Bandwidth-Extended Signal Generation Using Phase Vocoder and Value Copier
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional bandwidth extension methods, such as spectral band replication, often result in auditory artifacts and high computational complexity, especially when copying sinusoids near the border of low-frequency and high-frequency parts, and can lead to sparse spectra with high stretching factors, causing undesirable audio artifacts.
Innovation Solution
The use of a phase vocoder to generate a harmonically transposed version of the low-frequency signal content and a value copier to create a non-harmonically frequency-shifted version of the high-frequency signal content, reducing computational complexity and avoiding sparse spectral holes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If spectral band replication with QMF patching is used to generate high-frequency content, then bandwidth extension is achieved, but auditory artifacts occur particularly when sinusoids are copied near the border of low-frequency and high-frequency parts
Solution Approach 1:
The frequency spectrum is divided into multiple patches or bands. Instead of uniformly copying the entire low-frequency spectrum to high-frequency regions, the method segments the spectrum and selectively processes different patches. This segmentation allows avoidance of artifact-prone border regions while preserving quality in safe zones.
Solution Approach 2:
Different processing strategies are applied to different frequency regions. The patent identifies specific frequency ranges where patching causes artifacts and applies alternative methods (such as noise shaping or selective copying) only in those problematic local regions, while using standard patching in regions where it produces good results.
2Reliability
If conventional spectral band replication methods are used, then high-frequency content is generated, but computational complexity becomes comparatively high
Solution Approach 1:
The patent extracts only the essential spectral envelope information and selected spectral components from the low-frequency signal, rather than processing the complete spectral representation. This extraction approach reduces the amount of data that needs to be manipulated and copied, thereby lowering computational complexity while maintaining the bandwidth extension function.
3Productivity
If high stretching factors are applied in bandwidth extension, then the spectrum becomes very sparse, but this results in undesired audible audio artifacts
Solution Approach 1:
Before applying high stretching factors that would create sparse spectra, the method performs preliminary spectral envelope formatting and pre-processing of the low-frequency signal. This preliminary action ensures that the essential spectral characteristics are preserved and interpolated appropriately, so that even when stretching creates gaps, the perceptual quality remains acceptable.
Data Source
AI summary
An apparatus for generating a representation of a bandwidth-extended signal on the basis of an input signal representation includes a phase vocoder configured to obtain values of a spectral domain representation of a first patch of the bandwidth-extended signal on the basis of the input signal representation. The apparatus also includes a value copier configured to copy a set of values of the spectral domain representation of the first patch, which values are provided by the phase vocoder, to obtain a set of values of a spectral domain representation of a second patch, wherein the second patch is associated with higher frequencies than the first patch. The apparatus is configured to obtain the representation of the bandwidth-extended signal using the values of the spectral domain representation of the first patch and the values of the spectral domain representation of the second patch.