Breathy Vowels Research Articles

Voicing of glottal consonants and non-modal vowels

Variation in voicing is common among sounds of the world’s languages: sounds that are analyzed as voiceless can undergo voicing, and those analyzed as voiced can devoice. Among voiceless glottal sounds in particular, voicing is widespread: linguists often expect the voiceless glottal stop [ʔ] and fricative [h] to be fully voiced, especially between vowels. In this study, we use audio recordings from Illustrations of the International Phonetic Alphabet published in theJournal of the International Phonetic Associationto explore the extent to which glottal consonants and non-modal (breathy and creaky) vowels differ in terms of percentage voicing and voicing intensity in three phrasal positions. We find that voiceless [h] is only slightly less voiced than voiced [ɦ] in initial position. Between two vowels, both [h] and [ɦ] are as voiced as breathy vowels. Glottal stops and creaky vowels are both characterized by high percentages of voicing, but they differ in voicing intensity: in all phrasal positions, glottal stops generally have periods of strong and weak voicing, whereas creaky vowels are strongly voiced. In contrast, vowels described as ‘rearticulated’, ‘checked’, or ‘glottalized’ show similar drops in voicing intensity to glottal stops. We interpret these results through an articulatory lens: glottal consonants and non-modal vowels are both modulations in phonation resulting from laryngeal constriction and vocal fold spreading. We argue further that, because voicing during [ʔ] and [h] is largely predictable from respiratory and prosodic constraints, many cases of [ʔ] and [h] can be considered phonetically underspecified for voicing.

Distinguishing breathy consonants and vowels in Gujarati

Abstract Across languages, the acoustic and articulatory correlates of breathiness are similar whether they are associated with consonants or with vowels. This raises the question of whether breathy consonants are confusable with breathy vowels in languages in which a phonemically breathy vowel contrasts with a phonemically modal vowel that follows a breathy-aspirated consonant, e. g. Gujarati /ba̤ɾ/ ‘outside’ vs. /bʱaɾ/ ‘burden’, respectively. We investigate the perception of a minimal triplet of Gujarati words, with a breathy vowel vs. a breathy consonant vs. an all-modal sequence, via three tasks: free-sort, AX discrimination, and picture-matching identification. Results across the three tasks indicate that breathiness is indeed confusable across the association types. Specifically, while listeners do recognize the stronger breathiness in vowels following breathy consonants, they are not necessarily able to determine whether that breathiness is associated with the vowel or the consonant. Furthermore, they do not reliably recognize the subtler breathiness of breathy vowels, which often indicates that they are the same as or an acceptable realization of an all-modal sequence (/baɾ/ ‘twelve’). This suggests a potential perceptual merger in Gujarati, despite previously-reported evidence of a robust three-way contrast in production.

Breathy Vowels Are Not Phonemic in Kedang (Eastern Indonesia)

Breathy phonation refers to the laryngeal setting where the vocal folds are less tense and make less contact than in “modal” phonation, which consequently leads to continuous leaking of voiceless airflow, giving rise to the perception of breathiness in a speech sound. In Austronesian languages, contrastive breathy segments are very rare. For the Austronesian languages of Island Southeast Asia, only one language has been reported to have phonemically breathy vowels: Kedang, a language spoken on Lembata island, in eastern Indonesia. In this paper, we revisit the earlier analysis that in Kedang, breathiness distinguishes phonemic “breathy” from “modal” vowels. Presenting evidence of distributional, acoustic, and etymological nature, we argue that the so-called breathy onsetless vowels do not appear to be similar to breathy vowels described in the literature. Their “breathy” nature may have a historical source in initial glottal consonants that were lost, but is currently used as a phonetic strategy that is intended to enhance the perceptual contrast between syllables with a phonemic glottal onset versus onsetless syllables. We also suggest that the glottal stop in Kedang is phonemic in all positions and indicate a possible historical trajectory for its development.

Breathy Vowels Are Not Phonemic in Kedang (Eastern Indonesia)

Breathy phonation refers to the laryngeal setting where the vocal folds are less tense and make less contact than in "modal" phonation, which consequently leads to continuous leaking of voiceless airflow, giving rise to the perception of breathiness in a speech sound. In Austronesian languages, contrastive breathy segments are very rare. For the Austronesian languages of Island Southeast Asia, only one language has been reported to have phonemically breathy vowels: Kedang, a language spoken on Lembata island, in eastern Indonesia. In this paper, we revisit the earlier analysis that in Kedang, breathiness distinguishes phonemic "breathy" from "modal" vowels. Presenting evidence of distributional, acoustic, and etymological nature, we argue that the so-called breathy onsetless vowels do not appear to be similar to breathy vowels described in the literature. Their "breathy" nature may have a historical source in initial glottal consonants that were lost, but is currently used as a phonetic strategy that is intended to enhance the perceptual contrast between syllables with a phonemic glottal onset versus onsetless syllables. We also suggest that the glottal stop in Kedang is phonemic in all positions and indicate a possible historical trajectory for its development.

Using H1 instead of H1–H2 as an acoustic correlate of glottal constriction

Bickley (1982) proposed H1–H2 as an index of the stronger H1 observed for breathy vowels compared to modal ones. Because harmonic amplitudes also vary by overall sound pressure level (SPL), H1 amplitude was normalized to that of H2. Studies have also shown that H1–H2 is well perceived and correlates with glottal open quotient, and thus can be used to infer the articulations leading to changes in perceived quality. Yet in some cases, H1–H2 does not distinguish two voice qualities at statistically significant levels. We claim that this might be due not to the same open quotient between two voice qualities, but to the fact that H1–H2 can be especially noisy to estimate: for example, when f0 is irregular (as in creaky voice qualities), H1–H2 often has large variance that isn’t well modeled statistically. Using several corpora, we show that H1 alone (while controlling for SPL) allows for stronger statistical differentiation of voice categories than H1–H2. We further confirm that H1 can stand in place of H1–H2 as a correlate of glottal constriction. When differentiating voice qualities acoustically, researchers should consider measuring H1 (while controlling for differences in overall SPL) instead of, or in addition to, H1–H2.Bickley (1982) proposed H1–H2 as an index of the stronger H1 observed for breathy vowels compared to modal ones. Because harmonic amplitudes also vary by overall sound pressure level (SPL), H1 amplitude was normalized to that of H2. Studies have also shown that H1–H2 is well perceived and correlates with glottal open quotient, and thus can be used to infer the articulations leading to changes in perceived quality. Yet in some cases, H1–H2 does not distinguish two voice qualities at statistically significant levels. We claim that this might be due not to the same open quotient between two voice qualities, but to the fact that H1–H2 can be especially noisy to estimate: for example, when f0 is irregular (as in creaky voice qualities), H1–H2 often has large variance that isn’t well modeled statistically. Using several corpora, we show that H1 alone (while controlling for SPL) allows for stronger statistical differentiation of voice categories than H1–H2. We further confirm that H1 can stand in place of H1–H2 ...

Contrast in vowel quality in Burmese and Thai Mon varieties: Phonetic indicators towards a restructured language

The Mon language, a register language, has a number of vowels with on-gliding phases that could eventually develop into diphthongs. These have never been acoustically studied. The objective of this study was to clarify the vowel characteristics in order to assess the possibility for change toward a restructured language. The results showed that some F1, F2, and F3 values in Burmese Mon (BM) could be used to differentiate clear from breathy vowels, especially in the on-gliding phase. F3 values as well as a tendency to diphthongization could be crucial factors in how BM vowels differ from Thai Mon (TM) vowels. Nonetheless, it is likely to take some time to justify whether the Mon language will change and whether BM will ultimately become a different language type from TM. These Mon varieties seem to be on the continuum of language change.

Breathy voice during nasality: A cross-linguistic study

In some languages, there is a diachronic correspondence between nasal and breathy sounds, whose origin is often attributed to the acoustic similarities between nasal and breathy vowels. In this study, we test whether nasal consonants and vowels are also produced with breathier voice quality than their oral counterparts in three Yi (Loloish) languages: Bo, Luchun Hani, and Southern Yi. We analyzed oral vs. nasal vowels and consonants using electroglottographic and acoustic measures of phonation. Results indicate that nasal consonants are often breathier than laterals, as are vowels following nasals when compared to vowels following oral consonants. These findings support the assumption that at least some of these nasal-breathy sound changes involve a stage in which the two articulations co-occur. We claim that the production of breathy voice quality during nasals can arise through listener misperception or phonetic enhancement. These findings also contribute to the understanding of nasality as an abstract feature that involves multiple articulations.

Tone and registrogenesis in Quiaviní Zapotec

Tone and phonation type are known to show complex interactions. I argue that breathy vowels in one Central Zapotec variety, San Lucas Quiaviní Zapotec (Otomanguean, Mexico), has resulted from an original tonal contrast between the low and rising tones (registrogenesis), based both on language-internal and comparative evidence with other closely-related Central Zapotec varieties. The case of Central Zapotec is unusual in that the direction of the sound change is from a tonal contrast to a phonation contrast, while in other known cases the direction is usually the opposite.

Cepstral Peak Sensitivity: A Theoretic Analysis and Comparison of Several Implementations

The aim of this study was to develop a theoretic analysis of the cepstral peak (CP), to compare several CP software programs, and to propose methods for reducing variability in CP estimation. Descriptive, experimental study. The theoretic CP value of a pulse train was derived and compared with estimates computed for pulse train WAV files using available CP software programs: (1) Hillenbrand's CP prominence (CPP) software (Western Michigan University, Kalamazoo, MI), (2) KayPENTAX (Montvale, NJ) Multi-Speech implementation of CPP, and (3) a MATLAB (The Mathworks, Natick, MA, version R2014a) implementation using cepstral interpolation. The CP variation was also investigated for synthetic breathy vowels. For pulse trains with period T samples, the theoretic CP is 1/2+ε/T, |ε|<0.1 for all pulse trains (ε=0 for integer T). For fundamental frequencies between 70 and 230Hz, the CP mean±standard deviation was 0.496±0.002 using cepstral interpolation and 0.29±0.03 using Hillenbrand's software, whereas CPP was 35.0±3.8dB using Hillenbrand's software and 20.5±2.7dB using KayPENTAX's software. The CP and CPP versus signal-to-noise ratio for synthetic breathy vowels were fit to a logistic model for the Hillenbrand (R(2)=0.92) and KayPENTAX (R(2)=0.82) estimators as well as an ideal estimator (R(2)=0.98), which used a period-synchronous analysis. The findings indicate that several variables unrelated to the signal itself impact CP values, with some factors introducing large variability in CP values that would otherwise be attributed to the signal (eg, voice quality). Variability may be reduced by using a period-synchronous analysis with Hann windows.

Phonation type contrasts and vowel quality in Marathi

Phonation-type contrasts in consonants and vowels are typically associated with a number of acoustic differences. Phonemic breathy phonation, for instance, is fairly consistently associated with lower Cepstral Peak Prominence values than modal sounds, and with higher values for spectral measures such as H1-H2 and H1-A3. The effect of breathy phonation on vowel quality, however, is less consistent: lower first formant (F1) values have been found for breathy vowels in some languages, while others do not show consistent F1 differences based on phonation type. Furthermore, at present little work has investigated the effect of breathy voiced consonants on formant values in subsequent vowels. The current study presents data from Marathi, an Indic language with phonemically breathy-voiced obstruents and sonorants. Ten native speakers (five males and five females) produced real words that included the vowels [a] and [e] after both modal and breathy consonants. The effect of consonant phonation type on F1 and F2 values in subsequent vowels is reported in order to augment our understanding of the relationship between voice quality differences and vowel quality.

기식 등급에 따른 CPP (Cepstral Peak Prominence) 분석 비교

The aim of this study is to synthesize pathological breathy voice and to make a cepstral peak prominence (CPP) table following breathiness ranks by cepstral analysis to supplement reliability of the perceptual auditory judgment task. KlattGrid synthesizer included in Praat was used. Synthesis parameters consist of two groups, i.e., constants and variables. Constant parameters are pitch, amplitude, flutter, open phase, oral formant and bandwidth. Variable parameters are breathiness (BR), aspiration amplitude (AH), and spectral tilt (TL). Five hundred sixty samples of synthetic breathy vowel /a/ for male were created. Three raters participated in ranking of the breathiness. 217 were proved to be inadequate samples from perceptual judgment and cepstral analysis. Finally, 343 samples were selected. These CPP values and other related parameters from cepstral analysis are classified under four breathiness ranks (B0~B3). The mean and standard deviation of CPP is <TEX>$16.10{\pm}1.15$</TEX> dB(B0), <TEX>$13.68{\pm}1.34$</TEX> dB(B1), <TEX>$10.97{\pm}1.41$</TEX> dB(B2), and <TEX>$3.03{\pm}4.07$</TEX> dB(B3). The value of CPP decreases toward the severe group of breathiness because there is a lot of noise and a small quantity of harmonics.

The Perception of Pathologically-disordered Phonation by Gujarati, English, and Spanish Listeners

This study investigates the influence of linguistic experience on the perception of pathologically-disordered voices using 18 listeners of American English, which has allophonic breathiness, 12 listeners of Gujarati, which contrasts breathy and modal vowels, and 18 listeners of Spanish, which has neither allophonic nor phonemic breathiness. Listeners rated the similarity of pairs of pathologically-disordered voices. Multidimensional scaling was used to determine the properties that were most correlated with perception for each listener group. Results showed that Gujaratis' perception was correlated with the difference between the amplitude of the first (HI*) and second (H2*) harmonic (HI*-H2*), which is associated with the production of phonation in Gujarati. English listeners' judgments were correlated with the measure HI*-H2* and cepstral peak prominence, and Spanish listeners' judgments were correlated with HI*-H2 and HI*-AI* (the amplitude of the principal harmonic near the first formant). When compared to Esposito (2006), which asked the same listeners to rate the similarity of breathy and modal vowels from Mazatec, results showed that Gujarati listeners classified the pathologically-disordered stimuli in the same way that they classified the Mazatec stimuli, while English and Spanish listeners perceived the pathologically-disordered stimuli and the Mazatec stimuli in slightly different ways.

Contrastive breathiness on consonants and vowels.

While numerous languages exhibit contrastive breathy phonation either on consonants (e.g., Indic languages) or on vowels (e.g., Zapotec languages); very few languages preserve this contrast across both consonants and vowels. Two such languages are Gujarati and White Hmong. Given that breathiness on consonants is typically realized as a breathy-voiced release into the following vowel, how are the two types of breathiness distinguished in CV sequences, if at all? We examine acoustic and electroglottographic data collected from speakers of Gujarati and White Hmong to determine what properties reliably distinguish breathy and modal voice in potentially ambiguous CV sequences, and to explore the phonetic and phonological properties shared between these two genetically unrelated languages. Preliminary results from both languages are strikingly similar; modal vowels adjacent to breathy consonant releases strongly resemble phonemically breathy vowels, and only the exact timing and magnitude of the acoustic reflexes of breathiness phonetically distinguish the two sequences. Phonological evidence in the form of distributional constraints, however, can aid listeners of both languages in determining the source of breathiness: both languages prohibit the co-occurrence of breathy consonants and breathy vowels, and in White Hmong, tonal cues can also reveal the underlying phonation of an ambiguous CV sequence.

Breathy phonation in Gujarati: An acoustic and electroglottographic study.

Recent electroglottographic research on Gujarati [Khan (2009)] reveals that speakers consistently distinguish breathy and modal vowels (e.g., /bar/ “twelve” vs /bAr/ “outside;” capitalization represents breathiness) by both closed quotient and closing velocity. Despite this interspeaker uniformity measured at the voice source, the acoustic correlates of the contrast measured in the speech output vary considerably across speakers. While some speakers mark breathiness with a larger H1-H2 difference, others use lower periodicity (CPP) or greater changes in intensity (rms energy). It is unclear how breathy vowels are contrasted with modal vowels following breathy consonants (e.g., /bAr/ “outside” vs /Bar/ “burden”) or to what extent the preceding consonant manner can affect the spectral acoustics of the vowel (e.g., /pAr/ “mountain” vs /vAn/ “vehicle”). To better capture the variation in both vowel and consonant phonations, the current study examines acoustic and electroglottographic data collected from naturalistic productions of near-minimal sets from a larger pool of Gujarati speakers, using a wider range of vowel and consonant types. Preliminary results suggest that while all speakers distinguish phonation types in both consonants and vowels based on electroglottographic measures, the acoustic cues in the output are far more complex. [Work supported by NSF.]

An acoustic and electroglottographic study of breathy phonation in Gujarati.

While it has long been established that breathy and modal vowels in Gujarati can be reliably distinguished based on the H1-H2 measure [Fischer-Jorgensen (1967); Bickley (1982)], and that Gujarati listeners attend solely to H1-H2 when distinguishing phonation types in Gujarati and in other languages [Bickley (1982); Esposito (2006)], new research [Khan and Thatte (2009)] suggests that in more casual speech styles, H1-H2 may play a smaller role in distinguishing breathy-modal minimal pairs. In such speech styles, additional acoustic measures (e.g., H1-A3, CPP, and rms energy) are more effective in distinguishing breathy and modal vowels for some speakers. To more closely examine the phonetics of breathy vowels in casual speech, the current study examines both acoustic and electroglottographic data collected from naturalistic productions of breathy-modal minimal pairs. Preliminary data suggest that while all speakers distinguish modal and breathy vowels, the strategies used by each speaker to produce the contrast vary considerably. [Work supported by NSF.]

A produção de plosivas por crianças de três anos falantes do português brasileiro

OBJETIVO: apresentar as características fonético-acústicas de produção das consoantes plosivas vozeadas e não-vozeadas por crianças de três anos. MÉTODOS: análise qualitativa baseada na inspeção das características fonético-acústicas de eventos fônicos ocorridos no intervalo de obstrução, na liberação da plosão e na transição para a vogal e análise quantitativa das medidas do Voice Onset Time (VOT), de três repetições das seis consoantes plosivas do Português Brasileiro (PB), por quatro crianças, sendo três meninos e uma menina, falantes nativos do PB, por meio de análise fonético-acústica, em espectrograma de banda larga. Os achados foram comparados com estimativas de valores para a faixa adulta de falante nativo do PB. RESULTADOS: as crianças apresentaram produções semelhantes às que são encontradas na fala adulta, tanto para as plosivas vozeadas como para as não-vozeadas, bilabiais, alveolares e velares. Apresentaram também plosivas vozeadas com interrupção de barra de sonoridade e ocorrência de VOT positivo, e qualidade diferenciada de burst. Para as não-vozeadas, ocorreram a presença da breathy vowel no período de obstrução das plosivas e a presença de aspiração no segmento correspondente ao VOT positivo, para os três pontos de articulação pesquisados. CONCLUSÃO: as diferenças encontradas na produção das plosivas vozeadas, quanto às características da barra de sonoridade e qualidade do burst, a presença da breathy vowel nas plosivas não-vozeadas e aspiração, caracterizam a produção das plosivas nessa faixa etária. A diversidade das medidas acústicas, verificada pelos valores dos desvios-padrão sugerem que as crianças realizaram diversas tentativas, com diferentes ajustes para produzir tais sons.

Use of temporal information: detection of periodicity, aperiodicity, and pitch in speech

In this paper, we present a time domain aperiodicity, periodicity, and pitch (APP) detector that estimates 1) the proportion of periodic and aperiodic energy in a speech signal and 2) the pitch period of the periodic component. The APP system is particularly useful in situations where the speech signal contains simultaneous periodic and aperiodic energy, as in the case of breathy vowels and some voiced obstruents. The performance of the APP system was evaluated on synthetic speech-like signals corrupted with noise at various levels of signal-to-noise ratio (SNR) and on three different natural speech databases that consist of simultaneously recorded electroglottograph (EGG) and acoustic data. When compared on a frame basis (at a frame rate of 2.5 ms) the results show excellent agreement between the periodic/aperiodic decisions made by the APP system and the estimates obtained from the EGG data (94.43% for periodicity and 96.32% for aperiodicity). The results also support previous studies that show that voiced obstruents are frequently manifested with either little or no aperiodic energy, or with strong periodic and aperiodic components. The EGG data were used as a reference for evaluating the pitch detection algorithm. The ground truth was not manually checked to rectify or exclude incorrect estimates. The overall gross error rate in pitch prediction across the three speech databases was 5.67%. In the case of synthetic speech-like data, the estimated SNR was found to be in close proportion to the actual SNR, and the pitch was always accurately found regardless of the presence of any shimmer or jitter.

The role of linguistic experience on the perception of phonation

This study investigates the role linguistic experience has on the perception of phonation and the acoustic properties that correlate with this perception. Listeners from three languages (Gujarati, which contrasts breathy versus modal vowels, Italian, which has no breathiness, and English, which has allophonic breathiness) participated in two tasks. In the Visual Sort task, breathy and modal vowels from a variety of languages (e.g., Chong, Mon, etc.) were presented as icons on a computer screen. Subjects sorted the icons (the stimuli) into two groups based on perceived similarity of the talker’s voices. In the multidimensional scaling task, listeners heard pairs of Mazatec vowels, and moved an on-screen slider to indicate the perceived similarity of the vowels in each pair. Results will show that judgments were more uniform across subjects who had breathy categories present in their native language(s), and varied across subjects who lack a breathy category. The perceived similarity among the stimuli will correlate with a measurable acoustic property (H1-H2, H1-H2, H1-A3, H1-A1 or H1-A2, the average of H1-H2 compared to A1, and A2-A3). It is hypothesized that H1-H2 will be the most salient acoustic property for Gujarati listeners because this correlates with their production of breathy vowels.

A measure of aperiodicity content in a speech signal

Most of the current aperiodicity detectors measure aperiodicity indirectly, in which the absence of periodic component in a nonsilent region is termed as aperiodicity. Such indirect measurements are inadequate and can be misleading especially in cases like voiced fricatives or breathy vowels. This motivated us to develop a direct measure of aperiodicity which is independent of the periodicity measure. The speech signal is passed through a 60-channel gamma tone auditory filterbank. Average Magnitude Difference Function (AMDF) is computed on the envelope of each channel. The randomness in the distribution of the AMDF dips is the basis for the measure of the aperiodicity whereas the measure of the periodicity is based on the occurrence of the dips at multiple locations. The system was evaluated on the MOCHA database, which has simultaneous recording of the EGG data, and on the TIMIT database. Preliminary analysis shows that the aperiodicity and voicing accuracy on a per frame basis are 95% and 90.3%, respectively. In voiced fricatives and voiced stops, high aperiodicity and high periodicity were detected in 21.6% of the frames. Note that not all these sounds necessarily had both the sources prominent.

The timing of nonmodal phonation in vowels

In a language where breathiness or laryngealization is a contrastive property of vowels, such nonmodal phonation lasts longer and is more differentiated from modal phonation than in a language where nonmodal phonation results from the influence of preceding consonants. In Tagalog, breathy phonation occurs incidentally on vowels after /h/, and laryngealized phonation occurs after glottal stops. Mazatec, on the other hand, employs breathy and laryngealized vowels as separate phonemes that contrast with modal vowels. Several acoustic measures show that nonmodal and modal vowels are differentiated more strongly and over a longer duration in Mazatec than in Tagalog.An experiment examined words from six male and six female speakers of each of those languages, with corroborating modal and breathy vowels from four male and four female speakers of Chong, and modal and laryngealized vowels from one male speaker of Mpi. From each speaker, three vowels of each phonation type were analyzed. To determine the time course of phonation effects, measurements were made at 25 ms intervals through each vowel. The measurements were the amplitude differences between the first and second harmonic and between the first harmonic and the second formant, and cepstral peak prominence (a measure of periodicity).