Waveform Compression via Dynamic Scalar Vector Quantization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing waveform compression technologies for electronic musical instruments are inefficient due to the use of fixed quantizing systems, which do not adapt to the varying characteristics of musical instrument sounds, leading to suboptimal compression rates and data reduction.
Innovation Solution
A waveform compressing apparatus that dynamically selects the best quantizing system and codebook for each frame of waveform data based on the characteristics of the sound, using a combination of scalar and vector quantization methods to achieve the highest compression rate while maintaining an acceptable quantization error.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a fixed quantizing system is used for waveform compression, then the compression process is simple, but the compression rate is suboptimal because it cannot adapt to varying sound characteristics
Solution Approach 1:
The patent implements dynamic mode selection where the quantizing system automatically switches between scalar and vector quantization modes based on the characteristics of each frame of waveform data. The system evaluates correlation metrics and selects the appropriate quantizing mode in real-time, transforming a static compression process into a dynamic adaptive one that optimizes compression rate according to actual sound characteristics.
Solution Approach 2:
The patent changes the parameter of quantizing system type (scalar vs. vector) based on the correlation characteristics of the waveform data. By monitoring parameters such as inter-sample correlation and switching the quantizing approach accordingly, the system adapts to different sound conditions, achieving higher compression rates for correlated data while maintaining simplicity for uncorrelated portions.
2Productivity
If the vector quantizing system is adopted, then the compression rate can be improved for data with correlation, but it is difficult to achieve advantage because waveform data changes over time making it hard to find common characteristics
Solution Approach 1:
The patent divides the waveform data into multiple frames and applies different quantizing modes to each frame based on its local characteristics. Instead of applying a single vector quantizing system to the entire waveform, the system segments the data and selectively applies vector quantization only to frames where correlation analysis indicates it would be beneficial, thus adapting to time-varying sound characteristics.
Solution Approach 2:
The patent dynamically changes the quantizing parameter (mode selection) based on the correlation characteristics of each frame. By evaluating parameters such as inter-sample correlation and switching between scalar and vector quantization modes frame-by-frame, the system adapts to varying sound characteristics and achieves improved compression rates where applicable.
3Quantity of substance
If scalar quantization is used for all frames, then the processing is simple, but the data amount cannot be reduced significantly even though some frames have high correlation
Solution Approach 1:
The patent changes the quantizing parameter (mode selection) based on the correlation characteristics of each frame. By evaluating parameters such as inter-sample correlation and switching between scalar and vector quantization modes frame-by-frame, the system adapts to varying sound characteristics and achieves improved compression rates where applicable.
Solution Approach 2:
The patent implements a feedback mechanism where the system evaluates the correlation characteristics of each frame and uses this information to select the appropriate quantizing mode. The evaluation results feed into the mode selection process, creating a closed-loop system that continuously optimizes the compression approach based on actual data characteristics, thereby significantly reducing data amount where correlation exists.
4Productivity
If different quantizing modes are selected for each frame based on characteristics, then the compression rate is optimized, but the compression process requires significant time to evaluate and select modes
Solution Approach 1:
The patent applies a simplified correlation evaluation that checks only essential parameters to determine whether vector or scalar quantization is appropriate. Rather than performing exhaustive analysis, the system uses sufficient but not excessive evaluation criteria to make quick mode decisions, balancing compression optimization with processing time constraints.
Data Source
AI summary
In a waveform compressing apparatus, a trial mode selecting portion selects a trial mode having the highest compression rate from a plurality of candidate modes which have not been selected before as a trial mode for generating a residue code, the selected trial mode comprising a scalar quantization mode or a vector quantization mode. A waveform data compressing portion compresses a given data amount of original waveform data according to the selected trial mode so as to generate the residue code, the data amount being determined in correspondence with the selected trial mode. A waveform data restoring portion generates a restored waveform data from the compressed data using the generated residue code. A determining portion measures an evaluation value of a quantization error contained in the restored waveform data relative to the original waveform data, and determines whether the evaluation value is equal to or smaller than a predetermined allowable value. A mode change instructing portion outputs a mode change instruction for instructing the trial mode selecting portion to select another trial mode when the evaluation value is not smaller than the predetermined allowable value.