A loudspeaker, a tube and a microphone
The stage shows a loudspeaker cone at one end of a tube and a microphone at the other, wired to an oscilloscope. Press Play and the cone starts to pump in and out, and the sound sets off down the tube. Nothing reaches the microphone straight away, so the trace stays flat until the first compression arrives. Two things are altered so you can follow it. In everyday sounds the air moves back and forth by much less than a millimetre, so the movement is exaggerated, and everything runs in slow motion, several hundred times slower than reality at the starting frequency. The top-left chip gives the factor at 1× speed.
Squeezed and stretched air
Each dot stands for a small parcel of air. Where the cone pushes forward, parcels crowd together into a compression, a region of higher pressure, tinted warm and tagged C. Where the cone pulls back, they spread apart into a rarefaction, a region of lower pressure, tagged R. Watch the highlighted dot: it slides forward and back along the tube and always returns to its dotted rest line, while the C and R tags march steadily towards the microphone. The pattern travels; the air stays where it was. The gap between neighbouring compressions is one wavelength, marked under the tube. Check it against the scale: at 500 Hz in air it is 0.68 m, because .
Reading the trace
The oscilloscope plots the microphone's output, a potential difference that follows the air pressure there, against time. It is a graph, so a peak means a moment of high pressure at the microphone, not air that has moved upwards. The bright spot is the beam drawing that graph right now. Each sweep begins as the signal crosses the centre line going up, which holds a steady note still on the screen.
To find the frequency, first find the period: count the divisions across one complete wave and multiply by the timebase. At the start, one wave spans 4.0 divisions at 0.5 ms/div, so ms and Hz. For the size of the signal, count divisions from the centre line up to a peak and multiply by the Y-gain. If the peaks run off the screen, pick a Y-gain setting with more mV per division; if the waves are crammed together, pick a smaller timebase.
Pitch, loudness and what we can hear
Raising the frequency makes the note higher and packs more waves onto the trace. Raising the amplitude makes the note louder and the peaks taller, while the spacing stays put. People hear roughly 20 Hz to 20 kHz. Lower frequencies are infrasound and higher ones ultrasound; the slider marks both edges, and the Frequency readout names the band you are in. The limits are approximate, and the top one drops as people get older.
Other materials, and none at all
Water and steel pass sound on far faster than air, at about 1500 m/s and 5900 m/s here. The loudspeaker still sets the frequency, so the waves stretch out: at 500 Hz the wavelength becomes 3.0 m in water, matching . The stage can zoom out to fit the longer waves, so read the new scale under the tube. Choose Vacuum and the cone keeps vibrating, but with no particles to push, nothing reaches the microphone and the trace stays flat. For a slower look at how particles pass a longitudinal wave along, try the slinky simulation.