Harmonic analysis: key, resonances & tonal character

Harmonic analysis reads the musical and tonal content of your track — its key, its resonant peaks, and the harshness that makes a mix tiring. Unveil detects the key with a chromagram (Krumhansl-Schmuckler), finds narrow resonant peaks and sibilant/harsh build-ups, and tells you where the tonal balance is fighting itself.

Harmonics and why they define timbre

Almost every musical sound is a stack of frequencies. A note at 100 Hz also contains energy at 200 Hz, 300 Hz, 400 Hz and so on, each a whole-number multiple of the fundamental. These are the harmonics, and their relative strengths are what make a clarinet sound different from a violin playing the same pitch. Changing the harmonic balance changes the character of the sound without changing its note.

Harmonic analysis in mixing covers three related questions: what harmonics the sources contain, what harmonic content processing adds, and how the musical notes in different parts relate. All three affect whether a mix sounds full, clear, or in conflict with itself.

Saturation and distortion types

When a signal is passed through a non-linear process, new harmonics are created. The type of non-linearity decides which. Symmetrical clipping, as in many transistor and digital clipping stages, creates mostly odd harmonics (3rd, 5th, 7th). Asymmetrical processes, as in some tube and transformer circuits, add even harmonics (2nd, 4th), which are octaves and consonant. Real hardware mixes the two.

Tape saturation: gentle compression of peaks plus low-order harmonics, often with a soft high-frequency roll-off. Adds warmth and glue.
Tube saturation: tends toward even harmonics at low drive and richer content at high drive. Adds fullness.
Transistor and diode clipping: odd-harmonic heavy and brighter, with a harder character. Adds edge.
Digital hard clipping: abrupt and harmonically rich, which can alias if not oversampled. Useful for peak control in small amounts.
Bit reduction and sample-rate reduction: create inharmonic content that does not follow musical ratios. A deliberate lo-fi effect, not a subtle one.

Key and tonal centre

Every tonal piece has a key, the set of notes that sound stable. Key detection works by examining which pitch classes occur most, and which scale profile the distribution matches best. The result is usually good for clearly tonal music with a conventional structure and unreliable for atonal material, heavy percussion, or tracks with key changes. Treat a detected key as a hypothesis that you confirm by ear or against the score.

Knowing the key helps in two practical ways. It tells you which bass notes and sustained instruments should agree. And it tells you which frequencies carry the tonal centre, so you can see whether a resonance is musical or a problem. A boost near the fundamental of a tonic note adds weight, while a boost at an out-of-key frequency tends to sound wrong.

Clashing low-end notes and harmonic masking

In the low register, small pitch differences produce slow beating, which the ear hears as wobble or muddiness instead of a distinct dissonance. If a kick is tuned to G1 at about 49.0 Hz and the bass plays F#1 at about 46.25 Hz, the 2.75 Hz difference produces a pulsing roughly three times a second. Retuning the kick to the key, or choosing a different bass note on the downbeat, removes it.

Harmonic masking is the high-frequency counterpart. Instruments with strong harmonics at the same frequencies compete even if their fundamentals differ. A distorted guitar's upper harmonics can bury a vocal's presence region, and a bright pad can bury cymbals. The spectrum will show a dense plateau instead of distinct peaks.

Ways to resolve it: tune the kick and bass to the key, carve complementary EQ cuts, reduce drive on the less important part, and use arrangement changes, such as shifting a part up an octave, to put instruments in different registers.

How saturation adds perceived loudness and presence

Saturation does two helpful things at once. It compresses the peaks slightly, so the average level can rise with the same peak ceiling. And it adds harmonics, which spread energy across more frequencies. The ear weights mid and upper frequencies more strongly than deep bass, so extra harmonics at 1 to 4 kHz make a sound seem louder and more present at the same measured level.

This is also how bass is heard on small speakers. A 50 Hz note may be inaudible on a laptop, but its second and third harmonics at 100 and 150 Hz are not. The brain infers the missing fundamental from the harmonic pattern. Adding mild saturation to a sub bass often makes it audible across more systems without raising its low-frequency level.

The tradeoff is that harmonics add density. Too much saturation on everything produces a congested, harsh mix, and each stage compounds the next. Use it where an element needs weight or edge, and leave others clean.

Step-by-step workflow

Identify the key and the main tonal centres of the song.
Check the notes of the kick, bass and any sustained sub against the key. Fix tuning clashes first.
Look at the long-term spectrum for plateaus that suggest harmonic congestion, especially from 1 to 5 kHz.
Solo the suspects and find which carries the excess harmonic energy.
Apply saturation in small steps on elements that need presence. Match the output level so that louder does not fool you.
Re-check the spectrum and listen at matched loudness, on small speakers as well as large.

Worked example, mistakes and checklist

A trap track has an 808 at 41 Hz (E1 is 41.2 Hz) and sounds absent on phone speakers. Adding tube-style saturation at moderate drive creates harmonics at about 82, 123 and 164 Hz. The phone speaker, which rolls off below roughly 150 Hz, now reproduces the fourth harmonic and partly the third, and the bass part is audible. Output level is trimmed 1.5 dB so the full-range level does not rise, and the spectrum still shows the original fundamental intact.

Common mistakes: saturating the master as well as every channel; using distortion to fix a tonal problem that needs an EQ; forgetting that saturation also raises noise and sibilance; and leaving a bass note out of tune with a sustained pad.

Kick, bass and pads agree with the key.
No plateau of congested harmonics around 1 to 5 kHz.
Saturation applied with a purpose, level-matched.
Bass remains audible on small speakers.
Distortion does not alias or buzz at high frequencies.

What you get

Musical key detection so your corrective and creative moves stay in key.
Resonance and harshness detection — narrow peaks and 3–8 kHz build-ups that cause ear fatigue.
How tonal character ties back to the frequency spectrum and genre curve.
A free readout of resonances and harshness with the frequencies to tame.

Questions

How is the key detected?

A 12-bin chromagram is correlated against the 24 Krumhansl-Kessler major/minor key profiles. It returns a best-fit key, or nothing when the material is too atonal to be confident.

What causes harshness in a mix?

Usually a build-up in the 3–8 kHz presence region or narrow resonant peaks. Unveil flags these so you can de-ess or dip the specific band instead of dulling the whole top end.

Analyze your mix — free