Unconscious and Distinctive Control of Vocal Pitch and Timbre During Altered Auditory Feedback

Abstract

Vocal control plays a critical role in smooth social communication. Speakers constantly monitor auditory feedback (AF) and make adjustments when their voices deviate from their intentions. Previous studies have shown that when certain acoustic features of the AF are artificially altered, speakers compensate for this alteration in the opposite direction. However, little is known about how the vocal control system implements compensations for alterations of different acoustic features, and associates them with subjective consciousness. The present study investigated whether compensations for the fundamental frequency (F0), which corresponds to perceived pitch, and formants, which contribute to perceived timbre, can be performed unconsciously and independently. Forty native Japanese speakers received two types of altered AF during vowel production that involved shifts of either only the formant frequencies (formant modification; Fm) or both the pitch and formant frequencies (pitch + formant modification; PFm). For each type, three levels of shift (slight, medium, and severe) in both directions (increase or decrease) were used. After the experiment, participants were tested for whether they had perceived a change in the F0 and/or formants. The results showed that (i) only formants were compensated for in the Fm condition, while both the F0 and formants were compensated for in the PFm condition; (ii) the F0 compensation exhibited greater precision than the formant compensation in PFm; and (iii) compensation occurred even when participants misperceived or could not explicitly perceive the alteration in AF. These findings indicate that non-experts can compensate for both formant and F0 modifications in the AF during vocal production, even when the modifications are not explicitly or correctly perceived, which provides further evidence for a dissociation between conscious perception and action in vocal control. We propose that such unconscious control of voice production may enhance rapid adaptation to changing speech environments and facilitate mutual communication.

Spontaneous variability predicts compensative motor response in vocal pitch control

“…1 B). The amount of compensation was almost proportional to the amount of seven f o shifts (0, ± 25, ± 50, or ± 100 cents), as already shown in our previous study 36 . We calculated the compensation ratio for each participant, which was defined as a sign-inverted slope of a fitted line to compensation amounts as a function of introduced f o shifts (Fig.…”

“…We also assessed other factors that potentially affected the compensation process, such as the perceptual ability to detect a subtle difference in vocal pitch. For this aim, we estimated participants’ ability to detect the f o shifts induced in recorded own voices using a dataset from the listening tests performed in our previous study 36 . In this test, participants were asked to answer whether any pitch modification occurred in the second vocalization compared with the first one in each trial (Fig.…”

“…The dataset used in this study was originally obtained in our previous study 36 . The present study analyzed this in different ways to elucidate the relationship between the spontaneous variability and compensation behavior in vocal control.…”

Assessing Pain Research: A Narrative Review of Emerging Pain Methods, Their Technosocial Implications, and Opportunities for Multidisciplinary Approaches

“…Previous studies have shown language's utility in identifying ingested drugs and dosages ( 78 ), quantifying depression severity ( 79 ), dissociating and predicting mental illnesses ( 80 – 82 ), and tracking changes in neurological disease symptoms ( 83 – 85 ). Given many chronic pain conditions are also neurodegenerative ( 86 ) and linked to changes in cognitive functioning or emotional processing, researchers are investigating whether quantitative language features can also measure pain qualities, although qualitative research investigating pain and language has been around for some time ( 77 , 87 – 90 ). Unlike questionnaires, collecting patient language about their own experience is far less limiting, and whereas numeric PROs could be influenced by various cognitive processes, voice acoustic properties and lower-level linguistic and syntactic structures are harder to consciously influence in the same way ( 91 , 92 ).…”

Reliability and Efficiency of Pitch-Shifting Plug-Ins in Voice and Hearing Research

“…Some aspects were limited in the harmonizer, considering that the vocal harmonizer shifts the whole frequency spectrum (including associated vowel formants), has no formant correction parameter, and was developed for musical applications and not for voice research. After this product became discontinued, researchers started to use TC Helicon Voice One (Keough et al, 2013) and TC Helicon VoiceWorks Plus (Xu et al, 2020). Both TC Helicon hardware use a similar pitch-shifting algorithm but also have formant correction parameters that can provide additional control over pitch shifting.…”