Exploring music through data and algorithms

Exploring the potential of digitization and computational models for Brazil's musical cultural heritage
View project →
Goals and Challenges
View project →
Computational and Digital Approaches to Music Scholarship
View project →
Enabling interactive music visualization for a wider community
View project →This paper reports on an experiment extends a methodology to analyze and compare harmonic progressions at moments of formal segmentation in different styles based on corpora of Roman numeral annotations, and deploys it to the styles of Brazilian choro, Anglophone popular and jazz music.
What can we learn from the content of surviving medieval music manuscripts about their cultural context?
How many composers are we missing in RISM? What percentage of medieval sources of Gregorian chant have we already cataloged? How many differences in music prints do we expect to find between editions? How large is the coverage of songs from genres of a folk music tradition? And how close are we in estimating the size of the harmonic vocabulary of a large number of composers?
How much are we still missing?
We present a dataset of occurrences of so-called trope elements in manuscripts of European medieval chant.
We introduce a novel perspective on set-class analysis combining the DFT magnitudes with the music visualisation technique of wavescapes.
In this study, we employ a computational model that formalizes music-theoretic conceptualizations of tonal space, and use it to infer the most likely interval distributions for pieces in a large corpus of music, represented as so-called ‘bags of tonal pitch classes’.
We present a dataset of 264 annotated piano pieces of nine composers, composed in the long 19th century.
Here, we analyze more than 40 000 pieces of plainchant to better understand the evolution of modes and pitch distributions in a period of five centuries that saw the development of the Western modal practice.
We introduce phantom curves, a novel music-theoretical concept based on the discrete Fourier transform (DFT), and document the creative process that led to their discovery
We introduce wavescapes, a novel visualization method for tonal hierarchies that combines the visual representation of keyscapes with music analysis based on the discrete Fourier transformation (DFT) and illustrate it by analyzing compositions by Josquin, Bach, Liszt, Chopin, Scriabin, Webern, Coltrane, and Ligeti.
In this study, we use the Latent Dirichlet Allocation (LDA) topic model to discover tonal structures in a corpus of more than 2000 musical pieces from large historical range (ca. 600 years). and that the inferred topics have music-theoretically meaningful interpretations.