HRIR Filter Extraction for Low-Capacity Sound Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for localizing virtual stereo sound in 3D space require large capacity, high performance processors due to the increased number of head related impulse response (HRIR) filters and data, especially when multiple channels and listener movement are involved, making real-time interpolation challenging.
Innovation Solution
Extracting and using only important information from HRIR data, such as reflection sound waves, interaural level differences, and interaural time differences, to localize sound images with a small number of filter coefficients, enabling sound image localization using a small capacity low performance processor.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If measured HRIR filters are used for sound image localization, then localization accuracy is improved, but processor capacity and complexity increase
Solution Approach 1:
The patent extracts only the essential components from the complete HRIR filters that are necessary for sound image localization. Specifically, it separates and utilizes only the interaural time difference (ITD) and interaural level difference (ILD) information, discarding other redundant components. This extraction approach maintains localization accuracy while significantly reducing the number of filter coefficients and processor requirements.
Solution Approach 2:
The patent creates simplified copies of the original HRIR filters by generating synthetic ITD and ILD filters based on mathematical models rather than using complete measured HRIR data. These synthetic filters replicate the essential localization functionality with much fewer coefficients, enabling implementation on low-performance processors while maintaining acceptable localization accuracy.
2Reliability
If complete HRIR database is used for multiple channels, then localization performance is improved, but data capacity and processing requirements increase
Solution Approach 1:
The patent extracts only the critical ITD and ILD parameters from the complete HRIR database, discarding redundant information. By focusing solely on these two essential components, the system achieves reliable localization performance with a significantly reduced data capacity requirement, making it suitable for mobile devices with limited storage and processing capabilities.
3Adaptability or versatility
If interpolation algorithm is implemented in real-time, then listener movement is supported, but processing speed and performance requirements increase
Solution Approach 1:
The patent creates simplified synthetic ITD and ILD filters that can be rapidly computed in real-time, replacing the need for complex real-time interpolation of complete HRIR data. These simplified models enable the system to adapt to listener movement at lower processing speeds, making real-time tracking feasible on mobile devices with limited computational power.
Data Source
AI summary
A method and apparatus for localizing a sound image of an input signal to a spatial position are provided. The method of localizing a sound image to a spatial position includes: extracting from a head related impulse response (HRIR) measured with respect to changes in the position of a sound source, first information indicating a reflection sound wave reflected by the body of a listener; extracting from the HRIR second information indicating the difference between sound pressures generated in two ears, respectively, when a direct sound wave generated from the position of the sound source arrives at the two ears, respectively, of the listener; extracting third information indicating the difference between times taken by the direct sound wave to arrive at the two ears, respectively, from the HRIR; and localizing a sound image of an input signal to a spatial position by using the extracted information. According to the method and apparatus of the present invention, by using only important information having influence on sound image localization of a virtual sound source extracted from the HRIR, the sound image of the input signal can be localized to a spatial position with a small number of filter coefficients.