Formant Frequency Shift Research Articles

Simulation of acoustic effects of postural laryngeal oscillatory movement during vocalization

This paper addresses the contributions of low-frequency postural laryngeal kinematics to frequency and amplitude modulations of the vocal output signal. A one-dimensional Navier–Stokes solution for fluid transport and wave propagation is used along with a laryngeal source model to simulate voice production. Vocal fold adduction and vertical laryngeal posture are modulated sinusoidally at 2–10 Hz with various amplitudes to approximate kinematic phenomena observed with MRI during production of vocal vibrato and voice tremor. The acoustic modulations are quantified in terms of formant (resonance) and fundamental frequency shifts in the radiated output signal from the mouth. Postural oscillation of the larynx is a common kinematic component of vocal tremor, although less documented during production of vibrato.

Method for Improving Intelligibility of Elderly Speech Based on Formant Frequency Shift

In general, aging is progressing in developed countries. Elderly people have difficulty controlling their articulation accurately due to aging. We need to improve the quality of elderly speech for smooth communication. In this paper, we analyzed the 1st formant frequency (F1) and 2nd formant frequency (F2) between the more intelligible speech and less intelligible speech of Japanese elderly people. In addition, we improved the intelligibility of less intelligible elderly speech by using the formant frequency shift method. This method is the correcting by shift value of formant frequency based on LPC. The shift value is the magnification such as expanding the F1-F2 size of less intelligible speech.

Nonlinear frequency compression: Influence of start frequency and input bandwidth on consonant and vowel recognition.

By varying parameters that control nonlinear frequency compression (NFC), this study examined how different ways of compressing inaudible mid- and/or high-frequency information at lower frequencies influences perception of consonants and vowels. Twenty-eight listeners with mild to moderately severe hearing loss identified consonants and vowels from nonsense syllables in noise following amplification via a hearing aid simulator. Low-pass filtering and the selection of NFC parameters fixed the output bandwidth at a frequency representing a moderately severe (3.3 kHz, group MS) or a mild-to-moderate (5.0 kHz, group MM) high-frequency loss. For each group (n = 14), effects of six combinations of NFC start frequency (SF) and input bandwidth [by varying the compression ratio (CR)] were examined. For both groups, the 1.6 kHz SF significantly reduced vowel and consonant recognition, especially as CR increased; whereas, recognition was generally unaffected if SF increased at the expense of a higher CR. Vowel recognition detriments for group MS were moderately correlated with the size of the second formant frequency shift following NFC. For both groups, significant improvement (33%-50%) with NFC was confined to final /s/ and /z/ and to some VCV tokens, perhaps because of listeners' limited exposure to each setting. No set of parameters simultaneously maximized recognition across all tokens.

Perception of size-related formant information in male koalas (Phascolarctos cinereus)

Advances in bioacoustics allow us to study the perceptual and functional relevance of individual acoustic parameters. Here, we use re-synthesised male koala bellows and a habituation-dishabituation paradigm to test the hypothesis that male koalas are sensitive to shifts in formant frequencies corresponding to the natural variation in body size between a large and small adult male. We found that males habituated to bellows, in which the formants had been shifted to simulate a large or small male displayed a significant increase in behavioural response (dishabituation) when they were presented with bellows simulating the alternate size variant. The rehabituation control, in which the behavioural response levels returned to that of the last playbacks of the habituation phase, indicates that this was not a chance increase in response levels. Our results provide clear evidence that male koalas perceive and attend to size-related formant information in their own species-specific vocalisations and suggest that formant perception is a widespread ability shared by marsupials and placental mammals, and perhaps by vertebrates more widely.

Bee Threat Elicits Alarm Call in African Elephants

Unlike the smaller and more vulnerable mammals, African elephants have relatively few predators that threaten their survival. The sound of disturbed African honeybees Apis meliffera scutellata causes African elephants Loxodonta africana to retreat and produce warning vocalizations that lead other elephants to join the flight. In our first experiment, audio playbacks of bee sounds induced elephants to retreat and elicited more head-shaking and dusting, reactive behaviors that may prevent bee stings, compared to white noise control playbacks. Most importantly, elephants produced distinctive “rumble” vocalizations in response to bee sounds. These rumbles exhibited an upward shift in the second formant location, which implies active vocal tract modulation, compared to rumbles made in response to white noise playbacks. In a second experiment, audio playbacks of these rumbles produced in response to bees elicited increased headshaking, and further and faster retreat behavior in other elephants, compared to control rumble playbacks with lower second formant frequencies. These responses to the bee rumble stimuli occurred in the absence of any bees or bee sounds. This suggests that these elephant rumbles may function as referential signals, in which a formant frequency shift alerts nearby elephants about an external threat, in this case, the threat of bees.

Do ‘Dominant Frequencies’ explain the listener’s response to formant and spectrum shape variations?

Psychoacoustic experimentation shows that formant frequency shifts can give rise to more significant changes in phonetic vowel timber than differences in overall level, bandwidth, spectral tilt, and formant amplitudes. Carlson and Granström’s perceptual and computational findings suggest that, in addition to spectral representations, the human ear uses temporal information on formant periodicities (‘Dominant Frequencies’) in building vowel timber percepts. The availability of such temporal coding in the cat’s auditory nerve fibers has been demonstrated in numerous physiological investigations undertaken during recent decades. In this paper we explore, and provide further support for, the Dominant Frequency hypothesis using KONVERT, a computational auditory model. KONVERT provides auditory excitation patterns for vowels by performing a critical-band analysis. It simulates phase locking in auditory neurons and outputs DF histograms. The modeling supports the assumption that listeners judge phonetic distance among vowels on the basis formant frequency differences as determined primarily by a time-based analysis. However, when instructed to judge psychophysical distance among vowels, they can also use spectral differences such as formant bandwidth, formant amplitudes and spectral tilt. Although there has been considerable debate among psychoacousticians about the functional role of phase locking in monaural hearing, the present research suggests that detailed temporal information may nonetheless play a significant role in speech perception.

Female perception of size-related formant shifts in red deer, Cervus elaphus

Little is known about the role of specific acoustic cues in mammal vocal communication systems. Recent advances in bioacoustics allow us to use playback experiments in order to study the perceptual and functional relevance of individual acoustic parameters. In the current study we used re-synthesised male red deer roars in a habituation-discrimination paradigm to determine whether female red deer are sensitive to shifts in formant frequencies corresponding to the natural variation between the vocal tract lengths of a small and large adult red deer male. We found that hinds habituated to a given size variant show a significant dishabituation when they are presented with roars in which the formants have been modified to simulate the other size variant. The significant reduction in behavioural response to a final rehabituation playback shows this was not a chance rebound in response levels. Our results indicate that formants are salient for red deer hinds and that hinds can detect a shift in formant frequencies that has strong potential biological significance. We discuss the possible functions of formant perception in female red deer and more generally in nonhuman mammals.

Effects of auditory feedback deprivation on non-native French vowels produced by children with cochlear implants

The objective of the present study was to examine alternative explanations for changes in speech demonstrated by individuals with a cochlear implant, in response to auditory feedback deprivation. By employing French vowels, for which monolingual English speakers do not have internal representations, this study compared formant frequencies of vowels produced in the presence and absence of auditory feedback. It was hypothesized that the lack of auditory feedback should not lead to vowel formant frequency shifts for these novel sounds. However, if, in addition to formants frequency shifts, increases in speaking rate and fundamental frequencies are observed in the absence of auditory feedback, the data would support the idea that talkers attempt to speak more clearly to maximize information from other sensory channels. Four children with cochlear implants (7–11 years of age) and four children with normal hearing (mean age 8.6 years) repeated English (familiar) and French (unfamiliar) vowels presented through ...

Finite-element method analysis of acoustic characteristics of the vocal tract with the nasal cavity during phonation of Japanese /a/

For this study, the transfer functions and active sound intensities of a vocal tract model with and without a nasal cavity were computed using a three-dimensional finite-element method (FEM). The models were based on vowel data obtained by magnetic resonance imaging (MRI) of the vocal tract with a nasal cavity during phonation of the Japanese /a/. The oral cavity was also coupled with the nasal cavity in a three-dimensional volume of radiation. Effects of wall impedance were also examined. Coupling of the nasal cavity to the oral cavity indicated the following aspects. Additional peaks appeared below 3 kHz for the lossless condition. However, they disappeared in the simulation for the soft wall condition. The sound energy circulation did not occur in the simulation for the soft wall condition. Regarding effects of the wall boundary condition on the spectral envelope, three-dimensional simulations confirmed the upward shift of lower formant frequencies. However, disagreement of the formant frequencies between simulated and real speech should be investigated further by adjusting the wall boundary condition to a more realistic one.

Learning of auditory equivalence classes for vowels by rats

Four male Long-Evans rats were trained to discriminate between synthetic vowel sounds using a GO/NOGO response choice task. The vowels were characterized by an increase in fundamental frequency correlated with an upward shift in formant frequencies. In an initial phase we trained the subjects to discriminate between two vowel categories using two exemplars from each category. In a subsequent phase the ability of the rats to generalize the discrimination between the two categories was tested. To test whether rats might exploit the fact that attributes of training stimuli covaried, we used non-standard stimuli with a reversed relation between fundamental frequency and formants. The overall results demonstrate that rats are able to generalize the discrimination to new instances of the same vowels. We present evidence that the performance of the subjects depended on the relation between fundamental and formant frequencies that they had previously been exposed to. Simple simulation results with artificial neural networks could reproduce most of the behavioral results and support the hypothesis that equivalence classes for vowels are associated with an experience-driven process based on general properties of peripheral auditory coding mixed with elementary learning mechanisms. These results suggest that rats use spectral and temporal cues similarly to humans despite differences in basic auditory capabilities.

Vowel production with and without auditory feedback in children with cochlear implants

The effects of auditory feedback on vowel formant frequencies were investigated in the speech produced by three early-implanted (received cochlear implants before the age of 3 years) and three late-implanted children (after 5 years). The syllables /heed/, /hid/, /had/, /whod/, /hood/, and /hod/ were produced in a carrier sentence 12 times, each in both implant processor ON (auditory feedback present) and OFF (feedback absent) conditions. Comparisons of the vowel formant frequencies in processor ON versus OFF conditions revealed primarily second formant frequency (F2) shifts for a majority of the vowels, while showing minimal changes in the first formant frequencies (F1). In addition, a majority of the F2 shifts was towards the middle of the vowel space, suggesting centralization of vowels in the absence of auditory feedback. Further, most of the formant frequency changes were present in the vowels produced by late- compared to early-implanted children. These findings suggest that the late-implanted children may rely on auditory feedback to a greater extent than early-implanted children for speech production. Implications of the short-term effects of auditory feedback on speech production will be discussed in reference to the feedback and feed-forward models for speech production. [Work supported by NIH/NIDCD R03DC007052-01.]

Acoustic properties of vowels in clear and conversational speech by female non-native English speakers

Studies have shown that talkers can improve the intelligibility of their speech when instructed to speak as if talking to a hearing-impaired person. The improvement of speech intelligibility is associated with specific acoustic-phonetic changes: increases in vowel duration and fundamental frequency (F0), a wider pitch range, and a shift in formant frequencies for F1 and F2. Most previous studies of clear speech production have been conducted with native speakers; research with second language speakers is much less common. The present study examined the acoustic properties of non-native English vowels produced in a clear speaking style. Five female Cantonese speakers and a comparison group of English speakers were recorded producing four vowels (/i u ae a/) in /bVt/ context in conversational and clear speech. Vowel durations, F0, pitch range, and the first two formants for each of the four vowels were measured. Analyses revealed that for both groups of speakers, vowel durations, F0, pitch range, and F1 spoken clearly were greater than those produced conversationally. However, F2 was higher in conversational speech than in clear speech. The findings suggest that female non-native English speakers exhibit acoustic-phonetic patterns similar to those of native speakers when asked to produce English vowels clearly.

Analysis and recognition of whispered speech

In this study, we have examined the acoustic characteristics of whispered speech and addressed some of the issues involved in recognition of whispered speech used for communication over a mobile phone in a noisy environment. The acoustic analysis shows that there is an upward shift of formant frequencies of vowels as observed in the whispered speech data compared to the normal speech data. Voiced consonants in the whispered speech have lower energy at low frequencies up to 1.5 kHz and their spectral flatness is greater compared to the normal speech. In experiments on whispered speech recognition, results of our studies on adaptation of the whispered speech models have shown that adaptation using a small amount of whispered speech data from a target speaker can be effectively used for recognition of the whispered speech. In a noisy environment, the recognition accuracy decreases significantly for the whispered speech compared to the normal speaking of the same speech. A method to increase the SNR by covering the mouth with a hand has been shown to give a higher recognition accuracy for the whispered speech frequently encountered for private communication in a noisy environment.

Formant shift in nasalization of vowels

We measured the formant shifts of a vowel in the context of a nasal and investigated whether human perception is able to compensate for such shifts. According to the acoustic theory, nasal coupling causes a modification on the spectrum, including formant frequency shift. The first goal of this study is to confirm that the formant frequencies actually shift due to nasalization. Based on several measurements of formant frequencies of various vowels in nasal contexts, we confirmed that the first formant (F1) tends to shift in a more central direction when nasalized. In English, vowels should be perceived as the same phoneme regardless of nasalization. In other words, listeners might have the capability to compensate for such formant shifts. The second goal of this study is to eximine this compensation effect by a perceptual experiment. For stimuli, we synthesized a nonnasal vowel V0 that has the same formant frequencies as a nasalized vowel V1. A continuum was also synthesized between V0 and the nonnasalized version of V1. Results show V1 is more correctly identified than V0, which suggests the existence of the compensation effect.

A note on the acoustic-phonetic characteristics of non-native English vowels produced in noise

The Lombard reflex occurs when people unconsciously raise their vocal levels in the presence of loud background noise. Previous work has established that utterances produced in noisy environments exhibit increases in vowel duration and fundamental frequency (F0), and a shift in formant center frequencies for F1 and F2. Most studies of the Lombard reflex have been conducted with native speakers; research with second-language speakers is much less common. The present study examined the effects of the Lombard reflex on foreign-accented English vowel productions. Seven female Cantonese speakers and a comparison group of English speakers were recorded producing three vowels (/i u a/) in /bVt/ context in quiet and in 70 dB of masking noise. Vowel durations, F0, and the first two formants for each of the three vowels were measured. Analyses revealed that vowel durations and F0 were greater in the vowels produced in noise than those produced in quiet in most cases. First formants, but not F2, were consistently higher in Lombard speech than in normal speech. The findings suggest that non-native English speakers exhibit acoustic-phonetic patterns similar to those of native speakers when producing English vowels in noisy conditions.

Lab experiment using physical models of the human vocal tract for high-school students

Recently, the development of educational tools for acoustics has become popular in Japan. We believe that physical models of the human vocal tract are particularly useful for teaching acoustics. Formerly we proposed three models of the vocal tract corresponding to the Japanese vowels, /i/, /e/, /a/, /o/, and /u/. We presented cylindrical, nasalized, and plate type models. The models were made of transparent acrylic resin, enabling the configurations of the oral cavity to be seen from the outside of the model. In this presentation, we will discuss the results of a lab experiment in which we used these tools to teach the mechanism of vowel production to high-school students who had just finished studying basic acoustics. By manipulating the plates in the plate type model, students were able to simulate constrictions at nodes and antinodes, and they were able to hear the shift in formant frequencies. The exercise helped students to understand vowel production. We received positive feedback from those who participated in the experiment.

Effects of frequency shifts on vowel category judgments

To investigate the effects of fundamental frequency (F0) and formant frequency shifts on vowel identification, a high-quality vocoder (‘‘STRAIGHT’’) was used to process the syllables ‘‘bit’’ and ‘‘bet’’ spoken by an adult female talker. From these two endpoints a nine-step continuum was generated by interpolation of the time-varying spectral envelope. Upward and downward frequency shifts in spectral envelope (scale factors of 0.75, 1.0, or 1.33) were combined with shifts in F0 (scale factors of 0.5, 1.0, or 1.25). Downward frequency shifts generally resulted in malelike voices whereas upward shifts were perceived as childlike. Matched frequency shifts, in which F0 and spectral envelope (i.e., formant frequencies) were shifted in the same direction, had relatively little effect on phoneme boundaries. Mismatched frequency shifts, in which F0 was modified independently of spectral envelope or vice versa, resulted in systematic boundary shifts. The changes in the identifications functions were qualitatively consistent with predictions of a model trained using acoustic measurements derived from a database of naturally spoken vowel tokens from men, women, and children. The empirical and modeling results are consistent with the idea that vowel boundary shifts are a consequence of listeners sensitivity to the statistical structure of natural speech.

Acoustical-perceptual correlates of "whisper pitch" in synthetically generated vowels.

The purpose of this investigation was to clarify acoustical-perceptual relationships in identification of "pitch" during whispered vowel production. The experimenters systematically varied selected acoustic features of synthetically generated "whispered" vowels to control which formant frequencies were shifted (F1, F2, or F1&F2), the direction of formant frequency shifts (up or down), and the magnitude of formant frequency shifts (20 Hz, 40 Hz, 60 Hz). Two sets of stimuli were produced to simulate the resonance characteristics of the vowel /a/: one set for male talkers and one for female talkers. Ninety-four pairs of synthesized vowel tokens were randomly presented to 17 listeners who judged if the "pitch" of the second member of the pair was the same, higher, or lower than the "pitch" of the first member. The results showed an inverse relationship between the magnitude of formant frequency changes presented to the judges and the number of perceptual mismatches in "whisper pitch." Also, fewer mismatches in the identification of whisper pitch occurred when both F1 and F2 were changed simultaneously than when either F1 or F2 was changed individually. No differences were found between the perceptual responses to "male" and "female" vowel simulations. The primary implication of this study is that whisper pitch is more influenced by simultaneous changes in F1 and F2 than by changes in only one of the formants.

Context and identification of vowels in speech variation

Interpretation of a particular vowel pattern may vary according to the perceived speech patterns of the speaker [P. Ladefoged and D. E. Broadbent, J. Acoust. Soc. Am. 29, 98–104 (1957)]. Formants from a given speaker, using a specific speech style, reflect both anatomical dimensions of the speakers vocal tract and the speakers preferred vocal settings for the context of the utterance. Speech samples of a set of nine vowels in a standard CVC frame [G. E. Peterson and H. L. Barney, J. Acoust. Soc. Am. 24, 175–184 (1952)], recorded from local speakers, were analyzed and resynthesized, manipulating formant values to derive similar formant frequency ratios from different vowel bases. Thus if the listener identifies the vowels according to the formant patterns, the perceived vowel will differ from the base vowel from which it was derived, whereas if the overall formant frequency shift is detected, then the vowel should maintain its identity, differences in quality being attributed to a change of speaker or of speech style. Using listeners from the same speech community, monosyllabic word identification is contrasted with identification of excised vowels to ascertain the effect of word context on distinguishing the relatively long-term formant ratios of the speaker. [Work supported by NIDCD.]

Helium speech database and some new aspect of diver’s speech

Speech is distorted in pressurized helium-air. Many aspects, except for upward shifts of formant and pitch in the helium speech, however, have not been known yet. A helium speech database system including more than 5 000 words and messages of divers was developed. The vibration of the cheek, the neck, and the nose were also recorded for some utterances. They were taken at the simulation dives up to the depth of 300 m at JAMSTEC since 1982. The helium speech and the normal speech were compared in terms of (1) duration of phonemes, and (2) pitch frequencies and formant trajectories in two different hearing/speaking conditions: the first condition is a case when a diver spoke as he was listening to his own unscrambled speech, and in the other condition the diver was not hearing his unscrambled speech. As a result, the shorter duration of word-initial consonants, higher pitch by psychological stress, and rough movement of formant trajectories were observed in addition to the well-known formant and pitch frequency shifts. As a special property of helium speech, it is newly discovered that there exists a quasiformant below the normal first formant, and that it may be caused by the sound from the nostrils.