Parametric Audio Coding Spatial Parameter Block Distribution

The patent divides the spatial information parameters into multiple blocks and distributes them across multiple frames. Instead of coding all parameters at once for each short frame (which would require high bit rate), the parameter set is segmented and transmitted over several frames, reducing the bit rate requirement per frame while maintaining reconstruction quality through temporal accumulation of parameter information.

The patent applies non-uniform scalar quantization to the spatial information parameters, where different quantization step sizes are used depending on the parameter values. This allows finer quantization for parameters with smaller ranges or lower sensitivity requirements, and coarser quantization for parameters with larger ranges, thereby reducing the overall bit rate while maintaining adequate reconstruction quality. The quantization characteristics are adapted locally to the specific parameter properties rather than applying a uniform approach.

The patent extracts and codes only the necessary spatial information parameters required for acceptable stereo reconstruction, rather than extracting all possible parameters. By carefully selecting and coding only the essential parameters (such as inter-channel level differences and phase differences for specific frequency bands), the system achieves adequate stereo quality at reduced bit rate. The parameter extraction and coding is optimized to include only what is perceptually necessary.