Adaptive Bass Post-Filter Control for Speech and Music Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing bass post-filter in speech coders can degrade audio quality when not optimally controlled, particularly for signals with useful anti-harmonic components like music or noisy speech, as it either enhances or spoils the listening experience.
Innovation Solution
An adaptive bass post-filter control parameter is generated on the encoder-side using both encoded and decoded signals, and the original signal, which is then transmitted to the decoder to set a controllable bass post-filter characteristic, optimizing the attenuation of anti-harmonic noise using a minimum mean square error estimator.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the bass post-filter is activated to enhance clean speech quality, then speech clarity is improved, but audio quality deteriorates for music or noisy speech with useful anti-harmonic components
Solution Approach 1:
The bass post-filter is made dynamically controllable through adaptive control parameters (alpha, beta, gamma) that adjust the filter's attenuation characteristics based on signal analysis. The filter transitions from a static on/off state to a dynamic system that adapts its behavior to match the input signal type, enabling it to enhance speech while preserving music quality.
Solution Approach 2:
The invention changes the parameters controlling the bass post-filter from fixed pitch-based values to adaptive parameters (alpha, beta, gamma) that are optimized based on signal classification. By modifying these control parameters according to whether the input is speech or music, the system resolves the contradiction between speech enhancement and music preservation.
2Ease of operation
If the bass post-filter is controlled by pitch information only, then implementation is simple, but audio quality optimization is insufficient
Solution Approach 1:
The system introduces feedback mechanisms where the output of the bass post-filter is analyzed and used to adjust control parameters. The minimum mean square error estimator uses the difference between original and filtered signals to adaptively optimize parameters alpha, beta, and gamma, creating a closed-loop control system that improves audio quality while maintaining manageable complexity.
Solution Approach 2:
The invention replaces simple pitch-based mechanical control with a more sophisticated control mechanism based on minimum mean square error estimation. This substitution introduces signal analysis and adaptive parameter adjustment, achieving better audio quality optimization without excessive complexity increase.
3Object-affected harmful factors
If the bass post-filter is completely deactivated, then artifacts are avoided, but speech quality enhancement is lost
Solution Approach 1:
The system applies local quality control by treating speech and music signals differently through adaptive parameter adjustment. For speech segments, the bass post-filter parameters are optimized for maximum enhancement, while for music segments, parameters are adjusted to preserve anti-harmonic components. This localized optimization resolves the contradiction by applying different filter characteristics to different signal types within the same audio stream.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An Apparatus for processing an encoded signal (110), the encoded signal comprising an encoded audio signal having information on a pitch delay or a pitch gain, and a bass post-filter control parameter (101), comprises: an audio signal decoder (110) for decoding the encoded audio signal using the information on the pitch delay or the pitch gain to obtain a decoded audio signal (103); a controllable bass post-filter (112) for filtering the decoded audio signal (103) to obtain a processed signal (113), wherein the controllable bass post-filter (112) has the variable bass post-filter characteristic controllable by the bass post-filter control parameter (101); and a controller (114) for setting the variable bass post-filter characteristic in accordance with the bass post-filter control parameter (101) included in the encoded signal (102).