Community Noise Quality Assessment (Annoyance) by means of a Virtual Audio Environment

Résumé En

This study deals with community noise quality assessment (annoyance) of urban environments. It is based on a previous work that resulted in: a) a road-traffic vehicle audio signals extraction tool, allowing to estimate different sonic properties; b) a modeling tool that estimates the perceived annoyance level by adapting a current multiclass psychoacoustic model, given in the literature. The present work follows these outputs and aims at validating experimentally the annoyance estimations. A listening test is designed in order to be able to collect perceived annoyance in a virtual environment. Sound scenes made of urban soundscapes (background) and vehicles pass-by (foreground) are built and encoded in a 2D to multichannel algorithm, in order to create a listening experience as ecological as possible. Then, a consistent experimental protocol is designed in order to measure the perceived annoyance caused by each synthesized sound scene. The focus is put on the spatial audio environment that is used for the perceptual experiment. Especially, urban sound scene synthesis methods and tools are presented. Results from the perceived naturalness and immersion collected during the test are presented and analyzed. The global issue addressed by the use of virtual reality in laboratory experiments is finally discussed.