Case Study 2: Beneath the Headlines, Above the Mud — Aisha's Spoken-Word EQ Chain

Music gets most of this chapter's attention, but the single most common EQ job on Earth is a human voice, speaking, recorded in a room that was never meant for it. Aisha Brooks has published over a hundred episodes of Beneath the Headlines with no EQ at all — or rather, with the accidental EQ of a small office, a close-worked dynamic mic, and hope. This is the session where she built her chain: every move, every number, every verdict, and the two problems EQ refused to fix. It's a masterclass in spoken-word EQ precisely because nothing in it is exotic — one voice, one mic, six moves, and the law.

The Diagnosis Comes First

Aisha's relationship with her own sound has been improving on schedule. Chapter 3 taught her to find the hum; Chapter 8's calibration day taught her that her monitoring was lying before her gear was failing; Chapter 10 got absorption panels on the wall behind her desk and a heavy blanket over the window alcove, and her office stopped sounding quite so much like a bathroom. Chapter 21 cleaned up her gain: voice tracks now arrive around -18 dBFS average, no red lights, no whisper-takes rescued in post.

And yet. Every week she edits her episode, and every week, somewhere around the second segment, the same thought: why does my show sound like a podcast, and theirs sound like a studio?

So she does what Chapter 4 trained and Chapter 22 demands: she stops guessing and runs the comparison. In Logic, she drops a produced segment from a network-grade news show she admires onto the track below her own latest episode, pulls the reference down to match her level — louder is not better; she knows the law before this chapter formalized it — and toggles between them with her eyes closed, one band at a time, Chapter 4's five-pass discipline compressed to three questions. Her notes, typed in the episode doc:

• theirs: voice sounds CLOSE but not thick. mine: thick but somehow far away
• mine has a blanket around 200–300?? warm but cardboard. theirs has none, and you can still hear chest
• their consonants are crisp but never spitty. my S's literally hiss on earbuds
• low fog on mine when the chair moves / when the furnace kicks in. theirs is dead silent underneath

Read those notes against the chapter's vocabulary and the diagnosis writes itself: low-frequency freight (below ~80 Hz), proximity-effect mud (the 200 neighborhood), small-room boxiness (the 300 neighborhood), and a sibilance problem that — flag it now — is intermittent, which means a static EQ is the wrong scalpel. Four symptoms. The first three are today's work.

Why a Spoken Voice Gets Exactly These Problems

Worth one paragraph of mechanism, because Aisha's symptoms are nearly universal in home spoken-word recording — yours will rhyme with hers.

The freight is the room and the furniture: chair creaks, desk bumps transmitted up the mic stand, HVAC rumble, a truck two streets over. All of it lives below the speaking voice's lowest useful energy, and all of it is invisible until it stacks against a loudness-normalized platform or a compressor (next chapter) that faithfully amplifies everything she didn't remove. The mud is partly her technique: she works her dynamic broadcast mic close — four to six inches, the broadcast intimacy position — and Chapter 7 named the bill for that closeness: proximity effect, a directional mic's bass boost at close range, stacking extra warmth onto every word, centered for her voice and distance right around 200 Hz. A little is the "radio voice" everyone wants. Every episode, all episode long, it's a duvet. The boxiness is the office itself: small-room reflections and modal buildup in the low-mids (Chapter 10's physics), recorded into the take — less of it since the panels went up, but her back catalog is full of it, and the alcove still adds a quiet cardboard halo around 300. And the sibilance is the mic's presence region meeting consonants that only sometimes bite — a dynamic problem wearing a static disguise.

One more pre-flight note, straight from this chapter: Aisha's voice is an alto; her spoken fundamentals live roughly between 165 and 200 Hz. That number is about to matter, because the difference between "clean" and "shoebox" is whether her high-pass respects it.

The Chain, Move by Move

Aisha builds the chain in Logic's stock Channel EQ — every DAW ships an equivalent parametric (Appendix E has the translations), and the stock one is all this work needs. She loops a two-minute stretch of a real episode: her voice, a guest clip, and the intro music bed, because the chapter's no-solo rule applies to podcasts too — her voice doesn't have to sound perfect alone; it has to sit with the bed and read clearly against the guest. The Channel EQ's built-in analyzer stays on as a hypothesis machine; her eyes go elsewhere for every verdict.

Move 1 — high-pass, swept by ear, parked at 85 Hz, 18 dB/octave. The freight filter. She starts the cutoff low and sweeps up while a chair-creak-and-furnace passage loops: at 60 the fog thins, at 85 it's gone, at 120 — the lemming number for voice, repeated in a hundred tutorials — her own chest tone starts evaporating and the take goes thin and newsreader-cheap. Back down until whole: 85. Matched A/B verdict: better, immediately — kept. Her note: "the silence under my voice got expensive." The lesson generalizes exactly as the chapter promised: her voice earned its number; a deeper male voice in the same chair might earn 70, a brighter one 100. The method, not the number, is the discipline.

Move 2 — the mud cut: -3 dB at 210 Hz, Q 1.2, wide and polite. The proximity duvet. A broad bell centered where the close-mic buildup lives, set wide because the problem is wide — tone-scale, not surgical. At matched loudness the toggle is unambiguous: bypassed, the thickness reads for two seconds as "warm"; engaged, her words step forward and the thickness is revealed as upholstery. The verdict that matters comes from context: against the music bed, her cut voice no longer fights the bed's own low-mids — un-masking, the chapter's quiet dividend. Kept. She tries -4.5 for good measure; on the blind toggle it's thinner but not clearer — different-not-better, bypass wins — and she returns to -3. The doctrine in miniature: cut to solve, not to win.

Move 3 — the boxiness hunt: sweep-and-destroy, conviction at 320 Hz, -4 dB, Q 3. The cardboard halo is narrower than the mud and harder to point at, so she runs the chapter's searchlight: one bell, +9 dB, Q 5, dragged slowly up from 200 while a dry passage loops. At 320 the office jumps out of the speakers — the honk blooms, unmistakably the sound of the room she's spent months treating. Flip to a cut, widen slightly to Q 3, depth by ear: -4 dB is where the cardboard leaves and -6 starts hollowing her vowels, so -4 it is. Matched verdict: better — kept. Two honest footnotes in her doc. First: this cut is for the room in the take — episodes recorded since the panels went up need less of it, and someday, in a properly treated space, maybe none; the EQ is compensating for history, not replacing Chapter 10. Second: she went hunting for a second resonance because the sweep made everything sound guilty — caught herself mid-Swiss-cheese, re-read her own ears, stopped at one conviction. The searchlight nominates; only the law convicts.

Move 4 — the presence experiment: +2 dB at 4 kHz... revised to +1, wide. Now the additive temptation, taken honestly. Her reference's consonants are crisp; hers, after three cuts, are clearer but still a half-step behind. So: a wide bell, +2 dB at 4 kHz, the speech-intelligibility zone. It sounds fantastic — for one toggle. Gain-matched and blind, the +2 version edges toward the glassy side on her guest's brighter voice segments and adds a whisper of hiss from the office's noise floor. At +1, wide (Q 0.9), it survives every toggle: diction without glare. Kept, at half the ambition. The boost-bias trap, experienced firsthand and journaled: "the first version was louder-sounding even gain-matched. brightness is sneaky. halve the boost, keep the win."

Move 5 — the air decision: none. The chapter's air-shelf honesty, applied. A +2 dB shelf at 10 kHz auditions as instant polish — and brings the room's hiss and her mouth noise up with it, because a dynamic broadcast mic in a home office records exactly as much "air" as there is, which is: some voice, some HVAC. On the blind toggle the shine doesn't pay for the noise. Bypass wins by default. Her note: "air is for the version of me with the silent room. future Aisha's shelf."

Move 6 — the sibilance handoff: zero EQ, by decision. The S's remain — fine in most sentences, needles on earbuds when a phrase leans on them. She tries the static answer anyway, as an experiment with a predetermined verdict: a -5 dB cut at 6.5 kHz kills the painful S's and lisps every other consonant in the episode — her T's go soft, her F's blur, the diction she bought in Move 4 drains back out. Reverted in under a minute. The problem is dynamic — it only exists on certain syllables — so the tool must be dynamic: a de-esser, the frequency-aware automatic gain device this chapter previewed and Chapter 29 teaches properly. She writes it at the top of the parking list, where it joins the show's other unsolved torment: her levels still whiplash — she leans in, guests drift off-mic, intros jump. That one has an appointment too. It's called Chapter 23, and it's next.

The Chain on Paper, and the Caveat That Makes It Professional

Aisha's voice — final chain (Logic Channel EQ, one instance)

  HP 85 Hz        −3 dB @ 210     −4 dB @ 320     +1 dB @ 4 kHz
  18 dB/oct       Q 1.2           Q 3             Q 0.9
  ╱┄┄┄┄┄┄┄┄┄╮    ╭┄┄┄╮           ╭┄╮             ╭┄┄┄┄┄╮
 ╱  freight  ╲__╱ mud ╲_________╱box╲___________╱diction╲┄┄┄┄┄ (no air shelf;
     out         out      out                    in, gently    no sibilance cut —
                                                               de-esser → Chapter 29)

Caption: the finished spoken-word chain — three cuts, one modest boost, two refusals; total processing depth nowhere exceeds 4 dB.

Caption: the session ledger — six decisions, four keepers, and the two refusals that taught the most.

She saves the chain as a channel-strip preset — Chapter 6's template thinking — and then writes the caveat that elevates the whole exercise, the one most preset-collectors never learn: this chain is for her voice, her mic, her distance, her room. The guest track from today's episode gets its own pass with the same method and lands on completely different numbers — HP at 70 for his deeper voice, no boxiness cut at all because he recorded in a carpeted closet, a bigger mud cut because he ate the mic. The preset stores the workflow. The numbers get re-earned every time, by ear, against the law.

The verification ends the way every mixing session in this book ends: bounced before/after, gain-matched, slept on, then auditioned on the three systems her audience actually uses — earbuds on the morning walk, the car, the kitchen speaker. The verdict survives all three. Her note to herself, last line of the doc, is the one worth stealing whole: "It doesn't sound EQ'd. It sounds like my office finally shut up and let me talk."