Traffic noise: Annoyance assessment of real and virtual sounds based on close proximity measurements

- Annoyance judgments of real and virtual traffic sounds were compared.
- A methodology to generate auralized virtual sounds from close proximity tyre/road noise measurements was developed.
- Virtual stimuli can be used to assess traffic annoyance.
- A simple transformation model can be used to predict real annoyance rates from virtual sounds.

The negative impact of noise on human health is well established and a high percentage of environmental noise is related with traffic sources. In this study, we compared annoyance judgments of real and virtual traffic sounds. Virtual sounds were generated through an auralization software with input from close proximity tyre/road noise measurements and real sounds were recorded through a Head and Torso Simulator. Both groups had sounds generated at two speeds and from three urban pavement surfaces (asphalt concrete, concrete blocks and granite cubes). Under controlled laboratory conditions, participants rated the annoyance of each real and virtual stimulus.It was found that virtual stimuli, based on close proximity tyre/road noise, can be used to assess traffic annoyance, in spite of systematic lower rates than those found for real stimuli. The effects of type of pavement and speed were the same for both conditions (real and virtualized stimulus). Opposed to granite cubes, asphalt concrete had lower annoyance rates for both test speeds and higher rate differences between real and virtual stimuli. Additionally, it was also found that annoyance is better described by Loudness than by LAmax. This evidence is stronger for the virtual stimuli condition than for the real stimuli one. Nevertheless, we should stress that it is possible to accurately predict real annoyance rates from virtual auralized sound samples through a simple transformation model.The methodology developed is clearly efficient and significantly simplifies field procedures, allowing the reduction of experimental costs, a better control of variables and an increment on the accuracy of annoyance ratings.