Adaptive Reverberation Cancellation Using Sparse Signal Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sound field reproduction systems face challenges in reverberant environments due to the need for a large number of loudspeaker-microphone channels, difficulty in handling unknown reverberant room channels, and vulnerability to changes in ambient conditions, leading to poor sound localization and increased costs.
Innovation Solution
A signal processor using sparse methods to generate drive signals for loudspeakers, which reduces the number of microphones required, facilitates system convergence over a wide frequency range, and simplifies computation by assuming orthogonal sound functions and diagonal transfer functions, enabling adaptive reverberation cancellation in multi-zone sound field reproduction systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing sound field reproduction techniques are used in reverberant environments, then sound field reproduction is attempted, but the system requires a large number of loudspeaker-microphone channels and prior transfer function measurements, increasing device complexity and loss of time
Solution Approach 1:
The patent extracts and eliminates the requirement for prior transfer function measurements by implementing an adaptive filtering system that performs real-time room impulse response estimation. This removes the time-consuming pre-measurement step while maintaining sound field reproduction accuracy in reverberant environments.
Solution Approach 2:
The patent implements a feedback mechanism where the system continuously estimates the room impulse response using microphone signals and adaptively updates the loudspeaker drive signals. This closed-loop approach allows the system to converge to accurate sound field reproduction without requiring prior measurements or a large number of channels.
2Measurement precision
If prior transfer function measurement is performed for all loudspeakers, then accurate reverberation cancellation is achieved, but the process is time-consuming and vulnerable to environmental changes
Solution Approach 1:
The patent transforms the static pre-measurement approach into a dynamic adaptive system that continuously estimates and updates the room impulse response in real-time. This allows the system to adapt to environmental changes and converge to accurate transfer function estimates without requiring time-consuming prior measurements.
Solution Approach 2:
The system performs self-calibration by automatically estimating the room impulse response using its own loudspeaker outputs and microphone inputs. This eliminates the need for external measurement equipment or manual transfer function measurements, making the system self-sufficient and rapid to deploy.
3Adaptability or versatility
If adaptive filtering is applied in reverberant environments with low direct-to-reverberant-path power ratios, then reverberation cancellation is attempted, but the adaptive processes diverge and cause ill-conditioning problems
Solution Approach 1:
The patent changes the parameter representation by working in the wave domain rather than the time domain, and by using a diagonally dominant system matrix formulation. This parameter transformation ensures numerical stability and prevents ill-conditioning problems even in highly reverberant environments with low direct-to-reverberant ratios.
Solution Approach 2:
The patent applies preliminary regularization to the system matrix to ensure it remains diagonally dominant throughout the adaptive process. This preliminary action prevents ill-conditioning before it can occur, guaranteeing convergence of the adaptive filtering algorithm in reverberant environments.
4Productivity
If iterative calculation of pseudoinverse is performed in each iteration, then adaptive filtering is updated, but computational complexity increases and channel estimation errors occur
Solution Approach 1:
The patent extracts and eliminates the computationally intensive pseudoinverse calculation from each iteration by formulating the problem to work with a diagonally dominant system matrix. This allows the adaptive filter to be updated using simple, efficient operations without requiring complex matrix inversion, significantly reducing computational complexity.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A signal processor for determining a plurality of drive signals for driving a plurality of loud- speakers to cancel a reverberation effect in a listening area, wherein the signal processor is configured to determine from one or more measured audio signals a plurality of measured physical coefficients in a basis of physical sound functions, such that a sum of the physical sound functions, weighted with the plurality of measured physical coefficients approximates the one or more measured audio signals, wherein at least half of the plurality of measured physical coefficients are zero, determine a residual error between the plurality of measured physical coefficients and a plurality of desired physical coefficients, estimate a transfer function describing a transformation from the plurality of desired physical coefficients to the plurality of measured physical coefficients, based on the determined residual error, and update the plurality of drive signals based on the estimated transfer function, wherein the signal processor is configured to repeatedly carry out the above steps.