Automatic Chord Labelling: A Figured Bass Approach

Abstract

Automatic chord labelling is challenging, largely because the identification of chords directly from the musical surface can be ambiguous. Figured bass can potentially offer indications of harmonic rhythm and non-chord tones, thereby reducing this ambiguity. This paper proposes a series of four rule-based algorithms that automatically generate chord labels for homorhythmic Baroque chorales based on both figured bass annotations and the musical surface. These are applied to the existing Bach Chorales Figured Bass dataset, which consists of 139 chorales composed by Johann Sebastian Bach, and includes both the original music and figured bass annotations. Analysis of the chord labels produced by our algorithms reveals occasional discrepancies between the chords implied by the figured bass and the scored voices, something that provides a useful basis for exploring different chord interpretations. The chord annotations produced by our system are presented as the new Bach Chorales Multiple Chord Labels (BCMCL) dataset, which provides a choice of four parallel chord labels for each chorale. These range from one set of labels based only on the figured bass, which do not assume any music theoretical ideas proposed after the time the chorales were written, to a set of labels based on both the figured bass annotations and the full musical surface that considers the music from the perspective of modern tonal music theory. It is hoped that this dataset and the algorithms used to label it will be of interest for both future musicological research and research on automatic chord labelling systems. CCS CONCEPTS • Information systems → Music retrieval; • Applied computing → Sound and music computing; • Computing methodologies → Rule-based approaches.