High-Frequency Audio Encoding With Tonal Component Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio codecs face challenges in efficiently determining which tonal components in high frequency bands to encode, leading to redundancy and suboptimal coding quality due to limited coding bits.
Innovation Solution
An audio encoding method that adjusts the spectrum of high frequency band signals based on tonal component information, such as quantity, location, or existence, to optimize encoding and reduce redundancy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If bandwidth extension encoding is performed on high frequency band signal based on low frequency band signal, then coding bit rate is reduced, but tonal components in high frequency band are not accurately represented
Solution Approach 1:
The high frequency band is divided into multiple tiles, and tonal component detection is performed separately for each tile. This segmentation allows selective encoding of tonal components only in regions where they exist, rather than applying uniform encoding across the entire high frequency band, thus improving tonal representation accuracy while maintaining low bit rate.
Solution Approach 2:
Different encoding strategies are applied to different regions (tiles) of the high frequency band based on local characteristics. Tiles containing tonal components receive enhanced encoding treatment, while tiles without tonal components use standard bandwidth extension encoding. This local differentiation resolves the contradiction by improving tonal accuracy only where needed.
2Manufacturing precision
If tonal component encoding is added to high frequency band encoding, then coding quality is improved, but coding bit rate increases
Solution Approach 1:
Instead of encoding all tonal components in the entire high frequency band, the method performs partial encoding by detecting and encoding tonal components only in specific tiles where they are present. This partial action approach improves coding quality for tonal regions while avoiding the bit rate increase that would result from encoding the entire high frequency band.
Solution Approach 2:
The encoding process dynamically changes parameters based on detected tonal components. When tonal components are detected in a tile, the encoder switches to a more precise encoding mode for that region, using additional bits only where tonal accuracy is required, thus improving overall coding quality without proportionally increasing total bit rate.
3Productivity
If spectrum adjustment is performed based on tonal component information, then encoding efficiency is improved, but processing complexity increases
Solution Approach 1:
The high frequency band is divided into multiple tiles for independent processing. Tonal component detection and spectrum adjustment are performed on a per-tile basis, which localizes the computational complexity to small regions rather than requiring complex processing across the entire frequency band, thus improving encoding efficiency without excessive complexity increase.
Solution Approach 2:
Tonal component detection is performed as a preliminary step before spectrum adjustment and encoding. By identifying tonal components in advance and marking their locations, the subsequent spectrum adjustment process can be simplified and optimized, improving overall encoding efficiency while keeping the complexity of each individual processing stage manageable.
Data Source
AI summary
This application provides an audio encoding method and an encoding device. The method includes: obtaining a current frame includes a high frequency band signal and a low frequency band signal; performing first encoding on the high frequency band signal and the low frequency band signal, to obtain a first encoding parameter; performing second encoding on the high frequency band signal to obtain a second encoding parameter which indicates information about a tonal component of the high frequency band signal; adjusting, based on the information, a spectrum of a high frequency band signal obtained through bandwidth extension, to obtain an adjusted spectrum of the high frequency band signal; performing third encoding based on the adjusted spectrum of the high frequency band signal to obtain a third encoding parameter.