Improvising Duos - visual interaction collection

This dataset includes 24 items of audio-visual data in MP4 format, featuring recordings of live, improvised musical duo performances collected in Leipzig, Germany between April-July 2011 as part of the Improvising Duos research project. The primary aim of the project was to collect kinematic data via an optical motion capture system for use in a perceptual study. This perceptual study is reported in Moran, Hadley, Bader and Keller (2015), Perception of 'Back-Channeling' Nonverbal Feedback in Musical Duo Improvisation', PLoS ONE 10(6). Please refer to this publication for the data collection procedure. The musicians shown in these recordings were recruited for expertise in one of two styles of improvisation, either as jazz improvisers or as free improvisers. Both groups were instructed to play according to their respective styles, but following a formula which encouraged short, solo turn-taking episodes. In addition to the single-camera digital video data which comprises this particular collection, and the kinematic motion data, separate audio tracks were recorded for each musician using two Audio Technica AT 2035 condenser microphones; electric and digital instruments were recorded with a direct line in. This current collection, ‘Improvising Duos - visual interaction collection’, comprises a subset of the original improvised performances by the twenty-four musicians who contributed to the original project. The excerpts were selected (i) with respect to the usage permissions provided by the original participants, and (ii) for their suitability as a corpus to contribute to the work carried out as part of the AHRC-funded interdisiplinary research project, Interpersonal Entrainment in Music Performance (IEMP) (see https://www.dur.ac.uk/iemp/). Further details of this current data set and its corresponding data analysis are described in the manuscript, ‘Shared periodic performer movements coordinate interactions in duo improvisations’ (Eerola, Jakubowski, Moran, Keller and Clayton, under review) associated with this collection. The paper reports work based on 30 excerpts. Six of these 30 excerpts are not shared in this collection, respecting the specific permissions granted by the original performers. Dr Simone Tarsitani (Durham University) created the .MOV files used in the analysis and converted the selection to MP4 files with H264 MPEG-4 codec, for the purposes of sharing this dataset. These visual and audio recordings of live, improvised duo performances ar