Sound Quality Equivalent Modeling for Virtual Car Sound Synthesis

The pressure on development cycles in the automotive industry forces the acoustical engineers to create awareness of sound quality in the early stages of development, perhaps even before a physical prototype is available. Currently, designers have few tools to help them listen to their “virtual” models.

For the design of a synthesis platform of in-vehicle binaural sound, the sound should be modeled with almost identical sound quality perception. A concept is presented where the total sound of a vehicle is split in a number of components, each with its own sound characteristics. These characteristics are described in a signal model that allows the analysis of an existing sound into a limited number of signal components: orders-frequency spectra, time envelopes and time recordings. The introduction of this sound synthesis concept in a hybrid model consisting of networks of transfer paths, force inputs and noise sources will allow the evaluation of the impact of changes to structural components on the sound quality of in-vehicle sound.

This publication will report on the experience gained and the progress made in these fields.