Spatial Audio Rendering Using Real-to-Virtual Acoustic Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electronic devices render spatial audio signals with low quality and diminished immersion due to discrepancies between computed and real-world acoustic responses.
Innovation Solution
An audio processing method that adjusts acoustic features in virtual scenes based on conversion information derived from real scenes, aligning virtual acoustic features with real-world counterparts to enhance spatial audio quality and immersion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If acoustic feature extraction is performed based on virtual scene modeling, then spatial audio rendering can be achieved, but the acoustic features differ from real-world acoustic features resulting in low quality and diminished immersion
Solution Approach 1:
The patent introduces conversion information as an intermediary element that bridges the virtual scene acoustic features and real-world acoustic features. The conversion information contains correspondence relationships between acoustic features at the same position in virtual and real scenes, allowing the rendering device to adjust virtual acoustic features to better match real-world characteristics, thereby improving spatial audio quality and immersion without requiring direct access to real-world measurements at every position
Solution Approach 2:
The patent modifies acoustic feature parameters by adjusting them based on conversion information. The rendering device changes the acoustic feature parameters (such as reverberation time, early decay time, clarity) to align virtual scene characteristics with real-world characteristics, resolving the discrepancy between computed and real-world acoustic responses
2Measurement precision
If conversion information is used to adjust acoustic features at any position in the virtual environment, then acoustic feature accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent performs preliminary action by pre-establishing the conversion information that contains correspondence relationships between acoustic features at the same position in virtual and real scenes. This conversion information can be pre-computed or pre-measured and stored, allowing efficient adjustment of acoustic features during runtime without requiring complex real-time calculations for each position change
Data Source
Figure 1A~1B
Figure 1C~1D
Figure 2
AI summary
Embodiments of this application provide an audio processing method and an electronic device. The method includes: First, a source audio signal, a sound source position corresponding to the source audio signal in a virtual scene, and a first position corresponding to a user in the virtual scene are obtained, where the virtual scene is obtained through construction based on a real scene; and a first acoustic feature is obtained, where the first acoustic feature is obtained by converting, based on an acoustic feature at a second position in the real scene and an acoustic feature at the second position in the virtual scene, an acoustic feature at the first position in the virtual scene when the source audio signal is at the sound source position. Then, a spatial audio signal is generated based on the first acoustic feature and the source audio signal. In this way, the user can experience a spatial audio signal with high quality and a heightened sense of immersion.