Wind Noise Mitigation via Dual Microphone Signal Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing mobile communication devices struggle to effectively mitigate wind noise in speech signals without also removing desired speech, as current digital signal processing methods often corrupt the signal.

Innovation Solution

A method involving a primary microphone and a secondary microphone, where the secondary microphone is shielded from wind noise and generates an attenuated speech signal, which is then processed to replace the low-frequency portion of the primary microphone's signal, filtering and aligning it to mix with the high-frequency portion, thereby reducing wind noise without affecting the desired speech.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If digital signal processing methods are used to reduce wind noise, then wind noise is reduced, but desired speech is also removed

Engineering Contradiction:
Improvewind noiseVSAvoiddesired speech
Core Design Contradiction:
Object-affected harmful factorsVSLoss of information

Solution Approach 1:

The speech signal is divided into different frequency portions (low-frequency and high-frequency components). The low-frequency portion containing wind noise is processed separately from the high-frequency portion containing desired speech, allowing selective noise reduction without affecting speech quality

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The low-frequency portion of the signal contaminated by wind noise is extracted and replaced with a clean low-frequency reference signal obtained during a calibration phase when no wind noise is present. This extracts only the necessary clean frequency components while preserving the high-frequency speech content

Inventive Principle:
Principle #2Taking out (Extraction)

2Device complexity

If a single microphone is used, then device complexity is reduced, but wind noise mitigation is insufficient

Engineering Contradiction:
Improvemicrophone configurationVSAvoidwind noise
Core Design Contradiction:
Device complexityVSObject-affected harmful factors

Solution Approach 1:

A secondary reference microphone is introduced as an intermediary element during the calibration phase to capture an clean low-frequency reference signal free from wind noise. This reference signal serves as a mediator to replace the corrupted low-frequency portions in the primary microphone signal without requiring continuous multi-microphone arrays

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

A calibration phase is performed in advance when no wind noise is present to capture and store a clean low-frequency reference signal. This preliminary action prepares the necessary clean reference data before actual wind noise conditions occur, enabling effective noise mitigation without real-time complex processing

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3136385B1Method and device for mitigating wind noise in a speech signal generated at a microphone of the device
Publication Date: 2019.10.09 BLACKBERRY LTD
  • EP3136385B1 patent drawingFigure 1
  • EP3136385B1 patent drawingFigure 2
  • EP3136385B1 patent drawingFigure 3

AI summary

A mobile communication device and a method of mitigating wind noise in a speech signal generated at a microphone of the mobile communication device. A primary microphone of the mobile communication device receives speech and wind noise and generates a first speech signal based on the speech and the wind noise that is received. A secondary microphone of the mobile communication device receives attenuated speech and generates a second speech signal based on the attenuated speech that is received. The secondary microphone is shielded from the wind noise when the primary microphone receives speech and the wind noise. A processer of the mobile communication device processes the first speech signal to replace a low frequency portion of the first speech signal with the second speech signal in order to mitigate the wind noise in the first speech signal.