Overlapping Subband Audio Processing for Low-Aliasing Decimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current audio signal processing techniques face challenges in efficiently handling increasing bandwidths due to computational resource constraints, particularly in subband processing, where aliasing issues arise from non-ideal filtering, leading to degraded performance and increased complexity.
Innovation Solution
An audio signal processing apparatus that employs a filter bank with overlapping sub-filters to generate subbands, applies a frequency shift to at least one subband, and decimates the signals, ensuring the decimated sampling frequency is at least twice the bandwidth of each subfilter, thereby reducing aliasing and allowing for adaptive processing tailored to specific frequency intervals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If subband processing is applied to reduce computational resource usage, then processing efficiency is improved, but aliasing issues arise from non-ideal filtering
Solution Approach 1:
The audio signal is divided into multiple subbands using a filter bank, allowing independent processing of each subband. This segmentation enables reduced computational resource usage while managing aliasing through separate processing paths for each subband component.
Solution Approach 2:
A frequency shifter is introduced as an intermediary component between the filter bank and decimator. This frequency shifter modifies the spectral content of subband signals to prevent aliasing overlap, acting as a mediator that resolves the conflict between efficient decimation and aliasing prevention.
2Measurement precision
If the audio signal bandwidth is increased, then audio quality is improved, but computational resource usage increases substantially
Solution Approach 1:
The wideband audio signal is segmented into multiple subbands, each processed independently at lower computational cost. This allows the system to handle increased total bandwidth while maintaining reduced computational resource usage per processing unit.
Solution Approach 2:
Different processing parameters and filter characteristics are applied to different subbands based on their specific frequency characteristics. This local optimization allows efficient processing of each subband while maintaining overall high audio quality across the full bandwidth.
3Device complexity
If non-ideal filters are used in the filter bank, then device complexity is reduced, but subband aliasing increases
Solution Approach 1:
The frequency shifter serves as an intermediary that compensates for the aliasing introduced by simple non-ideal filters. By shifting frequencies before decimation, it prevents aliasing overlap without requiring complex ideal filters, thus maintaining low device complexity while reducing aliasing.
Solution Approach 2:
The approach converts the potential harm of aliasing from non-ideal filters into a manageable effect by using frequency shifting. The aliasing that would normally occur is transformed into a predictable frequency translation that can be controlled and compensated for in the synthesis stage.
4Productivity
If decimation is applied to reduce sampling frequency, then computational efficiency is improved, but aliasing artifacts are introduced
Solution Approach 1:
The frequency shifter is positioned as an intermediary between filtering and decimation, modifying the spectral content to prevent aliasing artifacts during the decimation process. This allows efficient downsampling while maintaining signal integrity.
Solution Approach 2:
Frequency shifting is applied as a preliminary action before decimation, preparing the signal in advance to prevent aliasing artifacts. This preliminary modification of the signal spectrum ensures that when decimation occurs, no harmful aliasing artifacts are introduced.
Data Source
AI summary
An audio signal processing apparatus comprises a receiver (403) receiving an audio signal sampled at a first sampling frequency, the audio signal having a maximum frequency below half the first sampling frequency by a first frequency margin. A filter bank (405) generates subband signals for the digital audio signal using overlapping sub-filters. A first frequency shifter (407) applies a frequency shift to at least one subband of the set of subbands and a decimator (409) decimates the subband signals by a decimation factor resulting in a decimated sampling frequency being at least twice a bandwidth of each of the overlapping sub-filters. The frequency shift for a subband is arranged to shift the subband to a frequency interval being a multiple of a frequency interval from zero to half the decimated sample frequency. The subband may be individually processed and the processed subbands may subsequently be combined to generate a full band output signal.