Detecting harmonic change in musical audio

Abstract

We propose a novel method for detecting changes in the harmonic content of musical audio signals. Our method uses a new model for Equal Tempered Pitch Class Space. This model maps 12-bin chroma vectors to the interior space of a 6-D polytope; pitch classes are mapped onto the vertices of this polytope. Close harmonic relations such as fifths and thirds appear as small Euclidian distances. We calculate the Euclidian distance between analysis frames n + 1 and n − 1 to develop a harmonic change measure for frame n. A peak in the detection function denotes a transition from one harmonically stable region to another. Initial experiments show that the algorithm can successfully detect harmonic changes such as chord boundaries in polyphonic audio recordings.

Are the Emotions Expressed in Music Genre-specific? An Audio-based Evaluation of Datasets Spanning Classical, Film, Pop and Mixed Genres

“…All features were extracted with the MIRtoolbox 2 ) using a framebased approach (Tzanetakis & Cook, 2002a Harte, Sandler, & Gasser, 2006), the centroid of an uncollapsed Chromagram], the frames were 2 s long, with an overlap of 50%, while for structural features (Repetition), frame length was 100 ms and overlap also at 50%. The results from the frames were then summarized by either the mean, the standard deviation, or the slope function.…”

“…A somewhat related measure, Spectral Entropy, is based on the spectrum collapsed into one single octave from which the entropy is calculated as an indication of simplicity or complexity of the chroma components. Harmonic Change Detection Function (HCDF) indicates the tonal diversity along the time (Harte et al, 2006) and Key Clarity, based also on chromagram, compares the ensuing tonal profile to preestablished key profiles representing major and minor keys (Gomez, 2006). Also there is a new measure of majorness , which is the amplitude difference between the best major score and minor score obtained by correlation with the key profiles.…”

A System for Automatic Chord Transcription from Audio Using Genre-Specific Hidden Markov Models

“…Harte and Sandler proposed a 6-dimensional feature vector called Tonal Centroid, and used it to detect harmonic changes in musical audio [13]. It is based on the Harmonic Network or Tonnetz, which is a planar representation of pitch relations where pitch classes having close harmonic relations such as fifths, major/minor thirds have smaller Euclidean distances on the plane.…”

Acoustic Chord Transcription and Key Extraction From Audio Using Key-Dependent HMMs Trained on Synthesized Audio

“…Recently, Harte et al propose a six-dimensional feature vector called tonal centroid and use it to detect harmonic changes in musical audio [2]. It is based on the harmonic network or Tonnetz, which is a planar representation of pitch relations where pitch classes having close harmonic relations such as fifths or major/minor thirds have smaller Euclidean distances on the plane.…”

“…A chroma vector is often a 12-dimensional vector, each dimension representing spectral energy in a pitch class in a chromatic scale. We also describe the use of the tonal centroid vector, which is a six-dimensional feature obtained from a 12-dimensional chroma feature as introduced by Harte et al [2], and compare it with the conventional chroma vector. We show that the tonal centroid features are more robust and outperform the conventional chroma features in chord recognition [3], [4].…”