Audio Spectrum Bandwidth Extension for Low-Bitrate Artifact Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio coding methods, such as MP3 and AAC, introduce artifacts like 'Birdies,' bandwidth limitation, tonal spike, and beating at low bitrates, degrading sound quality, and there is a need to detect and reduce these artifacts, particularly the effects of Spectral Band Replication (SBR) and bandwidth reduction.
Innovation Solution
An apparatus and method that separate and extend different portions of an audio signal's spectrum using specific parameters, apply anti-roughness filtering, and combine filtered signals to enhance audio quality, while also detecting characteristics related to SBR and artificial bandwidth limitation to inform post-processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If lossy coding operations like perceptually adapted quantization or parametric coding techniques are applied to reduce bitrate, then data transmission efficiency is improved, but audio signal quality deteriorates due to introduced artifacts
Solution Approach 1:
The patent detects coding artifacts (birdies, bandwidth limitation, tonal spikes, beating) introduced by lossy coding and converts this harmful effect into a benefit by using the detected artifact information to guide post-processing enhancement operations that specifically target and reduce these artifacts, thereby improving perceived audio quality while maintaining low bitrate
Solution Approach 2:
The patent implements a feedback mechanism where the decoded audio signal is analyzed to detect artifacts, and this detection information is fed back to control post-processing operations that adjust filtering and bandwidth extension parameters dynamically based on the detected artifact types and severity, creating a closed-loop system that continuously optimizes audio quality
2Length of stationary object
If spectral band replication is applied to extend bandwidth at low bitrate, then audio frequency range is improved, but artifact introduction increases due to overestimation of tonal-to-noise ratio
Solution Approach 1:
The patent performs preliminary detection of SBR processing and artifact characteristics before applying post-processing enhancement. By detecting the presence of tonal spikes and beating artifacts in advance, the system can pre-configurate appropriate filtering parameters and bandwidth extension settings to mitigate these artifacts before they become perceptually problematic
Solution Approach 2:
The patent dynamically adjusts processing parameters based on detected signal characteristics. When tonal spikes or beating artifacts are detected, the system modifies filtering parameters, bandwidth extension gain, and spectral replication parameters to reduce the energy of problematic tonal components while preserving natural audio content, thereby eliminating metallic sounds and roughness
3Productivity
If static low-pass filtering is applied to limit audio bandwidth at low bitrate, then coding efficiency is improved, but sound quality deteriorates due to dull and muffled sound impression
Solution Approach 1:
The patent replaces static low-pass filtering with dynamic bandwidth extension that adapts to the input signal characteristics. The system detects the presence and severity of bandwidth limitation artifacts, then dynamically adjusts the cutoff frequency and gain of post-processing filters to restore natural high-frequency content while maintaining coding efficiency, preventing the dull and muffled sound impression caused by fixed bandwidth limitation
4Object-affected harmful factors
If post-processing enhancement is applied to reduce artifacts, then audio quality is improved, but device complexity increases due to additional processing steps
Solution Approach 1:
The patent extracts only the essential artifact detection and mitigation functions from a full audio processing system. By focusing specifically on detecting major artifact types (birdies, tonal spikes, beating) and applying targeted post-processing operations for each, the system achieves significant artifact reduction without implementing complete audio enhancement pipelines, thereby limiting complexity growth while maintaining effective artifact suppression
Data Source
AI summary
An apparatus for processing an audio signal includes a separator for separating a first portion of a spectrum of the audio signal from a second portion of the spectrum of the audio signal, the first portion having a first signal characteristic and the second portion having a second signal characteristic. The apparatus includes a first bandwidth extender for extending a bandwidth of the first portion using first parameters associated with the first signal characteristic, for obtaining a first extended portion and includes a second bandwidth extender for extending a bandwidth of the second portion using second parameters associated with the second signal characteristic, for obtaining a second extended portion. The apparatus includes a combiner configured for using the first extended portion and the second extended portion for obtaining an extended combined audio signal.