High-Frequency Audio Bandwidth Extension With Adaptive Core Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for coding and decoding audio signals, particularly in the high-frequency band, face inefficiencies due to limited bit allocation, which affects the quality of spectral band replication (SBR) and hearing characteristics of humans.
Innovation Solution
A method involving a coding apparatus and decoding apparatus that generate and reconstruct audio signals by extending the bandwidth of the high-frequency band using a base signal derived from the low-frequency signal, employing techniques like CELP and MDCT for efficient bit allocation and energy quantization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a large number of bits are assigned to the low-frequency band to increase coding efficiency, then coding efficiency is improved, but the high-frequency band receives relatively few bits which degrades the quality of high-frequency signal representation
Solution Approach 1:
The patent applies spectral band replication by copying the low-frequency spectrum to generate the high-frequency signal. The low-frequency band signal is upsampled and used as a base signal, then energy information from the original high-frequency signal is copied and applied to the replicated spectrum, allowing high-quality high-frequency reconstruction with minimal bit allocation.
Solution Approach 2:
The patent segments the audio signal processing into distinct frequency bands (low-frequency band and high-frequency band) with different coding strategies. The low-frequency band is coded with high precision using available bits, while the high-frequency band uses a replication-based approach that requires fewer bits, optimizing overall bit allocation across frequency segments.
2Manufacturing precision
If spectral band replication is used to extend high-frequency bandwidth, then high-frequency signal quality is improved, but the complexity of the encoding and decoding process increases
Solution Approach 1:
The low-frequency signal serves multiple purposes: it is both the audio output for the low-frequency band and the source material for generating the high-frequency signal through upsampling and spectral replication. This self-service approach eliminates the need for separate high-frequency encoding, reducing overall system complexity while maintaining high-frequency quality.
Solution Approach 2:
The high-frequency base signal is generated in advance by upsampling the low-frequency signal before the actual audio coding process. This preliminary action prepares the high-frequency component that will later have energy information applied to it, streamlining the encoding process by pre-computing what would otherwise require complex real-time processing.
3Productivity
If limited bits are allocated to the high-frequency band, then overall bit usage is optimized, but the fidelity of the high-frequency audio output deteriorates
Solution Approach 1:
The patent applies different quality levels to different frequency bands according to human hearing characteristics. The low-frequency band receives high-quality coding with sufficient bits to preserve detail, while the high-frequency band uses a more efficient replication-based approach. This local quality differentiation optimizes overall bit usage while maintaining perceptually adequate fidelity across the full frequency range.
Data Source
Figure 1~2B
Figure 2C~2D
Figure 3~4
AI summary
Disclosed are an apparatus and a method for encoding/decoding for high-frequency bandwidth extension. The coding method comprises determining a core coding mode of a low-frequency signal of an input signal, based on characteristics of the low-frequency signal of the input signal, quantizing a linear prediction coefficient (LPC) from the low-frequency signal of the input signal. The method includes performing code excited linear prediction (CELP) coding on an LPC excitation signal of the low-frequency signal of the input signal when the core coding mode of the low-frequency signal of the input signal is determined to be a CELP coding mode. The method includes performing time-domain (TD) extension coding on a high-frequency signal of the input signal when the CELP coding is performed on the LPC excitation signal. The method includes performing audio coding on the LPC excitation signal when the core coding mode of the low-frequency signal of the input signal is determined to be an audio coding mode. The method includes performing frequency-domain (FD) extension coding on the high-frequency signal of the input signal when the audio coding is performed on the LPC excitation signal.