Audio Signal Compression Using Masking Curves for Hearing Loss

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional hearing aid processing methods, relying on linear filtering techniques, are inadequate for providing an enhanced audio experience for hearing-impaired individuals, especially when consuming audio content with limited dynamic range, as they fail to accurately simulate the sound perception of individuals with normal hearing.

Innovation Solution

A method that involves deriving a user's hearing profile through masking contour curve analysis, allowing for the customization of audio signals to match their specific hearing abilities by parameterizing a digital compression system to enhance audio perception, using both demographic data and hearing test results to create a target masking contour curve for multi-band compression.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If linear filtering techniques (equalization) are used to compensate for hearing loss, then frequency gain can be adjusted according to user's hearing profile, but the method is only applicable to conductive hearing loss and cannot accurately model sensorineural hearing loss

Engineering Contradiction:
Improveapplicability to different types of hearing lossVSAvoidaccuracy of hearing impairment modeling
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent changes the fundamental parameter of the processing approach from linear filtering to non-linear processing. Specifically, it uses masking contour curves and signal-to-noise ratio manipulation to transform the audio signal in a non-linear manner that accurately reflects the complex physiology of sensorineural hearing loss, thereby achieving both versatility across hearing loss types and reliability in modeling

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent substitutes the mechanical linear filtering approach with a psychoacoustic-based non-linear processing system. Instead of simply boosting frequencies, it replaces the mechanical EQ approach with a system that simulates the human auditory system's non-linear response to sound, including masking effects and dynamic range compression

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

2Manufacturing precision

If conventional hearing aid processing (wide dynamic range compression) is applied to recorded audio content, then faint sounds can be amplified, but the processing degrades the perceived quality of audio content with already limited dynamic range

Engineering Contradiction:
Improvedetection of faint soundsVSAvoidperceived quality of audio content
Core Design Contradiction:
Manufacturing precisionVSReliability

Solution Approach 1:

The patent applies partial compression only to frequency regions where masking occurs, rather than applying wide dynamic range compression across the entire audio spectrum. By selectively compressing only the problematic frequency bands identified through masking contour analysis, it provides benefit to faint sounds while preserving the quality of the overall audio content

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent applies different processing characteristics to different frequency bands based on the user's specific hearing profile and masking contours. Instead of uniform compression across all frequencies, it tailors the compression ratio and threshold locally to each frequency band, thereby improving faint sound detection in affected bands while maintaining audio quality in unaffected bands

Inventive Principle:
Principle #3Local quality

3Measurement precision

If linear equalization is used to boost frequencies for hearing loss compensation, then frequency response can be corrected, but the human auditory system's non-linear response is not accounted for

Engineering Contradiction:
Improvefrequency response correctionVSAvoidaccuracy of sound perception simulation
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent fundamentally changes the processing parameter from static linear gain to dynamic non-linear transformation. It uses masking contour curves to determine frequency-dependent compression ratios, transforming the audio signal in a manner that reflects the non-linear physiology of the human auditory system rather than applying simple linear frequency boosting

Inventive Principle:
Principle #35Parameter changes

4Ease of manufacture

If simple equalization techniques are applied, then the processing is computationally simple, but the method cannot provide beneficial sound personalization for sensorineural hearing loss

Engineering Contradiction:
Improvesimplicity of processingVSAvoideffectiveness for different hearing loss types
Core Design Contradiction:
Ease of manufactureVSAdaptability or versatility

Solution Approach 1:

The patent performs preliminary analysis to obtain the user's masking contour curve and hearing profile before applying the compression processing. By pre-characterizing the user's specific hearing loss pattern and masking behavior, the system can then apply targeted non-linear processing that is both computationally efficient and highly effective for that individual's type of hearing loss

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3641343B1Method to enhance audio signal from an audio output device
Publication Date: 2021.01.20 MIMI HEARING TECHNOLOGIES GMBH
  • EP3641343B1 patent drawingFigure 1
  • EP3641343B1 patent drawingFigure 2
  • EP3641343B1 patent drawingFigure 3

AI summary

A method of enhancing an audio signal from an audio output device is provided. For a frequency band, a user masking contour curve covering at least a part of said frequency band is obtained, a target masking contour curve is derived from the user masking contour curve, and a multi-band digital compression system is parameterized based on the sound level of the target masking contour curve at a given frequency and the sound level of the user masking contour curve at the same given frequency. The obtained parameters are outputted to provide an enhanced audio signal.