The contribution of temporal fine structure to the intelligibility of speech in steady and modulated noise

Abstract

Speech reception thresholds were measured with steady and amplitude-modulated noise maskers for signals processed to contain variable amounts of temporal fine structure (TFS) information. Subjects benefited more from TFS information for the modulated than for the steady masker. For both maskers, addition of TFS information up to 548 Hz improved performance, though the improvement was greater for the modulated masker. The addition of TFS information at higher frequencies improved performance further for the modulated masker only. These results are consistent with the idea that TFS information is important for listening in the dips of a fluctuating masker.

Consonant identification in noise using Hilbert-transform temporal fine-structure speech and recovered-envelope speech for listeners with normal and impaired hearing

“…A possible explanation for the presence of MR for HI listeners with T1 and T4 speech may arise from NH studies that show more robust MR for speech containing TFS cues compared to vocoded speech where such cues are removed (e.g., Hopkins and Moore, 2009). However, the importance of TFS cues for "listening in the gaps" has been called into question by several studies exploring this hypothesis (Oxenham and Simonson, 2009;Freyman et al, 2012).…”

“…The reduction in cochlear compression and decreased frequency selectivity that accompany sensorineural loss are another possible source of decreased MR (Moore et al, 1999;Oxenham and Kreft, 2014). A reduced ability to process the temporal fine structure of speech (TFS) (the rapid fluctuations in amplitude close to the center frequency of a narrow-band signal) as well as NH individuals (Lorenzi et al, 2006a;Lorenzi et al, 2009;Hopkins and Moore, 2009;Hopkins et al, 2008;Moore, 2014) has also been proposed as a factor in reduced MR. Support for this explanation lies in observed correlations between scores for understanding speech manipulated to degrade or convey TFS cues and the MR obtained with intact speech signals (Lorenzi et al, 2006a;Lorenzi et al, 2009;Hopkins and Moore, 2011;Hopkins et al, 2008). More recently, it has been suggested that the lack of MR in HI listeners may be based on their having less susceptibility than NH listeners to the random amplitude fluctuations present in steady-state noise (see Stone et al, 2012).…”

Level variations in speech: Effect on masking release in hearing-impaired listeners

“…Another hypothesis regarding the ability to make use of speech information present in the gaps is concerned with the ability to use temporal finestructure (TFS) cues. Hopkins and Moore (2009), for example, observed an increase in MR with an increase in the number of spectral channels in which TFS cues were present. However, other studies have found no direct link between TFS cues and MR (e.g., Oxenham and Simonson, 2009;Bernstein and Brungart, 2011;Freyman et al, 2012).…”

Benefit of temporal fine structure to speech perception in noise measured with controlled temporal envelopes

“…In the other condition (ENV&TFS), the stimuli have useful information in both their temporal envelopes and their TFS. Listeners perceive speech more accurately in the ENV&TFS condition (Qin and Oxenham, 2003;Gnansia et al, 2008Gnansia et al, , 2009Lorenzi et al, 2009;Hopkins and Moore, 2009). …”