Frequency Domain Signal Processor for Differential Microphone Arrays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional close-talking differential microphone arrays (CTDMAs) suffer from increased internal noise due to differential signal processing, which limits their noise cancellation effectiveness, especially at higher frequencies and in noisy environments, and also face challenges in maintaining signal quality due to phase mismatches and high pass filtering.
Innovation Solution
The proposed solution involves transforming microphone signals from the time domain to the frequency domain, allowing for separate processing of magnitude and phase information, where the differential operation is applied only to signal amplitudes, and the phase is restored from one of the original signals, thereby avoiding high pass effects and enhancing signal-to-noise ratio (SNR) across frequencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If differential signal processing is applied in a close-talking differential microphone array, then far field noise attenuation is improved, but internal noise is increased
Solution Approach 1:
The patent segments the signal processing by separating magnitude and phase components in the frequency domain. The differential operation is applied only to magnitude components, while phase components are preserved from the original signals. This segmentation allows noise reduction in the magnitude domain without the noise amplification that occurs with full differential processing of both magnitude and phase.
Solution Approach 2:
The patent introduces frequency domain transformation as an intermediary step between the time domain microphone signals and the final output. By converting to frequency domain, applying selective magnitude differential processing, and then transforming back to time domain, the system achieves noise attenuation without the internal noise amplification of conventional differential processing.
2Object-affected harmful factors
If differential summation is applied to microphone signals, then spatial enhancement of speech is improved, but signal quality is degraded due to noise amplification
Solution Approach 1:
The patent applies local quality by treating magnitude and phase components differently. The magnitude components undergo differential processing to enhance speech directionality, while the phase components are preserved unchanged. This selective local processing maintains signal quality by avoiding the amplification of noise that would occur if both components were processed differentially.
3Device complexity
If time domain signal processing is used, then processing simplicity is maintained, but frequency dependent noise cancellation is limited
Solution Approach 1:
The patent transitions from time domain processing to frequency domain processing, adding a dimensional transformation that enables frequency-dependent noise cancellation. By analyzing signals in the frequency domain, the system can apply different processing characteristics to different frequency components, achieving superior noise cancellation while maintaining practical implementation through efficient FFT algorithms.
Data Source
AI summary
A system and method for processing close talking differential microphone array (CTDMA) signals in which incoming microphone signals are transformed from time domain signals to frequency domain signals having separable magnitude and phase information. Processing of the frequency domain signals is performed using the magnitude information, following which phase information is reintroduced using phase information of one of the original frequency domain signals. As a result, high pass filtering effects of conventional differential signal processing of CTDMA signals are substantially avoided.