Harken Doppler lets you hear one train horn from three places at once. A high-speed train leaves Station A, races through Station B at about 384 mph, and brakes to a stop at Station C. The horn plays one steady pitch the whole way. What each person hears depends on where they are:
All three are heard together, so you can compare them side by side and hear how they sound against each other.
That is Test I. In Settings you can also choose Test II: everything doubled, a train that reaches the speed of sound (Mach 1, about 767 mph) as it passes Station B, with stations 3.64 km apart. The Observer On Platform hears the song climb octave after octave until it rises out of hearing, then a sonic boom.
Test III leaves sound behind and moves to space. A magic spaceship carrying a lamp leaves Planet A, passes a space station (Station B) at 3/5 of the speed of light, and brakes to a stop at Planet C. The lamp shines one steady colour, 0, the tonic colour. What changes now is the colour each observer sees: the Observer On Station, the Observer On Ship and the Observer In Distance. Each view also plays its colour as the matching Harken pitch, so you can hear it too.
Play is at the top of the page. Guide, About, Settings, Reset and the timeline are at the bottom. Reset stops the run and goes back to the start.
When the run is over, each view fades to black as its sound ends. A few seconds later everything resets to the start, ready to play again.
Observer In Distance (top). A side-on panorama of the whole line: City A on the left, Station B in the middle, City C on the right, with farmland, power lines and a nuclear plant between. The small figure on the Station B platform is the Observer On Platform. From 10.6 s, when the horn's sound reaches the Observer In Distance, coloured sound rings pulse out, paced and coloured by what they hear, just as the Observer On Platform's rings are by what the Observer On Platform hears. These rings do not come from the train. They come from a small coloured dot, the heard source: the place the train was when it sent the sound the Observer In Distance is hearing now. The dot appears at Station A, follows the train along the line, fast while the train comes toward them and slower after Station B, and joins the stopped train at Station C at the very end of the run.
Observer On Platform (lower left). What the Observer On Platform sees from the platform, looking up the track. At first they look toward Station A; the moment the train passes them they turn to follow it toward Station C. Their view zooms in on the train like a camera lens (the Zoom reading shows how much) and zooms out as it comes close. Coloured sound rings pulse out of the train: they come faster as the pitch they hear rises and slower as it falls, and they take the Harken colour of what they hear for 0. As the train goes by they also hear the rush of the pass: a real recording of a bullet train, heard only in their view, loudest at the very moment the train passes them.
Observer On Train (lower right). The view through the front window, with four readings: where the train is, its speed (mph and km/h), the distance travelled, and the train clock. The Observer On Platform appears at the edge of the Station B platform as the train arrives. The cab also has a low rumble, a drone on 0vv two octaves down. It follows the Tonic but not the Register, and no other setting touches it.
Observer On Station (upper left). They stand on a platform at Station B, a space station shaped like a pentagonal icositetrahedron (24 pentagonal faces), and look out through its opening toward Planet A. The ship does not stop at the station: it flies straight through, like an express train on the outer track. The ship's lamp first appears as a coloured point of light, which grows into the ship as it comes near. The moment it passes, they turn to follow it toward Planet C. Its light pulses out in coloured light rings, paced and coloured by what they see, just like the sound rings in Tests I and II.
Observer On Ship (upper right). The view through the ship's front window: Planet C straight ahead, growing as the ship nears it, and Station B ahead on the line. The ship flies in through one of the station's openings and out through another. The lamp glows above the window in the tonic colour; for them it never changes. The console shows where the ship is, its speed (as a fraction of the speed of light, c, and in km/s), the distance travelled (in light-seconds and millions of km), and the station clock.
Observer In Distance (below). The whole route side-on: Planet A on the left, Station B in the middle with the ship's line running through it, Planet C on the right, with stars and two faint far galaxies behind. They see the ship where it was when its light left it, so the ship in their view always runs late behind where it really is.
Under each view, Heard pitch shows how far the pitch is from 0, in steps of 12 TET, to two decimal places: +11.78 means almost an octave up; −6.95 means almost seven steps down. The coloured chips show the Harken number and colour of each voice as that person hears it. The Observer On Platform's and Observer In Distance's readouts stay blank until the sound reaches them. In Test II, as the Observer On Platform's song climbs toward the top of hearing, the chips pale toward white, and in the instant before the boom, once it has passed out of hearing, the readout says above hearing.
The chips name pitches on the TET grid you choose: whole numbers at 12 TET, + tweens at 24 TET (5+), and ° marks at 48 TET (5°, 5+, 5+°). At Continuous the chips name the nearest 48 TET pitch, so they follow the slide closely. At 96 TET and finer the chips show the nearest whole number. For the Observer On Train the reading is always 0.
In Test III the readout is called Seen colour. It works the same way: the shift in steps from 0, and chips showing the Harken number and colour. The lamp itself is not tied to any grid: its colour slides smoothly, whatever the TET.
Each view has its own volume slider and a speaker button. Press the speaker to mute that view: the speaker shows a cross, the slider turns grey and the picture fades. Press it again to bring the sound back. Use this to hear any view alone, or any two together.
The Observer On Platform starts at 80% volume, and the Observer On Train and the Observer In Distance at 60%.
Press Settings to open the panel; press Done to close it. Every change takes effect at once, even while the train is running.
| Setting | What it does |
|---|---|
| Test | Test I · Mach 0.5 (start-up): top speed 384 mph, stations 1.82 km apart. Test II · Mach 1: top speed 767 mph, stations 3.64 km apart, the Observer In Distance 6.31 km back. Both legs still take 21.25 s, so the pass lands on the same second. Test III · Relativity (0.6 c): the spaceship, top speed 3/5 of the speed of light (179,875 km/s), stops 6.375 light-seconds apart (1.91 million km), the Observer In Distance 11.04 light-seconds back. Each leg still takes 21.25 s. Changing the test stops the run and goes back to the start. |
| Tonic | Chooses the pitch that is 0 (12 choices; the start-up tonic is C). The numbers never move; the colours follow absolute pitch. |
| Register | Moves the horn up or down one octave: Lower, Middle (start-up) or Higher. All three views move together, chord and all. The Doppler shifts stay the same, and the cab's low drone does not move. |
| TET | Sets the grid the heard pitch snaps to: Continuous (start-up), 3072, 1536, 768, 384, 192, 96, 48, 24 or 12. Continuous uses no grid: you hear the pitch exactly as the physics gives it. At 3072 the steps are far too small to hear, so it sounds the same. At coarser settings you hear the pitch step from one TET pitch to the next. The Observer On Train never changes pitch, so TET does not affect them. |
| Instrument | The sound of the horn, one choice for all three views: Tuba (start-up), French Horn and other brass, saxes, woodwinds, organs, strings and synths. Only sustaining instruments are offered, so the tone holds for the whole run. |
| Harmonize | Builds a chord for the horn. Voice 1 is always 0. Press + Add Voice to add a voice, up to 7 in all, and × to remove one. Each voice has a pitch (0–11 at 12 TET; 0, 0+, 1 … 11+ at 24 TET or finer) and an octave: v (below), plain, or ^ (above). All three views play the chord, each through its own Doppler shift. |
| Time | What happens |
|---|---|
| 0 s | The train leaves Station A and speeds up evenly. |
| 5.3 s | The horn's sound reaches the Observer On Platform (sound takes 5.3 s to travel 1.82 km). They start to hear it, already rising, and their sound rings begin. |
| 10.6 s | The horn's sound reaches the Observer In Distance, 3.64 km from Station A. They start to hear it. |
| 21.2 s | The train passes the Observer On Platform at Station B at full speed, about 384 mph. Their pitch plunges from near +12 to near −7 in about 0.06 s. The train begins braking. |
| 42.5 s | The train stops at Station C and the horn stops. The Observer On Train's view fades to black. |
| 47.8 s | The last of the horn reaches the Observer On Platform; their pitch has risen back to 0. Their view fades to black. |
| 53.1 s | The last of the horn reaches the Observer In Distance. Their view fades, and the app resets. |
| Time | What happens |
|---|---|
| 0 s | The train leaves Station A, speeding up twice as hard as in Test I (about 1.65 g). |
| 10.6 s | The horn's sound reaches the Observer On Platform, 3.64 km away. They start to hear it. |
| 18.6 s | Their pitch reaches 12, one octave up. Each octave from here comes four times faster than the last. |
| 20.6 s | 24, two octaves up. |
| 21.1 s | 36, three octaves up. The horn thins into a pure whistle. |
| 21.2 s | The whistle races toward the top of hearing (the readout passes +46, then +60 a few hundredths of a second later), and the rings, wire glow and chips pale toward white. |
| 21.25 s | The song rises out of hearing; for an instant the readout shows "above hearing". The train passes Station B at Mach 1. The boom: everything still to come of the approach arrives at once, a chromatic cluster with a real sonic-boom recording under it. The view flashes white and the train streaks past in a blur. At the same moment the horn's first sound reaches the Observer In Distance, 7.29 km from Station A. |
| 21.5 s | The Observer On Platform's song lands near −12 (0v), one octave below where it began, and rises slowly back toward 0 as the train brakes. |
| 42.5 s | The train stops at Station C and the horn stops. |
| 53.1 s | The last of the horn reaches the Observer On Platform. |
| 63.7 s | The last of the horn reaches the Observer In Distance, and the app resets. |
| Time | What happens |
|---|---|
| 0 s | The lamp comes on and the ship leaves Planet A. The Observer On Ship sees 0 and always will. The other two see nothing yet: the lamp's light is still on its way. |
| 6.4 s | The light reaches the Observer On Station, 6.375 light-seconds away. The lamp appears as a point of light, already turning toward a higher colour. |
| 12.75 s | The light reaches the Observer In Distance, 12.75 light-seconds from Planet A. |
| 21.0 s | The Observer On Station sees the colour at its highest, +11.31, almost an octave up. |
| 21.25 s | The ship passes Station B at its top speed, 0.6 c, and begins braking. |
| 21.3 s | The Observer On Station sees it pass, a moment later, and turns to follow it. The colour plunges to −11.55 in about 0.2 s. |
| 21.6 s | The Observer In Distance sees their highest colour, +1.42. |
| 32.3 s | The Observer In Distance sees the ship pass Station B, 11 s after it really did. At that moment the ship moves square across their line of sight, and they see −3.86: time dilation alone. |
| 34.5 s | The Observer In Distance sees their lowest colour, −3.97. |
| 42.5 s | The ship stops at Planet C and the lamp goes out. The Observer On Ship's view fades to black. The Observer On Station still sees the ship moving, at −2.90: that light left it earlier. |
| 48.9 s | The Observer On Station sees the lamp go out. Their view fades to black. |
| 55.25 s | The Observer In Distance sees the lamp go out. Their view fades, and the app resets. |
| Setting | Start-up value |
|---|---|
| Test | Test I · Mach 0.5 |
| Tonic | C (0 in the octave nearest middle C) |
| Register | Middle |
| TET | Continuous |
| Instrument | Tuba |
| Harmonize | Voice 1 only (0) |
| Volume | Observer On Platform 80%, Observer On Train 60%, Observer In Distance 60%; all views sounding |
Harken Doppler grew out of a thought experiment: how fast must a train go for its horn to rise exactly one octave as it comes toward you? The answer is half the speed of sound. The app runs that train for real, with nothing faked or time-warped, and lets you hear it from three places at once. Test II asks the next question: what happens at the speed of sound itself? Test III asks the same question of light, in the spirit of Einstein's thought experiments: a lamp on a ship moving at 3/5 of the speed of light.
The train speeds up evenly from A to B, passes B at its top speed, then brakes evenly to a stop at C. The top speed is set to exactly half the speed of sound:
vmax = c / 2 = 171.5 m/s ≈ 383.6 mph ≈ 617 km/h
Speeding up evenly from rest over a distance D to that speed takes
T = 2D / vmax = 4D / c = 21.25 s
and braking to rest over the second leg takes the same time. The acceleration is vmax / T ≈ 8.1 m/s², about 0.8 g, far beyond any passenger train in service, which is why this stays a thought experiment.
A horn sends out sound waves at a steady rate. If the horn moves toward you, each wave starts a little closer than the one before, so the waves arrive bunched together and you hear a higher pitch. If it moves away, they arrive stretched out and you hear a lower pitch. For a moving source and a listener standing still, the heard frequency is
fheard = f · c / (c − v cos θ)
where v is the train's speed and θ is the angle between the train's direction and the line from the train to the listener, both taken at the moment the sound left the horn. Straight toward you, cos θ = 1; straight away, cos θ = −1; square across, cos θ = 0 and there is no shift at all.
In steps of 12 TET the shift is
n = 12 · log2( c / (c − v cos θ) )
At v = c/2, coming straight at you:
c / (c − c/2) = 2 → n = +12, one octave up
At the same speed going straight away:
c / (c + c/2) = 2/3 → n = −7.02, seven steps down (5v)
The rise and fall are not mirror images. A source moving toward you squeezes the waves more than the same speed moving away stretches them. This is the heart of the experiment: the octave up has no octave down to match it.
The Observer On Platform's actual extremes are +11.78 and −6.95, slightly short of +12 and −7.02. Just before the pass the train is not yet quite at top speed, and because they stand 2.8 m to the side, the horn is never exactly in line with them.
Standing only 2.8 m from the track, the Observer On Platform sees the angle θ swing from nearly straight-at-them to nearly straight-away-from-them while the train covers a few metres. At 171.5 m/s that takes a few hundredths of a second: the heard pitch falls from +11 to −6 in about 0.06 s. This is the familiar "neeee-yowww" of a fast train or race car going by.
After the pass, the train is moving straight away from the Observer On Platform, so cos θ stays at −1. What changes is the speed: as the train brakes, v falls steadily to 0, and the ratio c / (c + v) rises steadily back to 1. The pitch rises from about −7 back to 0 because the train is slowing down, not because it is turning or getting farther away.
What the Observer On Platform hears at any moment left the horn earlier, when the train was somewhere else. The app's clock is the Observer On Platform's clock. For each moment t it finds the moment the sound was sent, te, from
te + r(te) / c = t
where r is the distance from the horn to the Observer On Platform at te. The equation is solved by bisection (repeatedly halving the search interval), and the Doppler shift is then taken from the train's speed and position at te. Light is treated as instant, so the pictures show where the train really is now, while the sound arrives late.
The horn starts at A at 0 s and stops at C at 42.5 s, but each person hears it start and stop at a different time, because the sound has different distances to travel. A and C are the same distance from each of them, so the delay is the same at both ends:
| Who | Distance from A (and C) | Delay | Hears the start at | Hears the stop at |
|---|---|---|---|---|
| Observer On Train | 0 | 0 | 0 s | 42.5 s |
| Observer On Platform | 1822 m | 1822 / 343 = 5.3 s | 5.3 s | 47.8 s |
| Observer In Distance | 3644 m | 3644 / 343 = 10.6 s | 10.6 s | 53.1 s |
Until the sound reaches them, the Observer On Platform and the Observer In Distance hear nothing. Each view fades to black at its own stop time.
The Observer In Distance stands 3.16 km (1822 × √3) straight back from Station B. That places them exactly as far from Station A and from Station C as A is from C, 3644 m, so the three make an equilateral triangle with 60° at every corner. The sound's trip from A to them is as long as the train's trip from A to C. So far back, they see the train mostly from the side for the whole run, and their rise and fall are gentle: +1.90 at most, −1.71 at least.
Their picture keeps Stations A and C at its left and right edges and draws everything on and near the line small (about 58% of the size it would be at 1.82 km), to show how far away they are. This is an artist's way of showing distance, not true perspective.
Sound takes time to cross the fields, so what the Observer In Distance hears at any moment was sent some seconds earlier, from wherever the train was then. In their view a small coloured dot marks that place: the heard source. Their sound rings pulse out from the dot, not from the train.
The dot appears at Station A the moment their first sound arrives (10.6 s in Test I, 21.3 s in Test II); by then the train is already well down the line. It then travels along the track behind the train. While the train is coming toward them, the dot moves fast, because that part of the trip reaches them squeezed together; that is also when they hear the pitch highest. After the train passes Station B, the dot slows, because that part of the trip reaches them stretched out, and the pitch drops. The train stops at Station C first; the dot joins it there at the very end of the run (53.1 s in Test I, 63.7 s in Test II), when the last of the sound arrives.
Ask a blind observer standing there to point at the train, and they will be wrong and right at once: wrong about where the train is, right about where its sound is coming from. They point at the dot.
The dot is found the same way as the heard pitch: for each moment, the app solves for the moment the sound was sent, then places the dot where the train's horn was at that moment.
Test II doubles everything. The train reaches the speed of sound, c = 343 m/s (about 767 mph), as it passes Station B, and the stations are twice as far apart, D = 3644 m. Doubling both keeps the time the same: T = 2D / c = 21.25 s, exactly as in Test I. The push and the braking double too, to about 16 m/s², or 1.65 g. The Observer In Distance stays at the corner of the equilateral triangle, now 6.31 km back, so the sound from Station A reaches them at 21.25 s, the very moment the train passes Station B.
The climb without limit. Coming straight at you at speed v, the pitch is raised by c / (c − v). As v nears c, that has no top. Each extra octave needs the train to close half of its remaining gap to the speed of sound:
| Speed toward you | mph | Heard pitch | On the platform, at |
|---|---|---|---|
| c / 2 | 384 | 12 | 18.6 s |
| 3c / 4 | 576 | 24 | 20.6 s |
| 7c / 8 | 672 | 36 | 21.1 s |
| c | 767 | no limit | 21.25 s |
With the train speeding up evenly, each octave arrives four times faster than the one before. The song starts as a slow rise and ends as a zip.
Out of hearing. As a note climbs, its upper overtones pass out of hearing first, leaving little but its fundamental. So the instrument fades into a pure whistle at the same pitch (around 1,500 to 3,000 waves a second), and the whistle fades out near the top of human hearing (around 11,000 to 17,000). Above that the song keeps climbing in silence. A pitch no one can hear has no Harken colour, so the rings, wire glow and chips wash to white.
The boom is the song itself. Our train is silent, so what piles up is the horn's own song. In the last instant before the pass, every pitch of the climb from about 53 upward arrives together: a chromatic cluster, at whatever TET density you have chosen (20 pitches at 12 TET, 40 at 24 TET, a dense band at Continuous). It also carries all the sound's energy at once, which makes it a boom rather than a chord. The app plays the cluster with a real sonic-boom recording mixed in under it for the felt hit, flashes the view white (every pitch at once), and blurs the train, which crosses the Observer On Platform's view in about a third of a second.
Departing at Mach 1, c / (c + c) = 1/2: exactly one octave down. Right under the boom the song lands at −12 (0v), then rises back to 0 as the train brakes to a stop. Up without limit, down exactly an octave.
Who hears the boom. Only someone almost on the line of travel. The Observer On Platform, 2.8 m from the track, hears it. The Observer In Distance does not: at the instant of Mach 1 the train is moving across their line of sight, so its speed toward them never comes near c. Their song stays gentle, at most +4.04 (at 32.4 s) and at least −3.27 (at 49.3 s). The Observer On Train moves with the horn and hears a steady 0. The pile-up does travel on down the track toward Station C at the speed of sound, so anyone waiting at C would hear a boom about 10.6 s after the pass.
The Observer On Platform in Test II stands so close to a Mach 1 train that the blast of air would knock a person flat. Picture them as a humanoid robot, with camera eyes and microphone ears, bolted to the platform: it looks just like the person, and it can turn fast enough to follow the blur.
The setting. A magic spaceship carries a lamp from Planet A, passes Station B (a space station) at its top speed, and brakes to a stop at Planet C. The speed of light is c = 299,792 km/s, and distances are measured in light-seconds (ls): how far light goes in one second, 299,792 km. The top speed is 3/5 c, 179,875 km/s. Each leg still takes 21.25 s, so each leg is 0.3 c × 21.25 s = 6.375 ls (1.91 million km). The Observer On Station stands 0.05 ls (about 15,000 km) beside the path. The Observer In Distance stands 11.04 ls (6.375 × √3) straight back from Station B, at the corner of the equilateral triangle again: 12.75 ls from Planet A's stop and from Planet C's.
The light. Where the train carried a horn, the ship carries a light: a magic lamp that shines one pure colour, like a laser, but sends it out equally in every direction, like a very bright candle. (A real laser sends its light one way, in a narrow beam; ours is magic so that every observer, wherever they stand, can see it.) Its colour is 0, the tonic colour. The lamp itself never changes. What each observer sees depends only on where they are and how the ship is moving relative to them.
The shift. Light needs no air. Only the motion between the lamp and the observer counts, and the lamp's own clock runs slow, by the factor γ:
fseen = f / ( γ (1 − β cos θ) ), β = v / c, γ = 1 / √(1 − β²)
At 0.6 c, γ = 1.25. As before, θ is taken at the moment the light left the lamp, and the shift in steps is n = 12 · log2(fseen / f).
| Direction | Ratio | Steps |
|---|---|---|
| Coming straight on | 1 / (1.25 × 0.4) = 2 | +12, an octave up |
| Going straight away | 1 / (1.25 × 1.6) = 1/2 | −12, an octave down |
| Square across the line of sight | 1 / 1.25 = 0.8 | −3.86 |
The same both ways. Sound at Mach 0.5 rises an octave but falls only seven steps; at Mach 1 it climbs without limit and falls one octave. Light at 0.6 c rises an octave and falls an octave: 2 × 1/2 = 1. With no air to carry it, there is no favoured frame: the lamp moving toward you and you moving toward the lamp give the same answer.
The slow clock. Sound shows no shift at all when the source moves square across your line of sight. Light does: the moving lamp's clock runs at 0.8 the rate of the station clocks, so its light arrives 0.8 as often, −3.86 steps. This is time dilation, seen directly as a colour. The Observer In Distance sees it best: at 32.3 s they see the ship pass Station B, square across their line of sight, and their readout shows −3.86. The ship's own clock falls behind over the whole trip: at the stop it reads about 39.8 s, while the station clocks read 42.5 s. The app runs on the station clocks.
Seeing it late. Light takes time to travel too. Each observer sees the ship where it was when the light now reaching them left it, found by the same bisection as the sound. The lamp comes on at 0 s and goes out at 42.5 s, but each sees it later:
| Who | Distance from Planet A's stop (and Planet C's) | Sees the lamp come on | Sees it go out |
|---|---|---|---|
| Observer On Ship | 0 | 0 s | 42.5 s |
| Observer On Station | 6.375 ls | 6.4 s | 48.9 s |
| Observer In Distance | 12.75 ls | 12.75 s | 55.25 s |
So at 42.5 s, when the ship has already stopped, the Observer On Station still sees it moving toward Planet C, at −2.90.
What each observer sees. The Observer On Ship: 0, always. The Observer On Station: up to +11.31 just before the pass (standing 0.05 ls to the side, they never see it exactly head-on, and just before the pass the ship is not quite at top speed), a plunge to −11.55 within about 0.2 s as they see it go by, then a slow return toward 0 as it brakes. The Observer In Distance: gentle by comparison, at most +1.42 (at 21.6 s) and at least −3.97 (at 34.5 s). The low point is below −3.86 because just after the pass the ship is moving partly away from them, still fast.
The colour. The lamp glows in the colour each observer sees. Light has no grid, so the lamp's colour slides smoothly whatever the TET; the readout, the chips and the sound follow the TET as in the other tests. A Harken colour belongs to a pitch, not to an octave, so +12 and −12 show the same colour as 0.
The magic ship. To reach 0.6 c in 21.25 s the ship speeds up by about 8,460 km/s every second, about 860,000 g. No ship, and no rider, could survive that, so this ship is magic. The light it sends, and what each observer sees, follow the real physics.
The pictures. The planets are gas giants drawn at true size, each about the size of Jupiter (radius 0.233 ls, about 70,000 km), 0.8 ls beyond its stop and a little to the side of the path. The ship is a tall steel rocket with flaps and a cluster of engines, and its lamp is a beacon on top near the nose. Station B is a pentagonal icositetrahedron, a solid with 24 pentagonal faces, with an opening at each end for the ship to fly through without stopping. Both are far too small to see across these distances, so they are drawn far larger than life (the ship as if it were about 1,200 km across); their positions stay true. The station's look is artistic: the physics places the Observer On Station 0.05 ls from the ship's line, as before. The stars and galaxies are far enough away that they do not move. The light rings pulse as the sound rings do: 2 a second at 0, twice as fast an octave up.
The sound: an effect, not physics. In Tests I and II the sound is the experiment: the horn's waves travel through the air, and what each observer hears is real. Test III takes place in the vacuum of space, where there is no air or anything else to carry sound waves. The lamp makes no sound, and even if it did, nothing could carry that sound across empty space to the station or to the observer far away. In space, Test III would be completely silent.
What you hear is an add-on: each view plays its observer's seen colour as the matching Harken pitch, so your ears can follow the colour as it shifts. To keep it in the background, it plays at half the volume of Tests I and II (the volume sliders work as before). Here the light is the experiment; the sound only helps you follow it. Tonic, Register, TET, Instrument and Harmonize all apply, and the Observer On Ship keeps the low drone. Test III has no pass recording and no boom.
The physics gives a pitch that slides smoothly. At the start-up setting, Continuous, the app uses that pitch as it is, with no rounding. This is as close as a simulation can get: the pitch is worked out 400 times a second and the browser's audio clock joins the points in a smooth line, so the 0.06-second plunge still has about 24 points. Any other TET setting rounds the pitch to the nearest step of that equal division of the octave. At 3072 TET one step is 1200 / 3072 ≈ 0.39 cents, far finer than the ear can tell apart, so the slide sounds continuous. At 12 TET the pitch can only sit on whole steps, so you hear it jump from step to step. For coarse settings (96 TET and below) the app moves the pitch in hard steps so the jumps are clean.
Pitches are numbered from the tonic: 0 is the tonic, 1 to 11 are the steps above it, and + marks the 24 TET tween between two steps. Every number has a Harken colour bound to absolute pitch, so the chips and the sound rings show exactly what is being heard. The tonic is placed in the octave nearest middle C, so changing it never moves the sound more than half an octave. The Register setting then moves the horn one octave lower or higher if you wish; the cab's drone stays where it is.
The air is still: no wind, no temperature changes along the way, no echoes from the ground or buildings. The horn is a single point at the front of the train. Light is instant. Loudness does not change with distance, so each view can be heard clearly and compared at an equal level (the recorded rush of the pass is the one exception). None of these changes the pitches you hear.
The ideas inside the simulation, for anyone curious: no math needed.
Our train has a horn that plays one steady note, 0. It never changes.
Three people are listening at the same moment. The rider on the train hears 0 the whole trip. The observer on the platform hears the horn climb higher and higher as the train rushes toward them, all the way up to 12, a full octave. As it passes, the pitch plunges to about 5v, seven steps below 0. Then it slowly comes back to 0 as the train brakes to a stop at Station C. Far away in the fields, a third observer hears their own gentler version, and hears it late: the sound needs 10.6 seconds just to reach them.
So who is hearing the real note? All three of them. How fast the sound source is moving toward you or away from you changes what you hear. Scientists call this the Doppler effect, and you can hear it every time a siren races past.
The observer in the fields stands about 3.6 km from the train when it starts. When the train pulls away from Station A, they see it move at once. Light crosses that distance in about 12 millionths of a second, faster than a blink.
The sound is much slower. It crawls across the fields at about 343 metres every second, so the horn's song takes 10.6 seconds to reach them. For those 10.6 seconds they watch the train race along in total silence. And at the end, after the train has stopped at Station C, they still hear it singing for 10.6 seconds more, like hearing an echo of something that has already happened.
Light travels about 870,000 times faster than sound. You have seen this in a thunderstorm: the lightning flash comes first, and the thunder rolls in later. Count the seconds between them and divide by 3, and you know about how many kilometres away the lightning was. (Divide by 5 for miles.)
A horn makes sound by pushing the air in waves, many times every second. How many waves reach your ear each second decides the pitch you hear: more waves, higher pitch.
When the train is standing still, its waves spread out evenly in every direction. But when it rushes toward you, each new wave starts from a spot closer to you than the last one, so the waves bunch up and arrive more often. At 384 mph, half the speed of sound, they arrive exactly twice as often. Twice as many waves per second is a full octave: the 0 becomes 12.
Here is the surprise. When the train speeds away from you just as fast, the pitch does not drop a full octave. The waves only stretch out by half again, and the note falls about seven steps, to 5v. Coming and going are not mirror images. To drop a whole octave going away, the train would have to move at the speed of sound itself, and that is Test II.
A happy coincidence: 384 also appears in the Harken series of finer and finer tunings: 12, 24, 48, 96, 192, 384. That is only a trick of using miles per hour. The real link is simpler and deeper: the octave is a 2:1 ratio, and half the speed of sound gives exactly 2:1.
You cannot see sound, so the app draws it for you. Each observer's view has its own rings. On the platform they pulse out from the train's horn. Far out in the fields they pulse from a small dot that marks where the horn was when it sent the sound arriving now, because that is where the sound really comes from. They pulse twice a second when that person hears 0. When the pitch climbs to 12, they pulse twice as fast, four times a second. Their colour is always the Harken colour of the note being heard. So the rings let you see the song change, even with the sound turned off.
Real sound waves do something even stranger. Picture dropping pebbles into a pond while walking forward. Each pebble makes a circle of ripples, but each new circle starts a little further ahead than the last. In front of you the ripples crowd together. Behind you they spread apart. That is exactly what a moving horn does to the air, and it is why the observer ahead hears a higher note and the observer behind hears a lower one.
Test I tops out at half the speed of sound. Test II goes all the way.
At the speed of sound (Mach 1) the train keeps pace with its own song. Every wave it sends forward lands right on top of the one before. The song climbs octave after octave, each one arriving four times faster than the last, until it rises out of hearing. Then everything still to come of the approach arrives at once: BOOM. That is a sonic boom, what people mean by "breaking the sound barrier." Ours is made of music: a chromatic cluster of every pitch, all at once. Right after it, the song lands exactly one octave below where it started, on 0v, and drifts home to 0.
At twice the speed of sound (Mach 2) something even stranger happens: the train outruns its own song. Sounds the horn makes later reach you before sounds it made earlier. A train heading straight at you at Mach 2 would arrive in silence, then hit you with the boom, and then you would hear its approach song played backwards, in its true pitch, 0. As it raced away, the song would drop about 19 steps, to 7vv. So "playing the film backwards" is not only a movie trick. At Mach 2, physics really does it.
Stopping a train from 384 mph means getting rid of a huge amount of energy: for a 400-tonne trainset, about 6 billion joules in 21 seconds, roughly 300 megawatts, the power of a small town. Brakes that rub turn all of it into heat.
But it does not have to be wasted. Many electric trains, trams and electric cars use regenerative braking: the motors run backwards as generators while the train slows, turning motion back into electricity and feeding it into the line. Our magic track can do the same. For one stop from 384 mph that is about 1,640 kilowatt-hours, roughly two months of electricity for an average American home, collected in 21 seconds. And Leg 1 needed that same energy to get the train up to speed, so braking one train can help launch the next.
Is the heat that does escape lost to the universe? Never lost, only spread out. Energy cannot be destroyed, but once it becomes warm air drifting away from the rails, it is too thinly spread to do useful work again.
Could a loud horn make the power lines ring along with it? Almost, but not quite. A 300 m span of wire under tension has a very low natural note, about a quarter of a wave per second, with overtones packed so closely that some overtone always sits near any pitch the horn plays. But a cable weighing about 1.5 kg per metre is far too heavy for sound in the air to set moving, its own stiffness soaks up what little motion there is, and the horn's pitch keeps sliding, so nothing builds up. What really makes power lines vibrate is wind: steady wind sheds little whirls of air off the cable, which is why small weights hang near the towers to calm them.
So the faint glow on the wires near the train is artistic licence, like the sound rings: the wires "ring" in the colour of what each observer hears.
The sound physics in this app is real. The train is a little bit magic.
How hard it pushes. Test I speeds up and brakes at about 0.83 g, like a very hard emergency stop in a car, but held for 21 seconds. A real high-speed train brakes at about 0.1 g and would need about 15 km to stop from 384 mph. Test II doubles that, to about 1.65 g.
How it stops. Wheels alone cannot do it: the grip between steel wheels and steel rails is far too weak. So our train is made of a very light material, and the track itself does the braking: an active magnetic track, plus pneumatic clamps with diamond strips that grip the rails without wearing out, and a special system to carry away the heat.
Could you ride it? Yes, sitting down and belted in. Astronauts ride 3 to 4 g for minutes at launch. But anyone standing would be thrown down the aisle, seats should swivel to face backwards for the braking leg, and the instant switch from speeding up to braking at Station B would be a jolt. Coffee cups would not survive the trip.
Sound needs air to travel through, and that makes coming and going different. A train racing toward you at half the speed of sound raises its horn a whole octave, but racing away it lowers it only seven steps.
Light needs nothing to travel through. It always moves at the same speed, however fast you or the lamp are moving. So in Test III, coming and going come out even: the lamp coming toward you at 3/5 of the speed of light looks a full octave up, and going away a full octave down.
And there is a surprise that sound never gives. When the ship crosses your view sideways, neither coming nor going, its colour still shifts down, about four steps. A moving clock runs slow, and the lamp's light is a kind of clock: it ticks out its waves more slowly, so you see a lower colour. Albert Einstein worked this out in 1905, and it has been measured many times since. Satellite navigation would drift off course every day if engineers did not correct for it.
Watch the Observer On Station as the ship comes at them. They see the ship cross its 6.375 light-seconds in about 15 seconds, not the 21 it really takes. Why? Its light from far away left long ago; its light from close up left only just now. The light from the whole approach arrives crowded into a short time, so the ship seems to race in at up to 1.5 times the speed of light. Going away, the opposite happens: the light arrives spread out, and the ship seems to crawl away at less than 0.4 c.
Nothing really breaks the speed of light here. It is the same squeezing that raises the colour on the way in, and the same stretching that lowers it on the way out. Astronomers really see this, in jets of gas shot out from distant galaxies at nearly the speed of light: some seem to move several times faster than light, for exactly this reason.