Adaptive Sibilance Detection for Preserving Audio Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio signal processing systems fail to effectively distinguish between desirable short-term and long-term features and excessive sibilance, leading to degraded audio quality due to harshness caused by high energy in the 4 kHz to 12 kHz frequency range.
Innovation Solution
A system that adapts sibilance detection by using supervised or unsupervised machine learning-based classifiers to identify short-term features like impulsive and flat fricative sounds, and adjusts parameters to suppress sibilance using a multiband compressor, preserving desirable audio content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If traditional sibilance detection is used to suppress high energy in the 4 kHz to 12 kHz frequency range, then harshness is reduced, but desirable short-term and long-term features are also suppressed, degrading audio quality
Solution Approach 1:
The patent segments the audio signal analysis into short-term features (impulsive sounds, flat fricative sounds) and long-term features (smoothed spectrum balance), allowing the system to distinguish between desirable features and excessive sibilance by analyzing different temporal and spectral characteristics separately
Solution Approach 2:
The system dynamically adapts sibilance detection parameters based on the detected presence of short-term and long-term features. When these features are detected, the sibilance detection parameters are adjusted to avoid suppressing them, making the suppression process adaptive rather than static
2Reliability
If sibilance detection parameters are adapted to preserve short-term and long-term features, then audio quality is maintained, but the complexity of the detection system increases
Solution Approach 1:
The system performs preliminary detection of short-term features (impulsive sounds, flat fricative sounds) and long-term features (smoothed spectrum balance) before conducting sibilance detection. This preliminary analysis allows the system to pre-adapt the sibilance detection parameters to avoid suppressing desirable features, streamlining the overall processing
3Object-affected harmful factors
If aggressive sibilance suppression is applied to low-fidelity devices, then harshness is reduced, but the poor frequency response of low-quality microphones and speakers is exacerbated
Solution Approach 1:
The system changes the detection parameters based on the characteristics of the audio signal and the device being used. For low-fidelity devices, the sibilance detection parameters are adapted to account for poor microphone frequency response and low-quality speakers, allowing for more nuanced suppression that doesn't exacerbate the inherent limitations of these devices
Data Source
AI summary
A method is disclosed herein for adapting parameters of a sibilance detector. Time-frequency features are extracted from an audio signal being received and. Based on those time-frequency features, a determination is made of whether the audio signal includes a short-term feature or a long-term feature. In accordance with determining that the audio signal includes the short-term feature or the long-term feature, one or more parameters of a sibilance detector for detecting sibilance in the audio signal are adapted. Sibilance in the audio signal, is detected using the sibilance detector with the one or more adapted parameters.