Pitch is measured in hertz (Hz), a unit that counts how many times a sound wave vibrates per second. A sound vibrating 440 times per second has a frequency of 440 Hz, which corresponds to the note A above middle C, the international tuning standard. While hertz gives you the raw physical measurement, musicians, engineers, and scientists also use other scales and tools depending on whether they need to tune an instrument, analyze speech, or study how humans actually perceive sound.
Hertz: The Basic Unit of Pitch
One hertz equals one complete vibration cycle per second. When a guitar string vibrates back and forth 262 times in one second, it produces a frequency of 262 Hz, roughly the pitch of middle C. Double that frequency to 524 Hz and you hear the same note one octave higher. This relationship between frequency and perceived pitch is consistent: higher frequencies sound higher, lower frequencies sound lower.
The internationally recognized reference point is A4 = 440 Hz, reaffirmed by the International Organization for Standardization in 1955 and again in 1975. Orchestras, instrument manufacturers, and digital audio software all calibrate to this standard, though some ensembles (particularly in Europe) tune slightly higher, around 442 or 443 Hz.
Why Pitch Isn’t Measured on a Straight Line
Human hearing doesn’t process frequency in a linear way. The jump from 100 Hz to 200 Hz sounds like the same musical distance as the jump from 1,000 Hz to 2,000 Hz, even though the second gap is ten times larger in raw numbers. Both represent a doubling of frequency, which your ear interprets as one octave. This means pitch perception follows a logarithmic pattern: equal ratios of frequency produce equal perceived intervals.
This is why musicians don’t talk about pitch in hertz during performance. Saying a note is “6 Hz sharp” means something very different at 100 Hz than at 3,000 Hz. Instead, they use systems built around ratios.
Cents and Semitones
The most common unit for fine pitch measurement in music is the cent. One octave is divided into 1,200 cents, and one semitone (the distance between any two adjacent keys on a piano) equals 100 cents. This gives musicians a precise, ratio-based way to talk about how far a note sits from its target pitch.
Converting between cents and frequency ratios involves logarithms. An interval of x cents corresponds to a frequency ratio of 2 raised to the power of x/1200. You don’t need to do this math yourself. The important takeaway is that cents measure relative distance between pitches, not absolute frequency. A note that’s 15 cents flat is 15 cents flat whether you’re playing in the bass register or the upper treble.
Most trained musicians can hear a difference of about 5 to 10 cents. Differences smaller than that are generally imperceptible in a musical context, which is why tuner accuracy matters: a device that only resolves to 10-cent increments may not catch subtle tuning problems.
How Tuners Measure Pitch
Electronic tuners work by analyzing the incoming sound wave and calculating its fundamental frequency. The simplest clip-on tuners use LCD indicators spaced about 2.5 cents apart, so each tick mark on the display represents a small step toward or away from the target pitch. You play a note, and the tuner shows whether you’re sharp, flat, or in tune.
Strobe tuners are the most accurate type available. Traditional strobe tuners use a rotating disk backlit by lights that flash in sync with the input signal’s frequency. When the note is perfectly in tune, the pattern on the disk appears to stand still. If the note is sharp or flat, the pattern drifts in one direction or the other. Most modern strobe tuners replace the physical disk with software that simulates the same visual effect, offering accuracy within fractions of a cent.
Another common category uses LCD indicators spaced about 10 cents apart but achieves finer readings by lighting up two adjacent indicators simultaneously when the pitch falls between them. The green light comes on alone only when the pitch is correct. For any software-based tuner running on a computer or phone, accuracy depends on the quality of the device’s audio input. A sound card with an imprecise reference frequency will throw off the readings.
Pitch Detection in Software
When software needs to identify the pitch of a recorded sound, whether in a music app, a voice analysis tool, or a research lab, it relies on pitch detection algorithms. These algorithms take a raw audio waveform and estimate its fundamental frequency, which is the lowest vibrating component of the sound and the one your ear identifies as the pitch.
The most widely used approach is autocorrelation, which works by comparing a waveform to shifted copies of itself. When the shift matches the length of one vibration cycle, the waveform lines up with its copy, producing a strong correlation peak. The time between those peaks gives the period, and the inverse of the period gives the frequency. The YIN algorithm, one of the most established time-domain methods, refines this approach by computing what’s called an average magnitude difference function and selecting peaks as pitch candidates, reducing the kinds of errors that basic autocorrelation produces.
For visual analysis, spectrograms display pitch over time. A spectrogram plots time on the horizontal axis and frequency on the vertical axis, with brightness or color indicating how loud each frequency component is at each moment. This lets researchers and audio engineers see pitch contours, vibrato, glides between notes, and harmonic structure all at once.
Perceptual Scales: Mel and Bark
Hertz and cents measure physical properties of sound waves. But researchers studying how people actually hear pitch need scales that reflect the ear’s uneven sensitivity across the frequency range. Your ear is much better at distinguishing small pitch differences in the low and mid frequencies than in the high frequencies.
The Mel scale, developed in 1937 through listening experiments, maps frequency onto a scale where equal distances correspond to equal perceived pitch differences. Low frequencies are stretched out (because you hear fine distinctions there) and high frequencies are compressed. The Bark scale takes a different approach, dividing the audible range into 24 critical bands that model the physical response of the inner ear. Each band corresponds roughly to a section of the cochlea that responds as a group.
These perceptual scales are essential in speech recognition, audio compression, and any application where what matters isn’t the raw frequency but how a human listener would experience it.
Measuring Pitch in the Human Voice
The fundamental frequency of the human voice, often abbreviated F0, typically falls between 80 and 450 Hz. Adult males generally speak with a fundamental frequency around 85 to 180 Hz, adult females around 165 to 255 Hz, and children higher still. These ranges overlap, and individual variation is wide.
Voice analysis software extracts F0 by applying the same pitch detection algorithms used in music, searching for the lowest harmonic frequency in the voice signal. Clinicians use this to assess vocal health, linguists use it to study intonation and tone languages, and speech therapists track it during voice training. The search range is typically constrained to 80 through 450 Hz to avoid picking up noise or overtones as false pitch readings.
Measuring Pitch in Animal Sounds
Bioacoustics researchers use the same core tools (spectrograms, Fourier transforms, and pitch detection algorithms) to measure animal vocalizations, but the frequency ranges can be vastly different. Bat echolocation calls can exceed 100,000 Hz, while elephant infrasound rumbles drop below 20 Hz, both outside the range of human hearing.
To handle these signals, researchers use specialized microphones and high sample-rate recorders. The fast Fourier transform converts recorded audio into its frequency components, after which software can extract features like fundamental frequency, spectral centroid (a measure of the “brightness” of a sound), and harmonic structure. Studies on wild boar, for example, have categorized vocalizations into grunts, squeals, barks, and trumpets based partly on their fundamental frequency and frequency range. Researchers found that animals under stress produced longer, higher-frequency calls, making pitch measurement a useful indicator of animal welfare.
Mel frequency cepstrum coefficients, originally developed for human speech recognition, have also been widely adopted in animal bioacoustics for classifying species and call types from field recordings.