Asymmetric Audio Filterbank for Low-Delay Alias Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital filter banks face challenges in minimizing aliasing artifacts and achieving low system delay while maintaining near-perfect reconstruction properties, especially when modifying subband signals, due to limitations in prototype filter design and modulation techniques.
Innovation Solution
The development of an improved alias term minimization method for designing asymmetric prototype filters, specifically a 64-channel low delay complex-exponential modulated filter bank with a 640 coefficient prototype filter and 319 sample system delay, which optimizes for aliasing and pass band error rejection, using a composite objective function to iteratively determine filter coefficients.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If conventional cosine modulated filter banks are used, then implementation is simple and effective, but aliasing artifacts occur when subband signals are modified
Solution Approach 1:
The patent applies asymmetry by using asymmetric prototype filters instead of symmetric ones in the complex-exponential modulated filter bank. This asymmetric design allows for optimized aliasing suppression characteristics while maintaining implementation feasibility, directly addressing the contradiction between simple implementation and aliasing suppression.
Solution Approach 2:
The patent substitutes the conventional cosine modulation mechanism with complex-exponential modulation. This replacement transforms the filter bank from a real-valued system to a complex-valued system, enabling superior aliasing suppression properties while maintaining computational efficiency through the modified modulation approach.
2Object-affected harmful factors
If prototype filter length is increased to reduce aliasing, then aliasing suppression improves, but system delay increases
Solution Approach 1:
The patent changes the fundamental parameters of the filter bank by adopting complex-exponential modulation and asymmetric prototype filters. These parameter changes enable achieving low aliasing suppression with a relatively short prototype filter length (640 coefficients), thereby reducing system delay to 319 samples while maintaining effective aliasing suppression.
Solution Approach 2:
The patent introduces dynamics by optimizing the prototype filter coefficients iteratively using a composite objective function that balances aliasing suppression and delay constraints. This dynamic optimization process allows the filter design to adaptively find the optimal balance between aliasing suppression performance and system delay.
3Productivity
If subband signals are modified for spectral envelope adjustment, then audio coding efficiency improves, but aliasing artifacts are generated
Solution Approach 1:
The patent replaces the conventional cosine modulated filter bank with a complex-exponential modulated filter bank. This substitution fundamentally changes the modulation mechanism, enabling the filter bank to suppress aliasing artifacts even when subband signals are modified for spectral envelope adjustment in audio coding applications.
Solution Approach 2:
The patent converts the potentially harmful aliasing artifacts into a design constraint that drives the optimization of the prototype filter coefficients. By incorporating aliasing suppression into the composite objective function, the design process transforms the aliasing problem into a beneficial optimization goal, resulting in a filter bank that actively suppresses aliasing while allowing subband modifications.
4Manufacturing precision
If near-perfect reconstruction is achieved through iterative coefficient optimization, then reconstruction quality improves, but design complexity increases
Solution Approach 1:
The patent changes the optimization parameters by using a composite objective function that combines multiple performance criteria (aliasing suppression, passband error, delay). This parameter transformation simplifies the design process by providing a unified optimization framework that automatically balances multiple competing requirements, reducing design complexity while achieving near-perfect reconstruction.
Solution Approach 2:
The patent implements feedback through the iterative optimization process where the composite objective function continuously evaluates the filter coefficients and provides guidance for improvement. This feedback mechanism systematically drives the coefficients toward optimal values that achieve near-perfect reconstruction, making the complex design process more manageable and systematic.
Data Source
AI summary
An apparatus and method are disclosed for processing an audio signal. The apparatus includes an input interface, a digital filterbank having an analysis part and a synthesis part, a first phase shifter, a spectral envelope adjuster, a second phase shifter, and an output interface. The first phase shifter and the second phase shifter reduce a complexity of the digital filterbank, which includes both analysis and synthesis filters that are complex-exponential modulated versions of a prototype filter.