Vowel Context Research Articles

The effect of speaking rate, vowel context, and speaker intelligibility on the perception of consonant vowel consonants in noise

In this paper, we perform pilot experiments to evaluate the feasibility of a model to predict human recognition of speech sounds in the presence of noise at different speaking rates. CVC stimuli comprising a phonetically balanced set of 13 consonants and 3 vowels (/i/, /a/, /u/) were recorded in a sound proof booth by two talkers at two different speaking rates (fast and slow). Noisy stimuli were generated by adding babble noise at different levels to the quiet recordings. These stimuli were used to conduct perceptual experiments in which listeners were asked to listen and repeat back the CVC phrases presented in babble noise under 3 SNR conditions and both speaking rates. The data were transcribed by two trained linguists. Consonant confusion matrices were generated from these data and were analyzed by noise level, speaker, center vowel, and speaking rate. With the exception of /CuC/ stimuli, speaking rate had the most pronounced effect on perception with slow speech being more intelligible than fast speech in noise. /CaC/ stimuli were, on average, more robust than other stimuli in all conditions and one talker was significantly more intelligible than the other. Adetailed analysis of the results will be presented. [Work supported in part by the NSF.]

Signal detection of lipreading visemes using two dimensional and three dimensional images

The actual process by which the lipreader translates the lip movements they identify into a message is very complex. The lip movements observed represent only fragments of the complete message. The main purpose of this study is to investigate (1) the ability of lipreaders to use visual information alone to identify phonemes in varying contexts including nearby coarticulation effects and vowel neighborhoods; (2) lipreading responses using the effect of improved video presentation through 3D video, providing better and more realistic video presentation; and (3) the use of a novel measurement technique, i.e., a signal detection two-alternative-forced choice method of subject response that should provide measures of discrimination between phonemes including “visemes.” Video recordings were made in both 2D and 3D formats. This 3D image presented more realistically the movements such as lip-rounding and micro-movements of viewable articulators in three dimensions. Subjects with normal hearing were presented these video presentations. A Two-Alternative-Forced-Choice (2AFC) paradigm was used. The consonants were viewed with various vowel contexts. D-prime values were obtained for both the 2D and 3D videos. Particular consonant clusters were more discriminable in 3D.

Perceptual compensation with familiar and unfamiliar rounding coarticulation

We compared the integration of three kinds of contextual information in the perception of the fricatives [s] and [∫]. We asked American English listeners to identify sounds on an [s] to [∫] continuum and manipulated (1) the vowel context of the fricative ([Ce], [Co], [Cœ]), (2) the original fricative of the CV ([s] vs [∫]), and (3) the modality of the stimulus (audio-only, AV). There was a large compensation for coarticulation effect on perception—subjects responded with “s” more often when the following vowel was round. Interestingly, and perhaps significantly, perceptual compensation was not as great with the less familiar vowel [œ] even when listeners saw the face. Measurements of lip rounding in these stimuli show that [o] and [œ] have about the same degree and type of rounding over the CV. In a second experiment, we measured reaction time to audio-visual mismatches in these stimuli (again in a fricative identification task). We found that mismatches of audio and video consonant information slowed rea...

Semantic and Phonemic Listener Confusions in a Case of Isolated Congenital Aglossia

This article is the second in a series that examines the intelligibility of a person with congenital aglossia (PWCA). Specific factors examined in this study included (a) intelligibility for meaningful words versus nonsense words, (b) intelligibility for consonant-vowel-consonant words (CVCs) as a function of phonemic segment types, and (c) whether there is a correlation between intelligibility for these factors and the acoustic properties of the speech samples. Results revealed greater intelligibility for meaningful versus nonsense CVCs, greater intelligibility for back, low, and high-back versus front vowels embedded in CVCs, and greater intelligibility for productions as a function of phonemic variables, which demonstrated the following hierarchy: initial consonant > consonant vowel > vowel consonant > final consonant. Further results suggest that consonant recognition was consistently affected by “vowel context.” This suggests that movement sequencing appeared to be of importance for speech perception in productions of a PWCA.

Effect of temporal modification and vowel context on speech perception in individuals with auditory neuropathy spectrum disorder (ANSD)

Objectives: The study aimed to investigate the effect of time-scale modification on consonant perception in individuals with an auditory neuropathy spectrum disorder (ANSD). The study also evaluated the effect of vowel context on consonant perception. Study design: Eight individuals with ANSD in the age range of 15 to 24 years participated in the study. Their perception of eight stops and two liquids in the context of the vowels /a/, /i/ and /u/ were studied. The stimuli were stretched by 25%, 35% and 50% using pitch synchronous overlap and add algorithm. Results: It was found that 25% stretching resulted in significant improvement in perception in all three vowel contexts, while 35% stretching improved perception only in the context of /a/. However, 50% stretching resulted in deterioration in performance in all three vowel contexts. Furthermore, consonant perception in the context of /a/ and /u/ were significantly higher than that of /i/ in all the stretch conditions. Sequential information analysis revealed that perception of place and voicing were lowest in the context of /i/ compared to that of /a/ and /u/. Conclusion: Stretching the entire stimuli by a factor of 25% improves consonant perception in individuals with ANSD. In contrast, stretching beyond 25% resulted in either no improvement or deterioration in consonant perception. This information can be used to create a signal-processing algorithm that stretches the entire speech signal. Hence, without having to laboriously increase only the transition duration manually, improvement in speech perception can be brought about by stretching the whole signal. Additionally, consonant perception in individuals with ANSD is influenced by vowel context and hence two extreme vowels need to be used while studying the vowel context effect.

Comparison of Nasal Acceleration and Nasalance Across Vowels

The purpose of this study was to determine the performance of normalized nasal acceleration (NNA) relative to nasalance as estimates of nasalized versus nonnasalized vowel and sentence productions. Participants were 18 healthy speakers of American English. NNA was measured using a custom sensor, and nasalance was measured using the KayPentax Nasometer II. Speech stimuli consisted of CVC syllables with the vowels (//, /æ/, /i/, /u/) and sentences loaded with high front, high back, low front, and low back vowels in both nasal and nonnasal contexts. NNA showed a small but significant effect of the vowel produced during syllable stimuli but no significant effect of vowel loading during sentence stimuli. Nasalance was significantly affected by the vowel being produced during both syllables and sentences with large effect sizes. Both NNA and nasalance were highly sensitive and specific to nasalization. NNA was less affected by vowel than nasalance. Discrimination of nasal versus nonnasal stimuli using NNA and nasalance was comparable, suggesting potential for use of NNA for biofeedback applications. Future work to improve calibration of NNA is needed to lower intersubject variability.

The production and perception of coronal fricatives in Seoul Korean

This article presents new data on the contrast between the two voiceless coronal fricatives of Korean, variously described as a lenis/fortis or aspirated/fortis contrast. In utterance-initial position, the fricatives were found to differ in centroid frequency; duration of frication, aspiration, and the following vowel; and several aspects of the following vowel onset, including intensity profile, spectral tilt, and F1 onset. The between-fricative differences varied across vowel contexts, however, and spectral differences in the vowel onset especially were more pronounced for /a/ than for /i, ɯ, u/. This disparity led to the hypothesis that cues in the following vowel onset would exert a weaker influence on perception for high vowels than for low vowels. Perception data provided general support for this hypothesis, indicating that while vowel onset cues had the largest impact on perception for both high- and low-vowel stimuli, this influence was weaker for high vowels. Perception was also strongly influenced by aspiration duration, with modest contributions from frication duration and f0 onset. Taken together, these findings suggest that the ‘non-fortis’ fricative is best characterized not in terms of the lenis or aspirated categories for stops, but in terms of a unique representation that is both lenis and aspirated.

Articulatory phonetics of coronal stops in monolingual and simultaneous bilingual speakers of Canadian French and English

Previous studies of bilingual speech production have relied on individuals whose age of acquisition of their second language varies. In the proposed research, we take advantage of the unique multilingual environment of Montreal and examine speech production in individuals who have acquired two languages from birth and compare the results to monolingual speakers. Electromagnetic recordings of single-word productions were carried out on three groups of female Canadian speakers (French monolingual, English monolingual, French-English simultaneous bilingual). Spectral moment and formant transition analyses of coronal burst segments showed cross-linguistic differences across vowel contexts. Tongue place of articulation and shape likewise indicated cross-linguistic differences in static articulatory positions. Kinematic analyses further identified language-specific movement patterns that helped clarify certain results from the acoustic analyses, namely that spatiotemporal characteristics of coronal articulation help enhance vocalic dimensions important to the respective language. Similar patterns were observed among the bilingual subjects, with the notable exception that acoustic and kinematic spaces narrowed considerably, resulting in reduced overlap between languages. It appears that simultaneous bilingual speakers not only follow language-appropriate articulatory and acoustic patterns, but further minimize areas of cross-linguistic convergence otherwise found among monolingual speakers.

Effect of glottal dynamics in the production of shouted speech

In this paper characteristics of speech produced at different loudness levels are analyzed in terms of changes in the glottal excitation. Four loudness levels are considered in this study, namely, soft, normal, loud, and shout. The distinct changes in the excitation of the shout signal are analyzed using electroglottograph signals. The open and closed phases of the glottal vibration are distinctly different for shout signals, in comparison with those for normal speech. It is generally difficult to derive the glottal pulse information from the speech signal due to limitations in inverse filtering. Hence, the effects of changes in the excitation are examined by analyzing the speech signal using methods that can capture the temporal variations of the spectral features. In particular, the recently proposed methods of zero-frequency filtering and zero-time liftering are used in this analysis. It is shown that the closed phase behavior of the excitation at different loudness levels can be seen in the temporal variation of spectral energy in the low frequency (LF) (<400 Hz) region. The ratio of the LF to high frequency energy clearly discriminates the speech produced at different loudness levels. These distinctions in the excitation features are also observed in different vowel contexts and across several speakers.

On distinguishing articulatory configurations and articulatory tasks: Tamil retroflex consonants

Speech production can be described in multiple coordinate frames: articulatory configurations, gestural tasks, and acoustic patterns. Examination of the achievement of retroflex stops and liquids in Tamil suggests that we must consider separately the gestural task of apical pre-palatal constriction and the articulatory maneuver to achieve the task. The maneuver of the tongue during retroflex consonants varies across vowel contexts. Specifically, in the symmetrical intervocalic contexts between back vowels /a/ and /u/, an apical pre-palatal constriction is achieved by curling back the tongue. In the context of high front vowel /i/, a laminal pre-palatal constriction is achieved by bunching the tongue. However, the location of retroflex consonant constriction within the vocal tract is consistent across all of these vowel contexts, suggesting that the constriction task remains the same. Variation in the articulatory configuration of the retroflex in the two contexts was quantified through Gaussian curvature functions at fourteen points along the tongue, sampled at evenly spaced points throughout the vocal tract, on every other gridline of a polar-rectangular grid in every frame in each utterance. The empirical results support the notion that the articulatory configuration coordinate frame and the gestural task frame provide separate, but related, descriptions of speech production.

C-V coarticulation in velar plosives

We present 114 fps lingual ultrasound data of three speakers’ productions of words containing the initial velar plosive, and the following [a] and [i] vowels, in Mangetti Dune !Xung (N = 95 for [k] in the [a] context, N = 36 for [k] in the [i] context). We traced the midsaggital tongue edge at the frame just prior to the [k] release, and measured the tongue dorsum (TD) and tongue root (TR) constriction locations (CL’s). The TRCL was measured 1 cm below the [k] peak. Results show that [k] in the [i] context has a 1.1 cm further forward TDCL than [k] in the [a] context for one speaker, and 1.2 cm further forward for the second speaker. The results are similar to those found for English by Stevens and House (1963). The TRCL is retracted 0.3 cm in [k] in the [a] context compared with [k] in the [i] context for the first speaker, and 0.6 cm more retracted for the second speaker. A third speaker had a retracted TDCL and TRCL in both vowel contexts. These results confirm that [k] is less resistant to coarticulation. Results show that the tongue root is involved in dorsal-front vowel coarticulation.

A developmental study of vowels spoken in syllables and in sentence context

This study examined the effects of context on the production of vowels spoken by children of different ages in isolation and in a carrier sentence. Vowels were extracted from a database of hVd syllables produced by 207 native English talkers from the North Texas region, ranging in age from 5 to 18 years with approximately equal numbers of males and females. Preliminary analysis of a subset of the database (around 25% of talkers) showed a systematic reduction in vowel duration with increasing age for syllables in isolation. Vowels in sentence context were on average 30% shorter than in isolated syllables, and durations were less closely linked to age group. Formant frequencies (F1–F3) showed similar patterns for vowels in isolated syllables and sentences, and decreased as a function of age as expected. However, measures of formant movement across the vowel (from 20 to 80% of the vowel duration) revealed increased F1 and F2 movement for syllables in isolation compared to those produced in carrier sentences. A comprehensive analysis of the database will be presented and implications for vowel recognition will be discussed.

A comparative cross-linguistic study of vocal tract shaping in sibilant fricatives in English, Serbian and Mandarin using real-time magnetic resonance imaging

An articulatory study of sibilant fricatives is described, with the goal of describing variability in lingual articulation across languages. Real-time magnetic resonance imaging (rtMRI) data were collected from three speakers each of English and Mandarin and two speakers of Serbian and reconstructed at a rate of 22.4 frames per second. Parallel acoustic data were also collected and subsequently denoised. Subjects spoke the segments /s/ and /S/ in symmetrical vowel contexts (e.g., “pa sap” for English and “asa” for Mandarin). Articulation was analyzed using a semi-polar grid overlaid on the image plane and midsagittal distance functions were obtained by measuring cross distances at ~0.5 cm intervals from the glottis to the lips. Analysis shows that place of articulation for /s/ is more anterior compared to /S/ across languages. Apical articulation is observed for /s/ across languages, while /S/ is produced laminally in English and apically in Serbian and Mandarin. Patterns of tongue shaping variability differ, as well, across languages. For instance, higher standard deviation is observed anterior to the place of articulation for /S/ in Mandarin, compared to Serbian and English. [Work supported by NIH.]

Imitability of contextual vowel nasalization and interactions with lexical neighborhood density

This study investigates the imitability of contextual vowel nasalization in English. Unlike other phonetic features reported to be imitable [e.g., vowel formants (Babel, 2012), VOT (Nielsen, 2011)], vowel nasality is non-contrastive in English. Nasality is, however, systematically variable: words from dense lexical neighborhoods (high-ND words) are produced with greater nasality than words from sparse neighborhoods [Scarborough (2004), (2012)]. Two experiments were conducted to test (1) whether (experimentally manipulated) nasality can be imitated in a shadowing task, and (2) whether direction of manipulation (more or less nasality, enhancing or countering natural neighborhood-conditioned patterns) affects shadowing behavior. Subjects shadowed 16 high-ND words (which are naturally more nasal) containing a vowel-nasal sequence and modified by spectral mixing to exhibit either greater-than-natural (experiment 1) or less-than-natural (experiment 2) nasality. Both the increase and the decrease in nasality were imitated (though not overall degree of nasality, as our imitation model was more nasal in both conditions than any of our subjects). This change persisted into a post-shadowing task for just the less-nasal condition. These results indicate that speakers are sensitive to non-contrastive phonetic detail in nasality, affecting their subsequent production. Further, the naturalness of nasality (reflecting neighborhood-conditioned variation) may affect the pattern of imitation.

Modeling Consonant-Vowel Coarticulation for Articulatory Speech Synthesis

A central challenge for articulatory speech synthesis is the simulation of realistic articulatory movements, which is critical for the generation of highly natural and intelligible speech. This includes modeling coarticulation, i.e., the context-dependent variation of the articulatory and acoustic realization of phonemes, especially of consonants. Here we propose a method to simulate the context-sensitive articulation of consonants in consonant-vowel syllables. To achieve this, the vocal tract target shape of a consonant in the context of a given vowel is derived as the weighted average of three measured and acoustically-optimized reference vocal tract shapes for that consonant in the context of the corner vowels /a/, /i/, and /u/. The weights are determined by mapping the target shape of the given context vowel into the vowel subspace spanned by the corner vowels. The model was applied for the synthesis of consonant-vowel syllables with the consonants /b/, /d/, /g/, /l/, /r/, /m/, /n/ in all combinations with the eight long German vowels. In a perception test, the mean recognition rate for the consonants in the isolated syllables was 82.4%. This demonstrates the potential of the approach for highly intelligible articulatory speech synthesis.

Regional dialect leveling in Najdi Arabic: The case of the deaffrication of [k] in the Qaṣīmī dialect

AbstractThis study investigates the effect of linguistic and social factors (age, gender, and level of education) on the patterns of variation in the affrication of [] for [k] in the stem and suffix in the informal speech of 72 speakers of Qaṣīmī, a local dialect of Najdi Arabic, spoken in the Qaṣīm province in central Saudi Arabia. Findings indicate that affrication is significantly favored in the phonological context of front vowels, particularly the high front ones. Whereas suffix-based affrication is categorically used as [-], stem affrication is strongly correlated with the age, educational level, and gender of the speaker. In particular, older uneducated speakers from both sexes tend to maintain the use of the local variant [], whereas younger and middle-aged educated speakers, particularly women, increasingly shift toward the use of the supralocal variant [k]. The present findings are suggestive of patterns of variation that are typical in regional-dialect leveling, wherein the supralocal variant(s) associated with the major city dialect is (are) diffusing outward, at the expense of traditional and socially marked variant(s), by speakers of smaller towns' dialects. The substantial socioeconomic changes that Saudi Arabia has undergone in the last half century are suggested to have triggered and accelerated the linguistic shift.

Acoustic Characteristics and Distribution of Variants of /l/ in the Nanjing Dialect*

This paper reports acoustic characteristics of variants of /l/ and their distributions in the sound system of the Nanjing Dialect (NJ) in China. Prevocalic /l/ in NJ is reported to have different variants conditioned by the vowel context, yet no empirical evidence has been provided to support this claim. This study provides an analysis on temporal and spectral properties of allophonic variants of /l/ in the Nanjing Dialect, incorporating a set of statistical methods sequentially applied such as cluster analysis and discriminant analysis. Our approach is tested and proves valid in addressing the allophonic distribution of /l/ in the contemporary NJ. Statistical analysis and acoustic examination reveal two variants of /l/ in different vowel environments in this dialect.

Perception of Stop Onset Spectra in Chinese Children with Phonological Dyslexia

The ability to identify stop consonants from brief onset spectra was compared between a group of Chinese children with phonological dyslexia (the PD group, with a mean age of 10 years 4 months) and a group of chronological age-matched control children. The linguistic context, which included vowels and speakers, and durations of stop onset spectra were varied. Children with PD showed lower identification accuracy and exhibited a smaller vowel context effect for some stop-vowel combinations compared with the chronological age-matched control group. Further analyses revealed that the PD group had more variable response patterns, and their responses were less consistent with the acoustic characteristics of stop onset spectra. The results suggest that Chinese children with PD do not show greater sensitivity to allophonic acoustic variability compared with control children and exhibit a generally less robust response pattern to phonetic categories.

Study of the temporal pattern of nasalization in 4-12 year old cleft palate and normal Persian speaking children

Introduction : Hypernasality frequently occurs in children with cleft palate. The patterns of nasalization may differ in various languages. The one of contributive factor to oral-nasal resonance imbalance might be poor timing of velopharyngeal closure in relation to the activity of other articulators. Perceptual judgment has been regarded as the gold standard for rating hypernasality but it has been associated with poor reliability. There has been significant interest in developing acoustic measures to supplement perceptual judgment. The purpose of this study was to measure temporal patterns of nasalization differentiate 4-12 year old children with and without cleft palate and the extent to which vowel context contributes to these temporal differences. Methods and Materials : The study was conducted on 14 cleft palate with or without cleft lip (11 males and 3 females) and 14 normal children (11 males and 3 females). The total nasalization duration measures were obtained from acoustic waveforms and spectrograms acquired from speech samples (pimip, pamap and pumup) using Praat software. Data analysis was in the significant range of 95% (P < 0.05). Results: Total nasalization duration, as demonstrate by acoustic signals, showed different timing characteristics between children with cleft palate and without cleft palate and across vowel contexts (P ≤ 0.001). Conclusion: The timing parameters allow us to assume underlying temporal differences in cleft palate and normal Persian speaking children. The longer nasalization durations in children with cleft palate than normal children showed that temporal patterns delayed or deviance in children with cleft palate. Acoustic analysis of temporal pattern useful in survey resonance characteristics in cleft palate and normal Persian speaking children. Keywords: Temporal patterns, Nasalization, Cleft palate, Acoustic analysis , Velopharyngeal insufficiency

Perceptual cues in Korean fricatives

The current study explores the production and perception of two Korean fricatives - fortis [s’] and non-fortis (plain) [s]. Production data from 10 speakers were examined to investigate the acoustic cues that distinguish the two fricatives in different vowel contexts (high vowel /i/ vs. low vowel /a/). Measures included rise time, intensity, center of gravity (COG), F0, H1-H2, and CPP, as well as frication, aspiration, and subsequent vowel duration. COG and vowel duration consistently distinguished the two fricatives; additional cues varied across vowel contexts. For the /i/ context, intensity and F0 differed across the fricatives; for the /a/ context, rise time, H1-H2 and CPP differed across the fricatives. Four perceptual identification experiments were conducted with sixty native Korean listeners. Experiment 1 established that listeners can distinguish the two fricatives in intact natural syllables. In Experiments 2 and 3, listeners heard only the excised consonantal or vocalic segment. For the /a/ con...