Dyadic drumming with rhythmic distraction

This dataset contains MIDI and body motion data collected during an experiment investigating dyadic drumming with and without rhythmic distraction. Experiment design 25 pairs of adults (of which 1 pair was excluded for not following instructions) were tasked with drumming (on MIDI-enabled drums) a regular beat in synchrony with each other at a prescribed tempo (75 bpm). While drumming, they heard a metronome that they were instructed to ignore. The noise profile of the metronome was manipulated to contain long-range autocorrelations (LRC), short-range autocorrelations (SRC), or white noise (WN). The metronome's pitch was also manipulated to form a consonant (perfect 5th) or dissonant chord (diminished 5th) alongside the participants' drum tones. Each pair of participants performed 12 experimental trials with metronome manipulations and 1 baseline trial without any metronome. Data Rhythmic stimuli. Participant dyads were randomly assigned to one of three stimulus groups. For each group, four different stimuli were generated with each type of noise (LRC, SRC, WN). This meant that dyads could play with a different stimulus in every trial and never hear the same stimulus twice. Rhythmic stimuli files comprise lists of interbeat intervals (in miliseconds), which were fed into a metronome in Max/MSP. MIDI data. We collected MIDI data showing the timing of participants' drumstrikes and metronome tones in milliseconds, along with MIDI velocity values. MIDI files contain data from all trials (including the baseline, labelled as trial "0") and details about metronome distraction conditions. P1.txt contains data from participants who were randomly labelled as P1 P2.txt contains data from participants who were randomly labelled as P2 Metronome.txt contains data from the metronome Data from a given trial and dyad can be reconstructed based on the dyad and trial labels in the files. Motion data. Head and hand motion data were collected with a 12-camera infrared motion capture system (Qualisys). Motion files contain 3D position data for participants P1 and P2 along with details about metronome distraction conditions. Timestamps for MIDI and motion data. MIDI and motion data were aligned for each participant and trial by matching the sequence of MIDI interbeat intervals with the sequence of inter-strike intervals obtained from the participant's drumstick motion. Following this matching, we selected the 2-minute period with the most drumstrikes, then removed the first and last 5 seconds, resulting in 110 seconds of data. The MIDI and motion files in the uploaded dataset retain their original timestamps but have been trimmed to include only activity from this 110 second period.