Speech recognition in noise with active and passive hearing protectors: A comparative study

• Published in: The Journal of the Acoustical Society of America Vol. 129; no. 6; pp. 3702 - 3715
• Main Authors: Bockstael, Annelies, De Coensel, Bert, Botteldooren, Dick, D'Haenens, Wendy, Keppler, Hannah, Maes, Leen, Philips, Birgit, Swinnen, Freya, Bart, Vinck
• Format: Journal Article
• Language: English
• Published: Melville, NY Acoustical Society of America 01.06.2011; American Institute of Physics
• Subjects: Acoustic Stimulation; Acoustics; Adult; Amplifiers, Electronic - adverse effects; Analysis of Variance; Audiometry, Speech; Audition; Auditory Threshold; Biological and medical sciences; Ear Protective Devices - adverse effects; Equipment Design; Exact sciences and technology; Female; Fundamental and applied biological sciences. Psychology; Fundamental areas of phenomenology (including applications); Humans; Loudness Perception; Male; Materials Testing; Noise, Occupational - adverse effects; Noise: its effects and control; Perception; Perceptual Masking; Physics; Pressure; Psychology. Psychoanalysis. Psychiatry; Psychology. Psychophysiology; Recognition (Psychology); Sound Spectrography; Speech Intelligibility; Speech Perception; Time Factors; Hearing; Active system; Speech recognition; Hearing muffler; Noise pollution; Passive system; Background noise
• ISSN: 0001-4966; 1520-8524; 1520-9024; 1520-8524
• DOI: 10.1121/1.3575599

The perceived negative influence of standard hearing protectors on communication is a common argument for not wearing them. Thus, "augmented" protectors have been developed to improve speech intelligibility. Nevertheless, their actual benefit remains a point of concern. In this paper, speech perception with active earplugs is compared to standard passive custom-made earplugs. The two types of active protectors included amplify the incoming sound with a fixed level or to a user selected fraction of the maximum safe level. For the latter type, minimal and maximal amplification are selected. To compare speech intelligibility, 20 different speech-in-noise fragments are presented to 60 normal-hearing subjects and speech recognition is scored. The background noise is selected from realistic industrial noise samples with different intensity, frequency, and temporal characteristics. Statistical analyses suggest that the protectors' performance strongly depends on the noise condition. The active protectors with minimal amplification outclass the others for the most difficult and the easiest situations, but they also limit binaural listening. In other conditions, the passive protectors clearly surpass their active counterparts. Subsequently, test fragments are analyzed acoustically to clarify the results. This provides useful information for developing prototypes, but also indicates that tests with human subjects remain essential.