Variable Frequency Segmentation for Audio Spectral Coding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Traditional audio coding techniques use uniform frequency segmentation, which can be suboptimal as it does not account for varying spectral data intensity, leading to inefficient coding and increased bit-rate requirements.

Innovation Solution

The method involves variable-sized frequency segmentation of sub-bands, where finer segmentation is applied to regions with high spectral variance and coarser segmentation to more homogeneous regions, allowing for optimized coding by merging or splitting sub-bands based on intensity, tonality, and energy measurements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If uniform frequency segmentation is used, then device complexity is reduced and ease of operation is improved, but coding efficiency deteriorates and bit-rate increases

Engineering Contradiction:
Improveease of segmentationVSAvoidcoding efficiency
Core Design Contradiction:
Ease of operationVSProductivity

Solution Approach 1:

The spectrum is divided into multiple frequency sub-bands, and each sub-band is further segmented into sub-sub-bands based on spectral characteristics. This hierarchical segmentation allows the system to adapt to varying spectral data intensity, applying finer segmentation where needed and coarser segmentation where homogeneous, thereby improving coding efficiency without uniformly increasing complexity across the entire spectrum.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different segmentation granularities are applied to different frequency regions based on their local spectral characteristics. Regions with high spectral variance receive finer segmentation, while homogeneous regions receive coarser segmentation. This local adaptation optimizes coding efficiency for each region's specific characteristics rather than applying a one-size-fits-all approach.

Inventive Principle:
Principle #3Local quality

2Manufacturing precision

If finer segmentation is applied to high variance regions, then manufacturing precision of spectral representation is improved, but device complexity increases due to more sub-bands requiring configuration

Engineering Contradiction:
Improvespectral representation precisionVSAvoidsub-band configuration complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The segmentation structure is made dynamic and adaptive based on spectral characteristics. The system automatically adjusts the segmentation granularity for each sub-band according to measures of spectral variance, tonality, and energy. This dynamic adaptation allows high precision where needed while maintaining coarser structures elsewhere, balancing precision requirements against configuration complexity.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes key parameters including sub-band boundaries, sub-band sizes, and segmentation depth based on spectral characteristics. By dynamically adjusting these parameters according to measured spectral variance, tonality, and energy, the system achieves high spectral representation precision in critical regions while avoiding unnecessary complexity in homogeneous regions.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If variable sub-band sizes are used, then spectral data representation is optimized, but the number of bits required to code sub-band configuration increases

Engineering Contradiction:
Improvespectral coding efficiencyVSAvoidconfiguration overhead bits
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

Instead of applying fine-grained variable segmentation throughout the entire spectrum, the system applies variable sub-band sizes only where spectral characteristics warrant it. Homogeneous regions maintain coarser, more predictable segmentation, reducing the configuration overhead needed to describe the segmentation structure, while still achieving optimization in regions that require it.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The system performs preliminary analysis of spectral characteristics (variance, tonality, energy) before finalizing the segmentation structure. This preliminary action allows the system to pre-determine which regions will benefit from variable segmentation and which can use fixed segmentation, thereby optimizing spectral coding efficiency while minimizing the bits required to encode the configuration.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP1904999B1Frequency segmentation to obtain bands for efficient coding of digital media
Publication Date: 2017.03.08 MICROSOFT TECHNOLOGY LICENSING LLC
  • EP1904999B1 patent drawingFigure 1
  • EP1904999B1 patent drawingFigure 2
  • EP1904999B1 patent drawingFigure 3

AI summary

Frequency segmentation is important to the quality of encoding spectral data. Segmentation involves breaking the spectral data into units called sub-bands or vectors. Homogeneous segmentation may be suboptimal. Various features are described for providing spectral data intensity dependent segmentation. Finer segmentation is provided for regions of greater spectral variance and coarser segmentation is provided for more homogeneous regions. Sub-bands which have similar characteristics may be merged with very little effect on quality, whereas sub-bands with highly variable data may be better represented if a sub-band is split. Various methods are described for measuring tonality, energy, or shape of a sub-band. These various measurements are discussed in light of making decisions of when to split or merge sub-bands to provide variable frequency segmentation.