Abstract: With regard to the problem that sound images of all directions can not be realized perfectly by the fixed loudspeaker gain parameters, this paper proposes a design method of Ambisonic 3D sound images reconstruction based on multi-gain parameters. In this method, the spatially continuous directions of original sound sources are quantized by Spatial Localization Quantization Point（SLQP）. Then, sound images of quantized directions are reproduced by the corresponding gain parameters. The simulation results of reconstruction for Auro-10 configuration prove that the reproduced sound localization performance of the proposed method outperforms the conventional method with the fixed gain parameters.

Key words: ambisonics, sound localization, Spatial Localization Quantization Point（SLQP）, auro-10 configuration

摘要： 针对固定扬声器增益参数不能同时最优地实现全部方向声像重建的问题，提出一种基于多增益的Ambisonic三维（Three-dimensional，3D）声像重建方法。该方法将空间连续的声源方位进行空间格点量化（Spatial Localization Quantization Point，SLQP），根据量化方位所匹配的增益参数进行声像的重建。将所提方法应用于Auro-10扬声器配置中，客观测试结果表明，该方法重建声像定位质量优于传统固定增益参数方法。

关键词: 空间环绕声系统, 声像定位, 空间格点量化, 奥洛配置-10