Perceptual Evaluation of Materials for Virtual Sounds Reproduced by Linear Modal Synthesis Method

Abstract

Virtual reality is the most popular field in many countries, and there are so many technologies on visual expression, but to increase immersion, nowadays virtual sound is being taken center stage in virtual reality. Linear modal synthesis (LMS) has been investigated to synthesize virtual sound of rigid bodies, however, existing methods to synthesize virtual object sound have lacks of the real-time computations for perceptually satisfactory material parameters. In this study, We propose a technique based on feature extraction from one recorded audio clip to estimate material parameters of virtual objects for real-time sound rendering. A parameter estimation technique based on psychoacoustic principle is developed that incorporates material features to search the best material parameters for LMS. The perceptual test is implemented to compare similarity between the virtual sound with the real sound. The results show that our process synthesize perceptually realistic virtual sound.