QMF Audio Codec Filter Banks for Variable Sampling Rates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current audio processing technologies face challenges in efficiently resampling audio signals to accommodate different sampling rates, particularly at low bit rates, which requires additional computational resources and is not compatible with low-cost hardware, limiting audio bandwidth and increasing complexity.
Innovation Solution
A configurable audio signal processor with an analysis and synthesis filter bank system that adjusts the sampling rate by modifying the frequency domain representation, reducing the complexity of resampling through bandwidth changes and eliminating additional signal distortion, allowing for flexible sampling rate adaptation without the need for high-complexity resamplers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If additional resampling functional modules (interpolator, up sampler, down sampler) are used to adapt the output sampling rate, then the sampling rate compatibility is improved, but the computational resources and device complexity increase significantly
Solution Approach 1:
The patent merges the resampling function with the existing analysis and synthesis filter banks. The analysis filter bank decomposes the input signal into subbands, and the synthesis filter bank reconstructs the output signal, with the resampling ratio embedded in the filter bank structure itself. This eliminates the need for separate resampling modules (interpolator, up sampler, down sampler) and reduces computational complexity while maintaining sampling rate adaptability.
Solution Approach 2:
The analysis and synthesis filter banks serve multiple functions: they perform both signal decomposition/reconstruction and sampling rate conversion simultaneously. By configuring the filter banks with appropriate decimation and interpolation factors, the system can adapt to different sampling rates without requiring dedicated resampling hardware, thus reducing overall device complexity.
2Productivity
If the sampling rate is reduced at low bit rate operating points, then the bit rate efficiency is improved, but the audio bandwidth is limited to below the upper limit of human audible range
Solution Approach 1:
The patent changes the sampling rate parameter dynamically based on the bit rate operating point. At low bit rates, the system uses lower sampling rates (e.g., 16 kHz) to improve efficiency, while at higher bit rates, it transitions to higher sampling rates (e.g., 48 kHz) to expand audio bandwidth. This parameter adaptation allows the system to optimize between bit rate efficiency and audio quality according to the specific operating conditions.
Solution Approach 2:
The system dynamically adjusts the sampling rate and filter bank configuration based on the operating bit rate. The analysis and synthesis filter banks are reconfigured on-the-fly to match the desired output sampling rate, enabling smooth transitions between different audio quality levels and bandwidths without requiring separate fixed-rate processing paths.
3Loss of information
If uncommon sampling rates (not integer multiples of 16000Hz or 22050Hz) are employed to increase audio bandwidth, then the audio quality is improved, but the compatibility with low-cost hardware D/A converters is lost
Solution Approach 1:
The system uses parameter-adaptive filter banks that can be configured with different decimation and interpolation factors to generate various output sampling rates. By selecting appropriate integer ratios for the filter bank configuration, the system can produce standard sampling rates (16 kHz, 22.05 kHz, 48 kHz) that are compatible with low-cost hardware, while still achieving high audio bandwidth through efficient signal processing at higher intermediate rates.
4Adaptability or versatility
If multiple different output sample rates are supported, then the adaptability to different audio sources is improved, but the complexity of the resampling task increases
Solution Approach 1:
The analysis and synthesis filter banks are designed to handle multiple sampling rate conversions through a single unified structure. By adjusting the filter bank order and decimation/interpolation factors, the same hardware or software module can efficiently convert between various sampling rates (16 kHz, 22.05 kHz, 48 kHz, and others) without requiring separate dedicated conversion paths for each rate pair, thus reducing overall resampling complexity.
Data Source
Figure 1
Figure 2A~2C
Figure 3A~3B
AI summary
An apparatus for processing an audio signal is provided. The apparatus comprises a configurable first audio signal processor (110) for processing the audio signal (s0) in accordance with different configuration settings (conf) to obtain a processed audio signal (s1), wherein the apparatus is adapted so that different configuration settings (conf) result in different sampling rates (sr1) of the processed audio signal (s1). The apparatus furthermore comprises n analysis filter bank (120) having a first number (c1) of analysis filter bank channels, a synthesis filter bank (130) having a second number (c2) of synthesis filter bank channels, a second audio processor (140) being adapted to receive and process an audio signal (s2) having a predetermined sampling rate (sr2), and a controller (150) for controlling the first number (c1) of analysis filter bank channels or the second number (c2) of synthesis filter bank channels in accordance with a configuration setting (conf).