Choosing Which EQ Cut to Make First: A Cooking Salt Analogy for System Tuning

You've got a new PA rig in a room you don't know. The band is about to soundcheck, and the console has 32 bands of EQ per channel. Your hand hovers over the initial parametric filter. Where do you cut? That opening move is like grabbing the salt shaker before you've tasted the soup. Do it flawed, and you're chasing a bad seasoning job all night.

Here's the thing: most tuning guides treat EQ cuts as a laundry list—cut here, boost there—without explaining why the batch matters. But in practice, the sequence of your cuts reshapes the entire response. One well-placed cut can fix a boxy low-mid, tighten the kick, and even reduce vocal muddiness. Miss it, and you'll pile on half a dozen bands trying to clean up what one filter could have done. This article walks you through a cooking-inspired method: salt initial, then taste, then adjust. No jargon for jargon's sake. Just a way to think about EQ that keeps your mix clear and your audience happy.

Why the opening Cut Matters More Than Any Boost

The glitch with reactive tuning

Most engineers I watch start tuning the faulty way. They hear something ugly—a honk in the mids, a boxy low-mid—and reach for the nearest EQ band to slice it out. That feels productive. It solves the immediate irritation. But here is what nobody tells you at the start of the night: that initial cut becomes the lens through which you hear everything else. You notch out 250 Hz because the floor tom sounds thick. Now the kick drum sounds thin. So you boost 60 Hz. Now the bass guitar eats the kick. You chase a ghost. The issue isn't the frequencies—it's the batch you attacked them. Reactive tuning locks you into a chain of compensating moves that spread phase rotation across the spectrum and bury the original issue deeper. I have stood behind consoles where three hours of this produced a mix that sounded worse than the raw stack.

How bad initial moves compound

Think about what a filter does beyond gain reduction. Every cut introduces a phase shift—the signal is delayed slightly around the center frequency. One cut, tiny shift. Two cuts overlapping? The shift accumulates. By the phase you have carved out four or five resonances, the phase curve looks like a mountain range. Transient response blurs. Snare hits lose their crack. Kick drum beater attacks smear into a woolly thump. The real gut-punch: you can't undo this by adding more EQ. The damage is baked into the signal path. The only fix is to bypass everything and start over, which nobody does mid-set. So the framework stays muddy. That's why the sequence matters more than the frequency—the initial cut determines how much phase error you will carry for the rest of the night.

off queue compounds in another way too. You cut a resonance at 1.2 kHz. Now the ear hears a dip there and masks it by turning up the channel. But the resonance was only ringing at certain times—when the snare hit, when the vocalist leaned into a hard consonant. By cutting full-slot you dulled every snare hit and vocal sibilant. The boost you added later to compensate? That raises the noise floor and brings back the resonance in other material. The band sounds different on every song. Not flexible—broken.

Real-world stakes: a club gig gone off

I watched a friend tune a 300-capacity room in fifteen minutes before doors opened. One dominant resonance—about 160 Hz, from a corner-loaded sub stack. opening move: a wide bell cut, 4 dB. Clean. The kick sat right. The bass locked. Two hours later, the room filled and the resonance came back. Warmer air, more bodies reflecting energy—the room changed. But that primary cut was too broad and too deep. He had no headroom left to narrow it or adjust gain. The whole low end felt sucked out, and when he tried to boost his way out, the subs started farting. The set ended early. The promoter blamed the PA.

'The primary cut is not a surgical incision. It's a bet on what the room will do all night.'

— veteran framework engineer, after watching that gig fold

That night cost my friend a repeat booking. What he should have done: measure opening, find the narrowest possible band that tamed the ring, cut just enough to make it behave, then walk the room. Instead, he guessed wide and deep. The catch is simple—the initial cut sets the ceiling for everything after it. Make it too aggressive and you paint yourself into a corner you can't climb out of with more EQ. Make it too narrow and you miss the root cause. Make it at the off frequency and you mask the real glitch, which is often boundary interference, not the speaker itself. That's why the initial cut matters more than any boost. A boost is just volume. A cut is a structural change to your setup's timing.

The Cooking Salt Analogy: Taste Before Seasoning

Why chefs add salt early

A good cook doesn't dump salt on a finished plate. They season the water before pasta goes in, salt the steak an hour before it hits the pan. That early salt doesn't just sit on the surface—it gets pulled into the food, changing the structure of proteins, drawing out moisture, building flavor from the inside out. Taste before seasoning, every slot. The same logic applies to setup tuning: you don't reach for a boost until you've dealt with what's already flawed. Most engineers I know skip this step. They grab a parametric EQ, hear something harsh in the high-mids, and immediately add 4 dB at 3 kHz. That feels productive. It isn't.

The EQ equivalent: cut the loudest resonance opening

Every stack has a dominant resonance—a frequency that rings longer, sounds harsher, or eats up headroom before anything else can breathe. That resonance is the raw, unseasoned ingredient. If you boost before cutting it, you're not fixing the snag; you're adding gain to a signal that already has too much energy in that region. The result? Muddier low end, ear fatigue inside ten minutes, and a mix that falls apart when the room fills with people. Cut initial. Find that peak—sweep a narrow band filter with high Q, listen for the ugliest bump—and pull it down 3 to 5 dB. One filter. Instant clarity. I once fixed a club framework that sounded like a wet blanket over the whole PA. The engineer had boosted 80 Hz and 250 Hz trying to get more warmth. Both bands were already 8 dB too hot from boundary loading. We cut both by 5 dB and the stack opened up—no additional processing required. That's the power of subtraction.

„You can't polish a resonant mess. You have to scrape it off opening.“

— live sound engineer, after watching a rookie setup tech boost 400 Hz to „fix“ a boxy vocal mic

Field note: audio plans crack at handoff.

Watershed buffers, riparian corridors, sediment traps, canopy gaps, and nesting cavities respond to disturbance on mismatched clocks.

Fjords kelp basalt look wild.

Sail battens, reefing lines, winch handles, telltales, and tide tables punish skippers who trust apps alone.

Fjords kelp basalt look wild.

Field note: audio plans crack at handoff.

What happens if you boost before cutting

Thing is—boosting a frequency that's already dominating doesn't make the stack louder. It makes it smaller. You push the amplifier into limiting sooner, trigger driver protection circuits, and shrink your dynamic range. Worth flagging—this also introduces more phase shift at the filter's center frequency, which can smear transients and make the stack feel sluggish. The catch? You won't notice until you walk the room during soundcheck and the vocal sounds hollow, the kick drum loses punch, and the snare has no attack. That's not a tuning snag anymore. That's a salvage operation. I've watched engineers boost 5 kHz to get „air“ from a tweeter array that was already peaking at 6 kHz from a factory crossover glitch. The boost made the distortion worse. The fix? One cut at 6.3 kHz, −4 dB, and the framework finally had the top end they were chasing. off batch. Not yet. That hurts.

So the rule is simple: treat the room's acoustic signature like an ingredient that needs seasoning, not a canvas that needs paint. Cut the loudest thing initial. Taste. Then decide if you actually need a boost—and if you do, you'll need less of it. The stack will run cooler, the amps will breathe, and your mix will translate to a full dance floor instead of sounding like a car stereo in a concrete box.

What Happens Inside the Filter: Gain Reduction and Phase Shift

How a parametric cut reshapes the spectrum—and why it matters

When you dial in a cut on a parametric EQ, you aren’t merely turning down a frequency like a volume knob. Inside the filter, a carefully sculpted dip is carved into the signal, and that carve is never perfectly surgical. A 6 dB cut at 125 Hz with a wide Q (say, 0.7) pulls down not just the issue note but its neighbors—100 Hz, 160 Hz, often more. That sounds fine until you realise you just gutted the low-end punch of a kick drum that was sitting at 110 Hz. The trade-off is immediate: you fix a resonance but you also thin the instrument’s body. I have seen engineers chase a 200 Hz honk with a deep cut, only to wonder why the snare lost its weight. Look at the RTA after the filter: the dip is wider than you thought.

Narrow Q cuts promise precision, but they bring a different headache. A Q of 10 or higher produces a deep, skinny notch that can target a single ringing harmonic. The gain reduction is tight—good for stopping that 63 Hz mains hum without touching the bass guitar. However, the phase shift inside that narrow filter is abrupt. Signal components near the centre frequency get delayed relative to the rest of the spectrum. In a live framework with multiple channels, this phase rotation can create a subtle smear on transients. Most listeners won’t hear it on a snare hit alone, but stack three narrow cuts across the midrange and the whole mix starts to sound disconnected—like the drums are playing two feet behind the vocals. The catch is that narrow cuts are addictive: they fix one thing without ruining the whole band, yet cascading them silently degrades coherence.

‘A dozen narrow cuts on a graphic EQ often sound worse than two wide, gentle cuts. The phase pile-up is the culprit.’

— veteran framework tech, after tuning a festival FOH with 14 notches

Cascading cuts behave differently than boosts—here is why

There is a dirty secret about stacking filters: cuts interact non-linearly. Two 3 dB cuts at adjacent frequencies can sum to a 5.5 dB dip if the Q settings overlap, but they can also produce a small, unexpected bump where the phase responses collide. Boosts, by contrast, are more forgiving because the filter’s phase shift adds constructively with the original signal, masking some of the rotational mess. This asymmetry is why I always cut initial and boost later—and why I never stack more than three parametric cuts per channel unless the room forces me. The pitfall is visible on a transfer function measurement: cascaded cuts often create a wavy floor below 500 Hz that you didn't dial in.

What usually breaks initial is the monitor wedge. You cut a 160 Hz resonance in the vocal mic, then cut a 200 Hz build-up from the guitar amp bleed, then notch 250 Hz because the floor is flapping. Suddenly the vocal sounds hollow and phasey. The fix is brutal: reset the channel EQ, make one wide cut at 200 Hz with a Q around 1.8, and accept that the resonance is only partially suppressed. off batch. Not yet. The better move is to walk the room and identify the dominant resonance—usually the one that rings at the same frequency on every microphone—before touching the channel strips. That single, intentional cut at the setup level fixes more than three desperate notches on individual inputs.

One concrete anecdote: We tuned a small club with a persistent 80 Hz boom. The subs were stacked on a hollow stage, and every vocal mic picked up the thump. Instead of cutting each mic, we applied a single -4 dB cut at 80 Hz with Q 2.0 on the main output bus. The boom disappeared. The bass guitar, fundamental around 80 Hz on the low E string, lost a bit of weight—so we nudged the bass channel +1.5 dB at 80 Hz with a wider Q (0.8). The phase mismatch was negligible because the bus cut and the channel boost were opposite but not identical in shape. That's the asymmetry in action: cut wide, boost narrow if needed. The setup stayed coherent, the vocal mics stopped thumping, and we didn't reach for a dozen notches. Next phase you tune, watch the phase trace on the console’s built-in FFT. It will show you exactly where your cuts start to bite back. Use that data.

Step-by-Step: Tuning a Small Club stack with One Dominant Resonance

Identifying the issue frequency

Walk into the club. The room is empty, the subs are stacked on the floor, and the opening track you play has a low-end hum that feels like a fist pressing against your chest. Not the good kind. Before touching any EQ, you need to isolate the offender. I run a slow sine sweep from 30 Hz upward—just a few seconds per tone. The resonance hits hard around 63 Hz. The meter jumps 8 dB. The walls vibrate audibly. That's your dominant resonance. You could reach for a boost at 100 Hz to cover it—wrong queue. The catch is that boosting only masks the glitch and eats headroom. You fix the hum primary, then decide if the framework actually needs more low-end weight.

Applying the primary cut: frequency, Q, gain

Set your parametric EQ to 63 Hz. Start with a gain of –6 dB. Use a Q of 2.0—tight enough to grab the resonance without scooping out neighboring musical material. Listen. The hum drops, but the kick drum now sounds thin. That's the trade-off: over-cutting kills impact. Nudge the gain to –4 dB and widen the Q to 1.5. The resonance is tamed, and the kick retains its body. Most engineers skip this step—they dump the full –6 dB and move on. What happens then? Phase shift accumulates, and the stack sounds hollow. We fixed this by making the opening cut just enough to remove the irritation, not enough to sterilize the mix.

“Cut until the issue disappears, then back off 1 dB. That last dB is where the setup breathes.”

— old monitor engineer, explaining why he never trusts his ears after a long soundcheck

Layering additional cuts and the listening check

One dominant resonance is rarely alone. After the 63 Hz cut, sweep again. A secondary peak shows up at 112 Hz, about 4 dB hot. Apply a cut here: –3 dB, Q 1.8. Now play a full-bandwidth track—something with bass, vocals, and cymbals. The low end is clean, but the vocal has a honky, nasal quality. That's phase rotation from the initial two filters interacting. Add a gentle shelf cut at 250 Hz, –2 dB, Q 0.7. Not a notch—just smoothing. Now check the stack at three positions: center dancefloor, back bar, and near the stage. The resonance is gone everywhere? Good. If it returns near a wall, that’s boundary interference—handled in the next section. The point here: don't stack cuts blindly. Trust the listening check over the spectrum analyzer. I have seen engineers carve out four resonances on a graph, then walk away to a mix that sounds like a telephone. One dominant cut, a secondary trim, and a shelf. That's usually enough. That's the salt—applied, tasted, adjusted.

Seed starts, soil amendments, trellis tension, pollinator strips, and harvest windows punish vague calendars in wet seasons.

Ember nexus clamps seize overnight.

Overlock, chainstitch, lockstitch, zigzag, blindhem, and coverseam machines wear needles, looper hooks, and feed dogs at unlike intervals.

Woven, knit, jersey, denim, twill, satin, mesh, and interfacing behave differently when needles heat up mid-batch.

Fjords kelp basalt look wild.

Ember nexus clamps seize overnight.

Not every audio checklist earns its ink.

Not every audio checklist earns its ink.

When the Room Fights Back: Boundary Interference and Comb Filtering

Subwoofer Placement and the initial Cut Blind Spot

The cooking salt analogy assumes the issue is in the seasoning—the EQ. That works when the resonance is born inside the speaker or the room's natural mode. But a null from floor bounce? That's not a spike you can salt away. I have watched engineers burn four parametric filters trying to fix a 45 Hz suck-out that turned out to be a sub two feet from a side wall. Cutting at a cancellation frequency does nothing useful—you're pulling gain from a hole that's already empty. The room wins that round. What usually breaks opening is the realization that the null exists: you sweep a sine wave, the level drops 12 dB at 63 Hz, and your instinct says "cut that." Wrong move. You need placement, not EQ. Pull the sub away from the boundary, measure again, and suddenly the "resonance" you wanted to cut was never a resonance at all—it was a phase cancellation disguised as a mode.

A floor bounce creates a comb filter pattern. The initial deep notch depends on listener height and distance to the speaker. If you cut that notch with a narrow Q, you introduce a phase rotation that makes the adjacent frequencies smear in phase. The fix sounds worse than the glitch. Worth flagging—this is where the salt analogy needs a hard note: you can't season the dish if the pot has a hole in it. Move the pot.

Multiple Strong Modes: Which Resonance Gets the Knife opening

When the room has two or three dominant modes within an octave, the "cut the loudest one" rule gets messy. Apply a cut at 80 Hz—that shifts phase response at 105 Hz, which was your second-biggest snag. Now that mode rings longer. I have seen this cascade: a club stack with a 2 dB peak at 63 Hz and a 4 dB peak at 95 Hz. The engineer cut the 95 Hz initial. The 63 Hz peak then measured 5 dB because the previous cut altered the summed energy. That hurts. The trick is to measure with both subs running, apply the cut to the mode that causes the most audible ringing—usually the one with the longest decay slot, not the highest amplitude—and re-measure before touching the second peak.

If the decay times are nearly identical? Cut the lower mode initial. Lower-frequency modes have wider spatial influence and longer wavelengths. Cutting them early gives more stable results across the listening area. The catch is you might need a second pass after the upper mode settles. This is not a one-shot procedure—it's iterative, like adjusting salt in a reduction sauce. Taste, adjust, taste again.

“I cut the highest peak initial for years. Then I chased a moving target for three hours. Lower mode opening, every window now.”

— engineer after a brutal afternoon tuning a concrete block room with dual 18-inch subs

Systems with Limited EQ Bands—Where the Analogy Breaks Entirely

Not everyone has eight parametric filters. A three-band graphic EQ on a small PA rig forces brutal decisions. One cut eats your only low-frequency band. That means you can't also fix a high-mid harshness without sacrificing the sub region. The salt analogy still applies, but now you have one shaker with three holes. Do you use it on the low resonance or the vocal smear? I have faced this in a basement bar—one dominant 125 Hz boom, but the vocal mic was feeding back at 2.5 kHz. Cutting 125 Hz fixed the boom, but the feedback remained because the low-mid band was now shaping the 250 Hz range, which changed how the vocal wedge coupled with the mains. The answer was not a single cut—it was tilting the whole framework's gain structure. Sometimes the primary cut is actually a high-pass filter shift. Let the subs handle only what they must, roll off the mains higher, and the resonance disappears without touching the EQ bands you need for feedback control. That's not cutting primary. That is cutting differently. The analogy bends, but it doesn't break—you're still removing excess before adding anything. You just pick a different tool.

The Limits of Cutting opening: What This Approach Won't Fix

When the stack has a poor frequency balance overall

Cuts are surgical tools, not rebalancing wands. I have walked into clubs where the whole framework sounded like a wool blanket had been thrown over the mains — everything below 200 Hz boomed, everything above 8 kHz hissed, and the midrange barely registered. Making one strategic cut on a 63 Hz resonance might clean up the low end, sure, but the tonal imbalance remains. The vocal still disappears. The cymbals still bite your face off. That is not a resonance snag; that's a framework that was tuned with a blindfold and a prayer. What usually breaks initial in these scenarios is your own patience — you cut one peak, hear a new issue emerge, cut that, and suddenly the setup feels thin and lifeless. The catch: you have traded boom for anemia.

Worth flagging — broad-spectrum EQ grief requires a different initial move entirely. You start with a target curve that makes musical sense, then shape the setup to that curve using shelving filters or parametric bands set to gentle Q values. Cuts can still live inside that workflow, but they're no longer the opening act. Wrong batch, and you chase your tail until showtime.

Why cuts can't fix phase alignment

Here is where the salt analogy finally breaks down. You can oversalt a stew and still rescue it with potato chunks or dilution. You can't EQ your way out of a subwoofer that arrives 12 milliseconds after the top box. That destructive cancellation — the suckout around 100 Hz that makes the kick drum sound like a cardboard box — is not amplitude, it's timing. Cutting the offending frequency won't help; the cancellation depth and null position are fixed by path length difference. I fixed a conference room once where the engineer had carved a 6 dB hole at 80 Hz trying to fix a muddy low end. The real culprit? The subs were stacked 3 meters behind the mains. No cut fixes physics.

The tricky bit is distinguishing between a resonance that responds to reduction and a cancellation that demands delay alignment or physical repositioning. Quick test: walk the room. If the null moves or changes when you take two steps sideways, it's comb filtering, not a ringing mode. Don't reach for the EQ — reach for a tape measure and a DSP delay tool instead.

The danger of over-cutting and losing headroom

Every cut removes energy from the signal. That sounds obvious until you see a stack where someone has pulled down five or six narrow bands by 8 dB each — a thin, phase-shifted mess that still clips the amplifiers because the low-end rumble never got addressed.

'I cut every frequency that looked spiky in RTA. Now the framework sounds dead and the limiters still flash red.'

— A club owner’s frustrated text after a weekend of 'tuning'

Spec sheets, torque tolerances, pneumatic feeds, laminate rollers, and ultrasonic welders each demand separate maintenance cadences.

Letterpress quoins reward slow hands.

Spreading, layering, bundling, ticketing, shading, bundling, and nesting affect yield long before the operator touches pedal speed.

Letterpress quoins reward slow hands.

What actually happened: each cut reduced the average level, so they cranked the input gain to compensate. The peaks that originally triggered the limiter were untouched — just masked by the overall level drop. The result? Less headroom than when they started. I have seen this pattern repeat in small clubs and festival side-stages alike. The discipline is to cut only what audibly hurts, not everything that looks ugly on a graph. Leave the rest alone.

Flag this for audio: shortcuts cost a day.

Flag this for audio: shortcuts cost a day.

Reader FAQ: Common Questions About initial-Cut Strategy

Should I always cut before boosting?

Not always — but almost always. The exception is a stack that simply lacks energy above 10 kHz or below 50 Hz, where the drivers themselves are rolled off naturally. Boosting there is safe because there's nothing to cut opening. But inside the vocal range — 80 Hz to 4 kHz — cutting a dominant resonance before boosting anything else keeps your headroom intact. I have seen engineers boost 6 dB at 250 Hz to make a kick drum punchier, only to discover a 300 Hz room mode that now sounds like a cardboard box. By then they have already eaten up amp power and introduced phase rotation. Cut the resonance initial. Listen. Then decide if you still need a boost. Usually you won't.

How narrow should my initial cut be?

Narrow enough to fix the issue, wide enough to stay musical. A Q of 10 or higher can kill a single room mode without touching adjacent notes — that's fine for a pure 63 Hz hum from a ventilation fan. For a boxy resonance centered at 250 Hz that smears male vocals, start with a Q around 4. Too wide (Q of 1) and you suck the life out of every guitar chord; too narrow (Q of 20) and you might leave a hollow, ringing artifact that sounds worse than the original. The catch is that a narrow filter also introduces more phase shift around the center frequency. That shift can make transients sound smeared or "late." Worth flagging: a narrow cut on a midrange peak often sounds cleaner than boosting the adjacent region, but only if you test it at show volume, not at whisper level.

What if I have multiple snag frequencies at the same level?

Tackle them one at a time, starting with the lowest-resonant frequency that dominates the sound. Why? Low-frequency energy masks the harmonics above it. I once walked into a small club with a 70 Hz hump and a 110 Hz bump that both measured within 1 dB of each other. initial instinct — cut both. Wrong sequence. We cut only the 70 Hz peak opening, and the 110 Hz bump dropped 3 dB on its own because the subwoofer stopped over-excursing and unloaded the crossover region. The remaining 110 Hz peak then needed only a gentle 2 dB dip. Most teams skip this: they load twenty RTA bands and carve everything flat, then wonder why the setup sounds thin. Don't resemble a fishing net. Cut the lowest offender, re-measure, re-listen, then decide if the second peak still matters.

“I cut the room mode primary. The listener thought I replaced the subwoofer. I had only moved one fader.”

— veteran stack engineer, after a three-minute tuning session at a 400-capacity venue

Can I trust a measurement mic over my ears for the primary cut?

Trust both, but for different reasons. The mic sees what your brain filters out — that 50 Hz electrical hum you stopped noticing after the opening song, or the narrow 3.2 kHz ring that the vocalist hears as "ice pick." Your ears, however, tell you what actually hurts the mix: the mud that masks the snare, the honk that makes the piano sound like a toy. The pitfall is that beginners stare at a transfer function graph and start cutting every peak that exceeds the target curve, ignoring whether those peaks correlate with musical content. That hurts. A 6 dB spike at 180 Hz might look scary on screen, but if the bass player's open A string lives there, a deep cut kills the groove. Use the mic to locate the resonance — then listen for two full minutes before you touch the EQ. If the snag is real, you will want to fix it. If it's an artifact of the measurement position, you will hear the difference and leave it alone. That is the workflow: locate with data, decide with ears.

Takeaways: A Repeatable Workflow for Any stack

Listen, cut, reassess — in that sequence

Walk up to any framework and your hands will itch for the faders. Resist. The repeatable workflow I fall back on every time has three moves, and they never change. initial: listen to the problem without touching anything. Play a track you know — something with a kick drum, a vocal, and a ride cymbal. Walk the room. Find the one frequency that makes your teeth ache. Second: cut that single resonance by 3 dB with a Q no wider than 1.4. Third: listen again from the same spot. If the ache is gone but the system sounds dull, you overshot. Pull the gain back to -2 dB. If the ache remains, nudge the frequency up or down 10 Hz — room modes are rarely where the meter says they should be.

The catch? Most people cut twice before they listen once. I have seen engineers stack four EQ cuts from a measurement mic, step off the stage, and wonder why the room feels dead. The mic doesn't care about musicality — you do. That is why the third step is non-negotiable: reassess with your ears, not a graph. One cut, one listen, one adjustment. Then stop.

When to break the order — and why it costs you

Sometimes the room forces a deviation. A blown tweeter, a feedback loop that won't wait, a guest engineer who already boosted 5 kHz because "it sounds clearer." In those cases, cut the worst offender first — even if you haven't fully listened. Then fix the rest later. But here is the trade-off: every time you break the sequence, you lose the clean baseline. The next cut will be a guess, not a decision. I had to do this once in a club where the previous guy had left a +6 dB shelf at 60 Hz. We cut the resonance first, then had to pull the whole low end back — and still ended up with a phase shift that killed the kick's attack. Not ideal. The workflow works when you follow it. Deviate, and you're fixing someone else's mistakes on top of your own.

'A good first cut fixes the room. A bad first cut hides it — and you will chase that ghost all night.'

— overheard from a monitor engineer after a 32-band graphic EQ fight

One final test before the first song

After your last cut, play a full mix at show level — not a sine sweep, not pink noise, the actual music. Walk the perimeter of the listening area. If the system sounds balanced from the bar, the dance floor, and the back wall, you're done. If the vocal disappears in one corner, you missed a boundary reflection. If the kick turns spongy, your cut is too wide — the phase shift is smearing transient information. This is the test that most people skip. They chase a flat transfer function on a laptop and call it tuned. Wrong. Tuning is finished when the system disappears — when nobody in the audience thinks about the PA. They just dance.

One more thing: save that final EQ file under a date-stamped name. Next week when you walk into the same room, the resonance will have shifted because the humidity changed or someone moved a table. You will need that original cut as a starting point, not a permanent solution. Systems drift. Your workflow should not.