EQ

Every instrument has a few notes that ring louder than the rest. On an acoustic guitar it might be one string that booms through the body, or a whistle that leaps out whenever the melody touches it. On a voice it's often a hard edge on certain vowels. You don't want to lose the instrument. You want to lose the ring.

The EQ below has one narrow band, already boosted, sitting on the guitar. Press play and slowly move Frequency. Somewhere a note will jump out at you like it's been circled in red. That's the search. Now, before you do anything else, predict: if you flip Gain from +8 to −8 at the same spot, will the change be just as obvious?

Most people are surprised. The boost shouts. The cut, at the same size, usually sounds like the guitar relaxed a little. That asymmetry is the whole trick of corrective EQ.

Why does a boost shout and a cut whisper?

A narrow boost invents a peak that wasn't there, and your ear is built to notice peaks: a single loud frequency standing above its neighbours is exactly what a whistle, a feedback squeal or a ringing glass sounds like. A narrow cut does the opposite. It removes a sliver, and the frequencies on either side carry on almost unchanged, so your brain fills in the gap. The instrument keeps its identity.

So the shape on paper is symmetrical, but what you hear isn't. That's why engineers sweep with a boost (to find the problem fast) and fix with a cut (to remove it quietly).

It isn't magic, though. A cut through the body of a sustained note can still be obvious, and the asymmetry depends on what's around it. Always flip the band back to zero and compare.

What Frequency, Gain and Q do

The band in the lab is a bell, named for its shape. Three controls describe it:

• Frequency is where the centre of the change sits, in hertz.
• Gain is how much it boosts or cuts, in decibels.
• Q is how narrow it is. Higher Q means a thinner slice.

Try it in order. Find the jumping note with the boost, then flip Gain to a cut of 3 to 6 dB. Narrow it (raise Q) until it only catches the problem. Toggle the effect off and on. Does the ring come back? Does the guitar get something back too?

The goal isn't a flat spectrum. A guitar, a voice or a piano is defined by its peaks. Remove them all and you remove the instrument. You're choosing one change that helps this part in this song.

Don't let the search fool you

Here's the trap. Leave a big narrow boost on and sweep it across anything, and every frequency sounds like a problem. You're the one making the peak.

So before you cut, turn the boost back to zero and listen again. Is the ring still there, every time that note comes round? Then it's real. If it only exists while you're boosting, leave it alone. A good rule: find a candidate, confirm it at zero, fix it, stop. Endless sweeping makes a perfectly good recording sound suspicious.

Keep the listening level modest while you search. A narrow boost around 2 to 5 kHz can be piercing, because that's where human hearing is most sensitive. It's also why a slight bump there makes a voice "come forward" and too much makes it hurt.

Choose the shape for the job

A bell is one tool. Here's how the others feel, and what each is for:

What you hear	Shape to try	What it does
A narrow whistle or ring on certain notes	Bell cut or notch	Removes a thin slice; the rest of the note stays
Rumble or boom under everything	Low cut	Removes everything below a point, like a floor being swept
Too much weight, but the low notes matter	Low shelf	Turns the whole low region down by a set amount
Dull, but the sound was recorded well	High shelf	Lifts everything above a point, the "air" knob
Hiss or harshness at the very top	High cut	Rolls the top off, like a softer microphone

A shelf turns a whole region up or down and then levels off, like a step. A low cut (also called a high-pass filter) removes everything below its frequency. A high cut (a low-pass filter) does the opposite. A notch is an extremely narrow cut.

Here's the most common voice move in modern pop, ready to play: a low cut to clear rumble and breath thumps from below the voice, and a gentle high shelf for "air", the breathy sheen above about 8 kHz that makes a vocal sound expensive and close.

Cuts have a slope, measured in dB per octave: how steeply they fall away past their frequency. A steeper slope isn't automatically better. A steep low cut set too high takes away the chest of a voice along with the rumble.

An EQ can only turn up what's there

Boosting the top of a dull recording can't add detail that the microphone never caught. If the recording has nothing up there but hiss, a high shelf makes the hiss louder. EQ rebalances what exists. It doesn't invent.

That's the big difference between EQ and distortion, which does create new frequencies. When a bass is too dull to hear on a phone speaker, an EQ boost at 60 Hz won't help (the phone can't play 60 Hz), but a little saturation adds harmonics the phone can play.

Listen in context, at the same level

A correction that sounds great solo might not matter in the song, and a guitar that sounds thin alone might be exactly right next to the bass and drums. Make the final decision with everything playing.

And compare at the same loudness. Louder almost always sounds better, so a boost can fool you into thinking it helped. The EffectLab's bypass is level-matched: switch it and you're hearing the tone change, not the volume change.

Where a good correction breaks

Singers move. A resonance in one vowel shifts when the mouth changes shape, so a fixed narrow cut that fixes "ah" might scoop a hole in "ee". Listen across a whole phrase before you call it done. A wider, shallower cut, or fixing only the one offending word, often keeps more of the performance.

Same idea, different place: Timbre and harmonics explains why a piano and a violin playing the same note sound different. An EQ changes that same balance of harmonics, so every cleanup is also a small act of sound design.

Make it yours

Pick a different sound from the one you searched: your own voice, a bass line or some keys. Before touching a band, write one sentence about what bothers you: "the low note booms", "the S sounds sting". Choose the shape from the table and make the smallest change you can hear.

Check the lowest note, the highest note and the quietest passage. Then listen with the rest of the song. If it only helps solo, ask whether it's needed at all.

When it earns its place, open Parametric EQ from the lab and save the setting with a name that says its job, like "Acoustic, tame 1k ring". It won't fit every guitar, but next time you'll know where to start looking.

Quick check

1. You sweep a big narrow boost and every frequency sounds awful. Is the recording full of resonances? Probably not. The boost creates the peak you're hearing. Return to zero and only fix what you can still hear there.
2. The lowest bass note matters, but the low end is too heavy overall. Low cut or low shelf? Try a low shelf first. It turns the region down without removing the bottom note. A low cut high enough to help can take the note away with the weight.
3. A high shelf on a dull recording mostly adds hiss. Why? There was little musical detail up there to lift. An EQ can only rebalance what was recorded. For more presence, try a cut in the muddy low mids, or a little saturation to generate new upper harmonics.

Next steps

• Go deeper: Filters as instruments. What happens when the band doesn't stay put, and moves while you play?
• Use it: Making room with EQ. Why every track sounds fine alone and muddy together, and how small cuts fix it.