Audio Signal Processing for Crosstalk Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing audio signal processing technologies face challenges in effectively reducing cross-talk between loudspeakers, leading to narrow sweet spots and undesirable coloration due to ill-conditioning and sensitivity to errors, particularly in binaural audio reproduction.

Innovation Solution

The proposed solution involves decomposing left and right channel input audio signals into predetermined frequency bands to enhance accuracy of binaural cues and minimize complexity, using simple time delays and gains for low frequencies, and conventional cross-talk reduction for middle frequencies, while delaying or bypassing very low and high frequency components to avoid harmonic distortions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If inverse filters are used for cross-talk reduction, then cross-talk cancellation is achieved, but the system becomes sensitive to errors and produces narrow sweet spot

Engineering Contradiction:
Improvecross-talk cancellation accuracyVSAvoidsensitivity to errors
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent divides the frequency spectrum into multiple sub-bands and processes each sub-band separately with its own regularization value. This segmentation allows different parts of the frequency spectrum to be handled with appropriate error tolerance, reducing the overall sensitivity to errors while maintaining cross-talk cancellation effectiveness.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies frequency-dependent regularization values that change across different sub-bands. By adjusting the regularization parameter according to frequency characteristics, the system optimizes the balance between cross-talk cancellation and error sensitivity, achieving better reliability without excessive sensitivity to reproduction chain errors.

Inventive Principle:
Principle #35Parameter changes

2Loss of energy

If regularization is applied to control inverse filter gain, then dynamic range loss is reduced, but sweet spot becomes narrower

Engineering Contradiction:
Improvedynamic range lossVSAvoidsweet spot width
Core Design Contradiction:
Loss of energyVSArea of stationary object

Solution Approach 1:

The patent applies different regularization values to different frequency sub-bands based on their specific characteristics. This local quality approach ensures that each sub-band receives the appropriate amount of regularization to control dynamic range loss, while the overall system maintains a wider sweet spot through optimized frequency-dependent processing.

Inventive Principle:
Principle #3Local quality

3Device complexity

If sub-band division is used to lower complexity, then computational complexity is reduced, but spatial perception and sound quality are lowered

Engineering Contradiction:
Improveinverse filter design complexityVSAvoidspatial perception accuracy
Core Design Contradiction:
Device complexityVSManufacturing precision

Solution Approach 1:

The patent segments the frequency spectrum into multiple sub-bands and processes each with optimized regularization values. This segmentation reduces computational complexity by breaking down the full-frequency processing into manageable sub-problems, while maintaining spatial perception accuracy through frequency-dependent optimization in each sub-band.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the regularization parameter according to frequency sub-band characteristics. This parameter adaptation allows the system to achieve accurate spatial perception and sound quality by optimizing the processing for each frequency range, preventing the quality degradation that occurs with uniform sub-band processing.

Inventive Principle:
Principle #35Parameter changes

4Device complexity

If equal regularization values are applied to all frequencies, then computational simplicity is maintained, but low and high frequency components suffer from errors

Engineering Contradiction:
Improveregularization application simplicityVSAvoidfrequency component accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent changes the regularization value according to frequency sub-band characteristics. This parameter adaptation allows the system to achieve accurate spatial perception and sound quality by optimizing the processing for each frequency range, preventing the quality degradation that occurs with uniform sub-band processing.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3222058B1An audio signal processing apparatus and method for crosstalk reduction of an audio signal
Publication Date: 2019.05.22 HUAWEI TECH CO LTD
  • EP3222058B1 patent drawingFigure 1
  • EP3222058B1 patent drawingFigure 2
  • EP3222058B1 patent drawingFigure 3

AI summary

The invention relates to an audio signal processing apparatus (100) for filtering a left channel input audio signal (L) and a right channel input audio signal (R), a left channel output audio signal (X1) and a right channel output audio signal (X2) to be transmitted over acoustic propagation paths to a listener, wherein transfer functions of the acoustic propagation paths are defined by an acoustic transfer function matrix. The audio signal processing apparatus (100) comprises a decomposer (101), a first cross-talk reducer (103), a second cross-talk reducer (105), and a combiner (107). The first cross-talk reducer (103) is configured to reduce a cross-talk within a first predetermined frequency band upon the basis of the acoustic transfer function matrix. The second cross-talk reducer (105) is configured to reduce a cross-talk within a second predetermined frequency band upon the basis of the acoustic transfer function matrix.