Audio Coding Method Linear Predictive Parameter Modification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio coding methods struggle to increase audio bandwidth while maintaining a constant or slightly increased bit rate, leading to discontinuity in the spectrum between audio frames.
Innovation Solution
An audio coding method that determines a first or second modification weight based on the LSF differences between audio frames, modifying the linear predictive parameter accordingly to ensure a steady spectrum between frames, thereby allowing for wider bandwidth audio coding without significantly increasing the bit rate.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Length of moving object
If conventional coding manner is used to increase audio bandwidth, then audio bandwidth is improved, but bit rate greatly increases
Solution Approach 1:
The patent applies bandwidth extension technology where the high-band audio signal is generated by nesting the low-band signal through linear prediction. The high-band components are predicted based on the low-band components, allowing the bandwidth to be extended without proportionally increasing the bit rate, as the prediction coefficients are derived from the lower-rate low-band signal.
Solution Approach 2:
The patent changes the parameter representation by using linear predictive coding parameters (A0, A1, A2 coefficients) to represent the high-band audio signal. Instead of transmitting the full high-band signal directly, the system transmits prediction coefficients that can be applied to generate the high-band signal from the low-band signal, effectively changing the representation parameters to reduce bit rate requirements.
2Length of moving object
If linear predictive algorithm is used to calculate LPC coefficient, then bandwidth extension is achieved, but spectrum discontinuity occurs between audio frames
Solution Approach 1:
The patent introduces a feedback mechanism where the decoded low-band audio signal is fed back into the linear predictive coding process. The feedback coefficient is calculated based on the correlation between the current frame and previous frame, allowing the system to adjust and refine the prediction continuously. This feedback loop helps maintain spectrum continuity by correcting discontinuities that would otherwise occur between audio frames.
Solution Approach 2:
The patent makes the linear predictive coding parameters dynamic by calculating different coefficients for different audio frames based on the actual signal characteristics. Instead of using fixed parameters, the system adapts the LPC coefficients frame-by-frame, allowing smooth transitions between frames and maintaining spectrum continuity while still achieving bandwidth extension.
Data Source
AI summary
A method comprises determining a first modification weight according to linear spectral frequency (LSF) differences of the current frame and LSF differences of a previous frame of the current frame, when a signal characteristic of the current frame meets a preset modification condition, modifying the linear predictive parameter of the current frame according to the determined first modification weight, and coding the current frame according to the modified linear predictive parameter.