Slowed speech spreading into reverberant environments; steady-state suppression improves speech intelligibility

Abstract

Speech intelligibility is often decreased in reverberant environments. For improving speech intelligibility under reverberant environments, Arai et al. [Acoust. Sci. Technol. 23, 229–232 (2002)] suggested steady-state suppression, which suppresses steady-state portions of speech to reduce overlap-masking (causing degradation intelligibility) and improves speech intelligibility as preprocessing. Previous studies showed that speech intelligibility was improved through this processing under some reverberant environments [Hodoshima et al., J. Acoust. Soc. Am. 119(6), 4055–4064 (2006)]. On the other hand, speech intelligibility is known to be increased greatly by speaking slowly [Bolt and MacDonald, J. Acoust. Soc. Am. 21(6), 577–580 (1949)]. However, it is not optimal for improving speech intelligibility in terms of reducing overlap masking because vowels causing major parts of overlap masking are expanded by speaking slowly. Therefore, in the current study, we investigated the effects of steady-state suppression on speech perception with a decreased speech rate from 6 to 4 and 5 morae/s. Results showed that the slowest speech (4 morae/s) with steady-state suppression was the most intelligible. Also, steady-state suppression improved speech intelligibility at a speech rate of 4 morae/s in each reverberant condition (reverberation times of 1.5, 2.0, and 2.5 s). [Work partially supported by JSPS.KAKENHI (16203041).]