Using Spectrum Analysers to Improve Your Mix

What a Spectrum Analyser Actually Shows You

A spectrum analyser is a moving picture of where energy sits across the frequency range at any given moment. Instead of guessing whether a vocal sounds boxy or a kick drum feels thin, you get a visual readout: a line or filled curve running left to right from roughly 20 Hz up to 20 kHz, with the vertical axis showing level in decibels. Peaks are places where lots of energy is piling up. Dips are places where the mix is comparatively empty.

That information is useful precisely because our ears are unreliable in a small room. A resonant corner or a desk reflection can make 120 Hz sound enormous while the analyser shows a perfectly reasonable low end. The display is not the truth — your ears still decide — but it is a second opinion that never gets fatigued.

Setting Up the Display Sensibly

Before you read anything into the curve, make the curve readable. A few housekeeping habits save a lot of confusion:

• Use a gentle slope. Set the analyser to a 3 dB per octave tilt (sometimes labelled "pink" or "musical" mode). Pink noise then appears as a roughly flat line, which matches how we hear. A completely flat analyser setting makes almost every real mix look like it is falling off a cliff.
• Smooth it. One-third octave smoothing removes the fizz of individual harmonics and lets you see the broad shape. Use fine resolution only when you are hunting a specific ring or resonance.
• Watch an average, not a snapshot. Most tools offer an averaged curve alongside the live trace. The average shows the tonal balance of the whole section; the live trace shows what is happening right now.
• Check one source at a time. Solo the bass, then the vocal, then the drum bus. Context matters, but soloing tells you what each element is contributing.

Finding Mud in the Low Mids

The region between roughly 200 Hz and 500 Hz is where mixes most often turn to porridge. Look for a broad hump sitting under several instruments at once — guitars, snare body, piano left hand, the lower reach of a vocal. If three sources all peak around 300 Hz, the sum will be thick and undefined.

Rather than scooping the same band out of everything, decide which instrument owns that area. Give the bass and the kick the freedom below 150 Hz, let one mid-range element keep its 300 Hz warmth, and use narrow cuts of 2–4 dB with a moderate Q on the others. A cut of 3 dB at 250 Hz on a single guitar can clear more space than a 6 dB cut spread across the whole mix bus.

Also watch the very low end for problems the analyser reveals instantly: a steep roll-off below 40 Hz usually means rumble or handling noise rather than musical content, and a high-pass filter at 30–40 Hz will free up headroom without touching anything you can hear.

Taming Harsh Peaks in the Upper Mids

Between about 2 kHz and 5 kHz lives the presence that makes a vocal cut through — and the harshness that makes it tiring. The analyser is excellent here because offensive resonances often show as narrow spikes that stay put while the rest of the curve moves. Park the cursor on the spike, note the exact frequency, then apply a narrow, modest cut of 2–3 dB.

Be careful with cymbal and synth-heavy material: a bright, sibilant source can look alarming on screen while sounding perfectly fine, because a great deal of energy in the 8–12 kHz region is normal for hi-hats and airy vocals. Judge those by ear. Use the analyser to confirm what you suspect, not to invent problems.

If a de-esser is already working hard, that is a clue that the peak is dynamic rather than static. A static EQ cut will thin the sound for the whole performance; a dynamic tool that engages only on the loud syllables will keep the body intact.

Using Gaps to Shape Rather Than Cut

A spectrum analyser shows absences as clearly as it shows excess. Compare two instruments that occupy similar territory: if the guitar has a natural dip at 800 Hz and the snare has a peak there, you have a match made in heaven. Nudging the guitar down slightly in that dip, or lifting the snare a touch, creates separation without either sounding processed.

This is the heart of complementary EQ. The aim is not to make every track look flat on its own — that produces a mix of bland, identical curves. The aim is to make each part look slightly uneven in a way that interlocks with the others, like a jigsaw. Pull up two or three analyser instances side by side, or use a matching display, and look for where one curve's peak sits next to another's trough.

A useful exercise is to load a reference track you admire and compare its average curve to your own. Do not chase an exact match — arrangement, genre and mastering all differ — but if your mix has 8 dB more energy at 400 Hz than the reference, that is worth investigating.

Trusting Your Ears in the End

The analyser is a tool for asking better questions, not for making decisions. Use it to spot a suspicious build-up, confirm a suspicion about a resonance, or check that your low end is not wandering. Then switch the display off, listen at a sensible level, and ask whether the mix actually sounds better.

Eyes are easily fooled by a pretty curve, and a flat-looking spectrum can still sound dull, crowded or lifeless. The best results come from a loop: look, listen, adjust a little, listen again. Keep your moves small — 2 to 4 dB is usually plenty — and let the visual feedback guide you towards the problem rather than dictate the cure. Do that consistently and your mixes will get clearer, faster, with far less guesswork.