Development of novel signal processing techniques for audio reproduction using loudspeaker arrays

阵列技术是音频采集与再现领域的一个“热点”研究课题。音频和电信行业的主要国家和国际参与者（BBC， NHK，法国电信，德国电信等）开展的最新研究项目表明，这一领域的商业兴趣日益浓厚。使用紧凑型换能器阵列来产生私人聆听区域，与便携式设备一起使用是该技术的一个有前途的应用。ISVR在开发用于扬声器阵列的3D音频再现的DSP技术方面有着良好的记录。这有时被称为“跨耳音频”。虽然最近在这一领域取得了重大进展，第一批跨耳音频系统已经开始商业化，但仍有许多问题尚未解决，需要在学术水平上进行更多的基础研究。其中一些问题是：-扬声器最佳数量的定义-高架虚拟源的再现-漫射场的再现-个性化头部相关传递函数的重要性-低频系统不稳定性和阵列的最佳尺寸-高频的空间混叠和最佳扬声器间距的确定-多个听众（2个或更多）的再现-系统对环境条件（混响和背景噪声）的适应这个博士项目是分析这些不同的问题，并更好地了解它们的科学和工业相关性。预计该项目将重点关注其中的一个或两个方面，并在阵列设计和信号处理策略方面开发新的解决方案。预计这些新颖的解决方案将利用ISVR在阵列技术方面的内部专业知识，也可能基于新的DSP方法（例如压缩感知）和心理声学原理（ISVR在这方面也拥有强大的内部专业知识）。研究工作将包括理论调查和实验活动，并利用在S3A项目中开发的实验设备和方法。改进扬声器阵列的3D音频再现是epsrc支持的S3A项目资助的中心主题，该博士项目与此相关。该项目的成果也有可能通过涉及知识和技术的学术和工业传播的S3A活动产生重大影响。

Array technologies are a "hot" research topic in the field of audio capturing and reproduction. The latest research projects undertaken by the major national and international players in the audio and telecommunication industry (BBC, NHK, France Telecom, Deutsche Telekom, etc.) demonstrate the growing commercial interest in this field. The use of compact transducer arrays for the generation of private listening areas, to be used in conjunction with portable devices is a promising application of this technology. The ISVR has a strong track-record in the development of DSP technologies for 3D audio reproduction with loudspeaker arrays. This is sometimes referred to as "Transaural audio". Whilst significant advances have been achieved recently in this field and the first Transaural audio systems have started to be commercialised, a number of issues are still unresolved and require more fundamental research to be carried out at an academic level. Some of these issues are:- Definition of optimal number of loudspeaker- Reproduction of elevated virtual sources- Reproduction of diffuse fields - Importance of individualised Head Related Transfer Function- Low frequency system instability and optimal size of the array- Spatial aliasing at high frequency and determination of optimal loudspeaker spacing- Reproduction for multiple listeners (2 or more)- System adaptation to environmental conditions (reverberation and background noise)The aim of this PhD project is to analyse these different problems and gain a better understanding of their scientific and industrial relevance. It is then expected that the project will focus on one or two of these aspects and develop novel solutions in terms of array design and signal processing strategies. It is expected that these novel solution will leverage ISVR's in-house expertise in array technology and may also be based on novel DSP approaches (e.g. compressive sensing) and on psychoacoustical principles (in which ISVR has also strong in-house expertise). The research work will involve both theoretical investigation and experimental activity, leveraginbg also the experimental devices and methods developed in course of the S3A project. The improvement of 3D audio reproduction with loudspeaker arrays is a central topic of the S3A EPSRC-supported programme grant, to which this PhD project is associated. The outcomes of this project have significant potential for impact also through the S3A activity, which involves both academic and industrial dissemination of knowledge and technology.