eVADER: A Perceptual Approach to Finding Minimum Warning Sound Requirements for Quiet Cars.
Résumé
The eVADER project endeavors to enhance the technology of lectric-vehicles for the safety of pedestrians, especially the visually-impaired. The research presented here focused on eVADER's goal to isolate features of replacement sounds for electric-vehicles, maximizing pedestrian safety while minimizing noise pollution. It was predicted that 3 sound features (complexity, amplitude-modulation, frequency-modulation) could be the most informative to listeners, allowing for quieter emission. Sounds were synthesized using different combinations of the 3 (features) X 3 (levels). The 9 synthesized sounds were normalized (dBA), Doppler-shifted, and superimposed onto a binaural-recording of a moving electric-vehicle (prototype-sounds). Stimuli were embedded in a noisy background (traffic or traffic + rain), emulating a suburban soundscape. In this simulation, participants were virtual pedestrians, waiting to cross a street, using only sound. In a fractional, repeated measures design, 111 participants (29 visually-impaired) heard 11 stimuli (9-prototype-sounds, 1-electric-vehicle, 1-diesel) over 88 trials. Participants localized approaching cars (perpendicularly). Results indicated that detection of some prototype-sounds was equivalent to that of the diesel car. This was achieved without significantly increasing SPL. Despite normalization, detection varied among prototype-sounds via the 3 (feature) X 3 (level) combinations. This effect was magnified in noisier backgrounds. The practical and theoretical implications will be discussed.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...