![]() The most extensively investigated case is that of two concurrent synthetic vowels ( Assmann and Summerfield, 1990, 1994 Meddis and Hewitt, 1993 de Cheveigné et al., 1995 de Cheveigné et al., 1997 Assmann and Paschall, 1998).Īssmann and Summerfield (1990) proposed a segregation mechanism that separated competing sounds by sampling the internal representation of the sound mixture according to an F0-guided rule which, in their “place-time” model, was based on sampling the autocorrelation functions (ACFs) of each channel (where a channel is a particular filter in an auditory filterbank) at the delays corresponding to the periods of each sound. Although natural speech is not monotone, using monotone sounds to study Δ F0 benefit is advantageous at least in part because it allows for findings to be interpreted in the context provided by the literature on pitch perception and segregation of monotone harmonic sounds ( Assmann and Summerfield, 1990, 1994 Meddis and Hewitt, 1993 Micheyl et al., 2006 Micheyl et al., 2010 Micheyl and Oxenham, 2010 Wang et al., 2012). One common approach to investigating Δ F0 benefit is to use targets and maskers with monotone (i.e., non-time-varying) F0 contours and to measure speech intelligibility as a function of Δ F0 between target and masker ( de Cheveigné et al., 1995 Assmann, 1998 Assmann and Paschall, 1998 Deroche et al., 2014b, a Leclère et al., 2017 Madsen et al., 2017). However, understanding the causes of this deficit and the best ways to resolve it is difficult because the mechanisms underlying Δ F0 benefit are not yet completely understood in NH listeners. Studies of Δ F0 benefit have translational relevance because hearing-impaired (HI) listeners experience less Δ F0 benefit than normal-hearing (NH) listeners, a deficit which is thought to play a role in the difficulties HI listeners experience in understanding speech in the presence of competing talkers ( Summers and Leek, 1998 Oxenham, 2008). Hereafter, we refer to such improvements as Δ F0 benefit. ![]() A number of studies have demonstrated that as the F0 difference (Δ F0) between competing talkers increases, perceptual segregation of the talkers becomes easier ( Assmann and Summerfield, 1994 Assmann, 1998 Oxenham and Simonson, 2009 Deroche et al., 2014a Leclère et al., 2017). Moreover, voice pitch helps listeners better understand speech in challenging listening conditions. Voice pitch has important suprasegmental functions, such as conveying emotion and emphasis ( Frick, 1985), and is also a key indexical property of speech, conveying information about the talker's age and biological sex ( Assmann et al., 2006 Barreda and Assmann, 2018). Voiced speech has a clear pitch determined by the fundamental frequency ( F0) of vibration of the talker's vocal folds. Pitch is a salient perceptual dimension of many common sounds and plays a key role in the perception of speech and music ( Plack et al., 2005 Oxenham, 2018). Because this manipulation eliminated spectral overlap between target and masker components but preserved shared periodicity, the finding is consistent with the explanation for the lack of Δ F0 benefit at the octave based on spectral overlap, but not with the explanation based on harmonic cancellation. Removal of the masker's even harmonics when the target was one octave above the masker improved speech reception thresholds by about 5 dB. To assess these competing explanations, speech intelligibility of a monotonized target talker, masked by a speech-shaped harmonic complex tone, was measured as a function of Δ F0, masker spectrum (all harmonics or odd harmonics only), and masker temporal envelope (amplitude modulated or unmodulated). Alternatively, this effect can be explained by spectral overlap between the harmonic components of the target and masker. Earlier work suggested that an octave Δ F0 provided little benefit, an effect predicted by harmonic cancellation due to the shared periodicity of masker and target. A model of Δ F0 benefit based on harmonic cancellation proposes that a masker's periodicity can be used to cancel (i.e., filter out) its neural representation. Fundamental frequency differences (Δ F0) between competing talkers aid in the perceptual segregation of the talkers (Δ F0 benefit), but the underlying mechanisms remain incompletely understood.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |