Audio Frame Loss Correction Using Overlap-Add Noise Injection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current frame loss correction techniques in digital signal processing, particularly in audio signal decoding, often result in audible phase discontinuities and unpleasant auditory artifacts due to temporal inconsistency and inadequate periodization, leading to metallic sounds and energy discrepancies.
Innovation Solution
A method that generates a replacement signal structure using spectral components from valid samples, injecting window-weighted blocks with partial overlap to smooth energy transitions and avoid periodicity, incorporating pseudo-random parameter variability to eliminate perceptible repetition and maintain constant signal energy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If frame loss correction is performed by repeating the last decoded frame, then the decoder complexity is reduced, but audible phase discontinuities and temporal inconsistency occur
Solution Approach 1:
The replacement signal is divided into multiple blocks that are extracted from the residue and injected at different positions with varying parameters. This segmentation prevents simple repetition patterns while maintaining computational efficiency, resolving the contradiction between low complexity and avoiding phase discontinuities.
Solution Approach 2:
The injection parameters (position, timing, scaling) are made dynamic and variable rather than static and repetitive. By introducing variability in how blocks are injected into the replacement signal, the system avoids temporal inconsistency and phase discontinuities while keeping the overall process computationally efficient.
2Loss of time
If simple frame repetition is used for frame loss correction, then processing time is reduced, but energy discrepancies and periodicity artifacts occur
Solution Approach 1:
Blocks are extracted from the residue in advance and stored for later injection. This preliminary extraction allows the system to quickly assemble replacement signals without extensive real-time processing, reducing processing time while the controlled injection process prevents energy discrepancies.
Solution Approach 2:
The parameters of injected blocks (position, timing, scaling factors) are varied to match the temporal and spectral characteristics of the original signal. This parameter variation eliminates periodicity artifacts and energy discrepancies while maintaining efficient processing through the use of pre-extracted blocks.
3Manufacturing precision
If weighted noise injection is applied to correct frame loss, then signal quality is improved, but processing complexity increases
Solution Approach 1:
Instead of generating complex noise signals, the system copies and reuses blocks from the residue (which is already available from previous processing). This copying approach maintains signal quality by preserving authentic signal characteristics while avoiding the computational complexity of generating weighted noise from scratch.
Solution Approach 2:
The system uses transient blocks extracted from the residue as disposable elements for frame replacement. These blocks are processed and injected quickly without requiring complex long-term storage or sophisticated noise generation algorithms, achieving good signal quality with minimal processing complexity.
4Loss of time
If overlap-add is used with short windows for low-delay coding, then latency is reduced, but audible artifacts increase due to insufficient overlap
Solution Approach 1:
Different regions of the replacement signal receive different treatment through selective block injection. The overlap regions are carefully managed with appropriate weighting to ensure smooth transitions, while the latency is kept low by using short windows. This local optimization resolves the contradiction between low latency and artifact reduction.
Data Source
AI summary
A method for processing a digital signal, implemented during decoding of the signal, in order to replace a succession of samples lost during decoding, the method comprising steps of: generating a structure of a signal for replacing the lost succession, this structure comprising spectral components determined from valid samples received during decoding before the succession of lost samples; generating a residue between a digital signal available to the decoder, comprising received valid samples, and a signal generated from the spectral components; and extracting blocks from the residue, method in which window weighted blocks are injected into the structure using an overlap-add approach, the injected blocks partially overlapping in time.