Fig. 1: Environment and trial structure.

a Third-person view of the virtual environment. Participants were positioned behind a virtual grey screen displaying the target stimulus. During the trial, the screen progressed with smooth linear locomotion at a constant velocity, in line with a small walking guide (three-dimensional animated game object). The avatar shown is for illustrative purposes only and was not present during the experiment. b The visual detection task required participants to monitor a drifting circular annulus. Small target ellipses (~1.7 d.v.a, 20 ms duration, illustration not to scale) appeared with a variable inter-trial interval (ITI), responses were provided via right trigger click. c Example data from a single walking trial. The three-dimensional head position is recorded at 90 Hz (shown in magenta). Walking produces a stereotyped sinusoidal pattern of head motion on the vertical axis (head height, 2D projection on the back wall shown in grey). Peaks and troughs in head height correspond to the swing and stance phases of each step, respectively (see Methods). d (Left) Average detrended head height for each participant over their respective stride cycle. (Right) Distribution of average stride cycle duration across participants. Our primary interest was whether the timing of target onset relative to stride cycle phase would modulate task performance.