Spatial Audio Reproduction Using Subspace Analysis and Source Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current sound field reproduction technologies face limitations in efficiently reproducing spatial audio on available loudspeaker setups due to the need for a large number of loudspeakers at impractical positions and high costs, leading to reduced quality and intellectual property protection issues, especially in low-bandwidth applications and environments where spatial aliasing and localization errors occur.
Innovation Solution
A method and device that define a reproducible subspace based on the capabilities of the loudspeaker setup, perform spatial analysis in the spherical harmonics domain to extract localizable sources, and map remaining signals within this subspace for reproduction, using techniques like direction of arrival estimation and beamforming to enhance sound field reproduction precision and reduce the influence of strong interferers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large number of loudspeakers are used to reproduce complex sound scenes, then the spatial accuracy and quality of sound field reproduction is improved, but the device complexity and cost increase significantly
Solution Approach 1:
The patent extracts and separates localizable sound sources from the overall sound field using spatial analysis in the spherical harmonics domain. By identifying and isolating specific sound sources with defined positions, the system can reproduce these extracted sources using fewer loudspeakers while maintaining spatial accuracy, thus resolving the contradiction between spatial precision and system complexity
Solution Approach 2:
The sound field is segmented into different components: localizable sources and remaining sound field components. This segmentation allows the system to apply different reproduction strategies to each component, optimizing the use of available loudspeakers and reducing the total number required while preserving spatial accuracy for the extracted sources
2Loss of information
If spatial analysis is performed on the entire sound field, then complete spatial information is captured, but the influence of strong interferers and spatial aliasing increases
Solution Approach 1:
The patent applies local quality by performing spatial analysis specifically within the reproducible subspace rather than the entire sound field. By focusing analysis on the subspace where accurate reproduction is possible and mapping sources to the available loudspeaker configuration, the system captures necessary spatial information while avoiding the harmful effects of spatial aliasing and strong interferers that exist in the full sound field
3Ease of manufacture
If the sound field is reproduced on a limited loudspeaker setup, then the system is more practical and cost-effective, but the listening area and spatial performance are reduced
Solution Approach 1:
The patent transitions from reproducing the sound field directly in the spatial domain to representing it in the spherical harmonics domain. This dimensional transformation allows the system to analyze, extract, and map sound sources to a limited loudspeaker setup while preserving spatial characteristics, thereby maintaining practicality with fewer loudspeakers while extending the effective listening area through optimized spatial distribution
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
The invention relates to a method and a device for sound field reproduction into a listening area (5) of spatially encoded first audio input signals (1) according to sound field description data (2) using an ensemble of physical loudspeakers (3). The method comprises the steps of computing reproduction subspace description data (8) from loudspeaker positioning data (4) describing the subspace in which virtual sources can be reproduced with the physically available setup. Then, second (10) and third (12) audio input signals with associated sound field description data (11) (13) wherein second audio input signals (10) comprise spatial components of the first audio input signals (1) located within the reproducible subspace (6) and third audio input signals (12) comprise spatial components of the first audio input signals (1) located outside of the reproducible subspace (6). A spatial analysis is performed on second audio input signals (10) so as to extract fourth audio input signals (15) corresponding to localizable sources within the reproducible subspace (5) with associated source positioning data (13). Remaining components of second audio input signals (10) after spatial analysis are merged with third audio input signals (12) into fifth audio input signals (17) with associated sound field description data (18) for reproduction within the reproducible subspace (5). Finally loudspeaker alimentation signals (20) are computed from fourth (15) and fifth (17) audio input signals according to loudspeaker positioning data (4), localizable sources positioning data (16) and sound field description data (18).