We all know that feeling: the mix sounds a bit dull, so you boost some highs. Now the vocal is too bright, so you boost the guitar to keep up with it. A few rounds later everything is louder, the low end has a fog over it, and you’re not sure which move caused it. The habit that gets you there is boosting first.

I’ve been guilty of that habit in my early producing days, until a producer friend of mine, whose mixes always sounded clear, deep and open told me his simple secret which was using mainly subtractive EQ.

This article walks through a cut-first routine you can repeat on any session, including the part most guides skip: knowing when to stop cutting.

Why EQ boosts keep making your mix worse

Boosting makes a sound louder, and louder tends to sound better. Which is a trap. Your ears are getting fooled by the volume change, not the tonal change, so you walk away thinking you improved the track when you only made it louder.

Boosts have a side effect on the rest of the mix. When you lift a band on one track, that track takes up more space in that frequency range and covers parts of the other tracks sitting there. The covered tracks sound weaker, so you boost them too. This is masking, and it spirals: every track gets louder and bigger, nothing stands out, and the master bus runs out of headroom.

A boost also doesn’t only lift the band, it also leaves the thing that was bothering you. If a guitar has a boxy buildup at 350 Hz and you boost 3 kHz for clarity, the box is still there. It’s just now sharing space with a harsh top end.

The cut-first mindset: fix the problem before you add the sparkle

The habit is simple. Before any EQ move, ask what is wrong with the sound, and answer with a frequency range if you can. “It’s muddy” is a start. “There’s a buildup somewhere around 250 to 400 Hz” is a target you can act on.

Then cut that frequency range. Often, the part you wanted to boost turns out to have been there all along, buried under the problematic frequency. Take 3 dB out of the low mids on a dull-sounding acoustic guitar and the strings sound brighter, even though you never touched the highs.

Two rules keep this honest:

  • Level match every comparison. Cuts lower the volume slightly, and a quieter sound tends to feel worse. If you don’t compensate with the output gain of your EQ, you may get fooled by your ears (again).
  • Make the cut, then listen in context. A track that sounds great soloed can be the exact one clogging the mix, and the reverse is just as common.

Where the mud actually lives: the low mids and the other usual suspects

Mud is mostly a low-mid problem. Nearly every instrument has energy between about 200 and 500 Hz, so when you stack ten tracks, that range piles up fast. The ranges below are starting points, not laws, because every source is different.

Aa cluttered mixing desk and monitor speakers in a dim studio, with mud dripping out to the speaker cones, as if the sound itself were murky; A rat with sunglasses sitting on the top of the mixing desk between the speakers with its fingers in its ears. cool blue light from the screen behind. Japanes
  • Below roughly 30 to 40 Hz: subsonic rumble that most speakers can’t reproduce and that eats headroom. It shows up on synths, bass, room recordings and vocals (mic stand bumps, breath).
  • About 200 to 500 Hz: the classic mud zone. Pads, guitars, keys, vocals and bass all overlap here.
  • About 400 to 800 Hz: boxiness, the “cardboard” or “inside a cupboard” sound on drums, vocals and acoustic instruments.
  • About 2 to 5 kHz: harshness. This isn’t mud, but it’s the other place a cluttered mix hurts to listen to, especially with stacked guitars and synth leads.
  • About 5 to 8 kHz: sibilance on vocals, and hi-hats that sound spitty.

If your mix sounds cluttered and boomy but you can’t point to one instrument, assume it’s the sum of small low-mid buildups across many tracks. That’s why the routine below works track by track.

Sweep, find the ugly tone, then cut it

This is a core technique of subtractive EQing. Use a temporary boost as a flashlight, then flip it into a cut.

A hand turning a worn rotary EQ frequency knob on a hardware equalizer, with a small red LED glowing beside it, in warm low studio light with shallow depth of field, a pair of headphones resting on the desk beside it. A rat with sunglasses sitting on the hardware EQ with a big knife, cutting through
  1. Pick a track and loop the loudest section. Solo the track or leave it in context, whichever works best for you.
  2. Add a bell band with a fairly narrow Q, and boost it by 6 to 10 dB. This is temporary and only for spotting problems.
  3. Sweep slowly from about 150 Hz up to 1 kHz (wider if you suspect harshness higher up). Listen for spots where the sound gets ringy, honky, boomy or unpleasant. That’s your resonance.
  4. Stop on the worst spot and nudge the frequency until the ugliness peaks.
  5. Flip the gain to a cut. Start at around 2 to 3 dB down and widen the Q a little so it sounds natural rather than surgical.
  6. Toggle the band on and off with your output level matched. Ask whether the track sounds clearer and more natural with the cut on. If not, reduce it even more. If it still doesn’t help, remove it.

Don’t do this on every track by reflex. Do it on the ones that feel crowded, dull, muddy or harsh: the busiest instrument, the layered pad, the vocal that won’t sit. Sometimes a few well-placed cuts will do the trick. Sometimes a lot of small ones are the way to a transparent mix.

Carving space between kick and bass, and vocals and everything else

Some of the most effective cuts have nothing to do with a single track sounding wrong. They exist so two tracks that share the same frequency range don’t mask each other.

Kick and bass

Decide which one owns the very bottom. If the kick leads the sub region, cut a few dB from the bass around the kick’s fundamental, and let the bass carry the range just above it. If the bass leads, do the reverse. Either works, as long as you choose one and don’t let both pile onto the same frequencies.

EQing is only one way to unmask; another one is timing and dynamics. If you want to go further, my write-up on advanced sidechain compression techniques covers ways to make the two sit well in the mix.

Vocals and everything else

The vocal usually needs to be the clearest thing in the mix, so the instruments should make room for it, not the other way around. Two places to look:

  • Body, around 200 to 400 Hz: cut a little from guitars, keys and pads that live in the same range as the vocal’s chest.
  • Presence, around 2 to 5 kHz: a gentle 1 to 3 dB dip on competing instruments, like a rhythm guitar or a synth pad, leaves a gap for the vocal to cut through.

You’ll find the vocal need less boost once the tracks around give it enough space.

High-pass, bell cut or dynamic EQ: picking the right tool for the cut

Different problems call for different tools. These are the most common:

High-pass filter. Use it for content that doesn’t belong on the track at all, such as rumble under a vocal, or low-end energy on hi-hats, shakers, and most guitars and synths. A gentle slope, such as 12 dB per octave, sounds more natural. Steeper slopes remove more, but they can add a slight ringing around the cutoff. Set the cutoff by ear, raising it until the track starts to lose weight, then back off a touch. Leave kick and bass mostly alone, or filter only the unwanted subsonic part.

Bell cut. This is the right choice for a steady problem: a boxy snare, a constant honk in a vocal, a pad that sits heavily at 300 Hz the whole song. It’s the tool for the resonance hunt described above.

Dynamic EQ. Use it when the problem only appears sometimes. A vocalist who gets boomy on low notes or close to the mic, or an acoustic guitar with one note that rings, doesn’t need a permanent cut. A dynamic band only reduces the frequency when it crosses a threshold, so the rest of the time the track keeps its full tone.

A quick rule: if the problem is always there, use a static cut. If it comes and goes, use a dynamic one.

How to tell you’ve cut too much (and when a boost is totally fine)

Cutting has a cost, and most people only learn it by overdoing it once. Watch out for these signs:

  • The track sounds thin, hollow or small compared to when the EQ is bypassed (with levels matched).
  • The vocal or guitar sounds fine soloed but disappears in the full mix,.
  • You’ve stacked cuts of 5 or 6 dB or more on one track, or have four or five bands working on the same region. That usually means the source itself is the problem, and a different sound, a different take or a different mic position will do more than more EQ.
  • The mix has lost its warmth. Low mids are where a lot of body and fullness live. You want to remove a buildup, not the range.

A good habit is to check your cuts against the original in context every few moves and undo the last cut if it sounds worse. Also revisit them after a break, since fresh ears are better at catching overcutting.

And yes, boosts are totally fine. They just come second. A boost makes sense when:

  • You’ve made your cuts and something is genuinely missing, like no air on the top end. It’s difficult to cut your way to more air.
  • You want a broad, gentle lift, such as a small high shelf of 1 to 2 dB, rather than a narrow spike.
  • You’re using a boost and a cut together on purpose, for example dipping 300 Hz while lifting 4 kHz to move a vocal forward.

If you finish your cuts and find you no longer feel the urge to boost, that’s a good sign, not a missed step.

A 10-minute subtractive pass you can run on any session

Here’s a routine you can run on any mix. Set a timer, so you don’t wander off into endless tweaking.

A laptop and a MIDI keyboard and two monitor speakers on a tidy home studio desk, soft evening light from a desk lamp. The wall behind the desk is a brick wall. A rat with sunglasses standing on the MIDI Keyboard with a stopwatch in its hand, ready to push the start button. Japanese 80s Anime style.
  1. Minute 0 to 1: set up. Loop the busiest section of the song. Keep your monitoring level constant so you can match levels when you compare.
  2. Minutes 1 to 3: high-pass. Add a high-pass filter to every track that doesn’t need low end. Raise each one until it starts to thin, then back off.
  3. Minutes 3 to 5: low-mid check. Find the three or four most crowded tracks and take a few dB out of the 200 to 500 Hz area on each, using a wide bell. Listen to whether the mix opens up.
  4. Minutes 5 to 7: resonance hunt. Pick the two or three tracks that sound worst, usually the vocal, a lead and the snare, and run the sweep from earlier.
  5. Minutes 7 to 8: make room. Carve a little space between kick and bass, then cut a dB or two on supporting instruments where the vocal sits.
  6. Minutes 8 to 9: bypass check. Flip each EQ off and on at matched level. Pull back any cut that made things worse.
  7. Minutes 9 to 10: decide on boosts. Listen once more. If something still needs help, make a small, wide boost on one or two tracks and stop.

You won’t fix everything in ten minutes, but you’ll hear whether the mud was the issue. Most of the time it was.

Run the 10-minute subtractive pass on the mix you’re working on right now, then read How To Set An Equalizer For Clarity to see how the cuts fit into your wider EQ setup. Have fun with it, and go easy on the EQ boosts.

Leave a Reply

Your email address will not be published.