|Overview MSc project W. Bas de Haas|
|Bas de Haas studied Cognitive Artificial Intelligence at Utrecht University and two years of jazz guitar at the Utrecht Conservatory. In 2006 he joined the MCG as an intern. After a number of different activities within the MCG, Bas decided to write his masters thesis on the subject of groove and swing. Parts of the research were presented on the conference of the society for music perception and cognition (SMPC) in Montreal (Haas & Honing, 2007). Currently Bas is doing a PhD on music information retrieval at the Utrecht University.|
|Haas, W. Bas de (2007, June) The Role of Tempo in Groove and Swing Timing|
In music the notion of expressive timing, i.e., deliberately playing behind or before the beat, is very common. The question how expressive timing relates to tempo has been subject to research for quite a while. However, the effect of tempo on timing in jazz and pop performance was only scarcely investigated. This thesis investigates the relation between tempo and expressive timing in groove and swing drumming by analyzing the timing of three well-known professional drummers. Furthermore a model is presented that tries to capture this effect of tempo on expressive timing.
In a controlled experiment three expert drummers were asked to perform three musical fragments in sixteen repetitions in six different tempi on a complete midi drum kit. The results show that expressive timing largely depends on the style and tempo of a drum groove. In addition, it is found that expressive timing does not scale proportionally with tempo in a Swing and Funk drumming, but does scale proportionally with tempo in Shuffle drumming.
To model the relation between timing and tempo a knowledge representation is used that separates tempo and expressive timing as different aspects of musical time. The relation between timing and tempo is represented by three swing ratio models which are optimized to fit the newly acquired data.
The participating drummers were asked to play the following excerpts.
Some audio examples of the recordings are presented underneath the
score. Additionally, quantized versions of the score with a 2:1 and a
1:1 swing ratio are presented as well to highlight the differences in
An audio example of the swing excerpt quantized with a 1:1 swing ratio (mp3), a 2:1 swing ratio (mp3) and a human performance (mp3).
An audio example of the shuffle excerpt quantized with a 1:1 swing ratio (mp3), a 2:1 swing ratio (mp3) and a human performance (mp3).
An audio example of the funk excerpt quantized with a 1:1 swing ratio (mp3) and a human performance (mp3).
The collected data is freely available* as comma separated values text
Swing ride data (csv)
Shuffle hi-hat data (csv)
Funk hi-hat data (csv)
The data of other drums and cymbals is available on request. Please send an e-mail to bas.dehaas (at) cs.uu.nl.
|The Common Lisp source code of the knowledge representation described in Bas' Thesis is also freely available.* The GTF microworld (00 GTF.lisp) is based on a version of the GTF microworld in (Honing, 1995). The generalized timing functions framework (01 TFF.lisp) is based on (Honing, 2001). N.B. this software is provided "as is" and comes without warranty.|
Haas, W.B. de, &
Honing, H. (2007). Groove, swing and the role of tempo: A model
and some preliminary empirical evidence. Proceedings of the
Society for Music Perception and Cognition (SMPC)
Honing, H. (1995). The vibrato problem, comparing two solutions. Computer Music Journal, 19(3) 32-49.
Honing, H. (2001). From time to time: The representation of timing and tempo. Computer Music Journal, 35(3), 50-61.
* Free use only for research on condition a reference to one of the publications above and a text like "...data was made available by the Music Cognition Group and can be downloaded from https://www.mcg.uva.nl/haas/" is included. Re-distribution is not allowed.