Sound Basics: Frequency, the Spectrum and Decibels

Every mixing and sound design decision comes down to two questions: where is this sound in frequency, and how loud is it? Learning the vocabulary for both turns vague complaints ("it sounds muddy") into fixes ("cut 250 Hz on the pads").

Frequency

Sound is air pressure vibrating. Frequency is how many times per second it vibrates, measured in hertz (Hz). 440 Hz means 440 cycles per second — the A above middle C. Doubling the frequency raises the pitch by one octave: 220 Hz, 440 Hz and 880 Hz are all "A".

Human hearing runs from roughly 20 Hz to 20,000 Hz (20 kHz), and the top end shrinks with age; many adults hear little above 16 kHz.

Fundamentals and harmonics

A real instrument playing A440 isn't a single frequency. It produces the fundamental (440 Hz) plus harmonics at whole-number multiples: 880, 1320, 1760 Hz and up. The relative strength of those harmonics is timbre — why a piano and a violin on the same note sound different. It also means an instrument occupies far more of the spectrum than its note suggests: a bass guitar's fundamental may be 41 Hz, but the harmonics that make it audible on a phone speaker sit at 200–1000 Hz.

The spectrum, band by band

Engineers carve the spectrum into rough regions. The edges are fuzzy; the character of each is not.

Band	Range	What lives there	Too much sounds
Sub-bass	20–60 Hz	Kick and 808 weight, sub synths	Rumbling, eats headroom
Bass	60–250 Hz	Bass body, kick punch, warmth	Boomy
Low mids	250–500 Hz	Body of guitars, keys, vocals	Muddy, boxy
Mids	500 Hz–2 kHz	Core of most instruments	Honky, nasal
Upper mids	2–5 kHz	Presence, attack, intelligibility	Harsh, fatiguing
Highs	5–10 kHz	Sibilance, cymbal definition	Sharp, sizzly
Air	10–20 kHz	Sparkle, breath, room	Brittle, hissy

The ear is most sensitive around 2–5 kHz, which is why small boosts there sound dramatic and why harshness lives there. Most mix crowding happens in the low mids, where nearly every instrument has energy.

Decibels

A decibel (dB) is a ratio, not an absolute amount. Rules of thumb worth memorising:

• +6 dB ≈ double the signal's amplitude.
• +10 dB is roughly what people perceive as "twice as loud".
• 1 dB is about the smallest level change most listeners notice on music.
• −∞ dB is silence.

dBFS: the meter in your DAW

Digital meters read dBFS — decibels relative to full scale. 0 dBFS is the ceiling: the largest value the file can represent. Everything below is negative. Push a signal past 0 dBFS on an export and the waveform is chopped flat — clipping — which sounds like crackle or harsh distortion.

Practical targets while producing:

• Individual tracks: peaks around −18 to −10 dBFS.
• The master before mastering: peaks no higher than about −6 dBFS.
• A mastered file: peaks at −1 dBFS or lower, so streaming encoders don't clip it.

Peak level is not loudness

A meter showing peaks only tells you the highest instant. Two tracks peaking at −1 dBFS can sound very different in loudness: a sparse acoustic song with occasional loud hits feels quieter than a dense, compressed EDM track hitting the same peak constantly.

Perceived loudness depends on:

1. Average level over time, not the peak. This is measured in LUFS (loudness units relative to full scale), which streaming services use to normalise playback.
2. Frequency. The ear is less sensitive to bass and extreme treble at low volumes, so a mix checked quietly sounds thin, and a mix checked loudly sounds bigger and better than it is.
3. Duration. Sustained sounds read louder than short ones at the same peak.

This is why compression makes things sound louder without raising the peak: it raises the average.

Listening habits that follow from this

• Check balance at a moderate, consistent volume; loudness flatters everything.
• When comparing two versions or an effect on and off, match their loudness first, or you will just pick the louder one.
• Use a spectrum analyser to confirm what you hear, not to replace hearing it.