EFFECT-TIME CUE SCHEDULING FOR DISTRIBUTED GAZE-BASED MUSICAL PERFORMANCE

We present a distributed gaze-based musical performance system designed as a prototype for remote gaze-driven co-performance. The system coordinates visual targets with pre-cues and scheduled effect times, while audible MIDI events remain controlled locally through gaze-based grid selection. Three real-world conductor–performer sessions were analyzed through the same cloud-routed architecture: a co-located Korea–Korea trained-use reference session and two first-time remote sessions in India and South Africa. Cue-error p95 (95th percentile), defined here as participant-side cue-lifecycle timing error rather than audio-onset error, stayed near 10 ms while round-trip time (RTT) ranged from 164 ms to 357 ms. The main failures appeared in other layers. Gaze acquisition was lower in the first-time sessions and appeared sensitive to calibration familiarity, while reconnect conditions produced missing and duplicate cue-fire reports. The sessions suggest that distributed gaze performance should be evaluated by separating scheduled cue activation, gaze-driven sound generation, onboarding, and session-state management.