Tonality-Adaptive Audio Quantization for Musical Noise Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In very-low-bit-rate audio coding, musical noise artifacts occur due to insufficient bit allocation, particularly in stationary music or noise spectra, where transform lines are quantized to zero or not, leading to a tonal character in the decoded signal and inefficient bit usage, especially at low frequencies where the human auditory system is sensitive.

Innovation Solution

An audio encoder with a signal-adaptive dead-zone quantization scheme that modifies the dead-zone based on tonality indicating values for spectral lines or groups, allowing per-bin or per-band dead-zone selection before quantization, reducing complexity and side-information requirements, and optimizing audio quality by varying the dead-zone size based on frequency and tonality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a fixed dead-zone quantization is used, then the encoding complexity is low, but musical noise artifacts occur and audio quality deteriorates

Engineering Contradiction:
Improveencoding complexityVSAvoidmusical noise artifacts
Core Design Contradiction:
Device complexityVSObject-affected harmful factors

Solution Approach 1:

The patent applies dynamics by making the dead-zone size adaptive rather than fixed. The quantizer modifies the dead-zone size based on the tonality of each spectral bin, allowing the system to adapt to different signal characteristics. This resolves the contradiction by introducing signal-dependent adaptability that reduces musical noise while maintaining reasonable encoding complexity through efficient tonality detection and dead-zone adjustment mechanisms.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the dead-zone parameter dynamically based on tonality indication. By calculating tonality values for each spectral bin and adjusting the dead-zone size accordingly (larger dead-zones for non-tonal bins, smaller or zero dead-zones for tonal bins), the system optimizes quantization performance. This parameter adaptation eliminates musical noise artifacts while preserving encoding efficiency.

Inventive Principle:
Principle #35Parameter changes

2Object-affected harmful factors

If the dead-zone is adapted on the entire spectrum, then musical noise is reduced, but bit-rate increases due to side-information transmission

Engineering Contradiction:
Improvemusical noiseVSAvoidbit-rate
Core Design Contradiction:
Object-affected harmful factorsVSQuantity of substance

Solution Approach 1:

The patent applies local quality by adapting the dead-zone on a per-bin or per-band basis rather than uniformly across the entire spectrum. Each spectral bin or band receives a customized dead-zone size based on its local tonality characteristics. This localized adaptation reduces musical noise in non-tonal regions while preserving tonal components, and minimizes side-information requirements since only local tonality indicators need to be transmitted.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent segments the spectrum into multiple bins or bands and applies different dead-zone sizes to each segment based on its tonality. This segmentation allows independent optimization of each spectral region, reducing overall musical noise while controlling bit-rate through efficient representation of segment-specific parameters.

Inventive Principle:
Principle #1Segmentation

3Device complexity

If quantization is performed without tonality adaptation, then encoding complexity is low, but bits are wasted in spectral regions that could be better utilized

Engineering Contradiction:
Improveencoding complexityVSAvoidbit allocation efficiency
Core Design Contradiction:
Device complexityVSLoss of information

Solution Approach 1:

The patent introduces tonality-based parameter changes to optimize bit allocation. By calculating tonality indicators and adjusting dead-zone sizes accordingly, the system directs more bits to tonal spectral bins (preserving perceptually important information) and uses larger dead-zones for non-tonal bins (where fewer bits are needed). This efficient bit allocation improves audio quality without significantly increasing encoding complexity.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS11694701B2Low-complexity tonality-adaptive audio signal quantization
Publication Date: 2023.07.04 FRAUNHOFER GESELLSCHAFT ZUR FORDERUNG DER ANGEWANDTEN FORSCHUNG EV
  • US11694701B2 patent drawing
  • US11694701B2 patent drawing

AI summary

The invention provides an audio encoder for encoding an audio signal so as to produce therefrom an encoded signal, the audio encoder including: a framing device configured to extract frames from the audio signal; a quantizer configured to map spectral lines of a spectrum signal derived from the frame of the audio signal to quantization indices, wherein the quantizer has a dead-zone, in which the input spectral lines are mapped to quantization index zero; and a control device configured to modify the dead-zone; wherein the control device includes a tonality calculating device configured to calculate at least one tonality indicating value for at least one spectrum line or for at least one group of spectral lines, wherein the control device is configured to modify the dead-zone for the at least one spectrum line or the at least one group of spectrum lines depending on the respective tonality indicating value.