A chord distance metric based on the Tonal Pitch Space and a key-finding method for chord annotation sequences

  • Lucas Marques Federal University of Alfenas

Resumo


Music Information Retrieval (MIR) is a growing field of research concerned about recovering and generating useful information about music in general. One classic problem of MIR is key-finding, which could be described as the activity of finding the most stable tone and mode of a determined musical piece or a fragment of it. This problem, however, is usually modeled for audio as an input, sometimes MIDI, but little attention seems to be given to approaches considering musical notations and musictheory. This paper will present a method of key-finding that has chord annotations as its only input. A new metric is proposed for calculating distances between tonal pitch spaces and chords, which will be later used to create a key-finding method for chord annotations sequences. We achieve a success rate from 77.85% up to 88.75% for the whole database, depending on whether or not and how some parameters of approximation are configured. We argue that musical-theoretical approaches independent of audio could still bring progress to the MIR area and definitely could be used as complementary techniques.

Palavras-chave: Music Analysis and Synthesis, Music Information Retrieval

Referências

RIEMANN, H. Harmony Simplified: or, The Theory of the Tonal Fuctions of Chords. London: Augener Ltd., 1893.

Schoenberg, A. (2010). Theory of Harmony. University of California.

Carpenter, P. (1983). ”Grundgestalt” as Tonal Function. Music Theory Spectrum, 5, 15-38. doi:10.2307/746093. p.16-7.

Lerdahl F. (1982) Tonal Pitch Space. In Music Perception: An Interdisciplinary Journal, Vol. 5 No. 3, Spring, 1988; (pp.315-349) DOI: 10.2307/40285402.

Chew E. (2002) The Spiral Array: An Algorithm for Determining Key Boundaries. In Anagnostopoulou C., Ferrand M., Smaill A. (eds) Music and Artificial Intelligence. IC-MAI 2002. Lecture Notes in Computer Science, vol 2445. Springer, Berlin, Heidelberg.

Pauws S. (2004) Musical key extraction from audio. In Proceedings of the 5th ISMIR, Barcelona, Spain., pages 96–99, 2004.

Izmirli O. (2006) Audio key finding using low-dimensional spaces. In Proc. of the International Symposium on Music Information Retrieval (ISMIR), pages 127–132, Victoria, Canada, 2006.

Hu, D., & Saul, L. K. (2009) A Probabilistic Topic Model for Unsupervised Learning of Musical Key-Profiles. In ISMIR (pp. 441-446).

Krumhansl, C. L., & Kessler, E. J. (1982). Tracing the dynamic changes in perceived tonal organization in a spatial representation of musical keys. In Psychological Review, 89(4), 334-368.

de Haas W.B., Robine M., Hanna P., Veltkamp R.C., Wiering F. (2011). Comparing Approaches to the Similarity of Musical Chord Sequences. In: Ystad S., Aramaki M., Kronland-Martinet R., Jensen K. (eds) Exploring Music Contents. CMMR 2010. Lecture Notes in Computer Science, vol 6684. Springer, Berlin, Heidelberg.

ULTIMATE GUITAR TABS - 1,100,000 songs catalog with free Chords, Guitar Tabs, Bass Tabs, Ukulele Chords and Guitar Pro Tabs!. [online] Available at: https://ultimate-guitar.com [Accessed 12 Jun. 2019].

Cifra Club. (1996). Cifra Club - seu site de cifras e tablaturas. [online] Available at: https://www.cifraclub.com.br [Accessed 10 Jun. 2019].
Publicado
25/09/2019
MARQUES, Lucas. A chord distance metric based on the Tonal Pitch Space and a key-finding method for chord annotation sequences. In: SIMPÓSIO BRASILEIRO DE COMPUTAÇÃO MUSICAL (SBCM), 17. , 2019, São João del-Rei. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2019 . p. 136-140. DOI: https://doi.org/10.5753/sbcm.2019.10435.