Spatial Multizone Soundfield Reproduction Design
It is desirable for people sharing a physical space to access different multimedia information streams simultaneously. For a good user experience, the interference of the different streams should be held to a minimum. This is straightforward for the video component but currently difficult for the audio sound component. Spatial multizone soundfield reproduction, which aims to provide an individual sound environment to each of a set of listeners without the use of physical isolation or headphones, has drawn significant attention of researchers in recent years. The realization of multizone soundfield reproduction is a conceptually challenging problem as currently most of the soundfield reproduction techniques concentrate on a single zone. This thesis considers the theory and design of a multizone soundfield reproduction system using arrays of loudspeakers in given complex environments. We first introduce a novel method for spatial multizone soundfield reproduction based on describing the desired multizone soundfield as an orthogonal expansion of formulated basis functions over the desired reproduction region. This provides the theoretical basis of both 2-D (height invariant) and 3-D soundfield reproduction for this work. We then extend the reproduction of the multizone soundfield over the desired region to reverberant environments, which is based on the identification of the acoustic transfer function (ATF) from the loudspeaker over the desired reproduction region using sparse methods. The simulation results confirm that the method leads to a significantly reduced number of required microphones for an accurate multizone sound reproduction compared with the state of the art, while it also facilitates the reproduction over a wide frequency range. In addition, we focus on the improvements of the proposed multizone reproduction system with regard to practical implementation. The so-called 2.5D multizone oundfield reproduction is considered to accurately reproduce the desired multizone soundfield over a selected 2-D plane at the height approximately level with the listener’s ears using a single array of loudspeakers with 3-D reverberant settings. Then, we propose an adaptive reverberation cancelation method for the multizone soundfield reproduction within the desired region and simplify the prior soundfield measurement process. Simulation results suggest that the proposed method provides a faster convergence rate than the comparative approaches under the same hardware provision. Finally, we conduct the real-world implementation based on the proposed theoretical work. The experimental results show that we can achieve a very noticeable acoustic energy contrast between the signals recorded in the bright zone and the quiet zone, especially for the system implementation with reverberation equalization.