Release Burst Research Articles

Formulation and optimization of a modified microporous cellulose acetate latex coating for osmotic pumps.

A cellulose acetate (CA) latex was modified for use as a microporous coating for osmotic devices. Potassium chloride core tablets were coated with a CA latex formulation containing a plasticizer (triacetin) and a pore-forming agent (urea). To promote the coalescence of the cellulose acetate latex beads into a film on the surface of the tablet, it was necessary to cure the coated tablets, hereafter referred to as devices, at elevated temperatures. The objectives were to determine the effect of four formulation variables (plasticizer level, pore former level, cure time, and cure temperature) on the in vitro KCl release rate and coat burst strength using a full 2(4) factorial experimental design. Burst strength was measured as the number of grams force a depleted device could support before bursting. The results indicated that urea content was the most important variable, followed by triacetin content and cure time. Cure temperature did not influence the results. Response surfaces generated with the experimental values were used to predict a formulation which would have both a high release rate and a high burst strength. This formulation was prepared and tested both in vitro and in vivo in dogs. The in vitro release rate and burst strength results agreed with those predicted by the model. The in vitro and in vivo release rates were not statistically significantly different as determined by ALQ analysis.

Accounting for normal‐hearing listeners' perceptions of syllable‐initial /d/'s spoken by hearing‐impaired talkers

A substantial corpus of data on normal‐hearing listeners' perceptions of stop and glide consonants spoken by sensorineural hearing‐impaired talkers of widely varying speech production abilities has been collected. This report concerns the authors' first efforts to measure acoustic cues to consonants in these productions and to relate cue presence to perceptual judgments. The focus was on /d/, as produced in citation‐form /dVk/ syllables (V = /ɑ/, /æ), by ten severely through profoundly hearing‐impaired talkers representing the midrange of speech production abilities among our hearing‐impaired adults. The /dVk/ syllables, along with other syllables contrastive for the initial consonant, were presented to normal‐hearing listeners who made identification judgments for the syllable‐initial consonant and, at the same time, rated their confidence in each judgment. Acoustic analysis of the /dVk/ syllables generally revealed some form of measurable release burst at onset; however, the bursts were often abnormal for /d/ in either their onset characteristics, amplitude, duration, or spectrum. The strength and quality of the talkers' vocalic cues to /d/ idiosyncratically depended upon the identity of the syllable vowel. A token‐by‐token analysis of the perception data showed that some constellations of /d/ cues yielded relatively high identification accuracy, others yielded high confidence, but only a small subset yielded both. [Work supported by NIH and the Gallaudet Research Institute.]

Production of the English word-final /t/-/d/ contrast by native speakers of Mandarin and Spanish

Previous second-language (L2) research has shown that if the voicing of word-final obstruents exerts little or no effect on preceding vowel duration in a language, then native speakers of that language will produce a far smaller vowel duration difference in isolated English words like beat and bead than do native speakers of English. This might be due to phonetic interference, resulting from the perceptual identification of word-final consonants in English and the L1. If so, one would expect native speakers of languages without word-final obstruents to succeed better in acquiring the vowel duration cue to the word-final English /t/ vs /d/ contrast. To test this, native speakers of English, Spanish, and Chinese read lists of minimally paired /bVt/ and /bVd/ words. Native English-speaking listeners later identified far more final stops produced by native speakers of English than Spanish or Mandarin (95% vs 72%, 63%). When closure voicing and release burst cues were removed, the correct identification rates remained well above chance for the native but not the non-native speakers (88% vs 46%, 58%). This was probably because the vowel duration difference in words ending in /t/ and /d/ was far greater in words spoken by the native speakers of English than by the native speakers of Spanish and Chinese (87 vs 33, 34 ms). Also, F1 offset frequency was significantly higher in /t/- than /d/-final words spoken by the native but not by the non-native speakers. The results suggest that the inability of adult L2 learners to master the vowel duration cue to the English word-final /t/-/d/ contrast is not the result of cross-language phonetic interference. Some nonperceptual explanation must be found to account for this seeming limitation.

Acoustic correlates of the front/back vowel distinction: A comparison of transition onset versus ‘‘steady state’’

This study investigated whether F2 and F3 transition onsets could encode the vowel place feature as well as F2 and F3 "steady-state" measures [Syrdal and Gopal, J. Acoust. Soc. Am. 79, 1086-1100 (1986)]. Multiple comparisons were made using (a) scatterplots in multidimensional space, (b) critical band differences, and (c) linear discriminant functional analyses. Four adult male speakers produced /b/(v)/t/, /d/(v)/t/, and /g/(v)/t/ tokens with medial vowel contexts /i,I, E, ey, ae, a, v, c, o, u/. Each token was repeated in a random order five times, yielding a total of 150 tokens per subject. Formant measurements were taken at four loci: F2 onset, F2 vowel, F3 onset, and F3 vowel. Onset points coincided with the first glottal pulse following the release burst and steady-state measures were taken approximately 60-70 ms post-onset. Graphic analyses revealed two distinct, minimally overlapping subsets grouped by front versus back. This dichotomous grouping was also seen in two-dimensional displays using only "onset" data as coordinates. Conversion to a critical band (bark) scale confirmed that front vowels were characterized by F3-F2 bark differences within a critical 3-bark distance, while back vowels exceeded the 3-bark critical distance. Using the critical distance metric onset values categorized front vowels as well as steady-state measures, but showed a 20% error rate for back vowels. Front vowels had less variability than back vowels. Statistical separability was quantified with linear discriminant function analysis. Percent correct classification into vowel place groups was 87.5% using F2 and F3 onsets as input variables, and 95.7% using F2 and F3 vowel. Acoustic correlates of the vowel place feature are already present at second and third formant transition onsets.

Coarticulation effects in American English velar and alveolar stop consonants: An electropalatographical and acoustical analysis

Coarticulation effects of post-consonantal vowels on the place of tongue contact for American English velar and alveolar stop consonants were examined through simultaneous measurements with a new type of electropalatography (EPG) and with spectral analysis. The EPG system, developed by the author, has 49 electrodes, and is used to observe lingual contact with the soft palate, hard palate, and upper incisal teeth. It can record the palatograms at a 7.8-ms frame interval. The resulting palatograms can be compared with acoustic spectra made from simultaneous speech recordings. One American male subject produced VCV utternances in which the consonants were /k/, /g/, /t/, and /d/, and the vowels were /a/, /o/, /u/, /e/, and /i/. The results confirm that the tongue contact position for American English velar stop consonants is more strongly influenced by the vowels following it than for alveolar stop consonants. The value of F2 around the release burst was consistent with the strong coarticulation effects observed with the EPG.

Medial voicing distinctions in English trochees

Phonetic studies of distinctive consonantal voicing in English have focused on utterance‐initial and final positions. Much less attention has been given to medial stops following stressed and preceding unstressed vowels, as in pairs like tugging/tucking. The high frequency of occurrence of unaspirated voiceless stops in this context is commonly overlooked because of the usual emphasis on the aspiration of/p t k/in stressed syllables. The voiced stops in this context are likely to have glottal pulsing through much or all of the closure, unlike their utterance‐initial counterparts, which may not show glottal pulsing until the moment of release. For this study, a number of such word pairs were recorded in carrier sentences at a normal rate by three native speakers of American English: Three acoustic features were measured: voice timing, closure duration, and amplitude of release burst. Values of the first two features were clearly different for the two categories. With respect to the third feature, burst amplitude, the two categories differed much less clearly. To generate sets of stimuli for perception testing, the three features were covaried incrementally by waveform editing. Labeling data collected from a jury of native American English listeners will be presented to show the relative efficacy of these features for perception.

Native-language phonotactic constraints affect how well Chinese subjects perceive the word-final English /t/-/d/ contrast

A contrast between /t/ and /d/ exists in the initial but not the final position of Chinese words. This study examined identification of English word-final /t/ and /d/ tokens from which closure voicing and release burst cues had been removed. The performance of three Chinese groups was compared before, during, and after feedback training. The Cantonese subjects were expected to perform best because their L1 permits unreleased /p, t, k/ in word-final position, and the Mandarin subjects were expected to perform most poorly because their L1 permits no word-final obstruents. An intermediate level of performance was expected from the Shanghainese subjects, whose L1 permits /?/ word-finally. As predicted, the Cantonese subjects were significantly more sensitive to the English /t/–/d/ contrast than the Mandarin subjects, with the Shanghainese subjects showing an intermediate level of performance. The subjects in all three groups showed a significant increase in sensitivity as a result of the training. The increase in correct identifications averaged 25%. The results were interpreted to mean that native-language phonotactic constraints influence how syllables are processed. More specifically, it is concluded that the Cantonese subjects focused greater attention on the end of the consonant–vowel–consonant stimuli than the Mandarin subjects, which better enabled them to use remaining acoustic cues to the /t/–/d/ contrast such as F 1 offset frequency.

Acoustic properties and perception of stop consonant release transients.

This study focuses on the initial component of the stop consonant release burst, the release transient. In theory, the transient, because of its impulselike source, should contain much information about the vocal tract configuration at release, but it is usually weak in intensity and difficult to isolate from the accompanying frication in natural speech. For this investigation, a human talker produced isolated release transients of /b,d,g/ in nine vocalic contexts by whispering these syllables very quietly. He also produced the corresponding CV syllables with regular phonation for comparison. Spectral analyses showed the isolated transients to have a clearly defined formant structure, which was not seen in natural release bursts, whose spectra were dominated by the frication noise. The formant frequencies varied systematically with both consonant place of articulation and vocalic context. Perceptual experiments showed that listeners can identify both consonants and vowels from isolated transients, though not very accurately. Knowing one of the two segments in advance did not help, but when the transients were followed by a compatible synthetic, steady-state vowel, consonant identification improved somewhat. On the whole, isolated transients, despite their clear formant structure, provided only partial information for consonant identification, but no less so, it seems, than excerpted natural release bursts. The information conveyed by artificially isolated transients and by natural (frication-dominated) release bursts appears to be perceptually equivalent.

Storage temperature and time affect dormancy release and bud burst in one-year-old seedlings of Acer platanoides

Storage temperature and time affect dormancy release and bud burst in one-year-old seedlings of Acer platanoides

Location and classification of plosive consonants using expert knowledge and neural net classifiers

A rule-based segmentation and broad classification algorithm [R. A. Cole and L. Hou, Proc. ICASSP 88, 453–456 (1988)] located over 95% of segments labeled [b], [d], [g], [p], [t], [k], [ch], [jh], [dr], and [q] (glottal stop) before sonorants in utterances of the DARPA TIMIT database. Artificial neural net (ANN) classifiers were trained to discriminate among the labels using perceptually motivated features. In one condition, 37 feature measurements were used to describe (a) the averaged spectrum during the 15 ms following the release burst, (b) zero crossings and peak-to-peak amplitude contours in the region of the segment, (c) the duration of the segment, and (d) the amplitude of the plosive burst. In a second condition, an additional 16 features were used to characterize the averaged spectrum during the first 30 ms of the sonorant following the plosive. The ANN classifiers consisted of either 37 or 53 input units, 30 hidden units, and 1 output unit for each category. Classifiers were trained using backpropagation and tested on 2000 segments provided by 20 speakers. With different amounts of training, classification accuracy was consistently 3%–5% better when vowel spectra were used, suggesting that ANN classifiers are able to learn coarticulatory relationships between consonant and vowel spectra. Classification accuracy was 70% for the 10 plosive consonants.

Acoustic properties of isolated stop consonant release transients

This study focuses on the initial component of the stop consonant release burst, the release transient. In theory, the transient, because of its impulselike source, should contain much information about the vocal tract configuration at release, but it is usually weak in intensity and difficult to isolate from the accompanying frication in natural speech. For this investigation, a human talker produced isolated release transients of /b,d,g/ in nine vocalic contexts by “mouthing” these syllables. He also produced the corresponding CV syllables with regular phonation for comparison. Spectral analyses showed the isolated transients to have a clearly defined formant structure, which was not seen in natural release bursts whose spectrum was dominated by the frication noise. The formant frequencies varied systematically with both consonant place of articulation and vocalic context, with the latter, “coarticulatory” influence actually exceeding the former. The transient thus provides an acoustic snapshot of the vocal tract immediately after the release. [Work supported by NICHD.]

Release burst masking effects for temporal order identification

Nearly three decades ago, Hirsh [J. Acoust. Soc. Am. 31, 759–761 (1959)] found that the threshold for identification of temporal order of onset is significantly longer than the threshold for detection of onset asynchrony, and proposed that the order identification threshold might serve as a basis for the contrast between voiced and unvoiced aspirated stop consonants in initial position. In the last decade, a number of researchers have argued that temporal order identification threshold does not play a significant role in voicing contrast. One basis for this argument against a phonetic perception role is that the 15- to 20-ms range of typical order identification thresholds for simple stimuli are significantly shorter than VOT boundaries typical for synthetic stop consonants which include an initial release burst. The current research investigated the possible contribution of an initial noise burst analog as a masking stimulus. The initial noise burst shifts temporal order identification thresholds to significantly longer onset asynchronies which are equivalent in magnitude to VOT voicing contrast boundaries.

A differential effect of release bursts on the stop voicing judgments of native French and English listeners

Removing the release burst from original word-final English /g/ tokens influenced the stop voicing judgments of native French but not native English subjects. The results suggest that the French subjects gave greater weight to release burst information because stops are usually released in the word-final position of French words.

Acoustic analysis of amyotrophic lateral sclerosis speech

Seventeen victims of amyotrophic lateral sclerosis (ALS), were recorded as they produced /p t k b d g/ in a CVC syllable embedded in a carrier phrase. The duration of the stop consonant closure period, the release burst, and the aspirated phase of each of the stops, as well as the total duration of all pre‐ and post‐stop vowels was determined spectrographically. In general, the ALS speakers produced highly varied speech which included numerous instances of voicing throughout voiceless segments, frication of voiced and voiceless stops, and multiple release bursts. Comparisons of group mean durations, to those of normal speakers, indicates that the effects of decreased speaking rate in ALS are not evenly distributed across segments or portions of segments. Instead, some segment durations in ALS speech remain within normal limits while others increase relative to those seen in normal speech. Thus, the most sensitive acoustic/articulatory measure of the speech deterioration associated with ALS is achieved by comparing two or more durational measures with speakers. [Work supported by Kansas City Regional ALS Research Center, and NIH.]

Some observations on the development of anticipatory coarticulation.

The influence of vocalic context on various temporal and spectral properties of preceding acoustic segments was investigated in utterances containing [schwa No. CV] sequences produced by two girls aged 4;8 and 9;5 years and by their father. The younger (but not the older) child's speech showed a systematic lowering of [s] noise and [th] release burst spectra before [u] as compared to [i] and [ae]. The older child's speech, on the other hand, showed an orderly relationship of the second-formant frequency in [] to the transconsonantal vowel. Both children tended to produce longer [s] noises and voice onset times as well as higher second-formant peaks at constriction noise offset before [i] than before [u] and [ae]. All effects except the first were shown by the adult who, in addition, produced first-formant frequencies in [] that anticipated the transconsonantal vowel. These observations suggest that different forms of anticipatory coarticulation may have different causes and may follow different developmental patterns. A strategy for future research is suggested.

Identification of Hindi Stop Consonants in Syllable-Final Position

An experiment using CVC syllables was performed to study the relative importance of various acoustic cues for the identification of Hindi stop consonants in syllable-final position. The results indicate that the VC transition is not sufficient for place and voicing identification and that the voiced release burst of a voiced stop is a stronger perceptual cue than the voiceless release burst of a voiceless stop. All three cues (vowel formant transition, closure/voice bar, and release burst) are needed for the maximal recognition of unaspirated stops whereas, in the case of their aspirated counterparts, the removal of the VC transition from stimuli containing all three cues—vowel transition, closure/voice bar, and aspiration noise—causes an enhancement in correct scores due to an increase in the identifiability of the aspiration feature, the place scores remaining unaltered. It is also found that aspiration noise (including release burst) is a sufficient cue for the place identification of the aspirates. Voicing identification of voiced stops requires the presence of the occlusion interval filled with glottal pulsations.

The vowel-dependence of gross spectral cues to place of articulation of stop consonants in CV syllables

It has been argued by Stevens and Blumstein that the acoustic cues of stop consonant place of articulation residing in the gross spectrum, sampled at the release burst onset, are invariant across different vowel contexts. It is shown in this paper that the conclusion of in variance does not follow from the experimental findings of the proponents of the theory of acoustic invariance, and the present results indicate that the vowel exerts a reliable influence on the stop–vowel waveform from the very beginning in terms of gross spectral properties. Also, the consonant influence is seen to be strong throughout the 45 ms interval from the beginning of the CV syllable. Implications of these coarticulatory effects to theories of speech perception are discussed.

Release Burst and Formant Transition as Acoustic Cues for English Stop Consonants Perception by Native Speakers of Indian Languages

Acoustic cues, burst and formant transition, are investigated for English stop consonant perception by native speakers of Indian languages. Synthesized initial stop CV syllables are used. Burst is found to be usually of low intensity and nearly invariant for labials and apicals. It is not at all important in perception of labials, but carries a significant perceptual weight for velars, and is a sufficient cue for /di/. Formant transition though perceptually very important is not a sufficient cue for velars /gV/, /di/ & /de/. These results offer further evidence to perceptual importance of front cavity resonance, serving as a simplifying reference to articulation. Hence pointing at the functional equivalence of the two cues. Both providing a spectrally continuous change from the release into the following vowel. In general we suggest that burst carries perceptual weight only when its spectrum is near to the front cavity resonance of the vowel. And formant transition though very important, is a sufficient c...

The detection of French accent by American listeners.

The five experiments presented here examine the ability of listeners to detect a foreign accent. Computer editing techniques were used to isolate progressively shorter excerpts of the English spoken by native speakers of American English and French. Native English-speaking listeners judged the speech samples in one- and two-interval forced-choiced tests. They were able to detect foreign accent equally well when presented with speech edited from phrases read in isolation and produced in a spontaneous story. The listeners accurately identified the French talkers (63%-95% of the time) no matter how short were the speech samples presented: entire phrases (e.g., "two little dogs"), syllables (/tu/ or /ti/), portions of syllables corresponding to the phonetic segments /t/,/i/,/u/, and even just the first 30 ms of "two" (roughly, the release burst of /t/). Both phonetically trained listeners familiar with French-accented English and unsophisticated listeners were able to accurately detect accent. These results suggest that listeners develop very detailed phonetic category prototypes against which to evaluate speech sounds occurring in their native language.

Closure duration and release burst amplitude cues to stop consonant manner and place of articulation.

The perception of stop consonants was studied in a constant neutral [s-1] context. Truncated natural [p] , [t], and [k] release bursts at two intensities were preceded by variable silent closure intervals. The bursts, though spectrally distinct, conveyed little specific place information but contributed to the perception of stop manner by reducing the amount of silence required to perceive a stop (relative to a burstless stimulus). Burst amplitude was a cue for both stop manner and place; higher amplitudes favored "t," lower amplitudes favored "p" responses. The silent closure interval, a major stop manner cue, emerged as the primary place cue in this situation: Short intervals led to "t," long ones to "p" responses. All these perceptual effects probably reflect listeners' tacit knowledge of systematic acoustic differences in natural speech.