Module voices 

The voices layer: how many notes sound at once, and how a mono line joins its notes.

The count comes from the note list and each note’s held part. The join statistics come from the mono mix around each join. The method page is docs/src/analysis/voices.md.

Structs

Join

The join between two consecutive notes of a mono line.

VoicesInput

What the voices layer reads.

VoicesLayer

The voices layer of a stem.

Enums

VoicesError

Why the voices layer failed.

Constants

ATTACK_HIDES_JOIN

An attack longer than this hides a retrigger, so a join reads low.

CENTROID_SPAN

The centroid is read over this long on each side of a join.

CLUSTER_CENTS

Notes this close in pitch count as one voice: seven detuned copies of one pitch form one cluster.

DIP_RETRIGGER

A join that dips this far under the quieter neighbour is a retrigger.

JOIN_CONVENTION

The convention of a join.

JOIN_GAP_MAX

Consecutive notes join when the gap between them is under this.

JOIN_SPAN

The dip of a join is read this far either side of the onset.

OVERLAP_MIN

Two notes count as two voices when their held parts overlap for this long or more, and a later note starting this far before an earlier note’s held end stacks on it.

SUSTAIN_QUANTILE

Each neighbour’s level is this quantile of its track: a window of one period of the lower note still ripples on the higher one.

SUSTAIN_SPAN

Each note’s level beside a join is read over at most this long.

VOICES_CONVENTION

The convention of the count.