Chapter 31 Quiz

Q: The six jobs of mastering are: A) EQ, compression, limiting, dithering, encoding, uploading B) Final tonal balance, dynamics finishing, loudness to spec, sequencing and spacing, format deliverables, the last set of ears C) Recording, editing, mixing, balancing, bouncing, releasing D) Loudness, width, depth, punch, clarity, warmth

B. Count how many are about loudness: one. That table is the chapter's whole argument against "mastering = making it louder." A lists tools, not jobs; the tools serve the jobs and arrive in Chapter 32.

Q: The limiter sits *last* in the master chain because: A) It's the most expensive plugin and deserves the final slot B) Every move before it changes what it has to work with — tonal balance shifts which frequencies hit it hardest, glue changes the peaks' shape C) Limiters only work on full-loudness material D) Tradition from the tape era

B. Limit first and you limit problems into the record permanently, then EQ around the wreckage. Theo ran the second-to-last box of the chain with nothing in front of it — no gain-matched reference, no correction, no glue — which is why the price was so high.

Q: Every mastering before/after demo that impressed you was louder in the "after." This matters because: A) Louder files have more bass B) [Chapter 4](../../part-01-sound-fundamentals/chapter-04-listening/index.md) established that louder reads as better automatically, for everyone — so an unmatched comparison is rigged before it starts C) Loudness is the main thing mastering improves D) Quiet files can't be evaluated at all

B. Match the loudness and the magic shrinks to what it really is: a competent half-dB of tone, a touch of glue — real, valuable, and small. This single mechanism funds the entire demo culture of mastering, which is why matched-loudness comparison is law for every judgment in Part VII.

Q: The print spec calls for 24-bit WAV with no dither because: A) Dither only works on MP3 files B) Dither is unnecessary at 24-bit ([Chapter 2](../../part-01-sound-fundamentals/chapter-02-digital-audio/index.md)); it earns its keep later, at the final step-down to 16-bit — the last conversion, never sooner C) Dither makes files larger D) Mastering engineers add dither to every incoming file anyway

B. The ordering is the whole dither doctrine in one sentence: print stays 24-bit, dither waits for the 16-bit/44.1 kHz Redbook conversion if and when CDs happen. Stacking conversions casually along the way spends quality and buys nothing.

Q: Your session runs at 48 kHz. The print should be: A) 44.1 kHz, because that's CD standard B) 96 kHz, upsampled for extra quality C) 48 kHz — session native; rates change (if ever) once, at the deliverables stage, done right D) Whatever the distributor's upload page suggests

C. Nothing is gained by converting at the print, and a conversion is a processing step like any other — it happens once, at the end, with proper tools. The print is the source of truth for every future version of the record; keep it native.

Q: The single most common amateur print error, per the chapter: A) Printing at too low a sample rate B) The truncated reverb tail — bouncing a region that ends before the last tail rings out, which is unfixable downstream C) Forgetting to name the file D) Printing in mono

B. The session played the tail; the bounce cut it off — and you won't notice until the quiet listen you didn't do. A fade can always be applied later; audio that was truncated to nothing is gone. One second of head silence and fully rung-out tails cost nothing and prevent it completely.

Q: Mastering's working units start at 0.25 dB — far smaller than mixing's — because: A) Mastering plugins have finer knobs B) A half-dB shelf applied to the entire record, every element at once, for the whole runtime, is a lot of total change delivered gently C) Small numbers are a tradition of the profession D) Larger moves are illegal under broadcast standards

B. The scale of the move follows from the scale of the canvas. At the sum, there are no targeted moves — every decision hits everything — so a full decibel is a statement. Learning to hear at that resolution (exercise B5) is one honest reason the profession exists.

Q: Stem mastering, named honestly, is: A) A scam with no legitimate use B) The standard way all professional mastering is done C) A hybrid that buys back some element-level reach — and the moment the engineer can move your vocal against your track, it's a remix with your permission, at remix prices D) Mastering performed only on the drum stem

C. It's real, some engineers offer it, and it can rescue a borderline mix. But name the service you're buying: more rescue reach, less mix authority where you intended it, higher cost. The polish-not-surgery boundary hasn't moved — it's been crossed, knowingly.

Q: An EP's quiet acoustic closer measures well below the other tracks. A good mastering engineer will most likely: A) Limit it up to match the loudest track — consistency is the job B) Leave it sitting low by intent — its quietness is musical information, and level relationships between songs are a sequencing decision, not a meter-matching one C) Ask for the song to be remixed louder D) Move it earlier in the running order where quiet songs belong

B. Job 4 sets relative levels by intent. Albums you love do this constantly — the closer that sits low is part of the record's dynamics at the largest scale. Matching every track to the anthem destroys exactly the information sequencing exists to protect.

Q: A great master of a bad mix is a loud bad mix.

True. Mastering works on the sum: it cannot reach inside to fix balances, rescue arrangements, restore crushed dynamics, or add what the mix never had. Everything that makes mastering look like magic was decided at the mix — which is the chapter's point: the power never left your desk.

DataField.Dev

Chapter 31 Quiz

Closed book on the first pass — that's the honest score. Multiple choice first, then true/false with justification (the justification is most of the credit), then short answer, then one applied scenario: a "ready for mastering" folder that violates the spec in eight different ways, with you holding the checklist. Answers hide under each Verify fold; the scoring guide is at the end.

Section 1 — Multiple Choice (2 points each)

1. Mastering, defined — which is the chapter's anchor definition?

A) The process of making a mix as loud as commercial releases B) The final production stage, where the finished stereo mix is examined and finished — tone balanced, dynamics completed, loudness set, songs sequenced, formats delivered — by the last trained ears before the public C) A second mixing pass performed on better equipment D) The conversion of a mix into MP3 and other streaming formats

Verify

**B.** Note what the definition operates on: the *finished stereo mix* — two channels, every balance decision already baked together. That single fact generates almost everything else in the chapter, including every hard boundary. A is one job of six; C is what stem mastering quietly becomes; D is one slice of job five.

2. The six jobs of mastering are:

A) EQ, compression, limiting, dithering, encoding, uploading B) Final tonal balance, dynamics finishing, loudness to spec, sequencing and spacing, format deliverables, the last set of ears C) Recording, editing, mixing, balancing, bouncing, releasing D) Loudness, width, depth, punch, clarity, warmth

Verify

**B.** Count how many are about loudness: one. That table is the chapter's whole argument against "mastering = making it louder." A lists tools, not jobs; the tools serve the jobs and arrive in Chapter 32.

3. The one mastering job you structurally cannot perform on your own record is:

A) Final tonal balance — your EQ isn't good enough B) Sequencing — you're too attached to your favorite track C) The last set of ears — you hear your memory of the file, not the file D) Format deliverables — they require licensed software

Verify

**C.** Not a skill deficit — a freshness deficit. Four hundred passes built a prediction so strong it overwrites the signal, the same mechanism that stops an author proofreading their own novel. Every other job survives self-mastering at some discount; this one requires other ears or an engineered substitute (the next-day listen).

4. The limiter sits last in the master chain because:

A) It's the most expensive plugin and deserves the final slot B) Every move before it changes what it has to work with — tonal balance shifts which frequencies hit it hardest, glue changes the peaks' shape C) Limiters only work on full-loudness material D) Tradition from the tape era

Verify

**B.** Limit first and you limit problems into the record permanently, then EQ around the wreckage. Theo ran the second-to-last box of the chain with nothing in front of it — no gain-matched reference, no correction, no glue — which is why the price was so high.

5. A client asks their mastering engineer to "bring the vocal up." The only honest tool the engineer has is:

A) A vocal isolation algorithm built into mastering consoles B) Raising the master fader during vocal sections C) A push in the vocal's frequency range — which also pushes the guitars, hats, snare ring, and everything else alive in that zone D) Re-importing the vocal stem

Verify

**C.** Mastering EQ reaches *frequencies*, not *instruments* — the vocal and the cymbals share the same 3 kHz, inseparably. The engineer can tilt the whole photograph warmer or cooler; they cannot move a person who was standing in the wrong place. (D describes stem mastering — a different service at a different price, and a remix by another name.)

6. Per the element-versus-whole detector, which of these is a mastering note?

A) "The snare cracks too hard in the bridge" B) "The lead vocal disappears on quiet syllables" C) "Track 2 reads brighter than track 3 for no musical reason" D) "The 808 and the kick blur in the last chorus"

Verify

**C.** It names the whole — one track against another, a tonal-kinship observation, job 1 territory. A, B, and D all name elements, which makes them mix notes no matter what stage you noticed them at; they travel upstream to [Chapter 30](../../part-06-advanced-mixing/chapter-30-mix-troubleshooting/index.md)'s method.

7. Every mastering before/after demo that impressed you was louder in the "after." This matters because:

A) Louder files have more bass B) Chapter 4 established that louder reads as better automatically, for everyone — so an unmatched comparison is rigged before it starts C) Loudness is the main thing mastering improves D) Quiet files can't be evaluated at all

Verify

**B.** Match the loudness and the magic shrinks to what it really is: a competent half-dB of tone, a touch of glue — real, valuable, and small. This single mechanism funds the entire demo culture of mastering, which is why matched-loudness comparison is law for every judgment in Part VII.

8. The print spec calls for 24-bit WAV with no dither because:

A) Dither only works on MP3 files B) Dither is unnecessary at 24-bit (Chapter 2); it earns its keep later, at the final step-down to 16-bit — the last conversion, never sooner C) Dither makes files larger D) Mastering engineers add dither to every incoming file anyway

Verify

**B.** The ordering is the whole dither doctrine in one sentence: print stays 24-bit, dither waits for the 16-bit/44.1 kHz Redbook conversion if and when CDs happen. Stacking conversions casually along the way spends quality and buys nothing.

9. Your session runs at 48 kHz. The print should be:

A) 44.1 kHz, because that's CD standard B) 96 kHz, upsampled for extra quality C) 48 kHz — session native; rates change (if ever) once, at the deliverables stage, done right D) Whatever the distributor's upload page suggests

Verify

**C.** Nothing is gained by converting at the print, and a conversion is a processing step like any other — it happens once, at the end, with proper tools. The print is the source of truth for every future version of the record; keep it native.

10. The 2-bus rule at print time, both halves:

A) Strip everything from the master bus — deliver completely unprocessed audio B) Leave everything on — the engineer will sort it out C) Loudness tools (limiter, maximizer, clipper earning level) come off; sound-shaping glue and color the mix was balanced through generally stays D) Keep the limiter but bypass the EQ

Verify

**C.** The half everyone knows: loudness belongs to the mastering stage, so its tools come off. The half that gets missed: strip the [Chapter 26](../../part-06-advanced-mixing/chapter-26-saturation-harmonic-color/index.md) tape color and glue your balances were made *through*, and you're delivering a brighter, less dense cousin of your mix that nobody approved. Genuinely unsure? Print both, label both — and a hired engineer's published delivery spec outranks this book's defaults.

11. The single most common amateur print error, per the chapter:

A) Printing at too low a sample rate B) The truncated reverb tail — bouncing a region that ends before the last tail rings out, which is unfixable downstream C) Forgetting to name the file D) Printing in mono

Verify

**B.** The session played the tail; the bounce cut it off — and you won't notice until the quiet listen you didn't do. A fade can always be applied later; audio that was truncated to nothing is gone. One second of head silence and fully rung-out tails cost nothing and prevent it completely.

12. Mastering's working units start at 0.25 dB — far smaller than mixing's — because:

A) Mastering plugins have finer knobs B) A half-dB shelf applied to the entire record, every element at once, for the whole runtime, is a lot of total change delivered gently C) Small numbers are a tradition of the profession D) Larger moves are illegal under broadcast standards

Verify

**B.** The scale of the move follows from the scale of the canvas. At the sum, there are no targeted moves — every decision hits everything — so a full decibel is a statement. Learning to *hear* at that resolution (exercise B5) is one honest reason the profession exists.

13. Stem mastering, named honestly, is:

A) A scam with no legitimate use B) The standard way all professional mastering is done C) A hybrid that buys back some element-level reach — and the moment the engineer can move your vocal against your track, it's a remix with your permission, at remix prices D) Mastering performed only on the drum stem

Verify

**C.** It's real, some engineers offer it, and it can rescue a borderline mix. But name the service you're buying: more rescue reach, less mix authority where you intended it, higher cost. The polish-not-surgery boundary hasn't moved — it's been crossed, knowingly.

14. You release a single plus its instrumental. ISRCs required:

A) One — it's the same song B) Two — the code is assigned per recording, per version, and it's how plays and royalties find the right tally C) None until you press physical copies D) One per platform you distribute to

Verify

**B.** Per recording, per version: the single gets one, the instrumental its own, a future remix its own. The ISRC is the license plate royalties and chart data travel on — and the rule that matters is consistency: one version, one code, forever, never reused on different audio. [Chapter 35](../../part-08-music-business/chapter-35-distribution/index.md) goes deep.

15. An EP's quiet acoustic closer measures well below the other tracks. A good mastering engineer will most likely:

A) Limit it up to match the loudest track — consistency is the job B) Leave it sitting low by intent — its quietness is musical information, and level relationships between songs are a sequencing decision, not a meter-matching one C) Ask for the song to be remixed louder D) Move it earlier in the running order where quiet songs belong

Verify

**B.** Job 4 sets relative levels by intent. Albums you love do this constantly — the closer that sits low is part of the record's dynamics at the largest scale. Matching every track to the anthem destroys exactly the information sequencing exists to protect.

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

16. A great master of a bad mix is a loud bad mix.

Verify

**True.** Mastering works on the sum: it cannot reach inside to fix balances, rescue arrangements, restore crushed dynamics, or add what the mix never had. Everything that makes mastering look like magic was decided at the mix — which is the chapter's point: the power never left your desk.

17. Because mastering engineers work in accurate, full-range rooms, their EQ moves can target individual instruments that bedroom mixes got wrong.

Verify

**False.** The room buys *bandwidth and trust* — sub-octave truth, inter-song consistency, distortion onset, single-sample events — not element access. The EQ still reaches frequencies, not instruments; better monitoring means better-informed whole-mix moves, not surgical ones. (This is why "the room doesn't make their ears better — it makes their ears informed.")

18. If you plan to master the track yourself in Chapter 32, printing a 24-bit, ~-6 dB headroom mix and resting it is unnecessary ceremony — you can master straight off the 2-bus in the mix session.

Verify

**False.** The prep is identical whoever masters — that's why the decision tree's two branches converge on the same spec card. Mastering in the mix session means no fresh-ears gap, no posture change (the open session keeps you a mixer), and no clean source-of-truth artifact. The print, the rest, and the next-day listen are most of what makes self-mastering work at all.

19. The deciding variables in hire-versus-self are stakes, monitoring truth, and distance — not whether you're "good enough."

Verify

**True.** Skill isn't the binding constraint; perception and freshness are. A highly skilled self-masterer with a lying room and four hundred listens of familiarity still can't verify the bottom octave or perform the QC job. That's why the tree never asks "are you good enough?" — it's the wrong question.

20. Because a 32-bit float bounce can carry levels above 0 dBFS without clipping the file, the -6 dB headroom habit is now obsolete.

Verify

**False.** The float fact is true ([Chapter 21](../../part-05-mixing-foundations/chapter-21-gain-staging/index.md)'s sidebar) and the conclusion doesn't follow. Treat float headroom as an engineering footnote, not a workflow: deliver sane 24-bit headroom and nobody downstream — converter, plugin chain, engineer, or future-you — ever has to wonder what the file's levels mean. Headroom is a workflow, not a number you can technically evade.

Section 3 — Short Answer (4 points each)

21. Name the four "mastering cannot" boundaries and the single structural fact that generates all of them.

Verify

Cannot fix a balance (the buried vocal stays buried, with more expensive tonality). Cannot rescue an arrangement (seven banjos arrive as mud with a beautiful top end). Cannot restore what processing destroyed (no un-limiter, no un-clipper — only disguises). Cannot add what the mix never had (a shelf can only boost the rumor of missing lows; no tool builds depth behind a flat, dry mix). The generating fact: mastering works on the sum — two channels, every decision baked together, tools that reach frequencies rather than instruments.

22. Describe the next-day listen protocol — both passes and the note-writing standard — and name the three stacked mechanisms that make the same file sound different on Thursday than it did Wednesday night.

Verify

Protocol: print, stop, sleep — no evaluation on print night. Pass one, consume: main system, calibrated level, DAW closed, the whole record top to bottom, attention on the experience; phone or paper notes only. Pass two, translate: the [Chapter 30](../../part-06-advanced-mixing/chapter-30-mix-troubleshooting/index.md) lap (car, earbuds, phone, quiet listen), same notes file. Notes standard: timestamped, band vocabulary, symptom-first — written for future-you, per [Chapter 19](../../part-04-arrangement-production/chapter-19-collaboration-workflow/index.md)'s feedback rules. Mechanisms: adaptation (hearing renormalizes to whatever it receives; sleep resets the reference), expectation (you hear your intentions, not your file; distance weakens the prediction), and posture (no fader in reach turns the maker back into an audience).

23. State the revision discipline: the three triage buckets, what happens to each, and the stop condition. Then say what a problem list "as long as your arm" actually means.

Verify

Buckets: *problems* (objective, nameable, would-embarrass-you-later — these get fixed in one focused pass, every fix verified at matched loudness, then re-print v2); *preferences* (different-not-better — written down and let go; if you can't name the symptom a change cures, it doesn't reopen a finished mix); *mastering notes* (whole-record observations — into the brief, addressed at the next stage). Stop condition: when the remaining notes are all preferences, you're done — not out of ideas, *done*. A long, element-heavy problem list means the mix wasn't finished: that's not a revision, it's a return to [Chapter 30](../../part-06-advanced-mixing/chapter-30-mix-troubleshooting/index.md)'s method, better learned from your own next-day ears than from an engineer's polite email before release.

24. Give the four factors that argue for hiring a mastering engineer on a given release, and describe the hybrid play and what it teaches.

Verify

Factors: stakes (the statement EP/album; vinyl or CD, where errors are permanent objects); objectivity (you mixed it — you are the person least able to hear it fresh, and job 6 can't be self-performed); monitoring truth (if the genre lives in a bottom octave your room can't report, someone with full-range truth should sign off); and a second professional opinion on the mix itself (good engineers send mixes back — the cheapest consultation in audio). Hybrid play: self-master everything, hire for flagships, and when you hire, master your own version anyway, then A/B yours against the pro's at matched loudness. Where they differ is a surgical map of your blind spots — the fastest mastering education money can buy, purchased about once a year.

Section 4 — Applied Scenario (10 points)

25. A bandmate drops a folder called EP_MASTER_READY in your shared drive with the message: "All five songs prepped for mastering! Also — can the mastering engineer bring the lead vocal up a bit in the choruses on track 2? It gets buried." You open the folder and find: (a) all five songs as 320 kbps MP3s "to keep the upload small"; (b) the one WAV present is 16-bit/44.1 kHz, bounced from their 48 kHz session "because that's CD quality"; (c) peaks on every file sit at -0.2 dBFS, and their limiter is still on the 2-bus showing 6 dB of gain reduction "so it sounds finished"; (d) track 3's closing reverb tail cuts off dead at the final downbeat; (e) track 1's audio starts at the very first sample — no head silence; (f) the files are named "FINAL_MASTER_v2_REALFINAL"; (g) there are no reference tracks, no notes, no tech doc; (h) no instrumental or clean versions exist, "we'll bounce those later if anyone asks." Identify every violation, explain the specific harm of each, prescribe the fix — and answer the vocal question the honest way.

Verify

**The vocal question first, because it's the most expensive misunderstanding in the message:** "bring the vocal up" names an element — it's a mix note wearing a mastering costume. The engineer's only tool is a push in the vocal's frequency range, which drags the guitars, hats, and sibilance up with it. The buried-vocal fix happens in the mix session ([Chapter 30](../../part-06-advanced-mixing/chapter-30-mix-troubleshooting/index.md)'s method; probably a ride or a masking cut), *before* anything is printed. An honest engineer would send track 2 back — treat your bandmate to that consultation for free. **(a) MP3s:** lossy encoding permanently discarded audio ([Chapter 2](../../part-01-sound-fundamentals/chapter-02-digital-audio/index.md)); mastering an MP3 polishes the artifacts. Re-bounce everything as uncompressed WAV. **(b) 16-bit/44.1 from a 48 kHz session:** two needless quality losses — a bit-depth truncation (possibly undithered) and a sample-rate conversion done at the wrong stage. Print 24-bit at session-native 48 kHz; conversions happen once, at deliverables. **(c) -0.2 dBFS peaks with a limiter doing 6 dB GR:** the loudness stage has been pre-spent, badly — crushed dynamics don't come back, and the master chain has nothing to work with. Take the limiter off (its job is loudness — it belongs to the next stage), restore the mix's real peaks, and if those sit near zero, run the [Chapter 21](../../part-05-mixing-foundations/chapter-21-gain-staging/index.md) trim-gain audit to land around -6 dBFS without touching balances. **(d) Truncated tail:** the classic unfixable error — re-bounce with the tail ringing fully out. **(e) No head silence:** sloppy playback systems clip the start and the engineer can't hear the noise floor honestly; leave about a second of clean silence. **(f) "FINAL_MASTER_v2_REALFINAL":** it's neither final nor a master — [Chapter 19](../../part-04-arrangement-production/chapter-19-collaboration-workflow/index.md) naming rules: Artist_Song_MixPrint_v1_24-48.wav; "final" is a banned word. **(g) No brief:** add three references with what each demonstrates, intent notes, confessed concerns, and the tech doc — the brief aims the engineer's attention where the band's monitoring couldn't testify. **(h) No alts:** print the instrumental, a cappella, vocal-up, and clean versions *now*, same pass, same settings — three minutes each tonight, an archaeology project in eight months. Full credit: all eight violations with harms and fixes, plus the vocal answer routed upstream rather than promised downstream. Bonus point of understanding: nothing on this list requires talent — only the card. That's the point of the card.

Scoring

Section	Points available
Multiple choice (1–15)	30
True/False + justification (16–20)	10
Short answer (21–24)	16
Applied scenario (25)	10
Total	66

56–66: You know what mastering is, what it isn't, and what it needs from you — go execute the print and book the next-day listen. 44–55: Solid on the jobs, shaky on the boundaries or the spec; re-read "Polish, Not Surgery" and the spec card's line-by-line reasons, then redo exercises A4 and A6. Below 44: Re-read "The Six Jobs" and the threshold-concept block, then take this again in two days — after your own next-day listen, which will teach half the chapter by itself. The most expensive mastering mistakes are conceptual, not technical; this quiz is cheaper than all of them.