Pitch-Based Pre-Filtering for Audio Codec Noise Shaping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional low-delay transform audio codecs struggle to maintain audio quality due to insufficient frequency resolution of pitch harmonics in periodic signals, leading to audible distortion, despite efforts to reduce coding delay and complexity.
Innovation Solution
The implementation of a pitch-based pre-filter and post-filter system that adaptively filters audio signals based on estimated pitch periods to suppress or boost pitch harmonic peaks and spectral valleys, reducing distortion without increasing coding delay or significantly raising codec complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large transform window size (20-50 ms) is used in traditional audio codecs, then frequency resolution is improved and pitch harmonics can be resolved, but coding delay increases
Solution Approach 1:
The patent divides the audio signal processing into two separate stages: a low-delay transform codec handling time-critical processing with small window sizes, and a pitch-based post-filter handling frequency-resolution requirements. This segmentation allows each component to optimize for its specific function without compromise.
Solution Approach 2:
The pitch-based post-filter operates as a preliminary enhancement stage that processes the output of the low-delay codec. By applying pitch-adaptive filtering after the main decoding, the system pre-compensates for the frequency resolution limitations of the small-window transform, effectively resolving pitch harmonics without increasing coding delay.
2Loss of time
If a small transform window size (<10 ms) is used to achieve low coding delay, then coding delay is reduced, but frequency resolution becomes insufficient to resolve pitch harmonics
Solution Approach 1:
The pitch-based post-filter acts as an intermediary component between the low-delay transform codec and the final audio output. It mediates the frequency resolution deficiency by applying adaptive filtering that specifically targets pitch harmonic structures, effectively bridging the gap between limited transform resolution and required pitch analysis capability.
Solution Approach 2:
The system dynamically changes filtering parameters based on detected pitch period. By adapting the post-filter characteristics to match the actual pitch content of the signal, the system achieves effective pitch harmonic resolution without requiring fixed large transform windows, thus maintaining low coding delay while improving frequency resolution where needed.
3Manufacturing precision
If transform window size is increased to resolve pitch harmonics, then audio quality for periodic signals is improved, but codec complexity increases
Solution Approach 1:
The pitch-based post-filter employs dynamic parameter adaptation where filtering characteristics change based on real-time pitch detection. This allows the system to apply complex pitch-resolving filtering only when periodic signals are detected, rather than always using large transform windows, thus improving audio quality for periodic signals while maintaining lower average codec complexity.
Solution Approach 2:
Instead of applying full large-window transform processing to all signals, the system applies pitch-based post-filtering selectively and partially - only enhancing frequency resolution in the specific frequency regions where pitch harmonics are present. This partial action achieves the necessary audio quality improvement without the full complexity cost of large-window transforms across the entire spectrum.
Data Source
AI summary
Systems and methods for enhancing the quality of an audio signal produced by an audio codec are described herein. In accordance with the systems and methods, a pitch-based pre-filter adaptively filters an input audio signal to produce a filtered audio signal. An audio encoder encodes the filtered audio signal to generate a compressed audio bit stream. An audio decoder decodes the compressed audio bit stream to generate a decoded audio signal. A pitch-based post-filter adaptively filters the decoded audio signal to produce an output audio signal, wherein adaptively filtering the decoded audio signal comprises undoing at least part of a signal-shaping effect of the pitch-based pre-filter.