Virtual Speaker Audio Encoding for Lower HOA Data Volume
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high data volume and bandwidth requirements of higher order ambisonics (HOA) audio signals pose challenges in efficient transmission and storage, necessitating improved encoding and decoding methods to reduce data amount.
Innovation Solution
Selecting a first target virtual speaker from a preset virtual speaker set based on the current scene audio signal, generating a virtual speaker signal using attribute information, and encoding this signal to obtain a bitstream, rather than directly encoding the scene audio signal, thereby reducing data volume and improving encoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If HOA technology is used to record detailed sound scene information, then audio auditory effect is improved, but data volume increases
Solution Approach 1:
The patent extracts only the essential sound field information by selecting a target virtual speaker from a preset virtual speaker set based on the current scene audio signal. Instead of encoding all virtual speaker signals, only the signal from the selected target virtual speaker is encoded, thereby extracting the most important auditory information while discarding redundant data.
Solution Approach 2:
The patent segments the sound field information by representing it through individual virtual speaker signals rather than encoding the complete multi-channel HOA signal. The sound scene is divided into contributions from multiple virtual speakers, but only the dominant target virtual speaker is fully encoded, while others are represented more compactly.
2Reliability
If multi-channel encoding is used to encode each channel of audio signal, then encoding quality is maintained, but bandwidth occupation increases
Solution Approach 1:
The patent extracts and prioritizes the target virtual speaker signal that contributes most to the sound scene, encoding it with full quality using a core encoder. Other virtual speaker signals are represented more compactly, thereby maintaining overall encoding quality while reducing total bandwidth requirements.
Solution Approach 2:
The patent changes the encoding parameters dynamically by selecting different target virtual speakers based on the current scene audio signal. The encoding quality is adapted by adjusting which virtual speaker is selected as the target, allowing the system to maintain quality for the most important sound sources while using less bandwidth for less important ones.
3Measurement precision
If the quantity of channels increases to capture more sound field information, then spatial information accuracy is improved, but encoding complexity increases
Solution Approach 1:
The patent extracts the most significant spatial information by selecting a target virtual speaker that represents the dominant sound field characteristics. This approach captures the essential spatial accuracy needed for the current scene while avoiding the complexity of encoding all channel information.
Solution Approach 2:
The patent introduces dynamic selection of the target virtual speaker based on the current scene audio signal characteristics. The system adaptively changes which virtual speaker is selected as the target, allowing the encoding complexity to vary dynamically according to the scene requirements rather than always encoding all channels at full complexity.
Data Source
AI summary
An audio encoding and decoding method and apparatus, and a non-transitory readable storage medium are provided. The encoding method includes: selecting a first target virtual speaker from a preset virtual speaker set based on a current scene audio signal; generating a first virtual speaker signal based on the current scene audio signal and attribute information of the first target virtual speaker; and encoding the first virtual speaker signal to obtain a bitstream. According to the encoding method, an amount of encoded data is reduced, to improve encoding efficiency.