Audio Encoding Gain Reuse for Bitrate and Distortion Balance

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional encoding techniques for acoustic signals face challenges in reducing both quantization distortion and the number of bits in integer signal codes while maintaining efficient operation processing, especially when combining gain adjustment methods based on periodicity and non-periodicity.

Innovation Solution

The proposed solution involves processing the gain adjustment coder only for methods where the number of bits in the integer signal code is expected to be reduced, utilizing the gained gain in methods where bit reduction is not expected, and rearranging sample sequences to gather samples with similar magnitudes, thereby optimizing bit allocation and distortion reduction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If gain adjustment coding is performed for both periodicity-based and non-periodicity-based encoding methods, then quantization distortion is reduced, but operation processing amount increases

Engineering Contradiction:
Improvequantization distortionVSAvoidoperation processing amount
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by performing gain adjustment coding only for the non-periodicity-based encoding method in advance, and then reusing the resulting gain value for the periodicity-based encoding method. This avoids performing the same gain adjustment coding process twice, thereby reducing operation processing amount while maintaining quantization distortion reduction benefits.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent makes the gain adjustment coder multi-functional by having it serve both encoding methods through a shared architecture. The gain adjustment coder processes the non-periodicity-based method and its output is universally applied to both periodicity-based and non-periodicity-based encoding, eliminating redundant processing while maintaining effectiveness for both methods.

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

2Quantity of substance

If the number of bits in integer signal code is reduced, then bit rate is lowered, but quantization distortion increases

Engineering Contradiction:
Improvenumber of bitsVSAvoidquantization distortion
Core Design Contradiction:
Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent applies parameter changes by dynamically adjusting the gain value based on the bit allocation and signal characteristics. By optimizing the gain parameter in the gain adjustment coder, the system achieves better quantization performance for a given bit budget, effectively reducing quantization distortion without increasing the number of bits in the integer signal code.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If sample sequences are rearranged to gather samples with similar magnitudes, then variable-length encoding efficiency is improved, but processing complexity increases

Engineering Contradiction:
Improvevariable-length encoding efficiencyVSAvoidprocessing complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent applies segmentation by dividing the sample sequence into groups based on magnitude similarity before variable-length encoding. By segmenting the sequence into homogeneous segments, the variable-length encoder can more efficiently represent the data with fewer bits, improving encoding efficiency while keeping the segmentation logic relatively simple.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by creating local regions in the sample sequence where samples have similar magnitudes. This local homogeneity allows the variable-length encoder to work more efficiently on each local segment, improving overall encoding efficiency without requiring complex global processing.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3525208B1Encoding method, encoder, program and recording medium
Publication Date: 2021.07.28 NIPPON TELEGRAPH & TELEPHONE CORP
  • EP3525208B1 patent drawingFigure 1
  • EP3525208B1 patent drawingFigure 2
  • EP3525208B1 patent drawingFigure 3

AI summary

Between a method in which an integer signal code is obtained by using an encoding method based on periodicity and a method in which an integer signal code is obtained using an encoding method which is not based on periodicity using an index indicating a degree of periodicity of a sample sequence in a frequency domain, processing of a first coder is executed only in a method for which the number of bits of the integer signal code is expected to be reduced, and a gain obtained through the processing of the first coder in the method for which the number of bits of the integer signal code is expected to be reduced is utilized in a method for which the number of bits of the integer signal code is not expected to be reduced.