Pitch JND: How Small a Pitch Difference Can We Hear?

The just-noticeable difference for pitch, measured in cents and trainable with practice.

Short answer: the smallest pitch difference a listener can reliably tell apart, called the just-noticeable difference or JND, is usually only a few cents, though it varies widely by person, training, and register. Trained ears in the lab can resolve differences of about 3 to 5 cents under ideal conditions, which is roughly a 0.2 to 0.3 percent change in frequency; untrained listeners are often nearer 10 to 15 cents. This is a measure of musical pitch discrimination, a skill you can practice. It is not a hearing test.

Please note: everything below is about ear-training and musical listening practice. It is not a hearing test, cannot diagnose anything, and is not a substitute for professional advice. If you have any concern about your hearing, please see a qualified professional.

What "just-noticeable difference" means

The JND for pitch is the smallest change in a tone that a listener can reliably notice, the threshold below which two pitches sound identical. Psychophysicists usually pin it down with a forced-choice test. You hear two tones and decide which is higher; the computer nudges them closer together when you answer correctly and farther apart when you miss, and the gap settles at the point where you score about 71 to 75 percent correct. That settling point is your threshold. Because it is defined by reliable performance rather than a single lucky guess, the JND is a sturdier number than "the smallest gap I think I can hear."

Since pitch is perceived logarithmically, the JND is naturally expressed in cents, where 100 cents is one equal-tempered semitone, rather than in raw hertz. A gap of 3 cents means the same perceived distance whether you test it near middle C or an octave higher, even though the value in hertz doubles between the two. That is why every serious figure on this page is quoted in cents rather than hertz.

The numbers, with honest caveats

There is no single "human" JND, so it is best given as a range. Laboratory studies put highly trained listeners at roughly 3 to 5 cents for clear pure tones in the mid-range under ideal conditions, corresponding to a frequency difference of about 0.2 to 0.3 percent. Thresholds rise to something nearer 5 to 6 cents for more complex or shorter tones, and untrained listeners are commonly closer to 10 to 15 cents. One large-cohort study reported a median around 14.5 cents, with the keenest five percent of participants near 6 cents and the least sensitive five percent around 77 cents. That spread is enormous, and it is the single most important thing to understand about pitch acuity: it is deeply individual.

Sensitivity also depends on where in the range you listen. The ear is keenest in the region around 1000 to 2000 Hz, where the finest resolvable difference is on the order of 3 cents. Move toward the very low bass or the extreme treble and the threshold widens, because the inner ear encodes those frequencies less precisely. For everyday musical judgment, remember too that listeners in ordinary rooms, with ordinary attention, reliably notice pitch changes of only about 25 cents, well above the lab best. Real ears in real settings perform below their theoretical ceiling.

How the threshold changes with context

The reason published figures scatter so much is that several variables push the threshold around at the same time. The table below gathers typical values so you can see the direction each factor pulls.

Weber's law, and why it is only approximate

You may have met the idea that the JND grows in proportion to the base frequency, an application of Weber's law. For pitch this holds only roughly. Expressed as a fraction of frequency, the JND is fairly stable across the comfortable mid-range, which is exactly why cents work so well as a unit. But the fraction is not truly constant. It worsens toward the frequency extremes and for brief or quiet tones, so Weber's law is a useful approximation rather than a strict rule. The honest, well-supported version is simply this: the JND is a roughly fixed number of cents in the middle of your range, and it degrades at the edges.

Common misunderstandings

A few myths are worth clearing up. A small JND is not the same as "good hearing" in the medical sense; it reflects discrimination skill and attention, not the health of your ears. A single test result is not a fixed trait either, because the number moves with tone type, register, loudness, duration, and how rested and focused you are. And beating a friend's score proves nothing clinical; it mostly reflects who has practiced the task. Discrimination is a skill, and skills vary with training and conditions.

Training moves the needle

Pitch discrimination is not fixed. Trained musicians consistently resolve far smaller differences than untrained listeners, and focused, repeated listening tends to sharpen the skill over time. How far any one person improves varies, but the direction is well supported: deliberate practice narrows the gap you can hear. The practical route is many short, clean repetitions of exactly this comparison, with the difficulty tightening as you get better. That is what a structured ear-training game provides.

Practice your discrimination

Related reading

• What are cents in music? The unit every JND figure on this page is measured in.
• What is pitch? Why pitch is a perception, and how it relates to frequency.
• Tone memory training: the recall skill that supports careful discrimination.
• Cross-hub: Tempo JND and Weber's law, the same just-noticeable-difference idea applied to tempo.

Sources: Brian C. J. Moore, An Introduction to the Psychology of Hearing; Scientific Reports, Individual differences in pitch perception; Cochlea.eu, Psychoacoustics: Pitch; Wikipedia, Cent (music).