Localization and separation of acoustic sources by using a 2.5-dimensional circular microphone array.

Area of Science:

• Acoustics
• Signal Processing
• Array Signal Processing

Background:

• Circular microphone arrays (CMAs) are widely used in immersive audio but struggle with elevation angle resolution.
• Accurate 3D sound imaging requires precise localization of both azimuth and elevation angles.

Purpose of the Study:

• To propose and validate a novel 2.5-dimensional (2.5-D) microphone array system.
• To enhance the elevation angle estimation capabilities beyond traditional CMAs.
• To improve the quality of separated source signals in 3D audio environments.

Main Methods:

• A hybrid array combining a CMA and a vertical logarithmic-spacing linear array (LLA) was developed.
• Direction of Arrival (DOA) estimation using delay-and-sum beamformers on both arrays.
• Source amplitude extraction via Tikhonov regularization and convex optimization.
• Signal enhancement using the normalized least-mean-square algorithm with internal iteration.

Main Results:

• Experimental validation using a 3D-printed 24-element CMA and 8-element LLA.
• Demonstrated improved elevation angle resolution compared to standard CMAs.
• Objective and subjective listening tests confirmed enhanced source signal quality.

Conclusions:

• The proposed 2.5-D CMA effectively addresses the elevation angle limitation of traditional CMAs.
• This hybrid array design offers superior 3D sound imaging for advanced audio applications.
• The integrated signal processing techniques yield high-quality, separated audio signals.