Stereo Signal Processing via Exponential Weighting for Omni Microphones
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for generating stereo signals from omnidirectional microphones lack the ability to adaptively control stereo width, resulting in suboptimal stereo images that require precise microphone placement and orientation, and suffer from low signal-to-noise ratio and spectral defects.
Innovation Solution
The method introduces a stereo width control parameter, implemented through exponential weighting functions for differential signals, allowing for flexible adjustment of stereo width by modifying the power spectra and gain filtering, thereby improving left-right separation and noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If omni-directional microphones are used to reduce cost and device size, then manufacturing cost and device complexity are reduced, but stereo signal separation and spatial perception are degraded
Solution Approach 1:
The patent introduces an intermediary processing stage that converts omni-directional microphone signals into first-order differential signals through delay and subtraction operations. This intermediary transformation enables the system to achieve directional signal characteristics equivalent to expensive directional microphones while retaining the manufacturing advantages of omni-directional elements.
Solution Approach 2:
The patent replaces the mechanical solution of using physically separated directional microphones with a signal processing approach. By using delay elements and subtraction circuits to synthesize differential signals, the system achieves the same acoustic effect without requiring precise mechanical positioning or specialized microphone structures.
2Area of stationary object
If the distance between two omni-directional microphones is increased to improve stereo image width, then spatial separation is improved, but device size and complexity increase
Solution Approach 1:
The patent changes the parameter of microphone spacing from a fixed physical constraint to a flexible signal processing parameter. By adjusting delay times and weighting factors in the differential signal computation, the system can simulate various spacing effects without physically moving the microphones, thereby maintaining compact device size while achieving desired stereo image width.
3Manufacturing precision
If first-order differential signals are used to increase left-right separation, then stereo signal separation is improved, but signal-to-noise ratio deteriorates at low frequencies
Solution Approach 1:
The patent applies dynamic, frequency-dependent weighting to the differential signals. Rather than uniformly processing all frequencies, the system adaptively adjusts the gain and filtering characteristics based on frequency content, preserving the noise-rejection benefits of differential signaling at high frequencies while compensating for low-frequency noise amplification through selective attenuation.
4Reliability
If a fixed gain filter is applied to original microphone signals to improve SNR, then signal-to-noise ratio is improved, but adaptability to different stereo image requirements is lost
Solution Approach 1:
The patent creates a universal processing framework where a single omni-directional microphone pair can serve multiple stereo imaging purposes. By combining differential signal computation with adjustable gain filtering and frequency-dependent weighting, the system achieves both noise reduction and adaptive stereo width control, replacing the need for multiple fixed-configuration microphone arrays.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method (200) for determining an output stereo signal (Y1, Y2) comprising: determining (201 ) a first differential signal (x1) based on a difference of a first input audio channel signal (m1) and a filtered version of a second input audio channel signal (m2) and determining a second differential signal (x2) based on a difference of the second input audio channel signal (m2) and a filtered version of the first input audio channel signal (m1); determining (203) a first power spectrum (P1) based on the first differential signal (x1) and determining a second power spectrum (P2) based on the second differential signal (x2); determining (205) a first weighting function (W1) and a second weighting function (W2) as a function of the first power spectrum (P1) and the second power spectrum (P2); wherein the first weighting function (W1) and the second weighting function (W2) comprise an exponential function; and filtering (207) a first signal, which represents a first combination of the first input audio channel signal (m1) and the second input audio channel signal (m2), with the first weighting function (W1) to obtain a first output audio channel signal (Y1) of the output stereo signal (Y1, Y2), and filtering a second signal, which represents a second combination of the first input audio channel signal (m1) and the second input audio channel signal (m2), with the second weighting function (W2) to obtain a second output audio channel signal (Y2) of the output stereo signal (Y1; Y2).