To obtain the reference time and direction of each turn, two Optotrak Certus cameras (Northern Digital Inc., Waterloo, Canada) were placed in the room as shown in Figure 1 . This positioning was chosen to cover the maximum possible area around the turning zone. Six infrared markers were placed on the subject as shown in Figure 3 in order to measure head, shoulder and hip orientations at 100 Hz. While two markers per body segment are not sufficient to capture three-dimensional kinematics, they are sufficient to detect the turn onset [21 (link),28 (link)].
For each trial, two human experts were independently asked to manually determine the turn onset by examining the Optotrak-recorded head, shoulder and hip orientations simultaneously. Turn onset was defined as the moment when the turn actually begins occurring. If the difference between the two manually defined onsets was less than 200 ms, the mean of the two onsets was used as the reference turn onset for that trial. If the difference was more than 200 ms, the two experts were asked to re-examine the trial together and reach a consensus. This occurred in approximately 5% of all trials.
For each trial, two human experts were independently asked to manually determine the turn onset by examining the Optotrak-recorded head, shoulder and hip orientations simultaneously. Turn onset was defined as the moment when the turn actually begins occurring. If the difference between the two manually defined onsets was less than 200 ms, the mean of the two onsets was used as the reference turn onset for that trial. If the difference was more than 200 ms, the two experts were asked to re-examine the trial together and reach a consensus. This occurred in approximately 5% of all trials.