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

2025

Read Psychoacoustic assessment of synthetic sounds for electric vehicles in a virtual reality experiment (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-25593.txt.

Introduction

Pavlo Bazilinskyy, Md Shadab Alam, and Roberto Merino-Martinez assess electric vehicle sounds. Their 2025 paper uses psychoacoustics in virtual reality. That quietness raises detectability concerns for road users. Regulations mandate exterior sound signals at low speeds.

Synthetic sounds must balance noticeability with low annoyance. Urban environments are noisy, complicating sound design. Fifteen virtual reality scenarios featured a passing car. Sound signals included pure, intermittent, and complex tones. Varied frequencies were tested across those scenarios. Two rivals included diesel engine and tyre noise. Eleven-point standardized scales captured those ratings. Findings support tuning exterior sound design standards.

Catalog record

YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE23 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-25593

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Psychoacoustic assessment of synthetic sounds for electric vehicles in a virtual reality experiment Abstract The growing adoption of electric vehicles, known for their quieter operation compared to internal combustion engine vehicles, raises concerns about their detectability, particularly for vulnerable road users. To address this, regulations mandate the inclusion of exterior sound signals for electric vehicles, specifying minimum sound pressure levels at low speeds. These synthetic exterior sounds are often used in noisy urban environments, creating the challenge of enhancing detectability without introducing excessive noise annoyance. This study investigates the design of synthetic exterior sound signals that balance high noticeability with low annoyance. An audiovisual experiment with 14 participants was conducted using 15 virtual reality scenarios featuring a passing car. The scenarios included various sound signals, such as pure, intermittent, and complex tones at different frequencies. Two baseline cases, a diesel engine and only tyre noise, were also tested. Participants rated sounds for annoyance, noticeability, and informativeness using 11-point ICBEN scales. The findings highlight how psychoacoustic sound quality metrics predict annoyance ratings better than conventional sound metrics, providing insight into optimising sound design for electric vehicles. By improving pedestrian safety while minimising noise pollution, this research supports the development of effective and user-friendly exterior sound standards for electric vehicles. Keywords: electric vehicle, psychoacoustic sound metrics, road safety, listening experiment, exterior vehicle sound 1 Introduction The World Health Organisation reports that pedestrians account for 21% of the 1.19 million annual traffic deaths (roughly 250,000) [ 1 ] . A significant proportion of these casualties occur during road crossings [ 2 ] . Factors contributing to such incidents include misjudging the time required to cross, low visibility conditions, and visual obstructions that prevent the timely detection of approaching vehicles [ 3 , 4 , 5 ] . Enhancing auditory cues emitted by vehicles has been proposed as a potential strategy to mitigate these risks, particularly in the context of typically quieter electric vehicles (EVs) [ 6 ] and external Human-Machine Interfaces (eHMIs) for automated vehicles [ 7 , 8 , 9 , 10 ] . The primary purpose of adding synthetic sounds to EVs is to enhance their noticeability, particularly for vulnerable road users (VRUs), such as pedestrians (and especially individuals with visual impairments and hearing problems). Lai et al. indicated that sound level, temporal characteristics, and modulation patterns affect the speed with which pedestrians can detect an approaching vehicle [ 11 ] . Effective sound design should maximise noticeability at low speeds without being overbearing in quieter environments. Psychoacoustic metrics such as loudness, sharpness, roughness, and fluctuation strength provide valuable information to optimise detectability while maintaining perceived annoyance within acceptable limits [ 12 ] . Annoyance is a critical factor in designing synthetic sounds for EVs, as excessive noise levels can cause discomfort and increased environmental noise pollution. Studies have shown that the spectral content, duration, and repetition of sound signals influence perceived annoyance [ 13 , 11 , 14 ] . Intermittent sounds, for example, can be perceived as more intrusive than continuous ones [ 15 ] , whereas high-frequency sounds tend to be more irritating than lower-frequency ones [ 16 ] . Therefore, optimising EV sounds requires a balance between noticeability and minimal disturbance to urban residents [ 17 , 18 ] . An ideal EV sound should convey useful information to VRUs on vehicle behaviour, such as acceleration or deceleration. Informative sounds help users make more accurate judgments about vehicle proximity and direction of movement [ 12 ] . Complex sound structures, which integrate variations in pitch and modulation, can enhance informativeness while maintaining societal acceptance. Designing EV sounds that intuitively communicate vehicle dynamics can improve pedestrian interaction with EVs in urban settings [ 11 ] . This preliminary study aims to examine the psychoacoustic noise annoyance of various sounds of exterior vehicles to inform pedestrians of an approaching vehicle. In a virtual reality (VR) experiment, participants were asked to rate this parameter using the 11-point ICBEN scale. The open-source MATLAB toolbox SQAT (Sound Quality Analysis Toolbox) v1.2 [ 19 ] toolbox was then employed to calculate different psychoacoustic metrics to predict annoyance ratings. This analysis provides useful insights into optimising the design of external auditory communication for EVs. Section 2 discusses the methodology employed, including the synthetic sounds considered, the VR experiment, and the sound metrics used. The main results are presented and discussed in section 3 , whereas the conclusions and recommendations for future work are gathered in Section 4 . Finally, supplementary material is provided in Section 5 . 2 Methodology 2.1 Synthetic sounds Table 1 lists the 15 synthetic sounds that were used in the study, which can be categorised into four groups: (1) continuous pure tones at a single frequency, (2) intermittent pure tones with alternating 500-ms on/off intervals, (3) combined tones at a principal frequency plus secondary tones at frequencies \pm 90 Hz from the principal tone, and (4) double beeps at 1800–1900 Hz. The double beeps consisted of a 240 ms beep, a 10 ms pause, a 240 ms beep, and a 1000 ms pause. Furthermore, a diesel engine sound 1 1 1 https://youtu.be/watch?v=2Y33bTlAA-E was included as a reference case for internal combustion engine vehicles due to its distinctive and more familiar noise. To evaluate the sound profile of a quiet electric or automated vehicle, a stimulus containing only tyre noise 2 2 2 https://youtu.be/watch?v=X0wpizkkH_Q (i.e., without any synthetic sound added) was also incorporated. All evaluated sound stimuli were combined with a background noise recording of a quiet street 3 3 3 https://youtu.be/watch?v=6C-W_7BZBxQ . This list of sounds was previously used in an online crowdsourcing listening experiment (i.e. without VR) reported in [ 6 ] . This study considers a relatively simple sound source and observer geometry in a two-dimensional arrangement, i.e., the observer and the sound source are on the same plane; see Fig. 1 . For simplicity, the sound source is considered a point source with a constant velocity of 30 km/h (i.e., 8.33 m/s) in the positive x direction \mbox{\boldmath{$V$}}=(V_{x},\,0) , i.e., the source has a linear trajectory along a straight line at a distance y_{s} of 3 m from the observer, see Fig. 1 . The initial position of the sound source at is defined at \mbox{\boldmath{$r$}}(0)=(x_{s,0},\,y_{s}) and at a general instant \tau_{e} as \mbox{\boldmath{$r$}}(\tau_{e})=(x_{s,0}+V_{x}\tau_{e},\,y_{s}) . For the purpose of sound signal generation, the source moved from x_{s,0}=-60 m to x_{s}=60 m. Therefore, a signal length of 14.4 s was generated and used in each case. All sound stimuli (except for the case with only tyre noise) had an equivalent A-weighted sound pressure level ( \text{L}_{\text{p,A,eq}} ) of 65 dBA in the observer position. The VR environment employed enables binaural rendering of the sound accounting for the relative position of the sound source, the orientation of the participant’s head, and the Doppler-shift due to relative motion. The reflection of sound on the ground or any reflecting surfaces (such as buildings) was neglected for simplicity. Figure 1 : Scheme of the considered sound source and observer geometry [ 6 ] . Table 1 : Synthetic sounds used in the experiment. No. Characteristics 1 Pure tone, continuous, 350 Hz 2 Pure tone, continuous, 500 Hz 3 Pure tone, continuous, 1000

Read the complete text