What your ear already knows

"I'm tone-deaf" is something a lot of people say, and almost always wrongly. The hearing a musician needs is not a gift; it is a set of skills: telling a high sound from a low one, recognising a tune when it starts on a different note, catching a rhythm. Your brain does nearly all of this every day — when you recognise a voice on the phone, or a song from its first few notes. A musician does the same, only on purpose and more precisely.

Below are five experiments. In each one you listen and press something. There are no terms here: in this chapter hearing comes first, and the names can wait. Put on headphones or turn the speakers up — some of these sound worse on a phone's tiny speaker.

A staircase that never ends

Press Start. The sound climbs in steps, one step every half second. Listen, and follow where it goes.

Up it goes, up and up — and it never gets any higher. After twelve steps it is exactly where it began, yet your ear insists it is still climbing. It is the sound twin of the Penrose staircase in Escher's lithograph, where monks climb round and round forever.

The trick is in the picture above the buttons. Each sound here is not one note but ten copies of the same note stacked in storeys, each twice as high as the one below. All ten climb together. But the loudness of each is set by a bell that never moves: loud in the middle, quiet at the edges. A note that has climbed to the top edge fades away, and at the same moment a new one appears at the bottom. Your ear follows the movement and misses the swap.

Now narrow the bell with the Width slider. Once it is narrow, barely more than one note is left — and you will hear it drop back down every twelve steps. The illusion only works because your ear hears several storeys at once.

The same note, only higher

There was something odd in Shepard's staircase: ten notes sounded together and merged into one. Why does the ear glue together notes that differ by a factor of two so easily? Try it. In each pair the first note is always the same and the second is different every time. Find the pair where the second note sounds like the same note, only higher.

People usually find it first time, though nobody told them what to listen for. When one note has exactly twice the frequency of another, we hear the two as "the same". Most of the world's musical traditions give such notes the same name: a piano has a C at the bottom, in the middle and at the top of the keyboard. Musicians call this distance an octave; we will come back to it in chapter 1 and see why the waves in the picture line up.

A note without its lowest stripe

The sound of a note on a piano, on a synth or in a voice is not a single wave either. It is a stack of frequencies: the lowest one, then twice as high, three times, four times, and so on. In the picture each stripe is one frequency and its height is how loud it is. The pitch we hear is set by the lowest stripe. Or so it seems.

Play the tune, then move the "Remove below" slider to the right. The stripes on the left go dark: the sound gets thinner, but the tune stays the same. By 200 Hz not one note of the tune has its lowest stripe left. Compare it with "The lowest stripe alone": that button plays just this one frequency, nothing else — and its pitch matches what you hear in the trimmed sound. Your brain fills in the missing frequency from the spacing of the ones that remain: they are all multiples of it.

"Like a phone line" keeps roughly what a classic telephone line passed: frequencies from 300 to 3400 Hz. The lowest frequency of a man's voice is usually 90–150 Hz, and the line does not carry it. Yet the voice still sounds low, and you recognise your friend from the first word. This is called the missing fundamental, and people have been studying it for more than a century and a half, starting with experiments on sirens in the 1840s.

The same tune from another note

Take a familiar tune, "Frère Jacques", and listen to it from a different note. The "Start on" buttons shift the whole tune, and the keyboard shows which keys it lands on.

Whichever note you start from in "Exactly" mode, you recognise the tune at once, although every note in it is different. We remember not the notes but the steps between them: up so much, down so much. That is why a song is recognisable from any note, and why musicians move it freely to wherever it is easier to play.

Now switch on "White keys only" and start on D or A, say. The pattern on the keyboard looks almost the same — each note has moved by the same number of white keys. But the tune sounds different, sadder. The white keys are not evenly spaced: some pairs have a black key between them and some do not, so the steps of the tune have changed. Your ear noticed straight away. Where "sad" and "happy" come from is the subject of chapter 14.

A tune was played starting on D, with every note moved exactly two keys to the right, black keys included. What does someone who knows it from C hear?

The same tune, a little higher A different tune with a similar shape The same tune, but sadder

When every note moves by the same number of keys, black ones included, the steps between them stay the same. The tune is unchanged; it just sounds higher. It would turn sadder if it moved along the white keys only.

A song by its knock

The last experiment has no notes at all. You hear only the rhythm of a song, knocked out. Pick it from five titles; after your answer, you hear it with the notes.

With five titles to choose from it is not hard — but try tapping a song on the table for a friend and asking them to name it. In 1990 at Stanford, Elizabeth Newton ran exactly that experiment: some people tapped out well-known songs, others guessed. The tappers expected the listeners to get half the songs. The listeners got three out of 120. Whoever taps hears the melody in their head, while the listener hears only knocks. Rhythm tells you a lot about a song, but not everything, and hearing it apart from the notes is a skill of its own. Part three of the course is about it, starting with chapter 6.

You can already do this

Five experiments, five skills: hearing which way pitch moves, recognising "the same note" an octave apart, hearing a pitch that is not in the sound, recognising a tune from any note, and recognising a song by its rhythm. A musical ear is made of skills like these, and you have them. The course develops them: makes them faster, more precise, and gives them names.

Here is the very first of them as an exercise. Two notes play one after the other; is the second higher or lower? The notes are far apart, so it is hard to go wrong.

If you missed one, never mind; listen again. This exercise has a trainer: there the notes move closer together little by little, and after a week of five minutes a day you will start to tell apart what sounds identical today.

What to play on

In this course you play all the time. Any of these three will do.

• A MIDI keyboard. The best option: keys that feel how hard you press. Connect it to the computer over USB and press "Connect MIDI" in the bar at the top of the page. The browser will ask for permission. Web MIDI works in Chrome, Edge, Opera and Firefox; Safari does not support it, and no browser on an iPhone or iPad does.
• The computer keyboard. The middle row A S D F… are the white keys, the top row W E T Y U the black ones. Z and X shift by an octave, C and V change how hard the keys "press", and Space is the tap in rhythm exercises. The keyboard layout does not matter: AZERTY or Cyrillic works the same.
• The screen. The keys in the pictures respond to a mouse or a finger.

Press any key and see what comes through. The card shows the note, where it came from and how hard it was pressed. If your MIDI keyboard has a sustain pedal, a mod wheel or a pitch-bend wheel, you will see them move too.

You can get your bearings here: play anything, each note appears on a staff, and the keys are labelled. How to read that picture comes in chapter 10; for now, just notice that the higher the note, the higher it sits in the picture.

Sound delay

Between pressing a key and hearing it, time passes: the signal travels through the computer, the browser, the sound card and the headphones. On a good system with wired headphones that is a few hundredths of a second, and you will hardly notice. Wireless headphones often add 150–250 ms, and that you will notice: the click reaches you late, you tap along with it, and a perfectly timed tap gets counted as late.

To make the rhythm exercises fair, measure your delay. Press "Start calibration" and tap along with the clicks: Space, the Tap button or a MIDI key. The course remembers the average gap and subtracts it in every rhythm exercise. If you change headphones, calibrate again.

How the course works

Twenty-two chapters in eight parts lead from sound to writing music. Part one is about what sound is, how a synth makes it and what a MIDI keyboard sends. Then come the keyboard and note names, rhythm, reading music, the ear and playing. The last part teaches you to write: a motif, a melody, harmony, a bass line, an arrangement.

Every chapter starts with a wall — something you cannot yet play, hear or read — and builds a ladder over it. Almost every section has exercises: a card with a Start button, rounds, progress dots and a score. Your best result is saved, and the tick on the course map appears once you have read a chapter and passed most of its exercises.

The trainers page gathers every skill of the course, with levels. A trainer remembers where you go wrong and gives you more of exactly that. Five minutes a day there beats an hour once a week.

And there is the Notebook: a small studio in your browser that grows with the course. At first it simply records the notes you play; then they get lengths, a tempo, drums, a bass line and chords, and at the end you export your track as a MIDI file and open it in any music program. The Notebook lives in this browser; you can take it with you as a file or a link.

Six questions

The experiments in this chapter raised questions we cannot answer yet. Here are the six main ones. Each waits for its own chapter; the course map shows where the answer is.

Why do a high C and a low C sound like "the same note"? What makes a violin sound different from a flute on the same note? Why are there exactly twelve notes? Why do some tunes sound happy and others sad? Can someone really have no ear at all? What is a song that gets stuck in your head made of?

Next

Your ear works, your keyboard is connected, your delay is measured. Time to begin, with the first question: what does a "high" sound even mean, and how can two different notes sound like one? The answer lies in vibrations of the air, and in chapter 1 you will see them.

The Notebook is empty for now: your track will live here. The first notes go into it in chapter 3, where the Notebook learns to record what you play. If you can't wait, open the example — "Ode to Joy" with drums, bass and chords.