Novel low complexity coherence estimation and synthesis algorithms for parametric stereo coding

In this paper, we present novel low complexity coherence estimation and synthesis algorithms and their application to parametric stereo coding. Inter-channel correlation /coherence (IC) is an important parameter for parametric stereo coding as it represents the degree of similarity of the channels and is strongly related to the perception of width and diffuseness of the stereo image. It is relevant for most audio and music contents to re-generate ambience, stereo reverberation, source width, and other perceptions related to spatial impression. In the state of the art parametric stereo codec, the IC estimation and corresponding synthesis algorithms are very complex which prevents their use for complexity-constrained applications. Hence, we introduce novel low complexity coherence estimation and synthesis algorithms for stereo coding. A subjective listening test shows that with the proposed algorithms, the perceived quality for very low bit rate parametric stereo is improved with a limited computational complexity cost.