Time-Domain Filter Coefficient Updating for Adaptive Audio Response

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Time domain filtering in audio signal processing faces challenges in adapting filter coefficients in real-time to match the time-varying nature of audio signals, leading to suboptimal frequency responses.

Innovation Solution

A method and system for adaptively updating filter coefficients by analyzing audio signals to determine target gains for each subband, iteratively selecting and updating coefficients based on a least square method and a mapping matrix, allowing for real-time adjustment of filter coefficients to approximate target frequency responses.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If time domain filtering is used to achieve ultra-low latency and improved linear convolution, then processing speed and convolution quality are improved, but the filter coefficients cannot be adjusted in real-time to match time-varying audio signals

Engineering Contradiction:

Improveprocessing speedVSAvoidadaptability to time-varying signals

Core Design Contradiction:

SpeedVSAdaptability or versatility

Solution Approach 1:

The patent applies dynamics by making the filter coefficients adjustable and adaptive in real-time. The system transitions from static pre-trained coefficients to dynamic coefficients that can be updated based on current audio signal characteristics, allowing the filter to adapt to time-varying signals while maintaining time domain processing speed

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent changes the parameters (filter coefficients) based on analyzed audio signal properties. By computing target frequency responses from the audio signal and iteratively adjusting coefficients to match these targets, the system enables real-time adaptation while preserving the computational efficiency of time domain filtering

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If frequency domain filtering is used to easily adjust frequency response by changing subband gains, then adaptability is improved, but latency increases and linear convolution quality deteriorates

Engineering Contradiction:

Improvefrequency response adjustabilityVSAvoidprocessing latency

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent substitutes the frequency domain processing mechanism with a time domain mechanism. Instead of transforming to frequency domain, applying gains, and transforming back (which causes latency and convolution issues), the system directly adjusts time domain filter coefficients to achieve frequency response control, eliminating the need for FFT-based processing

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent achieves frequency response adjustment by changing time domain parameters (filter coefficients) rather than frequency domain parameters (subband gains). The coefficients are iteratively updated based on target frequency responses derived from the audio signal, providing adaptability without requiring frequency domain transformation

Inventive Principle:

Principle #35Parameter changes

3Device complexity

If filter coefficients are pre-trained offline with fixed values, then device complexity is reduced, but the filter cannot adapt to time-varying audio signals

Engineering Contradiction:

Improvefilter design complexityVSAvoidreal-time adaptation capability

Core Design Contradiction:

Device complexityVSAdaptability or versatility

Solution Approach 1:

The patent applies preliminary action by pre-computing and storing a mapping matrix that relates audio signal characteristics to optimal filter coefficients. This pre-computed mapping enables real-time adaptation without complex online optimization, maintaining low device complexity while achieving adaptability through the stored mapping relationship

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent enables the filter to self-adjust its coefficients based on the audio signal it processes. By analyzing the current audio portion and using the pre-computed mapping to determine appropriate coefficients, the system makes the filter self-adaptive without requiring external reconfiguration or complex training procedures

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS10332540B2Filter coefficient updating in time domain filtering

Publication Date: 2019.06.25 DOLBY LABORATORIES LICENSING CORP

AI summary

Example embodiments disclosed herein relate to filter coefficient updating in time domain filtering. A method of processing an audio signal is disclosed. The method includes obtaining a predetermined number of target gains for a first portion of the audio signal by analyzing the first portion of the audio signal. Each of the target gains is corresponding to a linear subband of the audio signal. The method also includes determining a filter coefficients for time domain filtering the first portion of the audio signal so as to approximate a frequency response given by the target gains. The filter coefficients are determined by iteratively selecting at least one target gain from the target gains and updating the filter coefficient based on the selected at least one target gain. Corresponding system and computer program product for processing an audio signal are also disclosed.