The stop-glide distinction: acoustic analysis and perceptual effect of variation in syllable amplitude envelope for initial /b/ and /w/

Amplitude change at consonantal release has been proposed as an invariant acoustic property distinguishing between the classes of stops and glides [Mack and Blumstein, J. Acoust. Soc. Am. 73, 1739-1750 (1983)]. Following procedures of Mack and Blumstein, we measured the amplitude change in the vicinity of the consonantal release for two speakers. The results for one speaker matched those of Mack and Blumstein, while those for the second speaker showed some differences. In a subsequent experiment, we tested the hypothesis that a difference in amplitude change serves as an invariant perceptual cue for distinguishing between continuants and noncontinuants, and more specifically, as a critical cue for identifying stops and glides [Shinn and Blumstein, J. Acoust. Soc. Am. 75, 1243-1252 (1984)]. Interchanging the amplitude envelopes of natural /bV/ and /wV/ syllables containing the same vowel had little effect on perception: 97% of all syllables were identified as originally produced. Thus, although amplitude change in the vicinity of consonantal release may distinguish acoustically between stops and glides with some consistency, the change is not fully invariant, and certainly does not seem to be a critical perceptual cue in natural speech.     

An auditory basis for the stimulus-length effect in the perception of stops and glides

To investigate possible auditory factors in the perception of stops and glides (e.g., /b/ vs /w/), a two-category labeling performance was compared on several series of /ba/-/wa/ stimuli and on corresponding nonspeech stimulus series that modeled the first-formant trajectories and amplitude rise times of the speech items. In most respects, performance on the speech and nonspeech stimuli was closely parallel. Transition duration proved to be an effective cue for both the stop/glide distinction and the nonspeech distinction between abrupt and gradual onsets, and the category boundaries along the transition-duration dimension did not differ significantly in the two cases. When the stop/glide distinction was signaled by variation in transition duration, there was a reliable stimulus-length effect: A longer vowel shifted the category boundary toward greater transition durations. A similar effect was observed for the corresponding nonspeech stimuli. Variation in rise time had only a small effect in signaling both the stop/glide distinction and the nonspeech distinction between abrupt and gradual onsets. There was, however, one discrepancy between the speech and nonspeech performance. When the stop/glide distinction was cued by rise-time variation, there was a stimulus-length effect, but no such effect occurred for the corresponding nonspeech stimuli. On balance, the results suggest that there are significant auditory commonalities between the perception of stops and glides and the perception of acoustically analogous nonspeech stimuli.     

The influence of noise on vowel and consonant cues

This study assessed the acoustic and perceptual effect of noise on vowel and stop-consonant spectra. Multi-talker babble and speech-shaped noise were added to vowel and stop stimuli at -5 to +10 dB S/N, and the effect of noise was quantified in terms of (a) spectral envelope differences between the noisy and clean spectra in three frequency bands, (b) presence of reliable F1 and F2 information in noise, and (c) changes in burst frequency and slope. Acoustic analysis indicated that F1 was detected more reliably than F2 and the largest spectral envelope differences between the noisy and clean vowel spectra occurred in the mid-frequency band. This finding suggests that in extremely noisy conditions listeners must be relying on relatively accurate F1 frequency information along with partial F2 information to identify vowels. Stop consonant recognition remained high even at -5 dB despite the disruption of burst cues due to additive noise, suggesting that listeners must be relying on other cues, perhaps formant transitions, to identify stops.     

Effect of burst amplitude on the perception of stop consonant place of articulation

We have examined the effects of the relative amplitude of the release burst on perception of the place of articulation of utterance-initial voiceless and voiced stop consonants. The amplitude of the burst, which occurs within the first 10-15 ms following consonant release, was systematically varied in 5-dB steps from -10 to +10 dB relative to a "normal" burst amplitude for two labial-to-alveolar synthetic speech continua--one comprising voiceless stops and the other, voiced stops. The distribution of spectral energy in the bursts for the labial and alveolar stops at the ends of the continuum was consistent with the spectrum shapes observed in natural utterances, and intermediate shapes were used for intermediate stimuli on the continuum. The results of identification tests with these stimuli showed that the relative amplitude of the burst significantly affected the perception of the place of articulation of both voiceless and voiced stops, but the effect was greater for the former than the latter. The results are consistent with a view that two basic properties contribute to the labial-alveolar distinction in English. One of these is determined by the time course of the change in amplitude in the high-frequency range (above 2500 Hz) in the few tens of ms following consonantal release, and the other is determined by the frequencies of spectral peaks associated with the second and third formants in relation to the first formant.     

The role of release bursts in the perception of [s]-stop clusters

The role of the release burst as a cue to the perception of stop consonants following [s] was investigated in a series of studies. Experiment 1 demonstrated that silent closure duration and burst duration can be traded as cues for the "say"-"stay" distinction. Experiment 2 revealed a similar trading relation between closure duration and burst amplitude. Experiments 3 and 4 suggested, perhaps surprisingly, that absolute, not relative, burst amplitude is important. Experiment 5 demonstrated that listener's sensitivity to bursts in a labeling task is at least equal to their sensitivity in a burst detection task. Experiments 6 and 7 replicated the trading relation between closure duration and burst amplitude for labial stops in the "slit"-"split" and "slash"-"splash" distinctions, although burst amplification, in contrast to attenuation, had no effect. All experiments revealed that listeners are remarkably sensitive to the presence of even very weak release bursts.     

Spectral structure across the syllable specifies final-stop voicing for adults and children alike

Traditional accounts of speech perception generally hold that listeners use isolable acoustic "cues" to label phonemes. For syllable-final stops, duration of the preceding vocalic portion and formant transitions at syllable's end have been considered the primary cues to voicing decisions. The current experiment tried to extend traditional accounts by asking two questions concerning voicing decisions by adults and children: (1) What weight is given to vocalic duration versus spectral structure, both at syllable's end and across the syllable? (2) Does the naturalness of stimuli affect labeling? Adults and children (4, 6, and 8 years old) labeled synthetic stimuli that varied in vocalic duration and spectral structure, either at syllable's end or earlier in the syllable. Results showed that all listeners weighted dynamic spectral structure, both at syllable's end and earlier in the syllable, more than vocalic duration, and listeners performed with these synthetic stimuli as listeners had performed previously with natural stimuli. The conclusion for accounts of human speech perception is that rather than simply gathering acoustic cues and summing them to derive strings of phonemic segments, listeners are able to attend to global spectral structure, and use it to help recover explicitly phonetic structure.     

Discrimination of frequency steps linked by glides of various durations.

Thresholds were measured for detecting steps in frequency linked by glides of various durations. The goals were to assess the relative importance of place and temporal information for this task, and to determine whether there is a mechanism for detecting dynamic frequency changes per se, as opposed to comparing the initial and final frequencies of the stimuli. Subjects discriminated a 500-ms sinusoid of constant frequency from a sinusoid with three parts: an initial part with constant frequency, a downward frequency glide, and a final part with constant frequency. The overall duration was 500 ms, and the glide duration was varied from 5 to 500 ms. In one special case, the portion of the stimuli when a glide might occur was replaced by a brief silent interval. The center frequency was fixed at 0.5, 1, 2, 4, or 6 kHz (condition 1), or varied randomly from one stimulus to the next over a 4-ERB range around the nominal center frequency (condition 2). The randomization impaired performance, but thresholds remained lower than the best that could be achieved by monitoring either the initial or final frequency of the stimuli. Condition 3 was like condition 2, but for each stimulus a glide in level was added at the time when a frequency glide might occur, so the initial and final levels differed; the glides in level varied randomly in extent and direction from one stimulus to the next over the range +/- 20 dB. This impaired performance, but thresholds remained lower than the best that could be achieved by monitoring changes in excitation level on only one side of the excitation pattern. Excitation-pattern models of frequency discrimination predict that thresholds should not vary across center frequency when expressed as the change in ERB number, delta E. For all conditions, delta E values increased at 6 kHz, suggesting a role for temporal information at lower frequencies. The increase was smallest for the longest glide duration, consistent with a greater relative role of place information when there was no steady state portion. Performance was better when a brief glide was present than when no glide was present, but worsened with increasing glide duration. The results were fitted well by a model based on the assumption that information from the steady parts of the stimuli (perhaps extracted mainly using temporal information) was combined with information from the glides (perhaps extracted mainly using place information).     

The role of temporal and dynamic signal components in the perception of syllable-final stop voicing by children and adults

Adults whose native languages permit syllable-final obstruents, and show a vocalic length distinction based on the voicing of those obstruents, consistently weight vocalic duration strongly in their perceptual decisions about the voicing of final stops, at least in laboratory studies using synthetic speech. Children, on the other hand, generally disregard such signal properties in their speech perception, favoring formant transitions instead. These age-related differences led to the prediction that children learning English as a native language would weight vocalic duration less than adults, but weight syllable-final transitions more in decisions of final-consonant voicing. This study tested that prediction. In the first experiment, adults and children (eight and six years olds) labeled synthetic and natural CVC words with voiced or voiceless stops in final C position. Predictions were strictly supported for synthetic stimuli only. With natural stimuli it appeared that adults and children alike weighted syllable-offset transitions strongly in their voicing decisions. The predicted age-related difference in the weighting of vocalic duration was seen for these natural stimuli almost exclusively when syllable-final transitions signaled a voiced final stop. A second experiment with adults and children (seven and five years old) replicated these results for natural stimuli with four new sets of natural stimuli. It was concluded that acoustic properties other than vocalic duration might play more important roles in voicing decisions for final stops than commonly asserted, sometimes even taking precedence over vocalic duration.     

Stimulus factors influencing the identification of voiced stop consonants by normal-hearing and hearing-impaired adults

The effects of mild-to-moderate hearing impairment on the perceptual importance of three acoustic correlates of stop consonant place of articulation were examined. Normal-hearing and hearing-impaired adults identified a stimulus set comprising all possible combinations of the levels of three factors: formant transition type (three levels), spectral tilt type (three levels), and abruptness of frequency change (two levels). The levels of these factors correspond to those appropriate for /b/, /d/, and /g/ in the /ae/ environment. Normal-hearing subjects responded primarily in accord with the place of articulation specified by the formant transitions. Hearing-impaired subjects showed less-than-normal reliance on formant transitions and greater-than-normal reliance on spectral tilt and abruptness of frequency change. These results suggest that hearing impairment affects the perceptual importance of cues to stop consonant identity, increasing the importance of information provided by both temporal characteristics and gross spectral shape and decreasing the importance of information provided by the formant transitions.     

Relative weights for frequency glide detection using narrowband noise

Narrowband noise stimuli were used to derive relative weights for detecting frequency glides in a yes/no procedure. One stimulus set was restricted to the duration of the glide. For the second stimulus set, there were fringe noise bands preceding and following the glide. For both sets, the center frequency of the linear glide was either fixed at 1000 Hz or randomly chosen on each trial from the range of 800-1200 Hz. Relative weights as a function of time were derived using a linear model and the linear classification method [A. Ahumada, J. Vis., 2, 121-131 (2002)]. Sensitivity was better for the fixed- than random-frequency conditions, and weight patterns from the random-frequency conditions were less reliable than those obtained from the fixed-frequency conditions. The magnitudes of the relative weights tended to be larger for the second half of the stimulus, suggesting that subjects paid more attention to the later than the earlier parts of stimuli. In the random-frequency conditions, the linear model failed to account for subjects' performance unless the stimuli were expressed in terms of relative changes in frequency rather than absolute frequency.     

Detection, direction discrimination, and off-frequency interference of center-frequency modulations and glides for vowel formants

Vowels are mainly classified by the positions of peaks in their frequency spectra, the formants. For normal-hearing subjects, change detection and direction discrimination were measured for linear glides in the center frequency (CF) of formantlike sounds. A CF rove was used to prevent subjects from using either the start or end points of the glides as cues. In addition, change detection and starting-phase (start-direction) discrimination were measured for similar stimuli with a sinusoidal 5-Hz formant-frequency modulation. The stimuli consisted of single formants generated using a number of different stimulus parameters including fundamental frequency, spectral slope, frequency region, and position of the formant relative to the harmonic spectrum. The change detection thresholds were in good agreement with the predictions of a model which analyzed and combined the effects of place-of-excitation and temporal cues. For most stimuli, thresholds were approximately equal for change detection and start-direction discrimination. Exceptions were found for stimuli that consisted of only one or two harmonics. In a separate experiment, it was shown that change detection and start-direction discrimination of linear and sinusoidal formant-frequency modulations were impaired by off-frequency frequency-modulated interferers. This frequency modulation detection interference was larger for formants with shallow than for those with steep spectral slopes.     

Infant discrimination of two- and five-formant voiced stop consonants differing in place of articulation

According to recent theoretical accounts of place of articulation perception, global, invariant properties of the stop CV syllable onset spectrum serve as primary, innate cues to place of articulation, whereas contextually variable formant transitions constitute secondary, learned cues. By this view, one might expect that young infants would find the discrimination of place of articulation contrasts signaled by formant transition differences more difficult than those cued by gross spectral differences. Using an operant head-turning paradigm, we found that 6-month-old infants were able to discriminate two-formant stimuli contrasting in place of articulation as well as they did five-formant + burst stimuli. Apparently, neither the global properties of the onset spectrum nor simply the additional acoustic information contained in the five-formant + burst stimuli afford the infant any advantage in the discrimination task. Rather, formant transition information provides a sufficient basis for discriminating place of articulation differences.     

The use of acoustic cues for phonetic identification: effects of spectral degradation and electric hearing

Although some cochlear implant (CI) listeners can show good word recognition accuracy, it is not clear how they perceive and use the various acoustic cues that contribute to phonetic perceptions. In this study, the use of acoustic cues was assessed for normal-hearing (NH) listeners in optimal and spectrally degraded conditions, and also for CI listeners. Two experiments tested the tense/lax vowel contrast (varying in formant structure, vowel-inherent spectral change, and vowel duration) and the word-final fricative voicing contrast (varying in F1 transition, vowel duration, consonant duration, and consonant voicing). Identification results were modeled using mixed-effects logistic regression. These experiments suggested that under spectrally-degraded conditions, NH listeners decrease their use of formant cues and increase their use of durational cues. Compared to NH listeners, CI listeners showed decreased use of spectral cues like formant structure and formant change and consonant voicing, and showed greater use of durational cues (especially for the fricative contrast). The results suggest that although NH and CI listeners may show similar accuracy on basic tests of word, phoneme or feature recognition, they may be using different perceptual strategies in the process.     

Effects of deafness on acoustic characteristics of American English tense/lax vowels in maternal speech to infants

Recent studies have demonstrated that mothers exaggerate phonetic properties of infant-directed (ID) speech. However, these studies focused on a single acoustic dimension (frequency), whereas speech sounds are composed of multiple acoustic cues. Moreover, little is known about how mothers adjust phonetic properties of speech to children with hearing loss. This study examined mothers' production of frequency and duration cues to the American English tense/lax vowel contrast in speech to profoundly deaf (N?=?14) and normal-hearing (N?=?14) infants, and to an adult experimenter. First and second formant frequencies and vowel duration of tense (/i/,?/u/) and lax (/I/,?/?/) vowels were measured. Results demonstrated that for both infant groups mothers hyperarticulated the acoustic vowel space and increased vowel duration in ID speech relative to adult-directed speech. Mean F2 values were decreased for the /u/ vowel and increased for the /I/ vowel, and vowel duration was longer for the /i/, /u/, and /I/ vowels in ID speech. However, neither acoustic cue differed in speech to hearing-impaired or normal-hearing infants. These results suggest that both formant frequencies and vowel duration that differentiate American English tense/lx vowel contrasts are modified in ID speech regardless of the hearing status of the addressee.     

Acoustic cues to tonal contrasts in Mandarin: implications for cochlear implants

The present study systematically manipulated three acoustic cues--fundamental frequency (f0), amplitude envelope, and duration--to investigate their contributions to tonal contrasts in Mandarin. Simplified stimuli with all possible combinations of these three cues were presented for identification to eight normal-hearing listeners, all native speakers of Mandarin from Taiwan. The f0 information was conveyed either by an f0-controlled sawtooth carrier or a modulated noise so as to compare the performance achievable by a clear indication of voice f0 and what is possible with purely temporal coding of f0. Tone recognition performance with explicit f0 was much better than that with any combination of other acoustic cues (consistently greater than 90% correct compared to 33%-65%; chance is 25%). In the absence of explicit f0, the temporal coding of f0 and amplitude envelope both contributed somewhat to tone recognition, while duration had only a marginal effect. Performance based on these secondary cues varied greatly across listeners. These results explain the relatively poor perception of tone in cochlear implant users, given that cochlear implants currently provide only weak cues to f0, so that users must rely upon the purely temporal (and secondary) features for the perception of tone.     

A comparison of three speech coding strategies using an acoustic model of a cochlear implant

Three alternative speech coding strategies suitable for use with cochlear implants were compared in a study of three normally hearing subjects using an acoustic model of a multiple-channel cochlear implant. The first strategy (F2) presented the amplitude envelope of the speech and the second formant frequency. The second strategy (F0 F2) included the voice fundamental frequency, and the third strategy (F0 F1 F2) presented the first formant frequency as well. Discourse level testing with the speech tracking method showed a clear superiority of the F0 F1 F2 strategy when the auditory information was used to supplement lipreading. Tracking rates averaged over three subjects for nine 10-min sessions were 40 wpm for F2, 52 wpm for F0 F2, and 66 wpm for F0 F1 F2. Vowel and consonant confusion studies and a test of prosodic information were carried out with auditory information only. The vowel test showed a significant difference between the strategies, but no differences were found for the other tests. It was concluded that the amplitude and duration cues common to all three strategies accounted for the levels of consonant and prosodic information received by the subjects, while the different tracking rates were a consequence of the better vowel recognition and the more natural quality of the F0 F1 F2 strategy.     

Effects of prosodic boundary on /aC/ sequences: acoustic results

This study presents various acoustic measures used to examine the sequence /a # C/, where "#" represents different prosodic boundaries in French. The 6 consonants studied are /b d g f s S/ (3 stops and 3 fricatives). The prosodic units investigated are the utterance, the intonational phrase, the accentual phrase, and the word. It is found that vowel target values, formant transitions into the stop consonant, and the rate of change in spectral tilt into the fricative, are affected by the strength of the prosodic boundary. F1 becomes higher for /a/ the stronger the prosodic boundary, with the exception of one speaker's utterance data, which show the effects of articulatory declension at the utterance level. Various effects of the stop consonant context are observed, the most notable being a tendency for the vowel /a/ to be displaced in the direction of the F2 consonant "locus" for /d/ (the F2 consonant values for which remain relatively stable across prosodic boundaries) and for /g/ (the F2 consonant values for which are displaced in the direction of the velar locus in weaker prosodic boundaries, together with those of the vowel). Velocity of formant transition may be affected by prosodic boundary (with greater velocity at weaker boundaries), though results are not consistent across speakers. There is also a tendency for the rate of change in spectral tilt moving from the vowel to the fricative to be affected by the presence of a prosodic boundary, with a greater rate of change at the weaker prosodic boundaries. It is suggested that spectral cues, in addition to duration, amplitude, and F0 cues, may alert listeners to the presence of a prosodic boundary.     

Vowel recognition via cochlear implants and noise vocoders: effects of formant movement and duration

Previous work has demonstrated that normal-hearing individuals use fine-grained phonetic variation, such as formant movement and duration, when recognizing English vowels. The present study investigated whether these cues are used by adult postlingually deafened cochlear implant users, and normal-hearing individuals listening to noise-vocoder simulations of cochlear implant processing. In Experiment 1, subjects gave forced-choice identification judgments for recordings of vowels that were signal processed to remove formant movement and/or equate vowel duration. In Experiment 2, a goodness-optimization procedure was used to create perceptual vowel space maps (i.e., best exemplars within a vowel quadrilateral) that included F1, F2, formant movement, and duration. The results demonstrated that both cochlear implant users and normal-hearing individuals use formant movement and duration cues when recognizing English vowels. Moreover, both listener groups used these cues to the same extent, suggesting that postlingually deafened cochlear implant users have category representations for vowels that are similar to those of normal-hearing individuals.     

Speech coding in the auditory nerve: V. Vowels in background noise

Responses of auditory-nerve fibers to steady-state, two-formant vowels in low-pass background noise (S/N = 10 dB) were obtained in anesthetized cats. For fibers over a wide range of characteristic frequencies (CFs), the peaks in discharge rate at the onset of the vowel stimuli were nearly eliminated in the presence of noise. In contrast, strong effects of noise on fine time patterns of discharge were limited to CF regions that are far from the formant frequencies. One effect is a reduction in the amplitude of the response component at the fundamental frequency in the high-CF regions and for CFs between F1 and F2 when the formants are widely separated. A reduction in the amplitude of the response components at the formant frequencies, with concomitant increase in components near CF or low-frequency components occurs in CF regions where the signal-to-noise ratio is particularly low. The processing schemes that were effective for estimating the formant frequencies and fundamental frequency of vowels in quiet generally remain adequate in moderate-level background noise. Overall, the discharge patterns contain many cues for distinctions among the vowel stimuli, so that the central processor should be able to identify the different vowels, consistent with psychophysical performance at moderate signal-to-noise ratios.     

Identification of resynthesized /hVd/ utterances: effects of formant contour.

The purpose of this study was to examine the role of formant frequency movements in vowel recognition. Measurements of vowel duration, fundamental frequency, and formant contours were taken from a database of acoustic measurements of 1668 /hVd/ utterances spoken by 45 men, 48 women, and 46 children [Hillenbrand et al., J. Acoust. Soc. Am. 97, 3099-3111 (1995)]. A 300-utterance subset was selected from this database, representing equal numbers of 12 vowels and approximately equal numbers of tokens produced by men, women, and children. Listeners were asked to identify the original, naturally produced signals and two formant-synthesized versions. One set of "original formant" (OF) synthetic signals was generated using the measured formant contours, and a second set of "flat formant" (FF) signals was synthesized with formant frequencies fixed at the values measured at the steadiest portion of the vowel. Results included: (a) the OF synthetic signals were identified with substantially greater accuracy than the FF signals; and (b) the naturally produced signals were identified with greater accuracy than the OF synthetic signals. Pattern recognition results showed that a simple approach to vowel specification based on duration, steady-state F0, and formant frequency measurements at 20% and 80% of vowel duration accounts for much but by no means all of the variation in listeners' labeling of the three types of stimuli.