The tool

Voice-leading explorer

Voice leading is easier to see than to describe. Pick a progression, a key, and four strings to play it on: this draws every voice as a line, loops the result, and counts the semitones — against the same chords in root position, which is what the neck gives you when nothing is choosing.

How to read it

Each column is one chord and each dot is one voice, colored by the job it is doing in that chord — root, third, fifth, seventh. The vertical axis is pitch, ten pixels to the semitone, so a line's slope is exactly how far that voice travels. A flat line is a note both chords contain, held rather than replaced. Copper marks the move that does the pulling: a seventh falling a half step onto the next chord's third.

Under the diagram the same voicings appear as fretboard cards, in the order the progression plays them. The two buttons on each card play the shape alone or over its root, and the two round buttons loop the whole progression the same two ways. There is deliberately no walking bass and no drums — a bass line of its own would cover the very moves the diagram is drawing — so the second button adds nothing but each chord's root, held underneath while the voices move above it. When a loop is running, a copper band follows the chord you are hearing.

Use that second button whenever a chord sounds like it might be a different chord. Most of these voicings are inversions, which means the lowest note you hear is not the root, and a chord in that position is genuinely ambiguous on its own — a bass note is what settles which of the possible readings it is. It is also what makes the comparison below a fair one.

The count, and what it is measured against

The number under the diagram adds up every semitone every voice moves. Switch the voicings to root position and it multiplies: same chords, same four strings, but taken as the shape with the root lowest every time — the first shape most players learn, and the one you land on by default when nothing is choosing between the inversions.

That comparison is the argument for voice leading, and here both sides are computed rather than asserted. Across every combination this page offers — four progressions, twelve keys, three string sets — the voice-led count is lower, never merely equal, and root position costs at least four times as much movement. No voice in a voice-led chain ever travels more than three semitones.

One caution about listening rather than counting. Root position puts the root at the bottom of every chord, and the voice-led inversions do not, so heard without a bass the root-position version has an advantage that has nothing to do with voice leading: it is easier to follow because each chord announces its own name. The second loop button takes that advantage away.

On the lower string sets it also puts a different problem in its place, and the problem is worth hearing. A chord voiced down there can land its lowest note exactly where the bass is playing, and the two stop sounding like two things — they thicken into one. On the bottom four strings that happens to two root-position chords out of every three, and to rather fewer of the voice-led ones, which sit higher by choice. On the top four it never happens at all, in any key, in either voicing — which is one of the plainer reasons guitarists comp up there when someone else is holding down the bottom.

One more thing is worth watching as you change key. For three of the four progressions the voice-led count is the same number in all twelve keys and on all three string sets: voice leading is a property of how the chords relate, not of where they sit on the neck. The minor cadence is the single exception, and it moves by two semitones at most. The root-position count, meanwhile, wanders from key to key, because root-position shapes fall differently on the strings and the jumps between them are an accident of tuning rather than a musical decision.

What the rule cannot decide

The chain is chosen by a rule: the smallest total movement, ties broken by the top voice and then by position. A rule is not taste. A melody note you want on top, a deliberate leap for effect, a register you are keeping clear for a singer — any of those beats the smallest move, and the article this grew out of makes that case with the ears rather than the arithmetic. What the rule is good for is showing the floor: the least the music has to move, so you can hear what you are trading away when you leave it.

The four progressions here are short on purpose. Voice leading is a picture you can read at a glance or not at all, and a twelve-bar form drawn this way is a wall of dots. The longer forms live on the progression pages, where the same chains are laid out as a chart with a band playing underneath.

Where to go from here

If the diagram raised a question the tool cannot answer, the reasoning is in the voice-leading article: why the sevenths fall, what the two moving voices have to do with the guide tones, and how the whole thing sounds when it is played badly on purpose. The shapes themselves come from the drop-2 family, and the voicing explorer holds all of them — every root, quality, inversion and string set — if you want to browse rather than compare.

Where Drop2 comes in

Choosing between inversions is what this page does; the app is where the shapes get into your hands. Drop2's Learn and Train screens drill the same drop-2 grips these chains are built from, and its mic grading listens to whether the chord you played is the chord you meant. Play-Along (part of Premium) puts them in a form with a band, which is the next thing to do once a progression stops needing the diagram.

What the app does not do is pick the voice leading for you — that decision stays yours, on this page or on the neck.

Next: The ii–V–I, in every key →