Audio Parameter Coding with Modulo Differentials and Shared Tables
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio coding systems, particularly object-based audio systems, face inefficiencies in encoding and decoding audio signals while maintaining quality, with methods like MPEG SAOC relying on assumptions and requiring high bitrates.
Innovation Solution
The proposed solution involves encoding and decoding vectors of parameters using modulo differential entropy coding, where each parameter is represented by an index value, and applying modulo N to differences between index values to reduce the number of possible symbols, sharing a common probability table for both elements, and encoding positions and values of upmix matrix elements efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If MPEG SAOC parametric coding is used to reduce bitrate, then bitrate is decreased, but coding complexity increases and quality deteriorates due to reliance on assumptions
Solution Approach 1:
The patent transforms the coding approach by changing parameters from assumption-based parametric coding to data-driven probabilistic coding. It uses probability tables and entropy coding with modulo differential coding to represent audio object parameters, replacing complex mathematical models with statistically optimized parameter representations that reduce both bitrate and complexity.
Solution Approach 2:
The patent creates simplified copies of audio objects through probabilistic modeling. Instead of transmitting full audio object data or relying on complex parametric assumptions, it transmits compressed probability tables and entropy-coded parameter representations that can be efficiently decoded to reconstruct audio objects with high fidelity at lower bitrates.
2Measurement precision
If separate probability tables are used for each element in parameter vectors, then coding precision is improved, but memory requirements increase
Solution Approach 1:
The patent merges separate probability tables into a shared common probability table that serves all elements in parameter vectors. This consolidation reduces memory requirements significantly while maintaining coding precision through the use of modulo differential coding, which exploits the statistical relationships between adjacent parameters to achieve accurate representation without element-specific tables.
Solution Approach 2:
The common probability table serves multiple functions simultaneously - it provides probability information for all parameter elements, enables entropy coding for the entire parameter vector, and supports modulo differential coding operations. This universal table replaces multiple specialized tables, reducing memory usage while maintaining or improving coding efficiency through shared statistical information.
3Ease of manufacture
If conventional audio coding schemes are used, then implementation is simpler, but audio quality and spatial accuracy deteriorate
Solution Approach 1:
The patent replaces conventional mechanical/audio coding approaches with information-theoretic methods. Instead of using traditional transform coding or parametric modeling, it employs entropy coding based on probability tables and modulo differential coding, substituting mathematical statistics and information theory for conventional audio coding mechanics to achieve superior quality.
Solution Approach 2:
The patent introduces probability tables as an intermediary between the audio signal and the coded representation. These tables serve as a statistical mediator that captures the inherent structure and correlations in audio data, enabling more efficient encoding than direct conventional methods while maintaining implementation feasibility through standardized entropy coding techniques.
4Manufacturing precision
If more parameters are transmitted to improve audio object reconstruction quality, then quality is improved, but bitrate increases
Solution Approach 1:
The patent extracts only the essential parameter information needed for high-quality reconstruction by using modulo differential coding. Instead of transmitting complete parameter values, it transmits only the differences from previous values encoded with entropy coding, extracting the minimal necessary information to maintain reconstruction quality while dramatically reducing bitrate.
Solution Approach 2:
The patent uses partial action by transmitting a subset of parameter information through entropy-coded differences rather than complete parameter sets. The modulo differential coding provides just enough information for accurate reconstruction without the excess of transmitting all possible parameter details, achieving optimal balance between quality and bitrate.
Data Source
AI summary
The present disclosure provides methods, devices and computer program products for encoding and decoding of a vector of parameters in an audio coding system. The disclosure further relates to a method and apparatus for reconstructing an audio object in an audio decoding system. According to the disclosure, a modulo differential approach for coding and encoding a vector of a non-periodic quantity may improve the coding efficiency and provide encoders and decoders with less memory requirements. Moreover, an efficient method for encoding and decoding a sparse matrix is provided.