Preference judgments of artificial processed and hearing-aid transduced speech

Area of Science:

• Audiology
• Signal Processing
• Psychoacoustics

Background:

• Understanding hearing aid preference is crucial for improving audiological devices.
• Signal processing algorithms and distortions significantly impact user satisfaction.
• Previous research has explored various factors influencing hearing aid selection.

Purpose of the Study:

• To assess the impact of signal processing algorithms and distortions on hearing aid preference.
• To compare preferences between normal-hearing and hearing-impaired listeners.
• To identify key acoustic features driving user choice in hearing aids.

Main Methods:

• Speech signals were manipulated using artificial processing (filtering, compression, clipping, noise) and real hearing aid recordings.
• Normal-hearing and hearing-impaired subjects evaluated sounds via pairwise comparisons.
• Multidimensional scaling and Principal Components Analysis were used to analyze preference data.

Main Results:

• Preference judgments were primarily based on intelligibility/clarity and signal versus background distortion.
• Hearing-impaired subjects preferred high-frequency emphasis, while normal-hearing subjects preferred original signals.
• Soft background noise was consistently disliked; hearing aid preference correlated with frequency cutoff characteristics.

Conclusions:

• Signal clarity and distortion type are key determinants of hearing aid preference.
• Hearing-impaired listeners' preferences differ significantly from normal-hearing listeners, particularly regarding frequency response.
• Acoustic features like frequency cutoff play a critical role in hearing aid selection for both groups.