Bandwidth Extension Encoder Energy Prediction Offset
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing bandwidth extension techniques in audio coding struggle with optimal high-frequency (HF) reconstruction, especially when supporting different bit rates, leading to suboptimal energy ratios between low-frequency (LF) and HF excitation.
Innovation Solution
The proposed solution involves an encoder that performs energy prediction for the extended-band signal using linear prediction coefficients (LPC) and encodes the residual with an offset dependent on the bit rate, allowing for flexible bit rate support and efficient energy coding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single energy coding scheme is used for different bit rates, then the coding structure is simplified, but the energy ratio between LF and HF excitation becomes suboptimal for certain bit rates
Solution Approach 1:
The patent applies dynamics by making the energy coding scheme adaptive to different bit rates. Instead of using a fixed coding scheme, the system dynamically selects or adjusts coding parameters based on the actual bit rate being used, allowing optimal energy ratio performance across multiple bit rates while maintaining a unified coding framework
Solution Approach 2:
The patent changes parameters by introducing bit rate-dependent offset values in the energy prediction. By adjusting the offset parameter based on the specific bit rate, the system optimizes the energy ratio between LF and HF excitation for each bit rate scenario, resolving the contradiction between coding simplicity and energy ratio accuracy
2Quantity of substance
If bandwidth extension transmits no or very few additional parameters, then the bit rate overhead is reduced, but the high-frequency reconstruction quality may be compromised
Solution Approach 1:
The patent uses copying by deriving HF excitation parameters from LF excitation parameters through prediction. Instead of transmitting independent HF parameters, the system copies the LF excitation structure and predicts HF characteristics based on the energy ratio, significantly reducing the number of transmitted parameters while maintaining reconstruction quality
Solution Approach 2:
The patent implements feedback by using the transmitted energy ratio information to adjust the HF excitation generation at the decoder. The decoded energy ratio serves as feedback that guides the reconstruction process, ensuring that the HF signal accurately reflects the original energy distribution even with minimal transmitted parameters
3Productivity
If the energy ratio between LF and HF excitation is optimized for one bit rate, then the coding efficiency is improved for that bit rate, but different bit rates cannot be supported optimally
Solution Approach 1:
The patent achieves universality by designing a single energy coding framework that can handle multiple bit rates. By incorporating bit rate-dependent offset adjustment, the unified coding scheme maintains optimal coding efficiency across different bit rates, eliminating the need for separate optimization for each bit rate
Solution Approach 2:
The patent applies dynamics by making the energy coding adaptive to different bit rates through offset adjustment. The system dynamically adjusts the energy prediction offset based on the operating bit rate, allowing the same coding scheme to achieve optimal efficiency across multiple bit rate scenarios
Data Source
Figure 1A~1B
Figure 1C
Figure 2~3
AI summary
Encoder (10) for coding a signal comprising a band-limited signal and an extended-band signal, the encoder (10) comprising: a calculator (12) configured to perform energy prediction of the extended-band signal based on LPC coefficients; and a coder (14) configured to encode a residual of the signal using the energy prediction and an offset (o); wherein the offset (o) is dependent on a bit-rate.