Audio Decoder With Modulo Differential Parameter Coding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing audio coding systems, particularly object-based audio systems, face inefficiencies in encoding and decoding audio signals while maintaining quality, as methods like MPEG SAOC rely on complex mathematical processes and assumptions about audio objects, leading to suboptimal bitrate and quality.

Innovation Solution

A method and apparatus for encoding and decoding audio signals using entropy coding with modulo differential encoding and shared probability tables for upmix matrices, reducing memory requirements and bit rate by aligning probability distributions and using a single Huffman codebook.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If MPEG SAOC parametric coding is used to reduce bitrate, then bandwidth efficiency is improved, but decoding complexity and quality increase due to complex mathematical processes and assumptions

Engineering Contradiction:
ImprovebitrateVSAvoiddecoding complexity
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent transforms the complex parametric coding parameters into a simplified probability distribution model. By representing audio object parameters as probability distributions and using entropy coding, the system maintains bitrate efficiency while dramatically reducing decoding complexity. The key is changing from direct parametric reconstruction to probability-based symbolic representation.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent replaces complex mathematical reconstruction processes with a probabilistic modeling approach. Instead of using heavy mathematical operations to reconstruct audio objects, the system uses entropy coding with probability tables, substituting mechanical/mathematical complexity with statistical modeling that is computationally lighter.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If separate probability tables are used for each audio object to maintain accuracy, then coding precision is improved, but memory requirements increase

Engineering Contradiction:
Improvecoding precisionVSAvoidmemory requirements
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent merges separate probability tables into a single shared probability table that serves all audio objects. By combining the probability distributions and using a unified Huffman codebook, the system reduces memory requirements from multiple separate tables to one shared table, while maintaining coding precision through the unified probabilistic model.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The single shared probability table serves multiple functions for different audio objects simultaneously. The universal probability distribution model can represent parameters for any audio object, making the memory structure multi-functional and eliminating the need for object-specific tables.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Measurement precision

If multiple Huffman codebooks are maintained for different parameter types to ensure accuracy, then coding accuracy is improved, but device complexity and memory usage increase

Engineering Contradiction:
Improvecoding accuracyVSAvoidcodebook management complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent combines multiple parameter-specific Huffman codebooks into a single unified Huffman codebook. By merging the codebooks and using a single probability distribution model for all parameters, the system reduces codebook management complexity from maintaining multiple separate codebooks to managing one unified codebook, while preserving coding accuracy through the universal probabilistic approach.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP4290510B1Audio decoder
Publication Date: 2025.11.05 DOLBY INTERNATIONAL AB
  • EP4290510B1 patent drawingFigure 1~2
  • EP4290510B1 patent drawingFigure 3~4
  • EP4290510B1 patent drawingFigure 5~6

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.