Multiband Audio Tuning with Band-Specific Crosstalk Cancellation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional audio playback systems face challenges in effectively reducing crosstalk across the entire frequency range of an audio signal, leading to distortion and harsh sound at lower frequencies or audible artifacts at higher frequencies due to inadequate crosstalk cancellation techniques.

Innovation Solution

The system separates the audio signal into multiple frequency bands (low, midband, and high-frequency feeds) and applies customized filters with delay and gain values to each band, using sensor data to adjust filters based on listener position and orientation for optimal crosstalk cancellation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If crosstalk cancellation is applied to the entire frequency range, then crosstalk reduction is maximized, but distortion and harsh sound occur at lower frequencies and audible artifacts occur at higher frequencies

Engineering Contradiction:
ImprovecrosstalkVSAvoiddistortion and audible artifacts
Core Design Contradiction:
Object-affected harmful factorsVSObject-generated harmful factors

Solution Approach 1:

The audio signal is divided into multiple frequency bands (low-frequency, midband, high-frequency) and processed separately. Different filter characteristics are applied to each band based on their specific requirements, avoiding the distortion and artifacts that occur when a single filter is applied to the entire frequency range.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different filter characteristics are applied to different frequency bands. The low-frequency band uses filters optimized for its characteristics, the midband uses filters optimized for crosstalk cancellation, and the high-frequency band uses filters optimized for its characteristics, ensuring optimal performance for each frequency range without compromising others.

Inventive Principle:
Principle #3Local quality

2Object-affected harmful factors

If conventional crosstalk cancellation techniques are applied, then crosstalk is reduced, but the listening experience is degraded due to distortion and harsh sound

Engineering Contradiction:
ImprovecrosstalkVSAvoidlistening experience
Core Design Contradiction:
Object-affected harmful factorsVSReliability

Solution Approach 1:

The audio signal is segmented into multiple frequency bands processed separately with appropriate filters for each band, preventing the distortion and harsh sound that degrade listening experience while maintaining effective crosstalk cancellation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Filter parameters such as cutoff frequencies, Q-factors, and delay values are adjusted according to the specific characteristics of each frequency band, optimizing the balance between crosstalk cancellation and preserving audio quality for the overall listening experience.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4583538A1Tuning of multiband audio systems executing crosstalk cancellation
Publication Date: 2025.07.09 HARMAN INT IND INC
  • EP4583538A1 patent drawingFigure 1
  • EP4583538A1 patent drawingFigure 2
  • EP4583538A1 patent drawingFigure 3

AI summary

In various embodiments, a computer-implemented method comprises generating, by an audio playback module from an input audio signal, a plurality of audio feeds that includes at least a midband feed and an additional feed, applying a crosstalk cancellation filter on the midband feed to generate a processed midband audio signal, applying an additional filter on the additional feed to generate an additional audio signal, where the additional filter applies at least a delay value to the additional feed, generating a plurality of processed audio signals based on both the processed midband audio signal and the additional audio signal; and transmitting, by the audio playback module, the plurality of processed audio signals to a plurality of speakers.