Sample Encoding Coefficient Groups to Reduce Quantization Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vector quantization methods, such as SVQ and AVQ, suffer from musical noise and block noise due to low amplitude quantization precision, which affects both frequency-domain and time-domain signals.
Innovation Solution
The technology involves outputting index information indicating a group of coefficients that minimize error between sample values and their quantized counterparts, allowing for reduced quantization error by adjusting decoded values with predetermined coefficients during decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If SVQ method quantizes samples in units of sub-bands with dynamic bit assignment, then quantization precision for sparse signals is improved, but musical noise occurs when energy is distributed across many frequencies
Solution Approach 1:
The patent segments the quantization process into two independent stages: (1) vector quantization of normalized samples to generate quantization indices, and (2) separate quantization of normalization values. This segmentation allows each stage to be optimized independently, preventing the musical noise issue that arises when a single quantization stage must handle both amplitude and frequency information under bit rate constraints.
Solution Approach 2:
The patent changes the quantization parameters by introducing separate quantization schemes for different signal components. Instead of using a uniform quantization approach, it applies vector quantization with codebooks for sample values and separate quantization for normalization values, allowing flexible bit rate allocation that maintains precision while avoiding musical noise artifacts.
2Loss of information
If AVQ method restores more frequency components, then completeness of decoded signals is improved, but amplitude quantization precision deteriorates causing musical noise
Solution Approach 1:
The patent segments the quantization process into two independent stages: (1) vector quantization of normalized samples to generate quantization indices, and (2) separate quantization of normalization values. This segmentation allows each stage to be optimized independently, preventing the musical noise issue that arises when a single quantization stage must handle both amplitude and frequency information under bit rate constraints.
Solution Approach 2:
The patent applies partial quantization action by using different quantization precision levels for different signal components. The vector quantization stage handles frequency component restoration with sufficient precision, while the separate normalization value quantization stage allocates remaining bits to maintain adequate amplitude precision, accepting that neither component receives full precision but both receive adequate precision to avoid musical noise.
Data Source
AI summary
In encoding, index information indicating a group of coefficients that minimizes the sum of the error between the value of each sample and the value is obtained by multiplying the quantized value of each of a plurality of samples by a coefficient corresponding to the position of the sample. The coefficient is selected from a plurality of groups of predetermined coefficients corresponding to the positions of the samples. In decoding, a plurality of values corresponding to an input vector quantization index are obtained as decoded values corresponding to a plurality of sample positions. With the use of a group of predetermined coefficients corresponding to the plurality of sample positions and indicated by input index information, the values obtained by multiplying the decoded values and the coefficients, corresponding to the sample positions are output.