High-Frequency Audio Encoding Using Low-Band Signal Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio signal encoding and decoding technologies are inefficient in handling high frequency signals, as they require significant bits for accurate representation, which is not necessary for human perception, and do not effectively utilize the relationship between low and high frequency signals.
Innovation Solution
The method involves linear prediction of high frequency signals to extract coefficients, generating signals using these coefficients and low frequency signals, and calculating energy ratios to encode and decode high frequency signals efficiently, thereby reducing the bit requirement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional encoding methods are used for high frequency signals, then encoding accuracy is maintained, but bit consumption increases significantly
Solution Approach 1:
The patent extracts only the essential characteristics of high frequency signals (energy values and spectral shape information) rather than encoding the complete signal. By separating the high frequency signal representation into minimal parameters that capture perceptually important features, bit consumption is dramatically reduced while maintaining encoding quality for human perception.
Solution Approach 2:
The patent applies partial encoding by focusing only on the most perceptually relevant aspects of high frequency signals. Instead of fully encoding all high frequency components, it selectively encodes energy values and spectral characteristics that matter most to human hearing, achieving acceptable quality with fewer bits.
2Quantity of substance
If spectral band replication is used to encode high frequency signals, then bit consumption is reduced, but utilization of low-high frequency relationship is insufficient
Solution Approach 1:
The patent uses feedback by deriving high frequency signal characteristics from the encoded low frequency signal. The spectral shape information and energy relationships are extracted from the low frequency portion and applied to reconstruct the high frequency components, creating an adaptive system that leverages the correlation between frequency bands.
Solution Approach 2:
The low frequency signal serves multiple functions: it is both decoded for audio output and used as a source for generating high frequency signal characteristics. This multi-functional approach allows the same low frequency data to drive both the bass response and inform the reconstruction of high frequency content, improving overall system efficiency.
3Measurement precision
If full high frequency signal encoding is performed, then audio quality is maintained, but compression efficiency decreases
Solution Approach 1:
The patent changes the representation parameters of high frequency signals from time-domain samples to frequency-domain characteristics (energy values, spectral shape). By transforming the signal into different parameter space that captures essential perceptual information more efficiently, compression efficiency is improved while maintaining audible quality.
Solution Approach 2:
The patent applies different encoding strategies to different frequency regions. High frequency components are encoded with simplified parameters focused on energy and spectral shape, while low frequency components receive more detailed encoding. This localized approach optimizes compression for each frequency band according to its perceptual importance and characteristics.
Data Source
AI summary
Provided are a method and apparatus for encoding and decoding a high frequency signal by using a low frequency signal. The high frequency signal can be encoded by extracting a coefficient by linear predicting a high frequency signal, and encoding the coefficient, generating a signal by using the extracted coefficient and a low frequency signal, and encoding the high frequency signal by calculating a ratio between the high frequency signal and an energy value of the generated signal. Also, the high frequency signal can be decoded by decoding a coefficient, which is extracted by linear predicting a high frequency signal, and a low frequency signal, and generating a signal by using the decoded coefficient and the decoded low frequency signal, and adjusting the generated signal by decoding a ratio between the generated signal and an energy value of the high frequency signal.