Hearing Aid Wind Noise Suppression via Power Ratio
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing hearing aids face challenges in effectively reducing wind noise, which can overload the microphone pre-amplifier and distort signal processing, particularly due to the computational cost and undesirable attenuation of low-frequency sounds in outdoor environments.
Innovation Solution
A computationally simple wind noise compensation method in hearing aids that determines the ratio of input signal power at low frequencies to total power, attenuating specific frequency bands to suppress wind noise while preserving music and speech quality, utilizing a multi-band signal processor with warped filters and adjustable gains.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If wind noise suppression is implemented using traditional spectral subtraction or multi-band attenuation methods, then wind noise reduction is achieved, but computational complexity increases and desired low-frequency sounds (music, speech) are undesirably attenuated
Solution Approach 1:
The patent extracts only the essential characteristic of wind noise (excessive low-frequency energy) and targets this specific feature for suppression. By using a simple low-pass filter to isolate low-frequency components and comparing their power to total power, the system selectively suppresses only the problematic wind noise portion without needing complex spectral analysis, thereby reducing computational complexity while maintaining effectiveness.
Solution Approach 2:
The patent changes the approach from complex spectral parameter analysis to a simple power ratio parameter. Instead of analyzing spectral shapes, harmonics, or using multiple frequency bands with different thresholds, the system uses a single parameter (ratio of low-frequency power to total power) that can be computed efficiently and adjusted dynamically to suppress wind noise while preserving desired sounds.
2Object-affected harmful factors
If aggressive low-frequency attenuation is applied to suppress wind noise, then wind noise reduction increases, but music and speech quality deteriorate due to unwanted attenuation of desired low-frequency sounds
Solution Approach 1:
The patent implements feedback by continuously monitoring the power ratio of low-frequency to total power and using this information to dynamically adjust the attenuation level. When the low-frequency power is excessively high (indicating wind noise), the system applies attenuation; when low-frequency power is normal (indicating music or speech), the system reduces or eliminates attenuation. This feedback mechanism ensures that desired sounds are preserved while wind noise is suppressed.
Solution Approach 2:
The patent makes the wind noise suppression system dynamic by continuously adapting the attenuation level based on real-time analysis of the power ratio. Instead of applying fixed aggressive attenuation, the system adjusts its suppression strength dynamically - applying strong attenuation only when wind noise is detected (high low-frequency power ratio) and reducing attenuation when music or speech is present (normal low-frequency power ratio), thereby preserving sound quality while suppressing wind noise.
3Device complexity
If simple attenuation methods are used to reduce wind noise, then computational cost decreases, but wind noise suppression effectiveness is insufficient
Solution Approach 1:
The patent replaces complex mechanical signal processing systems (spectral subtraction, multi-band filtering, harmonic analysis) with a simpler computational approach based on power ratio comparison. Instead of using computationally intensive spectral analysis or multiple filter banks, the system uses a single low-pass filter and power comparison, substituting complex mechanical processing with a streamlined computational method that is both efficient and effective.
Solution Approach 2:
The patent changes from complex multi-parameter spectral analysis to a single power ratio parameter. By focusing on the ratio of low-frequency power to total power, the system achieves effective wind noise suppression with minimal computation. This parameter change allows the system to capture the essential characteristic of wind noise (excessive low-frequency energy) without the computational burden of analyzing multiple spectral parameters simultaneously.
Data Source
AI summary
The present application relates to a hearing aid with suppression of wind noise wherein wind noise detection is provided involving only a single comparison of the input signal power level at first low frequencies with the input signal power level at frequencies that may include the first low frequencies whereby a computational cost effective and simple wind noise detection is provided. The determination of relative power levels of the input signal reflects the shape of the power spectrum of the signal, and the detection scheme is therefore typically capable of distinguishing music from wind noise so that attenuation of desired music is substantially avoided.