An interactive essay

Why Music
Sounds Like Music

A short tour of vibrating strings, ancient ratios, beautiful compromises, and the secret math hiding behind every song you've ever loved.

Scroll to begin ↓

I — PitchMusic begins with a wiggle.

Pluck a guitar string and look closely. It blurs. What you can't quite see is that it's moving back and forth — pushing the air, pulling the air, pushing the air — hundreds of times per second. That pressure ripple travels outward, finds your eardrum, and your brain interprets it as a note.

The faster the wiggle, the higher the note. We measure wiggles per second in hertz (Hz), and by international agreement the A above middle C wiggles exactly 440 times every second.

Drag the slider below. The wave will redraw itself in real time. Press play to actually hear it.

demo 01

Frequency → pitch

440 Hz ≈ A4 (concert pitch)
middle C ≈ 262 Hz

Slide low and the wave stretches; the tone deepens and starts to feel less like music and more like a felt rumble in your chest. Slide high and the wave compresses into a thin whistle. Below about 20 Hz you stop hearing it as pitch and start feeling it. Above 20,000 Hz you stop hearing it altogether — your dog, however, is still listening.

A small history Concert pitch wasn't always 440 Hz. In Mozart's day, an "A" could be anywhere from 415 to 470 depending on the city. Orchestras tuned higher and higher to sound brighter, until singers' voices started giving out. In 1939 a committee finally locked it at 440 — and people have been arguing about it ever since.

II — TimbreSame note, different voice.

If pitch were all there was to it, a violin playing an A would be indistinguishable from a flute playing an A would be indistinguishable from your dishwasher playing an A. Obviously, that's not the world we live in.

Here's the trick: when a real instrument plays a note, it isn't producing a single clean wave. It's producing the fundamental frequency plus a cascade of quieter, faster waves layered on top. These extras are called harmonics (or overtones), and they vibrate at exact integer multiples of the fundamental: 2×, 3×, 4×, and so on.

Different instruments emphasize different harmonics. A flute is mostly the fundamental — almost a pure sine wave. A violin is bright with high harmonics. A clarinet, weirdly, favors only the odd ones. The cocktail of harmonics is what your ears call timbre.

Click an instrument preset to start hearing it, then drag the harmonic sliders to morph the sound in real time. The waveform rebuilds itself as you go.

demo 02

Build an instrument from harmonics

fundamental: 220 Hz

The wave on screen is a portrait of a sound. The peaks and valleys aren't decorative — that exact shape is what's pressing on your eardrum thousands of times per second. Two waves with the same period but different shapes give you the same note, but a completely different voice.

Worth knowing In the 1860s the German physicist Hermann von Helmholtz built a wall of glass spheres called "resonators" — each one rang sympathetically with a particular harmonic. By holding them up to a sounding instrument, he could literally listen to its harmonic recipe, one ingredient at a time. He had built, by hand, a Fourier transform.

III — IntervalsWhen two notes meet.

One note alone is a sound. Two notes together are a relationship. And the strange, beautiful fact at the heart of music — the one the Greeks went a little mad over — is that some pairs of notes sound right together and others don't, and the difference is entirely in the ratio of their frequencies.

If one note is exactly twice the frequency of another, the two sound so similar we give them the same name. That doubling is called an octave. A4 is 440 Hz; A5 is 880 Hz; A6 is 1760 Hz. The ratio is 2:1. Sing "Some-where" from Over the Rainbow — that's an octave.

Ratios that are almost as simple — 3:2, 4:3, 5:4 — also sound pleasant, in descending order of simplicity. Ratios with bigger, uglier numbers sound rougher and more tense. This is not a cultural convention; it's physics. When two frequencies share a simple ratio, their waves line up periodically and the combined wave has a clean repeating shape. When they don't, the combined wave is a mess.

Click a preset interval to start playing, then drag the slider to slide the second note around. Listen to how the combined wave goes from clean and locked to chaotic and beating:

demo 03

Two waves, one ear

Ratio ≈ 3 : 2 — a perfect fifth
(low note fixed at 220 Hz; drag the slider to change the second)

Notice how the combined wave at the bottom locks into a clean, repeating shape at 2:1 or 3:2 — but turns into a chaotic, beating mess if you put the ratio at, say, 1.414 (which is √2, an irrational number, and the infamous "tritone" — the interval medieval theorists once called diabolus in musica, the devil in music).

Legend has it According to a charming and almost certainly false story, Pythagoras discovered ratios after walking past a blacksmith and noticing that hammers of weights in simple proportions made harmonious sounds. The math was right; the metallurgy was nonsense — hammer pitch doesn't actually work that way. But the legend stuck for 2,500 years, which is the highest compliment a legend can receive.

IV — The Sacred RatiosA short menu of consonance.

Every interval Western music is built on can be traced back to a small handful of small whole-number ratios. Here they are. Click each to hear it — clicking another will swap to the new one.

demo 04

Click an interval to hear it

(All built above the same A at 220 Hz, in pure just intonation.)

The further down that list you go, the bigger the numbers in the ratio and the more tension your ear hears. The major third (5:4) feels bright and happy. Lower its top note slightly to a minor third (6:5) and the same chord turns suddenly sad. That little 25/24 nudge — about 6%, a difference of one semitone — is the whole emotional axis of Western music. Major and minor: one tiny ratio, two completely different feelings.

Stack a few of these together and you get a chord. A "major triad" is a fundamental, a major third (5:4) above it, and a perfect fifth (3:2) above that. The frequencies relate to each other as 4 : 5 : 6. Three numbers. That's it. That's the chord that powers, depending on your taste, every pop song since 1957 or every hymn since 1607.

V — Why Twelve, and What's a Cent?Two pieces of vocabulary before the trouble starts.

Quick pause for housekeeping. We've been talking comfortably about ratios and intervals, but two ideas have been creeping in that deserve their own moment in the spotlight.

Why twelve notes?

A piano has twelve keys before the pattern repeats. Your guitar has twelve frets to the octave. Western music has twelve names for notes — C, C♯, D, D♯, E, F, F♯, G, G♯, A, A♯, B — and then it starts over. Why twelve? Why not seven, or ten, or twenty-three?

The honest answer: because we got lucky. Start on any note and go up by perfect fifths (multiplying by 3/2 each time). Every fifth lands on a new pitch class. After twelve fifths you've visited twelve distinct notes, and you're almost — almost — back where you started. Take any fewer steps and the loop doesn't close; take any more and you start landing on near-duplicates. Twelve is the smallest number that nearly works.

That "nearly" will turn out to be the source of all the trouble in the next section. But twelve, give or take a fraction of a hair, is the number our system settled on. Other cultures made other choices: roughly 5 notes per octave in much of East Asian pentatonic music, 17 in Turkish theory, 22 śruti in Indian classical music, 24 in Arabic maqamat, and a theoretically wonderful 53 if you really want to be precise about it.

And what's a cent?

A cent is a unit of pitch designed to be human-friendly. By definition, one octave equals exactly 1,200 cents. Each equal-tempered semitone — one piano key to the next — is exactly 100 cents. The smallest pitch change most people can hear is somewhere around 5–10 cents. So one cent is, in practical terms, a deviation almost too small to notice; ten cents is the audible edge; twenty cents is "that note sounds wrong."

Cents are logarithmic, like our hearing. A hundred cents low in the bass and a hundred cents high in the treble sound like the same step to us, even though the underlying frequency jumps are completely different. They give us a single ruler for measuring how far anything is from anything else, in any tuning system.

The ruler below shows one full octave (1200 cents) divided two ways at once. Above the line: the twelve equal-tempered semitones, each exactly 100 cents apart. Below the line: where the pure just-intonation ratios actually want to land. Click any mark to hear it.

demo 05

The cents ruler — equal steps vs pure ratios

Top row: equal-tempered semitones, every 100¢. Bottom row: just-intonation ratios. Notice how they don't quite line up.

This is, in compressed form, the whole problem we're about to solve. The pure ratios that nature gave us — 3:2, 5:4, 4:3, and friends — almost but don't quite land on a clean grid of 100-cent gridlines. Almost. That's the gap the rest of this story is about.

VI — A Crack in the UniverseThe ratios don't quite close.

Here's where it gets weird. If perfect fifths sound this good, why not build all of music out of them?

Start on any note. Go up a fifth (multiply by 3/2). Go up another fifth. Keep going. To keep all those notes in the same octave — the same "neighborhood" of pitches where we can compare them as notes — we'll do a small accounting trick: whenever the running pitch climbs past one octave, we'll divide by two, dropping it back down. (Halving the frequency moves you down exactly one octave, which in our system is the "same note" again.) The pitch class stays the same; we're just keeping our notes in arm's reach.

After twelve such fifths — once we account for all the octaves we've dropped — we should arrive right back at the note we started on, seven octaves above where the raw climb would have taken us. Twelve fifths upstairs, seven octaves upstairs, two paths to the same landing.

They don't arrive together.

demo 06

Climb the circle of fifths

Each step in the spiral is a perfect 3:2 fifth folded back into our starting octave. The 12th step should sit exactly on top of the first. It doesn't. It sits a tiny, audible distance past it. The spiral never closes.

Here's the same fact, shown a different way. Two ladders climbing in parallel: one in seven octaves (×2 each step), the other in twelve fifths (×3/2 each step). They should reach the same top rung.

demo 06b

Two paths up the same staircase

Seven octaves of A starting from 27.5 Hz should land on 3520 Hz. Twelve perfect fifths from the same A land on 3548 Hz instead.

That little gap has a grand name: the Pythagorean comma, about 23.46 cents — roughly a quarter of a semitone. Tiny, but very audibly wrong. The math is unforgiving: (3/2)12 = 129.746… and 27 = 128. Those numbers are made of different prime factors and they will never, ever be equal. You can stack pure fifths from now until the heat death of the universe and you will not land on a pure octave.

The comma is, in a real sense, the original sin of Western music.

Mathematical aside The Pythagoreans had a religious commitment to the idea that everything in the cosmos could be expressed as a ratio of whole numbers. So when one of them — supposedly Hippasus — proved that the square root of 2 was irrational, the legend goes that the others either expelled him from the brotherhood, or drowned him at sea. The comma is a smaller scandal, but a related one: the music of the spheres, it turns out, is slightly out of tune.

VII — Two Thousand Years of CompromiseSix historical answers to one stubborn problem.

The Pythagorean comma has been the central, low-grade scandal of Western music for about 2,500 years, and every era has invented its own way of dealing with it. Dozens of named tuning systems exist, with hundreds of small variations. What follows is a tasting flight of six of the most historically interesting — chosen because each one represents a different philosophy for solving the same fundamental impossibility.

We'll move through them in order. Each is a kind of statement of values: which intervals do we love most? Which keys do we sacrifice? How much error are we willing to spread, and where? Across each subsection you'll see the same diagram — a bar for every one of the twelve fifths around the circle, showing how much each one has been narrowed from a pure 3:2 to keep the math closing. The shape of the bars is the fingerprint of the tuning.

1.Pythagorean tuning

~500 BC – 1500 AD · the medieval world

The oldest answer, and the simplest. Build every fifth as a pure 3:2 ratio, exactly as Pythagoras would have insisted. Eleven of the twelve fifths come out pristine. The twelfth has to absorb the entire 23.46-cent comma on its own, and it sounds like an animal in distress.

That ugly twelfth fifth has a name: the wolf interval. Renaissance musicians called it that because it sounded like a wolf was loose in the chapel. As long as you played in keys that avoided it, everything was fine. Stray into the wrong key, and the wolf bit you.

11 × pure + 1 × −(Pyth. comma) total: −23.46¢
listen

Pure fifth vs. wolf fifth

A pure fifth (3:2)
220 Hz + 330.00 Hz
Locked. Effortless. Pleasant.
A wolf fifth
220 Hz + 325.55 Hz
All the error, in one place. Listen for the beating.

That slow wobble underneath the wolf fifth is the two waves almost agreeing and then sliding out of phase, over and over. The wobble has a name too — beats — and its rate is exactly the difference between the two frequencies.

What worked
  • Eleven perfectly pure fifths
  • Beautiful in keys that avoid the wolf
  • Mathematically elegant — only 3:2 ratios
What didn't
  • The wolf howls whenever you hit it
  • Major thirds are 22¢ sharp everywhere — noticeably tense
  • Whole keys are off-limits

For a thousand years, composers just lived with the wolf. They wrote in keys that avoided it; whole regions of the musical map were off limits. Imagine being a painter who could only use eight colors because the other four would make customers flee.

2.Quarter-comma meantone

~1500 – 1700 · Pietro Aron and the Renaissance keyboard

By the Renaissance, polyphonic vocal music had revealed something Pythagorean tuning couldn't deliver: pure major thirds matter more than pure fifths. A chord with a perfect 5:4 third sounds glowing and luminous; a Pythagorean third sounds clenched and bright in a wrong way. So Renaissance theorists made a striking trade — described in print by Pietro Aron in 1523 and standard for the next two centuries: narrow every fifth slightly so that the major thirds come out exactly pure.

The math is satisfying. Narrow each of eleven fifths by exactly 1/4 of a syntonic comma (about 5.4¢), and four such fifths stacked together land precisely on a pure 5:4 major third. The price: the twelfth fifth — still a wolf — becomes much worse than in Pythagorean. About 35.7 cents wider than pure. A genuinely vicious sound.

11 × −(syntonic / 4) + 1 × wolf total: −23.46¢
What worked
  • Major thirds are perfectly pure (5:4)
  • Triads in central keys sound luminous
  • Beloved by vocal and string ensembles
What didn't
  • Even worse wolf than Pythagorean (+35.7¢!)
  • Only ~8 keys actually usable
  • Modulation between keys is awkward at best, catastrophic at worst

As 17th-century music grew more harmonically adventurous — composers wanting to modulate freely, to startle the ear, to wander into remote keys — meantone's restrictions became suffocating. Something had to change.

3.Werckmeister III

1691 · Andreas Werckmeister, Germany

Andreas Werckmeister, a German organist and theorist, published a small but revolutionary book in 1691 called Musicalische Temperatur. In it he proposed several "well-tempered" tunings — the third of which became the most famous. The big idea: spread the comma across several fifths so that no fifth is unbearable, and every key is playable.

His recipe: split the Pythagorean comma into four equal quarter-pieces, and tuck them into four fifths concentrated near the central keys (C-G, G-D, D-A, and B-F♯). The other eight fifths stay perfectly pure. Total error: still 23.46¢, but no single fifth carries more than 5.9¢ of it.

4 × −(Pyth. comma / 4) + 8 × pure total: −23.46¢

For the first time, every key was usable. But each key sounded subtly different. C major, ringed by mostly pure fifths, sounded calm and clear. F♯ major, sitting in the pure-fifths region but built on stacked-up Pythagorean thirds, sounded sharper and more restless. The composer no longer just chose a key for pitch — they chose it for color.

listen

The same little tune, in different keys

Listen for the subtle differences in how each key "leans." The exaggeration here is mild — historical temperaments differ in the details — but you can feel that each key has its own weather.
What worked
  • Every key now playable
  • Each key has its own character / color
  • Almost certainly the tuning Bach had in mind
What didn't
  • No interval is perfectly pure anymore
  • Asymmetric — central keys "better" than remote ones
  • The differences are subtle to modern, equal-tempered ears

This is also, almost certainly, why Bach wrote The Well-Tempered Clavier: 48 preludes and fugues, one in every major and minor key, partly as a flex to show off the new tuning, partly as a love letter to the idea that every key was now habitable.

A composer's palette Across centuries, players and theorists wrote down what they thought each key "felt like" in well-tempered tuning. C major: innocent, simple. D major: triumphant, martial. E-flat major: heroic, the key of Beethoven's Eroica. B minor: patient sorrow. F-sharp minor: rage barely held in. Whether listeners can really hear those colors, or whether the composers made them hear them by association, is one of the great open questions in music perception.

4.Kirnberger III

1779 · Johann Philipp Kirnberger, a student of Bach

Bach's pupil J.P. Kirnberger had an obsession: he wanted the pure 5:4 major third on C — meantone's signature glory — but he also wanted every key playable. His third temperament, published nearly a century after Werckmeister's, is the most explicitly fractional of the well-temperaments.

Take the four fifths from C up to E (C-G, G-D, D-A, A-E) and narrow each by exactly 1/4 of a syntonic comma. After four such steps, you've narrowed by one full syntonic comma — which is exactly the difference between four Pythagorean fifths and a pure major third. So C-E is now a perfectly pure 5:4. The luminous meantone third, preserved at the center of the keyboard.

But the total tempering is only the syntonic comma (≈21.5¢). To close the circle of fifths back to C, you need to lose the full Pythagorean comma (23.46¢). The difference between those two — about 1.95¢ — has its own name: the schisma. Kirnberger tucked this tiny remainder into the F♯-C♯ fifth, and the circle finally closed.

4 × −(syntonic / 4) + 1 × −(schisma) + 7 × pure total: −23.46¢
What worked
  • C major has a perfectly pure 5:4 third
  • All 24 keys playable
  • Mathematically the most explicit recipe
What didn't
  • Strong asymmetry — remote keys sound noticeably "off"
  • The schisma fifth is a slightly awkward hack
  • Falling out of favor by 1800 in favor of more even systems

5.Vallotti

~1750 · Francesco Antonio Vallotti, Italy

The Italian Franciscan friar Francesco Antonio Vallotti described his temperament around 1750, though it wasn't published until well after his death. It's beloved by modern early-music performers for one reason: of all the well-temperaments, it's the most symmetric.

The recipe: take the Pythagorean comma, split it into six equal sixth-pieces (about 3.91¢ each), and apply them to the six "natural" fifths on the flat side of the circle — F-C, C-G, G-D, D-A, A-E, E-B. The other six fifths, on the sharp / remote side, stay perfectly pure.

6 × −(Pyth. comma / 6) + 6 × pure total: −23.46¢

The result feels remarkably modern. The natural keys (C, F, G, D) sound pleasant but not perfectly pure; the remote keys (F♯, C♯, A♭) sound a little bright and sharp, but never offensive. It's a clean middle ground between the colorful asymmetry of Werckmeister and the flat uniformity of equal temperament — and arguably, the most graceful well-temperament ever devised.

What worked
  • Beautifully symmetric distribution
  • Every key smooth, none harsh
  • Still in active use by Baroque ensembles today
What didn't
  • No interval is perfectly pure
  • Less character than Werckmeister or Kirnberger
  • Compromise of compromises — neither colorful nor uniform

6.Equal temperament

~1900 onward · the modern standard

By the late 19th century, music had outgrown even the well-tempered map. Composers wanted to modulate freely between any keys, mid-piece, without worrying about which thirds were pure today. The pragmatic, slightly brutalist solution finally won.

The idea is brutally simple: divide the octave into twelve exactly equal steps. Not equal by ratio (no simple ratio gets you twelve steps to an octave), but equal by the only thing that matters perceptually — equal in logarithm. Each semitone is the twelfth root of two: 21/12 ≈ 1.0594631, or precisely 100 cents. Every fifth is narrowed by exactly 1/12 of the Pythagorean comma — the same tiny 1.955¢ everywhere.

12 × −(Pyth. comma / 12) total: −23.46¢

The price: not a single interval other than the octave is pure anymore. Every fifth is slightly flat. Every major third is noticeably sharp. Every key sounds the same as every other, because the tuning is identical in every key.

The reward: you can play in any key, modulate however you like, retune nothing, and the wolf is dead.

listen

What we gave up to gain everything

The major third is the cruelest cut: equal-tempered is 14 cents sharper than pure. To a trained ear, that's a lot. To most modern listeners, raised on equal-tempered everything since birth — it just sounds like a major third.
What worked
  • Every key sounds identical (just transposed)
  • Free modulation between any keys
  • The wolf is permanently dead
  • Universal compatibility across instruments
What didn't
  • Nothing is in tune except octaves
  • Major thirds 14¢ sharp — the biggest casualty
  • Key character is gone
  • A subtle, universal mild detuning everywhere

This is the tuning of every piano, every guitar fret, every default keyboard preset. It's an act of universal mild detuning that nobody notices because everybody grew up inside it. We collectively decided to be slightly out of tune everywhere, in exchange for never being terribly out of tune anywhere.

It's one of the great pragmatic compromises in the history of art — and most people who love music have no idea it ever happened.

VIII — Familiar Songs, Two WaysThe compromise, in tunes you already know.

All of this would be a curiosity if it weren't for the fact that everything you've ever heard has been in equal temperament. Every recording. Every cover of "Happy Birthday." Every backing track. They're all sitting on a grid that the pure ratios never quite agreed to.

So what if you took a song you already know — a tune your ear has memorized — and tuned it back to the pure ratios our story started with? It would sound almost the same. But not quite.

Press play on the same melody in both tunings. The differences are small. Listen especially to the third and sixth scale degrees — those are where the gap is widest. If you turn on the drone (a sustained low note underneath), the difference jumps out: in just intonation, the chords lock; in equal temperament, they shimmer with a faint, restless beating.

demo 10

The same melody, two tunings

Just intonation tunes every interval to a pure ratio relative to C. Equal temperament rounds every interval to the nearest 100¢. The most-affected note in each tune is its major third (E) — 14 cents apart between the two systems.

If you can hear the difference, congratulations: you have officially trained your ears for about ten minutes longer than most adults ever will. If you can't hear it — that's also fine. It means equal temperament did its job. You grew up inside the compromise, and the compromise feels like home.

IX — An AsideSome people see music in color.

One more curiosity before we go.

A small percentage of people — somewhere between 1 in 200 and 1 in 2,000, depending on how strictly you define it — have a condition called chromesthesia, a form of synesthesia in which musical notes involuntarily produce sensations of color. For them, a C is not just a C; it's green. Or red. Or something the color of warm copper. The mappings are stable for each person but maddeningly different between people.

The composer Alexander Scriabin had it. He built a custom instrument called the clavier à lumières — a "keyboard of lights" — and wrote a symphony, Prometheus, that included a part for it. As the orchestra played, the room was supposed to flood with the colors of the keys. The premiere mostly failed because the technology of 1911 wasn't quite up to it, but the idea is wonderful.

Below is Scriabin's mapping — one of many — laid out across an octave. Click and hold a note to play it and to bathe in its color. There is no "right" answer for which color goes with which note. Every chromesthete will tell you something different. The mystery is not solved.

demo 11

One person's notes-as-colors (after Scriabin)

Click and hold for a longer note. The color is Scriabin's claim; the sound is just physics.
Newton, slightly fibbing Isaac Newton was the one who originally insisted the rainbow had seven colors — red, orange, yellow, green, blue, indigo, violet. Look closely and there really aren't seven distinct bands; indigo in particular is doing a lot of heavy lifting. He picked seven because he wanted the spectrum to mirror the seven notes of the diatonic scale. He was, in other words, doing music theory dressed up as optics.

X — CodaWhat you're left with.

You came in knowing nothing about music. You leave knowing that pretty much everything you've ever heard has been gently lying to you — that the piano in your living room is, by ancient standards, slightly out of tune in every key, and that this was a deliberate choice made by people who cared enormously about it.

If you remember anything, let it be this: music is what happens when humans have to negotiate between physics and beauty, between purity and possibility. The numbers don't quite cooperate, and so we compromise, and the compromise becomes the art.

You should now have a great many more questions than you started with. Here are a few to take with you:

⏐ end ⏐