "almost Proficient" is Developing.

The 70% gate. A weighted score of 2.8 out of 4.0 — seventy percent, the same line the book's quizzes draw — is the pass threshold. It's a gate, not a verdict on you: below it, the dimension scores tell you exactly which chapters reopen, the way the symptom table turns "my mix is bad" into "cut the 200–400 Hz pileup." This rubric is a diagnostic instrument wearing an assessment costume.

The floor rule. Any dimension scored Beginning caps the project at "not yet," whatever the weighted sum says. A record with one genuinely broken dimension isn't release-ready, and no amount of excellence elsewhere buys that back.

The Six Dimensions

1. Technical Quality

The "is the record clean" dimension: noise, edits, levels, artifacts — the layer Chapters 21 and 31–33 specify and Chapter 30 teaches you to interrogate. Listen on good headphones at attentive volume; this dimension hides from speakers across the room.

2. Translation

Does it hold up where listeners actually are? The car, the phone speaker, the earbuds, the quiet late-night listen (Chapter 30), the mono fold-down (Chapter 25) — and for spoken word, the noisy-room and low-volume tests, weighted heavily.

3. Craft Decisions

Were the choices deliberate? Arrangement discipline (Chapter 16), frequency pockets (Chapter 22), space architecture (Chapter 24), dynamics and movement (Chapters 23 and 27). The question here is not "is it tasteful" — that's Dimension 6. The question is whether what's audible was chosen.

4. Process & Documentation

The logbook dimension (Chapter 19): decisions traceable to reasoning, hygiene, versions — plus the collaboration documents on Project 2 and the cut list and licensing documentation on Project 3.

5. Release Readiness

Gates 4 through 6 as a dimension (Chapters 33, 35, 36): loudness QA, metadata, deliverables hygiene, and splits or rights where collaborators or licensed material exist.

6. Identity & Taste

The Chapter 40 dimension, weighted gently on purpose: does it sound like someone? Not "is it innovative" — coherent and recognizable beats novel and scattered.

Scoring & Weighting

Score each dimension 1–4 (Beginning = 1, Developing = 2, Proficient = 3, Professional = 4), multiply by its weight, and sum to a score out of 4.0.

A worked example: Technical Quality 3, Translation 3, Craft Decisions 3, Process & Documentation 4, Release Readiness 3, Identity & Taste 2. Weighted: 0.60 + 0.60 + 0.60 + 0.60 + 0.45 + 0.20 = 3.05 — a pass, with the gently-weighted 2 doing exactly its job: not sinking a releasable record, while naming the next mountain.

And the floor rule, restated because it's load-bearing: one Beginning anywhere caps the verdict at "not yet." The weights describe importance, not interchangeability.

A Note on "Professional"

Read the top row of those tables again and notice what isn't there. No console. No treated facility with a name. No budget line. Nothing in the Professional descriptors requires anything beyond what this book assumed in Chapter 1: a DAW, modest gear chosen with Appendix D sanity, a room improved with Chapter 10 labor, and trained ears.

That's not encouragement-poster optimism; it's the book's founding, documented fact. The debut album that won Album of the Year was produced in a bedroom by someone whose monitoring chain cost less than a used car — and what made it professional wasn't the room, it was that every decision on it was deliberate and the record translated everywhere (Chapter 5). Those are Dimensions 2 and 3. They are skills, not purchases.

So here's the honest bar this rubric actually sets. Professional means translation plus deliberateness: the work holds up on systems you didn't choose, in rooms you've never been, and every audible element is where it is on purpose — and you can say why. That's the whole test. Gear can make passing it more comfortable. It cannot pass it for you, and its absence cannot stop you.

Two honesty notes to close. First: Professional on Identity & Taste is rare on a first capstone, and that's correct — it mostly arrives with reps, somewhere along Chapter 40's ten-year roadmap, which is why it's weighted gently instead of pretended into reach. Scoring Proficient across all six dimensions is a genuine achievement and a releasable record; do not let the existence of a fourth row convince you the third row is failure. Second: this rubric never retires. Run it on your next release, and your tenth — self-assessment first, outside ears second, gates always. The score that matters isn't this project's. It's the trajectory.