Frequency-Domain Audio Processing for Broadband Noise Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing audio signal processing techniques are inadequate for reducing noise components in broadcasting receivers, such as radio and TV receivers, as they are optimized for mobile telephone applications and fail to effectively handle broadband audio signals with various noise sources, leading to insufficient noise reduction and increased circuit size and manufacturing costs.

Innovation Solution

An audio signal processing device and method that divides input signals into frames, converts them to frequency signals, detects and adjusts spectral levels within specific threshold ranges, and applies temporal smoothing to reduce noise without specifying noise components in advance, while also enhancing voice communication without deteriorating sound quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If spectral deduction method is used to reduce noise components, then noise reduction effect is improved, but computation amount increases significantly

Engineering Contradiction:
Improvenoise componentsVSAvoidcomputation amount
Core Design Contradiction:
Object-affected harmful factorsVSDevice complexity

Solution Approach 1:

The patent changes the parameter of noise spectrum estimation from minimum mean squared error method to a simplified method using only the current frame's spectral information. This parameter change reduces computation while maintaining noise reduction effectiveness by focusing on the most relevant spectral data.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent extracts only the essential spectral information from the current frame to estimate noise spectrum, rather than performing repeated computations over multiple frames. This extraction approach maintains noise reduction capability while significantly reducing computational load.

Inventive Principle:
Principle #2Taking out (Extraction)

2Object-affected harmful factors

If noise components are reduced using conventional methods, then noise reduction is achieved, but quantization noise increases when signal changes greatly

Engineering Contradiction:
Improvenoise componentsVSAvoidquantization noise
Core Design Contradiction:
Object-affected harmful factorsVSObject-generated harmful factors

Solution Approach 1:

The patent applies dynamic control to the noise reduction process by using temporal smoothing on the level control values. This dynamic adjustment prevents abrupt changes in signal processing, thereby reducing quantization noise while maintaining effective noise reduction throughout signal transitions.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent performs preliminary smoothing on level control values before applying noise reduction. This preliminary action prepares the control parameters to handle signal changes smoothly, preventing quantization noise from occurring during transient signal variations.

Inventive Principle:
Principle #10Preliminary action

3Ease of operation

If voice enhancement is applied to emphasize specific frequency bands, then voice clarity is improved, but sound quality balance deteriorates

Engineering Contradiction:
Improvevoice clarityVSAvoidsound quality balance
Core Design Contradiction:
Ease of operationVSManufacturing precision

Solution Approach 1:

The patent applies local quality enhancement by emphasizing specific frequency bands where voice components are present, while maintaining original sound quality in other frequency regions. This selective enhancement improves voice clarity without compromising overall sound balance.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP2172930B1Audio signal processing device and audio signal processing method
Publication Date: 2012.02.22 JVC KENWOOD CORP
  • EP2172930B1 patent drawingFigure 1
  • EP2172930B1 patent drawingFigure 2~3
  • EP2172930B1 patent drawingFigure 4

AI summary

In frequency signals obtained by converting input audio signals from time-domain signals to frequency-domain signals, a level control value setting unit 5 establishes a level control value for reducing the levels of spectrums at a noise-components level. A level control value smoothing unit 6 carries out a smoothing process of smoothing the level control value established by the level control value setting unit 5 temporally. A spectral adjustment unit 8 multiplies the level control value after the smoothing process by the frequency signals, performing a level control.