How this instrument works
Nothing about the wave itself speeds up. Sound leaves a moving horn at 343 m/s, exactly as it leaves a parked one, because the medium sets that speed and no source gets a vote. What motion changes is spacing. Any source chasing its own wavefronts crowds them ahead of itself and shortens each wavelength; any listener walking into those wavefronts simply sweeps them up at a higher rate. Both raise pitch, but by different arithmetic — which is precisely why source speed sits in the denominator here while observer speed sits in the numerator.
Christian Doppler published this idea in Prague in 1842, inside a paper about colours of double stars — an application that proved wrong, since stellar colour follows surface temperature and real stellar shifts are far too faint to see by eye. His principle outlived his bad example. Christophorus Buys Ballot tested it properly in 1845 by loading trumpeters onto an open railway carriage running between Utrecht and Amsterdam, posting musicians of perfect pitch along the track to write down what they heard. Hippolyte Fizeau soon carried it over to spectral lines, where it became the ruler astronomers still lay against receding galaxies.
Classical Doppler arithmetic holds only while the source stays slower than whatever wave it emits. Push source speed up to wave speed and the denominator collapses toward zero: wavefronts pile onto one cone and you get a shock — sonic boom, not infinite pitch. Three quieter assumptions hide inside as well. Speeds must be components along the line joining source and listener; the medium must be still, since wind alters effective wave speed; and every term here is Newtonian, so this expression does not describe light, which demands its relativistic cousin, complete with Lorentz factor and transverse shift.
- Enter Emitted frequency — pitch measured at the source itself — in hertz, kilohertz, or megahertz.
- Fill in Source speed towards observer and Observer speed towards source: positive while the gap closes, negative while it opens.
- Set Wave speed in the medium. Air at 20 °C carries sound at 343 m/s; fresh water manages roughly 1480 m/s.
- Read Observed frequency, then divide by your emitted value to see that shift expressed as a plain ratio.
Worked example — a 1000 Hz calibration tone
A loudspeaker on a tripod sends a 1000 Hz calibration tone across the still air of an empty field at 20 °C, where sound travels 343 m/s, and the technician holding the level meter stands perfectly still. Emitted frequency 1000 Hz; Source speed towards observer 0 m/s; Observer speed towards source 0 m/s; Wave speed in the medium 343 m/s. Substituting gives f′ = 1000 × (343 + 0) ⁄ (343 − 0) = 1000 × 1 = 1000 Hz exactly.
A null result, and a genuinely useful one — the first thing worth checking on any Doppler instrument, because both speed terms vanish, wave speed cancels against itself, and that ratio must land on unity. Now roll your speaker forward at 34.3 m/s, one tenth of sound speed, and this same tone reads 1111.11 Hz. Park it again and walk your technician forward at 34.3 m/s instead: 1100 Hz. Identical closing speed, answers eleven hertz apart. That asymmetry is physical, not some rounding artefact.
Questions
Why does moving the source give a different answer from moving the listener?
Because those two motions act on different things. A travelling source physically shortens the wavelength inside the medium, so its effect compounds: the factor is v ⁄ (v − v_s), which grows without bound as source speed nears wave speed. A travelling listener leaves wavelength untouched and merely meets wavefronts at a higher rate, giving the gentler linear factor (v + v_o) ⁄ v. At one tenth of sound speed those two differ by about 1%; at half of sound speed they diverge wildly.
Which way round do the signs go?
Positive means closing. Give Source speed towards observer a positive value when your source heads at whoever is listening, and do likewise for Observer speed towards source when your listener heads at whatever is emitting; either way the observed frequency rises. Retreat takes negative values. Only components along the straight line between them count, so a car passing at 30 m/s on a road 20 m away contributes well under its full speed near closest approach.
Does this apply to light, radar, or starlight?
No. Light has no medium, so there is no v to divide by and no way to separate source motion from observer motion — only relative velocity survives. The relativistic expression uses the Lorentz factor and predicts a transverse shift even when nothing approaches or recedes. At slow speeds classical numbers come close, yet error creeps in at order (v ⁄ c)², so reach for the relativistic Doppler instrument when handling starlight, satellite links, or police radar.
What happens when the source reaches sound speed?
Division by zero, and physics stops meaning anything — which is why this sheet refuses source speeds at or above wave speed. In reality wavefronts stop spreading ahead of that source and stack onto the Mach cone; nothing is heard until that cone sweeps past, then everything arrives at once as one boom. Faster still, and the source outruns its own sound, so events reach your ears in reverse order.
Why does a passing siren slide down in pitch rather than jump?
Because radial speed changes smoothly through that pass. Far off on approach, nearly all of the vehicle's speed is closing speed, so pitch sits near maximum. At closest approach radial speed is momentarily zero and pitch reads unshifted. Afterwards it settles at its low value. That glide occupies only a few seconds either side, and it steepens sharply as you stand nearer to the road.
What wave speed should I enter?
Whatever your medium genuinely carries waves at. Dry air at 20 °C gives 343 m/s, climbing roughly 0.6 m/s per degree Celsius, so a frosty morning sits nearer 331 m/s. Fresh water runs about 1480 m/s, steel around 5900 m/s. Wind counts too: sound travelling downwind effectively gains the wind speed, so outdoor work wants that component added, measured from emitter toward listener.