Multi-channel Object-Based Spatial Parameter Compression Approach for 3D Audio

摘要

To improve the spatial precision of three-dimensional (3D) audio, the bit rates of spatial parameters are increased sharply. This paper presents a spatial parameters compression approach to decrease the bit rates of spatial parameters for 3D audio. Based on spatial direction filtering and spatial side information clustering, new multi-channel object-based spatial parameters compression approach (MOSPCA) is presented, through which the spatial parameters of intra-frame different frequency bands belonging to the same sound source can be compressed to one spatial parameter. In an experiment it is shown that the compression ratio of spatial parameter can reach 7:1 compared with the 1.4:1 of MPEG Surround and S~3AC (spatial squeeze surround audio coding), while transparent spatial perception is maintained.