NMR Spin Echo

The sample's magnetisation in the plane the pulse tipped it into, seen turning with its line: each arrow is the spins in one slice of the field, a little faster or slower than the rest, so they fan out and their sum (the bright arrow, the note) shrinks. A π pulse (the pedal) flips the fan over, so the fast ones are behind and catch up: the arrows close again into the echo. The ring is what T₂ and diffusion leave, which no echo can pass.

Each sounding sample's signal, as a spectrometer records it, over the last few seconds: it starts at the pulse and fades as the spins dephase. The dashed line under each is what an echo could still bring back (T₂ and diffusion); a π pulse (a tick) turns the fade around, and the signal swells back to that line as the spins come back into step.

The spectrum a chemist would plot for this key's magnet: chemical shift in ppm, falling to the right, each line as wide as the field's spread and T₂ make it. Under it, the same lines in hertz: the notes you hear. On the low keys, weak magnets, J is no longer small beside the shifts' spacing, and the multiplets lean towards each other.

How much of each key's magnetisation points along the field, ready for its next pulse: full at rest, emptied by a 90° pulse (or partly, by a softer one), turned upside down by a π pulse, and growing back in T₁. A key struck again before it has grown back sounds quieter; struck while upside down, it sounds inverted.

Every note is a chord of its sample's chemical groups, struck at once and fading together; each echo lights the whole chord again. A multiplet's lines lie too close to part here: they are heard as a beat.