Vowel Acoustics Research Articles

Speech acoustics can be modulated by cognitive interference in a vowel-modified Stroop task

How are speech acoustics influenced by cognitive processes? In the current study, we used a novel variant of the Stroop test to measure whether the interference between color naming and reading could modulate vowel formant frequencies. Seventeen healthy participants named the color of words in three categories: (1) congruent words (e.g., “red” written in red), (2) color-incongruent words (e.g., “green” written in red), and (3) vowel-incongruent words with phonetic properties that partially matched their color (e.g., “rid” written in red). We hypothesized that the cognitive effort needed to inhibit reading—saying “red,” not “rid”—could affect vowel acoustics. For example, the correct spoken response (“red”) could more acoustically resemble the inhibited word “rid;” alternatively, the vowel could be influenced in the opposite direction, resembling “rad,” which would serve to accentuate the acoustic contrast between the spoken and inhibited words. As expected, participants were slower to produce words on color-incongruent trials than on congruent trials. Interestingly, vowel-incongruent trials were not significantly slower than congruent trials, but preliminary acoustic analyses of the first formant (F1) showed that some subjects systematically modulated their productions in the presence of incongruent vowels. This finding lends insight into how the brain integrates multiple pieces of information to produce speech.

The Impact of Contrastive Stress on Vowel Acoustics and Intelligibility in Dysarthria.

To compare vowel acoustics and intelligibility in words produced with and without contrastive stress by speakers with spastic (mixed-spastic) dysarthria secondary to cerebral palsy (DYSCP) and healthy controls (HCs). Fifteen participants (9 men, 6 women; age M = 42 years) with DYSCP and 15 HCs (9 men, 6 women; age M = 36 years) produced sentences containing target words with and without contrastive stress. Forty-five healthy listeners (age M = 25 years) completed a vowel identification task of DYSCP productions. Vowel acoustics were compared across stress conditions and groups using 1st (F1) and 2nd (F2) formant measures. Perceptual intelligibility was compared across stress conditions and dysarthria severity. F1 and F2 significantly increased in stressed words for both groups, although the degree of change differed. Mean Euclidian distance between vowels also increased with stress. The relative probability of vowels falling within the target F1 × F2 space was greater for HCs but did not differ with stress. Stress production resulted in greater listener vowel identification accuracy for speakers with mild dysarthria. Contrastive stress affected vowel formants for both groups. Perceptual results suggest that some speakers with dysarthria may benefit from a contrastive stress strategy to improve vowel intelligibility.

Lyrics-to-Audio Alignment by Unsupervised Discovery of Repetitive Patterns in Vowel Acoustics

Most of the previous approaches to lyrics-to-audio alignment used a pre-developed automatic speech recognition (ASR) system that innately suffered from several difficulties to adapt the speech model to individual singers. A significant aspect missing in previous works is the self-learnability of repetitive vowel patterns in the singing voice, where the vowel part used is more consistent than the consonant part. Based on this, our system first learns a discriminative subspace of vowel sequences, based on weighted symmetric non-negative matrix factorization (WS-NMF), by taking the self-similarity of a standard acoustic feature as an input. Then, we make use of canonical time warping (CTW), derived from a recent computer vision technique, to find an optimal spatiotemporal transformation between the text and the acoustic sequences. Experiments with Korean and English data sets showed that deploying this method after a pre-developed, unsupervised, singing source separation achieved more promising results than other state-of-the-art unsupervised approaches and an existing ASR-based system.

Cortical sensitivity to spoken acoustic variability in persons with aphasia: an MEG study

Cortical sensitivity to spoken acoustic variability in persons with aphasia: an MEG study

Bite block effects on vowel acoustics in talkers with amyotrophic lateral Sclerosis and Parkinson’s Disease

Amytrophic Lateral Sclerosis (ALS) affects predominantly the tongue and leaves the jaw relatively spared. By contrast, Parkinson’s disease (PD) is thought to affect the motor system globally with minimal jaw movements potentially further limiting tongue articulatory movements. However, currently only few studies have directly investigated articulator-specific contributions to deviant speech acoustics in talkers with dysarthria. In this study, a 10 mm bite block was used to decouple tongue and jaw and delineate their specific contributions on vowel acoustics in talkers with distinctly different impairment profiles. Acoustic vowel contrast was examined during sentence repetitions produced by talkers with ALS, PD, and controls under jaw-free and jaw-fixed conditions. Preliminary findings of seven speakers per group revealed that acoustic vowel contrast was significantly lower during the jaw-fixed than the jaw-free condition in ALS. However, in PD acoustic vowel contrast tended to be greater when the jaw was fixed. For controls, no difference was found between the two conditions. Further, compared to controls acoustic vowel contrast was significantly reduced during the jaw-fixed condition in ALS and during the jaw-free condition in PD. Outcomes suggest that the jaw promotes acoustic vowel contrast in ALS but may hinder it in PD.

Analysis of speech and tongue motion in normal and post-glossectomy speaker using cine MRI.

ABSTRACTObjectiveSince the tongue is the oral structure responsible for mastication, pronunciation, and swallowing functions, patients who undergo glossectomy can be affected in various aspects of these functions. The vowel /i/ uses the tongue shape, whereas /u/ uses tongue and lip shapes. The purpose of this study is to investigate the morphological changes of the tongue and the adaptation of pronunciation using cine MRI for speech of patients who undergo glossectomy.Material and MethodsTwenty-three controls (11 males and 12 females) and 13 patients (eight males and five females) volunteered to participate in the experiment. The patients underwent glossectomy surgery for T1 or T2 lateral lingual tumors. The speech tasks “a souk” and “a geese” were spoken by all subjects providing data for the vowels /u/ and /i/. Cine MRI and speech acoustics were recorded and measured to compare the changes in the tongue with vowel acoustics after surgery. 2D measurements were made of the interlip distance, tongue-palate distance, tongue position (anterior-posterior and superior-inferior), tongue height on the left and right sides, and pharynx size. Vowel formants Fl, F2, and F3 were measured.ResultsThe patients had significantly lower F2/Fl ratios (F=5.911, p=0.018), and lower F3/F1 ratios that approached significance. This was seen primarily in the /u/ data. Patients had flatter tongue shapes than controls with a greater effect seen in /u/ than /i/.ConclusionThe patients showed complex adaptation motion in order to preserve the acoustic integrity of the vowels, and the tongue modified cavity size relationships to maintain the value of the formant frequencies.

Acoustics of epenthetic vowels in Brazilian Portuguese

In this production study involving a task of real words and a task of nonce words, epenthesis in obstruent-obstruent clusters in Brazilian Portuguese was investigated and the acoustic properties of epenthetic and lexical [i] were compared. A number of linguistic factors were found to be significant predictors of the epenthesis rate: voicing of the consonants of the cluster, manner of articulation of the second consonant, place of articulation of both consonants, and position of the cluster with respect to the stressed syllable. Moreover, there was a higher epenthesis rate for nonce words than for real words. As for the quality of the epenthetic vowel, it was found to have the same F1 as the lexical [i] in the language, but it was significantly shorter in duration, and it had a significantly lower F2 in the nonce words, but not in the real words. Finally, we found a relatively high rate of deletion of lexical [i], conditioned by the same factors that conditioned epenthesis rate, but in the opposite direction. It is concluded that epenthesis is a phonological phenomenon in Brazilian Portuguese, but that there is also acoustic reduction in the language that affects both lexical and epenthetic vowels.

Acoustic characteristics of vowels in two Totonacan languages

This study examines the acoustics of vowels from two Totonacan languages, Upper Necaxa Totonac (UNT) and Huehuetla Tepehua (HT). Both languages have five-vowel systems consisting of the qualities /aeiou/ as well as phonemic quantity distinctions (short, long), and lexical stress. In addition, UNT makes use of contrastive phonation on vowels, while HT vowels may be produced with allophonically non-modal phonation when adjacent to glottalic segments and glottal stops. Data from four speakers (two female and two male) of each language are reported. Acoustic measures are based on multiple repetitions of each vowel in a variety of lexical items, elicited within a frame sentence. Measurements are taken from stressed vowels surrounded by obstruents wherever possible. A variety of analyses are undertaken. Traditional visualizations of the vowel spaces of male and female speakers using normalized F1-F2 plots are combined with comparisons of dynamic formant trajectories across the time course of vowel production. V...

The relationship between fundamental frequency and within-speaker vowel reduction

The contribution(s) of fundamental frequency (F0) to vowel production has potential theoretical and clinical implications. The sufficient contrast hypothesis [Diehl et al., J. Phon. 24, 187–208 (1996)] suggests that a high F0 is associated with effects of spectral undersampling, and thus may require more exaggerated formant frequencies to maintain the perceptual distinctiveness of vowels. Nonetheless, studies on the interplay between F0 and vowel acoustics would appear inconclusive [Byrd, J. Acoust. Soc. Am. 92, 593–596 (1992)] [Weirich & Simpson, J. Acoust. Soc. Am. 134, 2965–2974 (2013)]. This study evaluated the sufficient contrast hypothesis as a within-speaker mechanism, with an overarching hypothesis that formant frequency exaggeration associated with spectral undersampling, if any, occurs within a given production system (i.e., speaker). To this end, formant frequencies of vowels and F0 were obtained for the same speaker across different speaking tasks. Within speaker, it was hypothesized that vowel formant frequencies as well as F0 change across tasks. More importantly, average F0 is hypothesized to be negatively correlated with the degree of vowel formant frequency reduction within-speaker across tasks. That is, a vowel system produced with a higher F0 would be less prone to reduction because the spectral peaks are better defined. The contribution(s) of fundamental frequency (F0) to vowel production has potential theoretical and clinical implications. The sufficient contrast hypothesis [Diehl et al., J. Phon. 24, 187–208 (1996)] suggests that a high F0 is associated with effects of spectral undersampling, and thus may require more exaggerated formant frequencies to maintain the perceptual distinctiveness of vowels. Nonetheless, studies on the interplay between F0 and vowel acoustics would appear inconclusive [Byrd, J. Acoust. Soc. Am. 92, 593–596 (1992)] [Weirich & Simpson, J. Acoust. Soc. Am. 134, 2965–2974 (2013)]. This study evaluated the sufficient contrast hypothesis as a within-speaker mechanism, with an overarching hypothesis that formant frequency exaggeration associated with spectral undersampling, if any, occurs within a given production system (i.e., speaker). To this end, formant frequencies of vowels and F0 were obtained for the same speaker across different speaking tasks. Within speaker, it was hypothesized that vowe...

Vowel variability in elicited versus spontaneous speech: Evidence from Mixtec

This study investigates the influence of speech style, duration, contextual factors, and sex on vowel dispersion and variability in Yoloxóchitl Mixtec, an endangered language spoken in Mexico. Oral vowels were examined from recordings of elicited citation words and spontaneous narrative speech matched across seven speakers. Results show spontaneous speech to contain shorter vowel durations and stronger effects of contextual assimilation than elicited speech. The vowel space is less disperse and there is greater intra-vowel variability in spontaneous speech than in elicited speech. Furthermore, male speakers show smaller differences in vowel dispersion and duration across styles than female speakers do. These phonetic differences across speech styles are not entirely reducible to durational differences; rather, speakers also seem to adjust their articulatory/acoustic precision in accordance with style. Despite the stylistic differences, we find robust acoustic differences between vowels in spontaneous speech, maintaining the overall vowel space pattern. While style and durational changes produce noticeable differences in vowel acoustics, one can closely approximate the phonetics of a vowel system of an endangered language from narrative speech. Elicited speech is likelier to give the most extreme formants used by the language than is spontaneous speech, but the usefulness of phonetic data from spontaneous speech has still been demonstrated.

The contribution of vowels to auditory-visual speech recognition and the contributions of Diane Kewley-Port to the field of speech communication

Throughout her career, Diane Kewley-Port has made enduring contributions to the field of Speech Communication in two ways—through her research on vowels and through her mentoring. Diane has contributed greatly to current knowledge about vowel acoustics, vowel discrimination and identification, and the role of vowels in speech recognition. Within that line of research, Richie & Kewley-Port (2008) investigated the effects of visual cues to vowels on speech recognition. Specifically, we demonstrated that an auditory-visual vowel-identification training program benefited sentence recognition under difficult listening conditions more than consonant-identification training and no training. In this presentation, I will describe my continuing research on the relationship between auditory-visual vowel-identification training and listening effort, for adults with normal hearing. In this study, listening effort was measured in terms of response time and participants were tested on auditory-visual sentence recognition in noise. I will discuss the ways that my current work has been inspired by past research with Diane, and how her mentoring legacy lives on.

A multi-modal imaging system for simultaneous measurement of speech articulator kinematics for bedside applications in clinical settings

A critical step toward a neurological understanding of speech generation is to relate neural activity to the movement of articulators. Here, we describe a noninvasive system for simultaneously tracking the movement of the lips, jaw, tongue, and larynx for human neuroscience research carried out at the bedside. We combined three methods previously used separately: videography to track the lips and jaw, electroglottography to monitor the larynx, and ultrasonography to track the tongue. To characterize this system, we recorded articulator positions and acoustics from six speakers during production of nine American English vowels. We describe processing methods for the extraction of kinematic parameters from the raw signals and methods to account for artifacts across recording conditions. To understand the relationship between kinematics and acoustics, we used regularized linear regression between the vocal tract kinematics and speech acoustics to identify which, and how many, kinematic features are required to explain both across vowel and within vowel acoustics. Furthermore, we used unsupervised matrix factorization to derive "prototypical" articulator shapes, and use them as a basis for articulator analysis. These results demonstrate a multi-modal system to non-invasively monitor speech articulators for clinical human neuroscience applications and introduce novel analytic methods for understanding articulator kinematics.

Vowel intelligibility and the second-language learner

Diane Kewley-Port’s work has contributed to our understanding of vowel perception and production in a wide variety of ways, from mapping the discriminability of vowel formants in conditions of minimal uncertainty to vowel processing in challenging conditions, such as increased presentation rate and noise. From the results of these studies, we have learned much about the limits of vowel perception for normal-hearing listeners and the robustness of vowels in speech perception. Continuously intertwined with this basic research has been its application to our understanding of vowel perception and vowel acoustics across various challenges, such as hearing impairment and second-language learning. Diane’s work on vowel perception and production by second-language learners and ongoing research stemming from her influence will be considered in light of several factors affecting communicative success and challenge for second-language learners. In particular, we will compare the influence of speaking style, noise, and syllable disruption on the intelligibility of vowels perceived and produced by native and non-native English-speaking listeners.

Control of spoken vowel acoustics and the influence of phonetic context in human speech sensorimotor cortex.

Speech production requires the precise control of vocal tract movements to generate individual speech sounds (phonemes) which, in turn, are rapidly organized into complex sequences. Multiple productions of the same phoneme can exhibit substantial variability, some of which is inherent to control of the vocal tract and its biomechanics, and some of which reflects the contextual effects of surrounding phonemes ("coarticulation"). The role of the CNS in these aspects of speech motor control is not well understood. To address these issues, we recorded multielectrode cortical activity directly from human ventral sensory-motor cortex (vSMC) during the production of consonant-vowel syllables. We analyzed the relationship between the acoustic parameters of vowels (pitch and formants) and cortical activity on a single-trial level. We found that vSMC activity robustly predicted acoustic parameters across vowel categories (up to 80% of variance), as well as different renditions of the same vowel (up to 25% of variance). Furthermore, we observed significant contextual effects on vSMC representations of produced phonemes that suggest active control of coarticulation: vSMC representations for vowels were biased toward the representations of the preceding consonant, and conversely, representations for consonants were biased toward upcoming vowels. These results reveal that vSMC activity for phonemes are not invariant and provide insight into the cortical mechanisms of coarticulation.

Neural decoding of spoken vowels from human sensory-motor cortex with high-density electrocorticography.

We present the first demonstration of single-trial neural decoding of vowel acoustic features during speech production with high performance. The ability to predict trial-by-trial fluctuations in speech production was facilitated by using high-density, large-area electrocorticography (ECoG) combined with an adaptive principal components regression. In experiments from two human neurosurgical patients with a high-density 256-channel ECoG grid implanted over speech cortices, we demonstrate that as much as 81% of the acoustic variability across vowels could be accurately predicted from the spatial patterns of neural activity during speech production. These results demonstrate continuous, single-trial decoding of vowel acoustics.

Vowel acoustics in dysarthria: speech disorder diagnosis and classification.

The purpose of this study was to determine the extent to which vowel metrics are capable of distinguishing healthy from dysarthric speech and among different forms of dysarthria. A variety of vowel metrics were derived from spectral and temporal measurements of vowel tokens embedded in phrases produced by 45 speakers with dysarthria and 12 speakers with no history of neurological disease. Via means testing and discriminant function analysis (DFA), the acoustic metrics were used to (a) detect the presence of dysarthria and (b) classify the dysarthria subtype. Significant differences between dysarthric and healthy control speakers were revealed for all vowel metrics. However, the results of the DFA demonstrated some metrics (particularly metrics that capture vowel distinctiveness) to be more sensitive and specific predictors of dysarthria. Only the vowel metrics that captured slope of the second formant (F2) demonstrated between-group differences across the dysarthrias. However, when subjected to DFA, these metrics proved unreliable classifiers of dysarthria subtype. The results of these analyses suggest that some vowel metrics may be useful clinically for the detection of dysarthria but may not be reliable indicators of dysarthria subtype using the current dysarthria classification scheme.

Vowel Acoustics in Dysarthria: Mapping to Perception

The aim of the present report was to explore whether vowel metrics, demonstrated to distinguish dysarthric and healthy speech in a companion article (Lansford & Liss, 2014), are able to predict human perceptual performance. Vowel metrics derived from vowels embedded in phrases produced by 45 speakers with dysarthria were compared with orthographic transcriptions of these phrases collected from 120 healthy listeners. First, correlation and stepwise multiple regressions were conducted to identify acoustic metrics that had predictive value for perceptual measures. Next, discriminant function analysis misclassifications were compared with listeners' misperceptions to examine more directly the perceptual consequences of degraded vowel acoustics. Several moderate correlative relationships were found between acoustic metrics and perceptual measures, with predictive models accounting for 18%-75% of the variance in measures of intelligibility and vowel accuracy. Results of the second analysis showed that listeners better identified acoustically distinctive vowel tokens. In addition, the level of agreement between misclassified-to-misperceived vowel tokens supports some specificity of degraded acoustic profiles on the resulting percept. Results provide evidence that degraded vowel acoustics have some effect on human perceptual performance, even in the presence of extravowel variables that naturally exert influence in phrase perception.

Interspeaker Variability in Hard Palate Morphology and Vowel Production

Differences in vocal tract morphology have the potential to explain interspeaker variability in speech production. The potential acoustic impact of hard palate shape was examined in simulation, in addition to the interplay among morphology, articulation, and acoustics in real vowel production data. High-front vowel production from 5 speakers of American English was examined using midsagittal real-time magnetic resonance imaging data with synchronized audio. Relationships among hard palate morphology, tongue shaping, and formant frequencies were analyzed. Simulations were performed to determine the acoustical properties of vocal tracts whose area functions are altered according to prominent hard palate variations. Simulations revealed that altering the height and position of the palatal dome alters formant frequencies. Examinations of real speech data showed that palatal morphology is not significantly correlated with any formant frequency but is correlated with major aspects of lingual articulation. Certain differences in hard palate morphology can substantially affect vowel acoustics, but those effects are not noticeable in real speech. Speakers adapt their lingual articulation to accommodate palate shape differences with the potential to substantially affect formant frequencies, while ignoring palate shape differences with relatively little acoustic impact, lending support for acoustic goals of vowel production.

Interspeaker variability in relative tongue size and vowel production

The tongue varies across speakers in terms of the proportion of the overall speech production apparatus that it occupies. Differences in tongue size have the potential to result in speaker-specific articulatory strategies for shaping the vocal tract area function and, in turn, individual patterns of vowel acoustics. The present study examines the interplay between relative tongue size and vowel production using real-time magnetic resonance imaging with synchronous audio. Two populations of native American English subjects are considered, one containing healthy adult speakers with no relevant pathologies, and another containing speakers who had undergone glossectomy as treatment for tongue cancer. All subjects were imaged in the midsagittal plane while reading phonetically balanced English sentences. The size of the tongue and the speech production apparatus were quantified from an overall average posture, and their ratio was correlated with the shape of the vowel space in terms of acoustics (e.g., formant frequencies), constrictions (i.e., location and degree of minimum constriction), and parameterized vocal tract cross-distance functions. Results indicate that relative tongue size can be used to explain and predict observable interspeaker differences in vowel production.

Effects of dialect on vowel acoustics and intelligibility

A great deal of recent research has focused on phonetic variation among American English vowels from different dialects. This body of research continues to grow as vowels continuously undergo diachronic formant changes that become characteristic of certain dialects. Two experiments using the Nationwide Speech Project corpus (Clopper & Pisoni 2006a) explored whether the Midland dialect is more closely related acoustically and perceptually to the Mid-Atlantic or to the Southern dialect. The goal of this study was to further our understanding of acoustic and perceptual differences between two of the most marked dialects (Mid-Atlantic and Southern) and one of the least marked dialects (Midland) of American English. Ten vowels in /hVd/ context produced by one male talker from each of these three dialects were acoustically analyzed and presented to Midland listeners for identification. The listeners showed the greatest vowel identification accuracy for the Mid-Atlantic talker (95.2%), followed by the Midland talker (92.5%), and finally the Southern talker (79.7%). Vowel error patterns were consistent with vowel acoustic differences between the talkers. The results suggest that, acoustically and perceptually, the Midland and Mid-Atlantic dialects are more similar than are the Midland and Southern dialects.