Research project: Shaping the future of systems for personalized audio
We are developing Signal Processing strategies for the enhancement of audio quality in systems for private sound focusing.
Project Overview
Imagine yourself in a crowded place. You may want to listen to your favourite music with your portable device (like a smart phone) without bothering people nearby. You can certainly do it by using headphones. However, headphones might be uncomfortable after prolonged use. Wouldn't it be great if your portable device delivered sound to you and not to other people nearby without the need for headphones? Think of it like being in a private sound zone.
The good news is that this technology is available and it is called personalized audio [1]. This is made possible by using a number of loudspeakers arranged in a given geometry (the so-called loudspeaker arrays) driven by purposefully designed input signals. These input signals are designed so that the sound reproduced by the array driven with these signals undergoes constructive acoustic interference where the listener is located (thus producing sound) whilst minimizing sound radiation elsewhere.
The availability of cost-effective hardware for multi-channel audio and the high computational power provided by portable device is making systems for personalized audio very attractive for a number of Academic institutions as well as big companies. Over the last two decades, researchers have focused on ways to improve the directivity performance of arrays, so that the system can deliver highly directive sound radiation in space. However, the audio quality at the listener's position and the power consumption needed to perform highly directive sound radiation are issues that need to be address to meet market needs.
Ideally, we would like high directivity performance, high audio quality at the listener's positions, and low power consumption. The bad news is that these performance requirements are conflicting with each other [2]. This research project sponsored by HUAWEI aims at the development of signal processing strategies for systems for personalized audio for the control of the trade-off between directivity, audio quality, and power consumption. Controlling this trade-off would then allow to optimize system performance based on, for example, the characteristic of the loudspeaker array used (e.g., the form factor, the transducer mounted, etc) and the targeted applications (e.g., music reproduction, private communications, multi-zone sound field reproduction).
Ferdinando Olivieri
[1] W.F. Druyvesteyn and J. Garas, Personal sound, Journal of the Audio Engineering Society, Vol. 45(1997), No. 9, p. 685-701
[2] Elliott, S. J., Cheer, J., Choi, J. W., & Kim, Y. (2012). Robustness and Regularization of Personal Audio Systems. IEEE Transactions on Audio, Speech, and Language Processing, 20(7), 2123–2133.