Evaluation of Loudspeaker-Based Virtual Sound Environments for Testing Directional Hearing Aids

Area of Science:

• Audiology
• Acoustics
• Hearing Aid Technology

Background:

• Laboratory hearing aid (HA) benefit assessments often lack real-world accuracy.
• Virtual sound environments (VSEs) offer more realistic acoustic scenarios for HA testing.
• The impact of VSE accuracy on subjective performance measures is not fully understood.

Purpose of the Study:

• To validate two common methods for creating VSEs for hearing aid outcome assessment.
• To compare the accuracy of model-based versus recording-based VSEs.

Main Methods:

• A cocktail-party sound scene was created in a real room and reproduced in an anechoic chamber using a loudspeaker array.
• Two VSEs were generated: one using room acoustic modeling and another using microphone array recordings.
• 18 participants with mild-to-moderate hearing loss completed speech intelligibility and acceptable noise level tests in all three environments.

Main Results:

• Subjective performance trends were similar across real and virtual environments for directional hearing aid algorithms.
• Directional benefits were slightly reduced in the model-based VSE and further reduced in the recording-based VSE.
• The general behavior of subjective performance was preserved.

Conclusions:

• Virtual sound environments are suitable for testing directional hearing aids.
• A slight reduction in sensitivity and potential underestimation of directional benefit were observed in VSEs compared to real environments.
• The high ecological relevance of VSEs may outweigh minor limitations, improving HA outcome measures.