Audio Decoding with Non-Uniform Dequantization for Lower Bit Use
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio coding systems face challenges in minimizing bit stream information while maintaining audio quality, as high compression leads to poor sound quality due to complex calculations or information loss, and low compression results in capacity issues.
Innovation Solution
The implementation of a non-uniform scalar quantization scheme for audio parameters in parametric spatial coding, where smaller step-sizes are used for ranges of high human sound perception sensitivity and larger step-sizes for less sensitive ranges, along with a scaling function to adjust quantization based on perceived sound characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If high compression is applied to minimize bit stream information, then bit stream capacity is improved, but audio quality deteriorates due to information loss and complex calculations
Solution Approach 1:
The patent applies non-uniform quantization where different step-sizes are used for different ranges of quantized parameters based on human sound perception sensitivity. Smaller step-sizes are used for ranges where human perception is most sensitive, and larger step-sizes are used for less sensitive ranges. This local differentiation allows efficient bit allocation that maintains audio quality in critical regions while reducing bit consumption in less critical regions.
2Device complexity
If uniform quantization is used for simplicity, then device complexity is reduced, but audio quality deteriorates due to insufficient perceptual optimization
Solution Approach 1:
The patent changes the quantization parameter (step-size) based on the value of the quantized parameter and human sound perception sensitivity. Instead of using a fixed uniform step-size, the system dynamically adjusts the step-size according to the specific range and perceptual importance, achieving perceptually optimized quantization without requiring complex vector quantization or model-based approaches.
Data Source
AI summary
The present disclosure provides methods, devices and computer program products for non-uniform quantization of parameters. The disclosure further relates to a method and apparatus for reconstructing an audio object in an audio decoding system taking the non-uniformly quantized parameters into account. According to the disclosure, such an approach renders it possible to reduce bit consumption without substantially reducing the quality of the reconstructed audio object.