A song’s voices are a grid
A melody is a line of numbers. Stack a few of those lines on top of each other and you have a grid — one row per voice, one column per moment in time:
# two voices, read across eight moments
lead: 1 3 5 8 · 1 3 5 8 # a major arpeggio, climbing
bass: 1 · 1 · 1 · 1 · # the root, keeping the pulse
Read each row through a voice and the grid becomes an arrangement: the lead steps up a major arpeggio — do, mi, sol, do — while the bass holds the root underneath. Two voices, sounding together, written as one block of data.
It’s the same idea the rest of the site is built on — a line of numbers read through a context — but now in two dimensions at once. Across is time; down is the voices that play together. A single melody is one row of this; a whole multi-voice arrangement is the grid.
And because the arrangement is just a grid, every move you’d make on one melody is available to the whole thing at once — including the simplest. Flip the grid end to end and every voice reverses with it: hear it reversed »