Abstract
Vehicle driving simulators enable the recreation of real driving experiences by simultaneously reproducing multimodal information. Here, we performed the perceptual evaluation of a virtual reality (VR)-based driving simulator in three experiments with experienced drivers with a special focus on comparing headphone-based and loudspeaker-based sound reproduction. First, we assessed how sound source localizations vary with three sound reproduction methods: Higher order ambisonics (HOA), the matrix inversion method, and binaural recordings played back over headphones without incorporating head tracking. The reproduction methods employing loudspeakers showed better performance in terms of perceived horizontal angle and distance than the reproduction over headphones without head tracking. Second, we assessed how the multimodality affected the perceived spatial immersion and powerfulness of the driving simulator. Visual stimuli were presented together with driving noise through either loudspeakers or headphones, and we controlled vibration and motion stimuli: vibration only, motion only, vibration and motion. While the inclusion of motion and vibration led to significantly higher ratings for spatial immersion and powerfulness, the choice of sound reproduction method did not have a significant impact on the ratings. Third, adding motion to the reproduced scenarios had a greater influence on immersion and powerfulness ratings than adding vibration alone. The results highlight the importance of carefully considering the multimodal information to optimize the driving simulator.
Original language | English |
---|---|
Article number | 110137 |
Journal | Applied Acoustics |
Volume | 224 |
Number of pages | 14 |
ISSN | 0003-682X |
DOIs | |
Publication status | Published - 5 Sept 2024 |
Keywords
- Binaural reproduction
- Higher order ambisonics
- Matrix inversion
- Multi-modal perception
- VR-based automotive simulator