EN

Wavelength, frequency and speed of sound

A sound wave is described by its frequency — the pitch, in hertz — and its wavelength — the distance travelled during one period. The two are linked by the speed of sound in the medium: λ = c / f. In air, that speed depends mostly on temperature.

This is the basic calculation of acoustic treatment: a porous absorber only really works from a thickness close to a quarter of the target wavelength — hence the λ/2 and λ/4 values returned with the result.

Medium and known quantity

Propagation medium

Change the medium then run the calculation again to switch the fields.

c = 331.3 · √(1 + θ/273.15) — about +0.6 m/s per degree.

Calculation

The audible range runs from roughly 20 Hz to 20 kHz.

Frequency, period and wavelength: three ways to describe a wave

A sound wave is one motion, counted three ways. Frequency f (hertz) is how many cycles pass each second — that is pitch. Period T is the duration of one of those cycles: T = 1/f. Neither number cares which medium you are in.

Wavelength λ does. It is the distance the wave travels during one period, so it belongs to the medium as much as to the note. In air near 20 °C (c ≈ 343 m/s), concert A at 440 Hz stretches about 78 cm — roughly a forearm. Drop the same A into water and λ jumps to a couple of metres, because the wave is simply faster.

λ = c / f   ·   T = 1 / f   ·   c = λ · f

In fact the three quantities lock together: pick any two and the third is already decided. That is the whole calculator.

Why the speed of sound depends on temperature, not pressure

A stubborn myth: “sound is slower up a mountain because the air is thinner.” At a fixed temperature it is not. The Newton–Laplace speed is √(γP/ρ). Raise the pressure and the density rises in the same proportion, so the ratio — and the speed — stay put. Sea-level or plateau, same θ, same c.

Temperature is the real lever. It sets how fast the molecules already jostle, and that agitation is what carries the compression. Warmer air, quicker molecules, higher c. The page uses the dry-air square-root law; a pocket rule that tracks it closely is the linear fit below.

c = 331.3 · √(1 + θ / 273.15)   ≈   331.5 + 0.6 · θ   (θ in °C)

At a comfortable 20 °C you land near 343 m/s. At 0 °C you are closer to 331 m/s. A 10 K swing is about 6 m/s — enough to shift a 1 kHz wavelength by a couple of millimetres, which is why a studio that quotes millimetres also quotes the room temperature.

Humidity and altitude: small but real effects on celerity in air

Water vapour is lighter than the N2/O2 mix it displaces (18 g/mol against 29). Moist air is therefore a shade less dense, and sound a shade faster. At 20 °C, dry to saturated is typically 1 to 2 m/s. You will see it on a precision rangefinder. You will not hear it.

Altitude is mostly temperature in disguise. The standard lapse rate in the lower atmosphere is about 6.5 °C per kilometre, so at 4,000 m you are looking at something near 320 m/s — not because the pressure fell, but because it got cold. Well, the calculator’s air mode only asks for θ, which is the honest way to put it.

Table of speeds of sound by medium (gases, liquids, solids)

Sound is a compression wave. Stiffer and lighter media run faster; floppy, heavy ones crawl. Air is slow. Water is already four times quicker. Metals are another jump again, because the elastic modulus is huge compared with the density.

MediumTypical c (m/s)Note
Air, 20 °C343Everyday reference
Helium, 20 °C1007Light gas — the squeaky-voice gag
Fresh water, 20 °C1480Four times air; this page’s preset
Sea water1522Salt raises both density and bulk modulus
Wood, along the grain3800Very anisotropic; across the grain is slower
Concrete3400Structure-borne sound in buildings
Glass5300Type and temper move the number
Steel5100Alloy and heat treatment; 5,000–5,900 is the usual band
Aluminium6420Stiff and light — a favourite structural path

The presets on this page are round working values, not certified lab figures. Pick “Custom speed” the moment you have a measured c for the actual sample.

Wavelength benchmarks: 20 Hz, 100 Hz, 1 kHz, 10 kHz

Keep four numbers in your head, all in air at 343 m/s. They tell you, at a glance, whether a room “looks large” or “looks small” to a given note.

  • 20 Hz (deep bass) — λ = 17.15 m, taller than a two-storey house.
  • 100 Hz (low end) — λ = 3.43 m, about a room height.
  • 1 kHz (midrange) — λ = 34.3 cm, a ruler.
  • 10 kHz (treble) — λ = 3.43 cm, smaller than a thumb.

Below a hundred hertz the wave is furniture-sized. That is why bass is a placement problem and treble is a surface problem.

The quarter-wavelength: useful thickness of a porous absorber, air gap behind a panel

A porous absorber — mineral wool, fibreglass, open-cell foam — dissipates energy where the particle velocity is high. On a rigid wall that peak sits a quarter-wavelength out. Thickness ≈ λ/4 is therefore the classic target, not a superstition. At 500 Hz, λ ≈ 69 cm, so you want something like 17 cm of wool, or a wool-plus-air-gap that adds up to the same depth.

The same λ/4 shows up behind a perforated facing or a slat panel: the cavity is a resonant trap. A ported loudspeaker box is the same physics with a different job — the port is a Helmholtz neck on a tuned volume, not an absorber, but the length still scales with wavelength.

duseful ≈ λ / 4 = c / (4f)

Thin foam on a wall is a mid/high treatment. It will not touch 80 Hz, however expensive the catalogue photograph.

Room modes and bass wavelengths

Two parallel walls a distance L apart support a standing wave when L is a multiple of λ/2. Those axial modes are the boomy notes and the dead spots you walk through. First axial frequency:

fn = n · c / (2L)   (n = 1, 2, 3, …)

A 5 m room, n = 1: f ≈ 343 / 10 = 34 Hz. Below roughly 150 Hz, λ is already larger than most domestic dimensions, so the “wave” is more a pressure oscillation of the whole volume than a travelling beam. That is why a single T60 never explains a bass problem — T60 averages the decay; modes are places.

Diffraction: when an obstacle is small compared with λ

Waves bend around things that are small compared with their wavelength. A 5 cm mullion is invisible to 100 Hz (λ ≈ 3.4 m) and a hard edge to 10 kHz (λ ≈ 3.4 cm). Bass goes around the door; treble casts a shadow. Same doorway, two different geometries.

The rule of thumb is blunt and useful: obstacle ≪ λ, the wave barely notices; obstacle ≫ λ, you get reflection and a shadow. Screens, baffles, even a head between two ears — all of them are size-versus-λ problems before they are “acoustic products”.

Resonators (Helmholtz, quarter-wave) and pipe lengths

Blow across a bottle: the cavity is a spring, the neck is a mass. That Helmholtz frequency can sit well below anything the bottle’s own size would suggest as a λ/2 pipe. The wavelength “belongs” to the resonance, not to the object’s outline.

A pipe closed at one end is simpler. It resonates when its length is an odd multiple of λ/4. Open at both ends, multiples of λ/2. A 1 m closed tube in air: fundamental ≈ 343 / 4 ≈ 86 Hz.

closed: L = (2n−1) · λ/4   ·   open: L = n · λ/2

Organ pipes, car exhaust silencers, and the port on a bass-reflex cabinet all live on that arithmetic. End corrections (the extra “virtual” length at a flanged opening) are why a measured pipe is always a bit shorter than the naïve λ/4.

Frequency and musical pitch: equal temperament and A 440

Western equal temperament splits the octave into twelve equal semitones, each a factor of 21/12 ≈ 1.05946. Double the frequency, up an octave. The usual concert reference is A4 = 440 Hz (some orchestras sit a few hertz higher). C4, middle C, then lands near 261.6 Hz.

λ for that 440 Hz A is 78 cm in air — close to a tuning-fork length, which is not an accident of furniture design. Change the air temperature and the fork still rings 440; the wavelength, and any wind instrument’s tuning, will not. Players warm up for that reason, not for folklore.

The Doppler effect: what happens to the perceived frequency

Source and listener in relative motion: the number of wavefronts you meet per second changes. For a stationary listener and a source moving at speed v along the line of sight:

approaching: f′ = f · c / (c − v)   ·   receding: f′ = f · c / (c + v)

A car at 72 km/h (exactly 20 m/s) with a 440 Hz horn: about 467 Hz on the way in, 416 Hz on the way out. Sharper, then flatter. Swap who moves and the formula changes (the denominator and numerator trade roles). Radar, astronomy, medical Doppler ultrasound — same bookkeeping, different waves.

Worked example: sizing an absorbing panel for 125 Hz

A 125 Hz room boom is a common complaint. Walk the numbers before you order foam.

  1. Wavelength. λ = 343 / 125 ≈ 2.74 m.
  2. Porous λ/4. 2.74 / 4 ≈ 0.69 m. A 70 cm wool slab would work. Few rooms can spare that depth on a wall.
  3. Resonant alternative. A membrane absorber (plywood or mass-loaded vinyl over a sealed cavity) or a Helmholtz perforated panel can be tuned to 125 Hz in 15–25 cm of depth by choosing the mass and the cavity. Bandwidth is narrower than a thick porous stack — several tunings, or a damped membrane, broaden it.
  4. Sanity check. You are fighting a three-metre wave. There is no 20 mm miracle tile. Either you spend depth, or you spend resonance.

Interference, path difference and comb filtering

Two arrivals of the same tone, delayed by a path difference Δ, add or cancel according to how many wavelengths sit in that delay. Δ = λ/2 is a cancellation; Δ = λ is a reinforcement. A loudspeaker plus its floor reflection is the textbook comb: a series of peaks and notches whose first notch sits at f = c / (2Δ).

A 1 m extra path in air: first notch ≈ 172 Hz, then 3, 5, 7… times that. Move the speaker or the listener a few tens of centimetres and the comb slides. That is placement, not EQ, for the first attack on a hollow midrange.

Why 343 m/s is a convenient round number

343 m/s is dry air at about 20 °C, not a constant of nature. 340 is the mental-arithmetic cousin (λ in metres ≈ 340/f). 331 is 0 °C. Outdoor winter work at 5 °C is closer to 335. Quote the temperature with the speed, or the millimetre you just cut on a port is a guess.

The linear rule 331.5 + 0.6·θ is plenty for site work. This page’s square-root form is the one to use when you actually type a number.

Speaker arrays and the λ/2 spacing rule

Put two drivers on a baffle and they start to beam once their centre-to-centre spacing is no longer small compared with λ. A common design fence is spacing < λ/2 at the highest frequency they both play. At a 2 kHz crossover, λ/2 ≈ 8.6 cm — which is why a 6.5-inch woofer next to a dome already argues with the crossover choice.

Line arrays and column speakers play the same game in the vertical plane. The calculator will not design the box, but it will tell you, in centimetres, whether a given frequency still “sees” the pair as one source.

Infrasound, ultrasound and the limits of hearing

Call 20 Hz–20 kHz the usual hearing band and you have already picked two wavelengths in air: 17 m down to 1.7 cm. Infrasound (a few hertz) has λ of tens of metres; walls and windows become small obstacles, which is why distant blasting is felt more than “heard”. Ultrasound at 40 kHz — a parking sensor — has λ ≈ 8.6 mm, so it beams, reflects cleanly, and dies on a bit of cloth.

Same formula the whole way. Only the size of the world, relative to λ, has changed.

λ = c / f is only as honest as c. Air at 20 °C, fresh water and steel differ by a factor of fifteen. A wavelength typed for a loudspeaker in a living room is nonsense in a swimming pool, and the other way around.

Related calculators