engleski

Simple dynamic measurement system for testing IMU sensor precision in spatial audio

Spatial sound is used in a variety of systems these days. The application of high-quality audio systems with real-time processing ranges from hi- fi systems to systems for live sound and virtual reality. The data needed to (re)create realistic surround sound audio is often collected using Inertial Measurement Unit (IMU) sensors. With technological progress, the number of such sensors in many devices is constantly increasing. By combining the data from all sensor devices, a dynamic virtual audio field can be created that provides an immersive experience for the user. When processing the collected raw data, attention must be paid to the precision and tolerance of the sensors. The IMU sensors used in this type of measurement usually consist of an accelerometer, a gyroscope and sometimes a magnetometer, all of which are triaxial. When processing and editing the measured data, the effects of low precision and accuracy, parameter drift and latency can greatly affect the measurement uncertainty and the overall user experience. Sensor devices used for spatial audio fall into three groups: VR headsets, smartphones and DIY sensor systems. However, the cost of a device does not necessarily guarantee the accuracy of the measured data. This paper presents the results of dynamic measurements on sensors used in embedded systems. Several different types of sensors with different shapes and sizes are used. Some are built on PCB boards with processing units, while others are implemented as modules on dedicated PCBs. A microcontroller is used for data processing. The most widely used open-source platform Arduino is used for the measurements. To avoid the magnetic influence of the moving parts in the setup, a simple aluminium/plastic pendulum is used. To minimise the measurement error, all measurements are taken simultaneously. The measurement results are compared, and the quality of the measured sensors is evaluated.

IMU sensor accuracy, binaural synthesis, spatial audio

nije evidentirano