EQ Basics: Cut First, Boost Last

What EQ Actually Does (and What It Can't)

An equalizer (EQ) is a tool that turns the volume of specific frequency ranges up or down. That's the whole job. It doesn't add warmth, magic, or "analog character" in any meaningful sense a beginner needs to worry about — it makes some frequencies louder and others quieter.

That sounds almost too simple to be the most important mixing tool you own, but it is, for one reason: every instrument in your session is fighting for the same frequency ranges. A voice, an acoustic guitar, a piano, and a synth pad all pour energy into the midrange at the same time. When two sources occupy the same frequencies at similar levels, the louder one hides the quieter one. Engineers call this masking, and it's the main reason home mixes sound cluttered and muffled while professional mixes sound clear.

EQ is how you resolve those fights. And the most reliable way to use it — especially while you're learning — is to cut before you boost.

One thing EQ cannot do: fix a bad recording. If the source is thin, boomy, or harsh because of mic placement or a bad room, EQ can reduce the damage but never remove it. Ten minutes moving the mic beats an hour of surgical EQ every time. That's not a slogan; it's how the physics works. EQ can only reshape what's there.

Why Cut First: The Case for Subtractive EQ

Subtractive EQ means shaping a sound primarily by removing frequencies you don't want, rather than boosting the ones you do. There are three practical reasons this approach wins for home recordists.

1. Your ears lie about "better"

Human hearing has a well-documented bias: louder almost always sounds better, at least for the first few seconds. Boosting any frequency band makes the track louder overall, so a boost nearly always sounds like an improvement in the moment — even when it isn't. You reach for a boost, it sounds "bigger," you keep it, and twenty boosts later your mix is a loudness arms race with no space left in it.

Cuts work against that bias. When you cut a frequency range and the track still sounds good — or sounds clearer — you know the improvement is real, not just a volume trick. A useful habit: after any EQ move, compensate the track's output gain so the overall loudness matches the bypassed version, then A/B them. If the move only sounds better because it's louder, you'll catch it.

2. Cuts preserve headroom

Headroom is the space between your signal's level and the point where the system distorts (0 dBFS in a digital system — decibels relative to full scale, where 0 is the maximum). Every boost pushes signal level up and eats headroom; stack boosts across 20 tracks and your mix bus is slammed before you've done anything deliberate with the level. Cuts move you the other direction. A mix built on subtraction arrives at the mix bus with room to breathe, which makes every downstream decision — compression, limiting, mastering — easier.

3. Cuts solve masking directly

Masking is a "too much stuff in one place" problem. Boosting the vocal to fight a piano that's crowding it just makes the pile bigger. Cutting the piano in the range where the vocal lives actually removes the conflict. Subtraction is the only move that makes space.

Is "never boost" a rule? No — and honest engineers disagree about how strictly to apply the cut-first principle. Plenty of great mixers boost freely and get superb results, because they've spent years calibrating their ears against the loudness bias. The point isn't that boosting is wrong. It's that cutting is the safer default while your ears are still in training, and it fails more gracefully when you get it wrong.

The Controls: What Every Knob on an EQ Means

Most stock DAW EQs are parametric, meaning you control three things per band. Learn these once and every EQ plugin ever made becomes familiar.

• Frequency — where on the spectrum the band operates, measured in hertz (Hz, cycles per second). Human hearing spans roughly 20 Hz to 20 kHz. Low numbers are bass; high numbers are treble.
• Gain — how much you turn that range up or down, in decibels (dB). Cuts are negative values, boosts positive.
• Q (bandwidth) — how wide or narrow the affected range is. High Q = narrow, surgical. Low Q = broad, gentle. Broad moves sound more natural; narrow moves are for isolating specific problems.

And the common band shapes:

• Bell (peak) — boosts or cuts around a center frequency, tapering off on both sides. Your general-purpose tool.
• High-pass filter (HPF) — removes everything below a chosen frequency and lets highs pass through. Also called a low-cut. The single most-used EQ move in mixing.
• Low-pass filter (LPF) — the mirror image: removes everything above a frequency.
• Shelf — boosts or cuts everything above (high shelf) or below (low shelf) a frequency by a fixed amount. Gentler and broader than a bell; good for overall tilts like "a touch less dark."
• Notch — an extremely narrow, deep cut for killing one specific problem, like electrical hum.

A graphic EQ — the row of fixed sliders you see on live consoles — does the same job with less flexibility. In a DAW there's rarely a reason to prefer it.

A Practical Subtractive Workflow

Here's a repeatable order of operations for EQing any track. Work with the song playing, not the track soloed — more on that below.

Step 1: High-pass the junk

Almost every track that isn't a kick drum or bass instrument carries low-frequency energy that contributes nothing: mic stand rumble, room resonance, traffic, plosive thumps, the proximity effect from close-miking. Individually inaudible, collectively it's mud that eats headroom and makes your low end lumpy.

Put a high-pass filter on the track and raise its frequency slowly while the full mix plays. Stop when you hear the track start to thin out, then back off slightly. On a vocal that might land somewhere around 80–120 Hz; on a bright acoustic guitar strumming in a dense mix, considerably higher. Let context decide, not a number you read somewhere — including here.

One honest caveat: don't high-pass everything reflexively at a high setting. Stripping the low mids out of every track is a classic route to a thin, small-sounding mix. Filter what's genuinely junk; keep what's body.

Step 2: Find and cut the problems

For resonances — specific frequencies that ring out unpleasantly, often from the room or the instrument itself — the classic technique is the boost-and-sweep: create a narrow bell, boost it hard (8–10 dB), and sweep it slowly across the spectrum. When a frequency jumps out as ugly, honky, or ringing, you've found a candidate. Then flip the boost to a cut, start narrow and shallow, and deepen only as far as needed.

Use this technique with an honest warning attached: everything sounds bad boosted 10 dB narrow. Sweeping will tempt you to cut a dozen frequencies that were never actually a problem in context. The fix is discipline: after each sweep-found cut, bypass it while the full mix plays. If you can't hear the difference, delete the cut. Most tracks need zero to three real cuts, not ten.

Typical problem zones: boxiness around the low mids, honk or nasality in the midrange, harshness in the upper mids. The table below gives the map.

Step 3: Carve space between fighting tracks

This is where subtractive EQ earns its keep. Find pairs of tracks that mask each other — vocal vs. guitars, kick vs. bass, piano vs. everything — and cut one to make room for the other. The move is usually a broad, gentle cut (1–4 dB, low Q) in the supporting instrument, right where the featured instrument's core energy lives. Cutting a few dB of midrange from thick guitars to clear a lane for the voice does more for vocal clarity than any boost on the vocal channel. This carving is most of what making a vocal sit in a mix actually consists of.

Step 4: Boost last, broad and small — if at all

Once the junk is filtered, the problems cut, and the space carved, listen again. Often the track already sounds "boosted" — because removing mud and masking is what clarity sounds like. If it still needs character — air on a vocal, snap on a snare — boost, but broad (low Q) and modest (1–3 dB). If you want more than about 4–6 dB of boost, the problem is usually the recording or the arrangement, not the EQ.

Frequency Guide: What Lives Where

Ranges below are approximate and overlap — real instruments spread energy everywhere, and no chart substitutes for sweeping and listening. But as a starting map, this is the mainstream consensus:

Two ranges deserve special respect from home recordists. The 200–500 Hz zone is where small untreated rooms dump resonance into every close-miked track — it's the most common cut in home mixing. And 2–5 kHz is where our ears are most sensitive, so both harshness problems and intelligibility live there; small moves have big effects.

EQ and Compression: Which Comes First?

EQ and compression interact, because a compressor reacts to whatever signal hits it. If a vocal has a boomy low end and the compressor comes first, the compressor clamps down on the boom — meaning it responds to a problem you were going to cut anyway. The common-sense default: put your corrective, subtractive EQ (high-pass filter, problem cuts) before the compressor so it reacts to the cleaned-up signal, and put any character boosts after it so the compressor doesn't fight them. This is a default, not a law; engineers break it constantly for specific reasons. If you're still fuzzy on what a compressor is actually doing to the signal, read compression explained without the jargon first — the two tools make far more sense as a pair.

Habits That Make EQ Decisions Trustworthy

• EQ in context, not in solo. A track soloed and a track in the mix are different listening problems. Solo is for hunting a specific resonance; every keep-or-delete decision happens with the whole song playing. Tracks that sound wrong soloed are often perfect in the mix, and vice versa.
• Match loudness before judging. Bypass-compare every move at equal loudness, or the louder version wins by default.
• Monitor at moderate levels. Your ears' frequency response changes with playback volume — bass and treble seem to recede at low levels (the equal-loudness effect), so mixing loud tricks you into thin, bright mixes. Moderate level is also a hearing-safety issue: sustained loud monitoring causes permanent, cumulative hearing damage, and ringing or dullness after a session is a warning sign, not a badge of effort. Conversational-level monitoring, most of the time, protects both your mix decisions and the only pair of ears you get.
• Take breaks. Ears adapt to any tonal balance within minutes. After a long stretch, everything sounds "normal" — including your mistakes. Step away, come back, and the harsh 3 kHz buildup you stopped noticing will announce itself.
• Count your bands. If a track has eight active EQ bands, the odds are high that several are canceling each other out or doing nothing. Periodically bypass bands one at a time and delete the passengers.
• Don't mix with your eyes. Spectrum analyzers are useful for confirming what you hear (especially below 100 Hz, where small speakers lie). But if a curve looks dramatic and sounds like nothing, it is nothing.

The one-line takeaway: EQ is a space-making tool, not a sweetening tool — cut what's cluttering the mix first, and most of your boosts will turn out to be unnecessary.