Chapter 3 — Exercises

How to use these. Difficulty runs ★ (warm-up) to ★★★ (stretch). Time estimates are honest averages — slower is fine, skipping is not. Parts A and D need paper or a notes file; Part B needs headphones and the named tracks (any streaming service); Part C needs your actual setup and your DAW. Do Part B with fresh ears, not at the end of a long session. Selected answers and worked solutions live in the back matter (Answers to Selected Exercises); exercises marked ✎ have an entry there.

Part A — Conceptual (about 45 minutes total)

A1. ★ The ten stages, cold. ✎ Without looking at the chapter, write the complete signal chain from source to ear, in order. Mark the two border crossings (analog→digital, digital→analog) and label each stage's domain: acoustic, analog electrical, or digital. Then check against the chapter diagram and note which stage you forgot — almost everyone forgets one, and it's usually the room. (5 min)

A2. ★ Count the gain stages. ✎ A vocalist sings into a dynamic mic, which feeds an interface's mic input (gain knob). The DAW track has clip gain on the recorded region, an EQ plugin with an output control, and a channel fader, routed to a master fader. The interface has a monitor volume knob feeding powered speakers, each with a rear input-trim knob. The singer steps closer to the mic on the quiet verses. How many gain stages can you count in this description? Defend your count — especially whether the singer's movement qualifies, and why. (5 min)

A3. ★★ The boost that can't win. Your friend recorded a podcast interview peaking at -40 dBFS and plans to "fix it with +26 dB of gain in the edit." Explain, in terms of what was captured versus what can be changed later, why the result will hiss — and why the same +26 dB applied at the preamp during recording would have been clean. Use the phrase signal-to-noise at least once, correctly. (7 min)

A4. ★★ Two mismatches, two opposite symptoms. Plugging a microphone into a line input produces near-silence. Plugging a synth's line output into a mic input at normal gain produces distortion. Explain both symptoms from one underlying idea: what each input expects to receive, and what gain it applies based on that expectation. (6 min)

A5. ★★ Balanced audio for a ten-year-old. ✎ In four sentences or fewer, with no jargon (you may not use "polarity," "differential," or "common-mode"), explain why a three-pin mic cable can run across a room full of power cables and arrive clean, while a guitar cable doing the same picks up hum. Metaphors encouraged. Test it on an actual non-audio human if one is available; revise until they nod. (8 min)

A6. ★★ Phantom power triage. For each mic scenario, state: does it need 48V, does it ignore 48V, or does it call for caution — and why. (a) A condenser mic that records nothing, meter flat. (b) An SM58 on a guest's voice, phantom accidentally left on. (c) A vintage ribbon mic borrowed from a collector. (d) A modern active ribbon. Add the one ritual you should perform before toggling phantom in any scenario. (6 min)

A7. ★★★ The tradeoff essay. ✎ "Every gain decision is a tradeoff between noise floor and headroom." Argue this with two concrete cases: (a) recording a whisper-quiet fingerstyle folk performance, and (b) recording a screaming punk vocalist who gets louder every take. For each: where do you set gain, what are you protecting, what are you risking, and what symptom appears if you choose wrong in each direction? One paragraph per case. (10 min)

A8. ★★ Decisions versus data. A recording-chain failure damages audio; a monitoring-chain failure damages decisions. Explain why the second is often more expensive in the long run, using Jaylen's Chapter 1 car-stereo disaster as your evidence. Identify which specific monitoring stage lied to him and what decision the lie produced. (6 min)

Part B — 🎧 Listening Lab (about 60 minutes total)

Ear training is non-negotiable in this book — trained ears with cheap gear beat the reverse, every time. Use headphones for B1–B5. Listen at a moderate, consistent volume except where instructed. Keep your listening journal (started in Chapter 1) open; each exercise ends with a written observation.

B1. ★ The noise floor of 1975 — Queen, "Bohemian Rhapsody." The famous a cappella opening was bounced across tape generations, and every bounce re-recorded the previous tape's noise floor along with the voices. Listen to the first minute at a generous volume. Attend under the voices: the steady tape hiss, and how the texture of the floor shifts as stacked layers enter. You are hearing gain stages — dozens of them, analog, accumulated. Journal: describe the floor in one sentence, and name the noise species. (8 min)

B2. ★ Gain riding made audible — The Beatles, "A Day in the Life." Listen to the final piano chord — the last 50 seconds or so. As the chord decays, the story goes, the engineers pushed the faders up to follow it down into silence, so the studio's noise floor rises into audibility along with the chord's tail: listen for the room becoming present — air, hiss, small mechanical sounds — as the gain climbs. This is the level window from the chapter, performed in real time: signal sinking toward the floor, gain dragging the floor up after it. Journal: at what point in the decay do you first notice the room? (7 min)

B3. ★★ How quiet a bedroom can be — Billie Eilish, "when the party's over." Recorded in a bedroom — and one of the quietest noise floors in modern pop. Listen in the gaps: between vocal phrases, before entrances, after endings. How much hiss? How much room? Notice the mouth sounds and breath detail that a close, quiet chain captures — intimacy and exposure are the same decision viewed from two sides (there's your tradeoff theme). Journal: contrast this floor with B1's in two sentences. The half-century between them is the history of the chain getting quieter. (8 min)

B4. ★★ The room as a stage in the chain — Led Zeppelin, "When the Levee Breaks." Two microphones in a stairwell at Headley Grange; the building is most of the drum sound. Listen to the drum intro twice: first for the hits themselves, then only for the space around them — the bloom after each snare, the size the echoes imply. The chapter said the chain starts acoustic: this is stage 1's environment doing more for the record than any processor could. Journal: in one sentence, what does the room add that a dead room would have lost? (8 min)

B5. ★★ Everything must whisper — Nirvana, "Something in the Way" (Nevermind). A near-whispered performance: when the source is this quiet, every stage of the chain is working close to its floor, and there's nowhere for noise to hide. Listen for what sits under and around the vocal and cello — how close the quiet elements feel, and any floor you can detect in the spaces. Journal: estimate where this performance forced the engineers' gain decisions (high gain and trust the room, or low gain and lose the whisper?) and what that risked. (8 min)

B6. ★★★ Floor hunting in your own library. Find two recordings you already love: the noisiest fade-out you can locate (hunt in pre-1985 releases — end-of-song fades are where analog floors surface) and the cleanest, blackest silence you can find (post-2015 productions are good hunting grounds). Level-match them by ear — adjust volume so the music feels equally loud — then A/B the final ten seconds of each. Journal: three sentences — name each floor's species, and say which recording you'd have guessed was "better made" before this chapter, and whether that's changed. (10 min)

B7. ★★ Noise as an aesthetic choice — Jimi Hendrix, "The Star-Spangled Banner" (Live at Woodstock). A cranked tube amp chain, outdoors, at maximum gain: the noise floor isn't hidden here — it's part of the instrument. Listen to the quieter moments and the spaces between phrases, and try to name every species you can pick out of the amp's roar using the chapter's taxonomy (hiss from the gain stages, hum and buzz from the electrics). Then sit with the tradeoff: this chain breaks half the rules in this chapter and produced one of the most famous guitar performances ever recorded. Journal: two sentences on when noise is a flaw and when it's a feature — and who gets to decide. (8 min)

Part C — 🎛️ DAW Lab (about 100 minutes total; requires your own setup)

These are the exercises that turn this chapter from trivia into reflexes. C1 and C2 are this chapter's core homework — they feed the Project Checkpoint directly.

C1. ★ The silence print. ✎ Set your normal recording gain on your main input. Record ten seconds of your room with nobody performing. Duplicate the region, boost the copy heavily (+24 dB or more, or normalize it), and listen on headphones. Identify every species you hear using the chapter's taxonomy table: hiss, hum, buzz, whine, crackle. Note the original (un-boosted) peak level of the floor from the meter. File the recording in your project under audits/ with today's date. This file is your "before" — you'll re-print it after every fix and after Chapter 10's room treatment. (10 min)

C2. ★★ The full chain map. Execute Move 1 from the chapter on your real setup: two columns (capture path, monitor path), every device, every cable labeled by type and balanced/unbalanced status, every gain stage circled and its current position noted. Then the audit questions: How many gain stages total? Which single stage applies the most gain? Which run is your longest unbalanced cable, and what does it pass near? What OS-level audio processing did you find and disable? Tape the finished map where you can see it from your chair. (25 min)

C3. ★★ The headphone-knob proof. ✎ A direct cure for the most common gain superstition. Set proper input gain (peaks around -18 to -12 dBFS while performing at real level). Record take one. Now change only your headphone/monitoring volume — crank it noticeably louder — and record take two of the same material at the same distance and intensity. Compare the two waveforms and their peak readings. They should be essentially identical. Write the moral in your own words somewhere you'll see it. (Aisha's version: "the headphone knob is not the record knob.") (10 min)

C4. ★★ Gain-redistribution shootout. ✎ Record the same source twice — same performance energy, same position: (a) healthy preamp gain, peaks around -15 dBFS; (b) preamp gain very low, peaks around -40 dBFS. In the DAW, boost (b) until both sound equally loud. Now solo the gaps — the silences between phrases — in each take, headphones up. Describe the difference in the floors. You have just heard why gain belongs at the earliest stage, and no forum argument will ever un-teach it. (15 min)

C5. ★★★ The half-split drill. Run the protocol even if your rig is currently clean — drills are for before the fire. Write your monitor chain's midpoint (for most setups: headphones plugged straight into the interface). Then script your hunt: for a hypothetical hum in your monitors, list the exact sequence of tests you'd run, what each result would prove, and which half each test eliminates. Execute at least the first two splits for real: (1) headphones-at-interface test; (2) laptop-on-battery test (if a laptop powers any part of your chain). Log results. Bonus: if you have powered monitors on a different outlet from your interface, move everything to one strip and re-record your C1 silence print. Compare. (15 min)

C6. ★★★ The wrong-hole lab (only with gear you own — keep levels low). If you have a passive electric guitar or bass: record a 20-second riff DI'd into the Hi-Z/instrument input, then the same riff into a line input (gain matched as closely as you can). A/B at matched loudness and describe what the wrong input did to the top end — you're hearing the impedance sidebar. No guitar? Substitute: plug a phone playing a reference track into a line input, record 10 seconds, then into a mic input with the gain near minimum; compare cleanliness and tone. Before each test: monitors muted, gain at minimum, raise slowly. Journal one sentence per comparison. (15 min)

C7. ★★ The border-crossing toll. If your interface has a direct-monitoring control, record a short vocal or instrument pass twice: once monitoring through the DAW (software monitoring), once with direct monitoring engaged. Feel the difference — the first routes your live signal across both borders (A/D, buffers, D/A) before you hear it; the second taps the signal before the borders and skips the round trip. Then write the tradeoff in one line each: what direct monitoring gives you, and what it takes away (hint: where do your plugins live?). No direct-monitor switch? Demonstrate the round trip instead: raise your buffer size to maximum and play a virtual instrument live, then drop the buffer low and repeat — log both latencies by feel and what the low setting cost (Chapter 2's tradeoff, now located on the chain map). (10 min)

Part D — Synthesis (about 45 minutes total)

D1. ★★★ Three chains on paper. ✎ Design a complete signal chain (capture and monitor paths) for each scenario, listing every stage in order, the level (mic/instrument/line) at every connection, and balanced/unbalanced for every cable: (a) a podcaster interviewing two in-person guests around a table; (b) a bedroom rapper recording over purchased beats; (c) a duo recording acoustic guitar and vocal simultaneously in a living room. For each chain, mark the single biggest noise risk with a ⚠ and state the one place you'd spend extra money first — with the tradeoff you're accepting by not spending elsewhere. (20 min)

D2. ★★★ The wiring doc Aisha wishes she'd had. Write a one-page "studio re-setup sheet" for your own room: what plugs into what (from your C2 map), every gain knob's home position, the phantom-power ritual, the OS settings that must stay off, and a five-line checklist titled "After moving the desk." Keep it with your session templates. You are writing to the future you who re-plugs everything in a hurry between errands — be kind but specific. (10 min)

D3. ★★★ The earliest-stage reply. A friend's recordings hum, and they're about to buy an AI-powered noise-removal plugin (Chapter 38 will weigh those properly). Write them a 150-word message applying the earliest-stage doctrine: acknowledge the tool is real, explain what the late fix costs, and give them a three-step cheap hunt to try first — without being a jerk about it. Tone matters; you're going to be the friend people send mixes to for the rest of your life. (8 min)

D4. ★★ Your tradeoff ledger. List five chain decisions already embodied in your current setup — things like: bus-powered interface (portability vs. power cleanliness), monitors against the wall (space vs. bass accuracy — Chapter 10 will explain), one long unbalanced cable (budget vs. noise exposure), consumer headphones (cost vs. honest monitoring), recording near a window (light vs. street noise). For each: what it buys you, what it costs you, and whether you're keeping the tradeoff on purpose now that you can name it. (7 min)

Part M — Mixed Practice (interleaving Chapters 1–2; about 25 minutes)

M1. (Ch 1 + 3) ★★ A guitarist plays a single note whose fundamental is 220 Hz. List the frequencies of its first three harmonics above the fundamental. Now: which signal-chain failure from this chapter adds harmonics the player never produced, and at which two stages is that failure most likely to occur? ✎ (5 min)

M2. (Ch 2 + 3) ★★ At which numbered stage of the ten-stage chain does the sample rate take effect, and what must happen to frequencies above half the sample rate before that stage — and why? Name the ugly artifact if it isn't done. (5 min)

M3. (Ch 1 + 2 + 3) ★★ Three level scales appeared across three chapters: dB SPL, dBFS, and the informal mic/instrument/line levels. Assign each scale to the stages of the chain where it meaningfully applies, and identify one stage where none of the acoustic scales applies at all. ✎ (5 min)

M4. (Ch 2 + 3) ★★ Overdubbing through the DAW, you hear yourself noticeably late. Trace the round trip: which chain stages and which Chapter 2 mechanism create the delay? You lower the buffer size and the delay shrinks — what did that cost, and what symptom appears if you push it too far? Then name the interface feature that sidesteps the round trip entirely, and its cost. (6 min)

M5. (Ch 2 + 3) ★★★ Sixteen-bit audio puts roughly 96 dB between its format noise floor and the clipping ceiling — and Chapter 2 said that's usually plenty for delivery. Your C1 silence print showed your room-plus-preamp floor far, far higher than any format's floor. Write three sentences drawing the practical conclusion: where is the real noise bottleneck in a bedroom studio, what should you spend effort on first, and why does this make the 24-bit-vs-16-bit forum war mostly irrelevant to your actual recordings? (4 min)

Done? The quiz is shorter than it looks and the case studies pay off everything above — Fourteen Gain Stages follows one vocal from larynx to a stranger's earbuds, and the Hum-Hunt watches Aisha run Part C5 for real, with stakes.