Acoustic Delay Measurement Using Adaptive Filter and Programmable Buffer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for measuring audio latency in audio systems are susceptible to ambient noise and multi-modal reverberation, and adaptive filters often struggle with echo delays that exceed their capacity, making precise delay measurement challenging.
Innovation Solution
A converged adaptive filter coupled with a programmable delay buffer is used to measure audio latency by introducing a programmable delay into the audio stream, allowing the filter to converge and compute the delay, which can then be added to the buffer to provide an accurate measurement of the system's inherent delay.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional cross-correlation or time-domain reflectometry methods are used to measure audio delay, then measurement can be performed, but the measurement precision deteriorates due to ambient noise and multi-modal reverberation
Solution Approach 1:
The patent introduces a trained neural network model as an intermediary between the noisy audio signal and the delay measurement process. The neural network processes the corrupted audio signal and extracts delay information by learning the underlying patterns, effectively filtering out ambient noise and reverberation effects while preserving the true delay measurement.
2Adaptability or versatility
If the adaptive filter size is increased to handle larger echo delays, then the filter capacity improves, but the device complexity and computational burden increase
Solution Approach 1:
The patent transforms the problem from the time domain to the frequency domain by applying FFT (Fast Fourier Transform). This dimensional transformation allows the system to analyze and cancel echoes across different frequency components simultaneously, effectively increasing the filter's capacity to handle large delays without proportionally increasing the time-domain filter size.
Solution Approach 2:
The patent changes the representation parameters of the audio signal by transforming it into the frequency domain. This parameter change enables more efficient processing of long-delay echoes through frequency-selective filtering, reducing the computational complexity compared to using a very large time-domain adaptive filter.
Data Source
AI summary
An acoustic delay measurement apparatus measures an audio delay introduced by an audio system. A programmable delay buffer, which introduces a programmable delay into the audio stream, receives an audio stream from an audio source and outputs a reference signal representing the audio stream. An adaptive filter is responsive to the reference signal to generate an estimate signal to match on convergence of the adaptive filter an audio signal output by the audio system, which is the audio stream delayed by an amount representative of the audio delay introduced by the audio system. A processor including a coefficient analysis block reads coefficients in the adaptive filter after convergence, computes a delay introduced into said estimate signal by the adaptive filter, and adds the computed delay to the programmable delay buffer to provide a measurement of the audio delay.