Presence vs. Precision: Spatial Audio Latency in Virtual Opera

Sammanfattning

Audiovisual synchronization significantly impacts the viewing experience in streaming environments, yet research on latency sensitivity toward spatial audio in Virtual Reality (VR) remains limited. This study conducted an audiovisual synchronization experiment in a virtual opera house, where 18 participants performed multisensory integration tasks using the Meta Quest Pro. The results indicated a substantial increase in the Just Noticeable Difference (JND) for spatial audio, reaching 461.23 ms, compared to 224.14 ms under the stereo audio condition. Similarly, the Point of Subjective Simultaneity (PSS) shifted from 22.47 ms in the stereo condition to 71.17 ms in the spatial audio condition. Subjective questionnaire results revealed that participants perceived spatial audio as more natural and requiring less cognitive effort, despite the observed decrease in temporal precision. Eye-tracking data further showed that gaze anchors remained concentrated on the conductor’s face, with no apparent differences in gaze distribution between audio conditions, indicating that auditory presentation did not fundamentally alter visual attention strategies. Overall, while the data showed a numerical trend toward decreased user temporal sensitivity in the spatial audio condition, this effect did not reach statistical significance. This study provides practical implications for the design of future immersive streaming systems and contributes to the development of latency thresholds that better balance user experience and system performance.