Comprehensive design cycle for car engine sound: from signal processing to an integrated audio system inside the vehicle

摘要

Sound Design of a car engine can be achieved using two methods: either a classical approach based on mechanical design and prototyping; or using a pure acoustic method based on signal processing, psychoacoustics, sound design, digital audio algorithm, audio and loudspeakers devices. We will be focusing on the second method, which is a multi-step process. First, signal processing requires a highly accurate order analysis. From this analysis, real-time audio synthesis is performed and is directly compared to the original recording, in order to check the precision of the order analysis. The synthesis algorithms will be described. This sound comparison is also used to validate the audio synthesis algorithm's realism and precision. Sound synthesis is controlled by various parameters from the car: vehicle speed, accelerator pedal position, RPM and Torque applied on the wheels. Several sound design samples will be showed and played-back. Finally, an embedded audio device allowed us to integrate and test the Sound Design while driving real cars - both those with internal combustion engines and those with electric motors. The audio device acquires the RPM and other information through the CAN bus of the vehicle. The output is then played through the native car's audio system and s loudspeakers with no detectable hearing artifacts. Additionally, this audio device can also be used to generate the external warning sound to alert pedestrians of the presence of electric or hybrid vehicles.