Ensemble timing: a theoretical model and practical demo of a virtual ensemble training tool - Alan Wing and Min Li (University of Birmingham)

Wing et al (2014 http://dx.doi.org/10.1098/rsif.2013.1125) proposed that phase correction underlies ensemble timing in classic quartet playing. Each player corrects the timing of their next note in proportion to the summed weighted asynchronies of their previous note onset with the note onsets of the other three players. They described two professional chamber music quartets whose correction gain values were close to .25, a value which is optimal in the sense of minimising the between player asynchrony variance. In our presentation we will describe use  of linear phase correction in simulation of quartet performance.

We will present results of a new study of virtual quartet performance to determine the sensitivity of one performer to changes in phase correction gain of a second performer with respect to the first performer. We will also describe the effects of musical performance experience and role of the assigned part (first vs second violin).

In the study we simulated first order linear phase correction of note onset times of three virtual members of a quartet “performing” a 44-note Haydn excerpt, with note onset times of the fourth performer (the participant) determined by tapping the index finger on a midi drum board. We ran two separate experiments; the first involved musicians (more than 10 years playing experience) and non-musicians, the second only non-musicians. In the first experiment (musician and non-musician) participants performed the first violin part, in the second (non-musician) participants performed the second violin part. The gain (relative to the participant) and timing SD of the other (virtual) violin were systematically varied. The gain was set either lower (0.1) or higher (0.4) than fixed gains (0.25) used by virtual viola and virtual cello with respect to the virtual violin. The timing SD was set either lower (10 ms) or higher (30 ms) than the 20 ms SD of the viola and cello.

Across the two experiments the results showed participants (i) were able to “play” with the simulated virtual players (ii) used phase correction gains that (a) complemented the gain of the other (virtual) violin (b) were insensitive to the timing variability of the other violin. In the first experiment musicians used lower phase correction gains than non-musicians and their asynchrony variability was lower. Comparison of the results of the non-musicians in the first and second experiment did not showed any consistent effect of playing the first vs second violin part.

Following our research talk we will present a 30-minute interactive workshop on the linear phase correction simulation of ensemble (quartet) timing. Workshop participants will be able to experience the effects on their timing of setting different phase correction gains and timing variability in the virtual quartet.

The simulator we describe was produced as part of an EPSRC funded project on Augmented Reality Music Ansemble (https://arme-project.co.uk/ PI Max Di Luca). Following the simulator demonstration we will conclude with a brief review of the aims and achievements of the ARME project and possible future directions for research.

Biographies

Alan Wing. Alan discovered Psychology as an option course when doing physics as an undergraduate at Edinburgh University. He followed it up with a PhD at McMaster University in Canada and a postdoc at Bell Labs in New Jersey. From there he went to Cambridge APU and then to Birmingham as Professor of Human Movement in Psychology at the University of Birmingham where he leads the Sensory Motor Neuroscience group. Alan is an amateur cellist.

Min Li.  Susan studied for a BSc Psychology at the University of Kent and later a Research Masters degree at the University of Birmingham, where she also completed her PhD on multisensory perception in 2018. After working with high-field fMRI in Glasgow on predictive information and illusory perception in the visual system, she returned to Birmingham to work, first on a BBSRC funded project on multisensory touch and then on the ARME project. Susan recently took up a position as Research Fellow in cross-sensory cognition in Computer Science at the University of Bristol. Susan is an amateur violinist and pianist.