Psychoacoustic assessment of synthetic sounds for electric vehicles in a virtual reality experiment

Bazilinskyy, Pavlo; Alam, Md Shadab; Merino-Martinez, Roberto · 2025 · Crossref

DOI: 10.61782/fa.2025.0710

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of designing synthetic exterior sounds for electric vehicles (EVs) that balance pedestrian safety with acoustic comfort. As EVs operate more quietly than internal combustion engine vehicles, regulations mandate audible signals to enhance detectability for vulnerable road users. However, these signals risk causing excessive noise annoyance in urban environments. The research investigates how psychoacoustic metrics predict perceived annoyance compared to conventional sound metrics, aiming to optimize sound design for high noticeability and low disturbance. The researchers conducted a virtual reality (VR) experiment with 14 participants, who evaluated 15 audiovisual scenarios featuring a passing vehicle. The stimuli included various synthetic sounds: continuous pure tones, intermittent tones, combined tones with secondary frequencies, and double beeps, across frequencies of 350 Hz to 2000 Hz. Two baseline cases—a diesel engine and tyre noise only—were also tested. All synthetic sounds were normalized to an equivalent A-weighted sound pressure level of 65 dBA. Participants used 11-point ICBEN scales to rate annoyance, noticeability, and informativeness. The study utilized the SQAT toolbox to calculate conventional metrics (e.g., maximum sound pressure level) and psychoacoustic metrics (loudness, sharpness, roughness, fluctuation strength, and tonality), combining the latter into a global psychoacoustic annoyance (PA) metric. Results indicated that perceived annoyance increased with tonal frequency, with pure tones at 2000 Hz rated as most annoying (mean ~7/10) and tyre noise as least annoying (mean ~1/10). Pure tones were generally more annoying than intermittent or combined tones. Correlation analysis revealed that psychoacoustic metrics significantly outperformed conventional metrics in predicting annoyance. The global PA metric showed the strongest correlation ($\rho \approx 0.89$), followed by loudness ($N_5$, $\rho \approx 0.88$) and effective perceived noise level ($EPNL$, $\rho \approx 0.86$). Conventional metrics like maximum sound pressure level ($L_{p,max}$) showed no significant correlation. Sharpness ($S_5$) also demonstrated a strong correlation ($\rho \approx 0.77$), confirming that high-frequency content drives annoyance. The findings suggest that psychoacoustic parameters provide a more accurate basis for evaluating EV exterior sounds than traditional noise assessment metrics. This implies that sound design should prioritize optimizing psychoacoustic qualities, such as minimizing sharpness and managing loudness, rather than merely controlling sound pressure levels. The study concludes that balancing acoustic comfort with detectability is critical for the social acceptance of EVs. Future work will analyze noticeability and informativeness data and expand the study to include participants with visual or hearing impairments and higher-fidelity experimental setups.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• auditory warnings