Psychoacoustics for Composers: How the Ear Builds Music
linglin · 2026-11-18 10:00:00
Psychoacoustics is the study of how we perceive sound — not what the air is doing, but what we hear. The gap between the two is large, and it is exactly where composers work. Here are the ideas I think every composer and producer should know.
Loudness is not volume
Our hearing runs from about 20 Hz to 20 kHz (less as we age), but it is not equally sensitive across that range. The ear canal resonates around 3 kHz, so we are most sensitive roughly between 2 and 5 kHz — the region of consonants, baby cries, and the "presence" of a voice. Very low and very high frequencies need much more energy to sound equally loud.
These equal-loudness contours (first measured by Fletcher and Munson in the 1930s, now standardised as ISO 226) also change with level: at quiet volumes the bass and extreme treble seem to vanish. That is why a mix that sounds full when loud can sound thin when quiet, and why you should check mixes at low volume.
Pitch is a guess
Play the harmonics 200, 300, 400 and 500 Hz together with no 100 Hz component at all, and most listeners will still hear a pitch at 100 Hz. This is the missing fundamental (or virtual pitch). The brain sees the pattern of harmonics and infers the fundamental that would explain it.
This is why you can recognise a bass line on a phone speaker that physically cannot reproduce it, and why producers add distortion or upper harmonics to a sub-bass to make it "translate" to small speakers. Pitch is not measured; it is reconstructed.
Critical bands and roughness
The cochlea works something like a bank of overlapping filters. Two tones that fall inside the same filter — the same critical band — interfere with each other. If they are very close, you hear slow beating. A bit further apart, the beating becomes fast and turns into roughness, a buzzy, grating quality. Further apart still, the tones separate and sound smooth.
Plomp and Levelt (1965) found that roughness peaks when two pure tones are about a quarter of a critical band apart. In the middle register a critical band is roughly a minor third wide; in the bass it is much wider in musical terms. That single fact explains a lot:
- A major third sounds sweet in the middle of the piano and muddy two octaves lower.
- Consonance depends partly on register, not only on interval.
- Rich tones with many harmonics are consonant when their harmonics line up (simple ratios like 3:2) and rough when many harmonics collide.
Masking
A loud sound can hide a quieter one near it in frequency — that is masking. Masking spreads more upward than downward: a loud low sound covers a lot of the frequencies above it, but a loud high sound hides little below it. It also happens in time: a loud hit can mask quieter sounds a few milliseconds before and after it.
For arrangers, masking is the hidden reason parts "disappear" in a mix. If two instruments fight for the same band at the same moment, one of them loses. MP3 and AAC compression use masking models to throw away the sounds you wouldn't have heard anyway.
Auditory scene analysis
Your ears receive one combined pressure wave. Somehow you hear a singer, a guitar, a car outside. Albert Bregman called this auditory scene analysis: the brain groups sound into "streams" using cues like
- Common onset — partials that start together belong together.
- Harmonicity — partials that fit one harmonic series belong to one source.
- Pitch proximity — close notes in sequence sound like one line.
- Timbre and location — same colour and same place suggests same source.
Alternate a high note and a low note slowly and you hear one melody jumping. Speed it up and it splits into two melodies, one high and one low. Baroque composers exploited this: in Bach's solo violin and cello works, a single instrument implies two or three voices at once, a technique called compound melody.
Space and the precedence effect
When the same sound reaches you directly and then from a reflection a few milliseconds later, you hear one sound, located where the first arrival came from. This precedence (Haas) effect lets us localise sources in reverberant rooms, and lets producers widen a sound with short delays without it seeming doubled.
Illusions that teach
The Shepard tone seems to rise forever without getting higher — octave-spaced tones fade in at the bottom and out at the top. It shows that pitch has two dimensions: height (high/low) and chroma (which note name). Composers use the same principle for endless risers and for music that feels like it is ascending without arriving.
The takeaway
Everything a listener calls "clear", "muddy", "sweet", "harsh", "full" or "thin" is a perceptual judgement. You are not composing for a spectrum analyser; you are composing for a cochlea and a brain with specific habits. In the next post I turn these ideas into practical writing techniques.
Comments (4)
Character posts are example voices written by LingLin.Art to get discussions going — they are not real members.
The critical band section explains so many mixing problems. Sending this to every client who asks why their mix is muddy.
Great explanation of the missing fundamental. In games we rely on it constantly.
The Shepard tone 'height vs chroma' explanation finally made it click for me.
Bookmarking this for our new arrangers at the studio.
← All posts