Spectral Center Of Gravity Research Articles

Impact of cognitive load and frustration on drivers’ speech.

Secondary tasks such as cell phone calls or interaction with automated speech dialog systems (SDSs) increase the driver’s cognitive load as well as the probability of driving errors. This study analyzes speech production variations due to cognitive load and emotional state of drivers in real driving conditions. Speech samples were acquired from 24 female and 17 male subjects (approximately 8.5 h of data) while talking to a co-driver and communicating with two automated call centers, with emotional states (neutral, negative) and the number of necessary SDS query repetitions also labeled. A consistent shift in a number of speech production parameters (pitch, first format center frequency, spectral center of gravity, spectral energy spread, and duration of voiced segments) was observed when comparing SDS interaction against co-driver interaction; further increases were observed when considering negative emotion segments and the number of requested SDS query repetitions. A mel frequency cepstral coefficient based Gaussian mixture classifier trained on 10 male and 10 female sessions provided 91% accuracy in the open test set task of distinguishing co-driver interactions from SDS interactions, suggesting—together with the acoustic analysis—that it is possible to monitor the level of driver distraction directly from their speech.

Effects of initial position versus prominence in English

This study investigates effects of three prosodic factors—prosodic boundary (Utterance-initial vs. Utterance-medial), lexical stress (primary vs. secondary) and phrasal accent (accented vs. unaccented)—on articulatory and acoustic realizations of word-initial CVs (/nε/, /tε/) in trisyllabic English words. The consonantal measures were linguopalatal Peak contact and Release contacts (by electropalatography), Seal duration, Nasal duration and Nasal energy for /n/, VOT, RMS burst energy and spectral Center of Gravity at the release for /t/; and the vocalic measures were linguopalatal Vowel contact, Vowel F1, Vowel duration and Vowel amplitude. Several specific points emerge. Firstly, domain-initial articulation is differentiated from stress- or accent-induced articulations along several measures. Secondly, the vowel is effectively louder domain-initially, suggesting that the boundary effect is not strictly local to the initial consonant. Thirdly, some accentual effects can be seen in secondary-stressed syllables, suggesting that accentual influences spread beyond the primary-stressed syllable. Finally, unlike domain-initial effects, prominence effects are not cumulative. Thus we conclude that, at least for the kind of word-initial syllables tested here, different aspects of prosodic structure (domain boundary vs. prominence) are differentially encoded.

Timbre perception.

An electroencephalographic study dealing with timbre perception has been performed. Timbre is next to tone and chroma is the basic element of sound and is perceived as a multidimensional space in which perceptual coordinates correlate with acoustic parameters. To show these correlations and their associations with different brain areas, sounds all constant with respect to their fundamental pitch of 311 Hz were varied in timbre. Due to these variations of acoustical parameters clear differences could be found concerning the perceptional pathway and corresponding brain areas. The modification applied to the sounds included variations of attack time, spectral center of gravity, and even harmonic attenuation. Next to other findings a hierarchy of perception was found where variations of the attack time was overwritten by the even harmonic attenuation and the spectral centroit variations, but only if all three were varied simultaneously. In all other cases the attack time variations caused clear perceptional delays in the prefrontal lobe.

A multidimensional acoustic analysis of vowels in two Polish dialects

In this paper, we attempt to examine the acoustic (oral) vowel spaces of two main Polish dialects: the Polish spoken in Cracow and that spoken in the wider region of Warsaw. A semi‐spontaneous corpus consisting of four speakers (two male, two female) of each dialect was recorded with approximately 2000 vowel occurrences for each speaker. The segmentation was performed manually and was subsequently corrected with an intensity‐based algorithm. The tokens were submitted to acoustic analyses, and two multidimensional representations were then prepared, one based on formants (accompanied with bandwidths and amplitudes) and the other on critical bands. Both representations included measurements for duration, F0, spectral center of gravity, skewness, and kurtosis. These values served as input for discriminant and principal component analysis which allowed us: (a) to compare the relative weight of each cue, (b) to isolate those parameters that allow a differentiation of the members of a vowel system (in a language such as Polish, F1 and F2 are traditionally the sole cues regarded as relevant), and (c) to detect any (dis)similarities between the two dialects in the three aforementioned aspects.

Spectral Integration of Dynamic Cues in the Perception of Syllable-Initial Stops

The present experiments examine the potential role of auditory spectral integration and spectral center of gravity (COG) effects in the perception of initial formant transitions in the syllables [da]-[ga] and [t^ha]-[k^ha]. Of interest is whether the place distinction for stops in these syllables can be cued by a ‘virtual F3 transition’ in which the percept of a frequency transition is produced by a dynamically changing COG. Listeners perceived the virtual F3 transitions comparably with actual F3 transitions although the former were less salient a cue. However, in a separate experiment, static ‘virtual F3 bursts’ were not as effective as actual F3 bursts in cueing the alveolar-velar place distinction. These results indicate that virtual F3 transitions can provide phonetic information to the perceptual system and that auditory spectral integration (completed by the central auditory system) may play a significant role in speech perception.

Salience of virtual formants as a function of the frequency separation between spectral components

The center-of-gravity (COG) hypothesis proposed by Chistovich and others for the perception of static vowels suggests that auditory spectral integration may occur when two or more formants fall within a 3.5 bark bandwidth. While several studies have examined the bandwidth limits of such integration, this study examines the extent to which spectral integration is uniform within this putative 3.5-bark range. We examine the perceptual salience of virtual formants produced by modifying the spectral COG of two closely spaced narrow-bandwidth resonances. Three different vowel series were created: [i-■], [■-■] and [■-■]. A second set of vowels was then created in which one of the formants (F1 in [i-■], F2 in [■-■] and F3 in [■-■]) was replaced by a virtual formant whose COG matched that of the formant that had been removed. The frequency separation between the two component resonances was then systematically varied between 1.5 and 3.5 barks and a singleinterval 2AFC vowel identification task was used to obtain estimates of vowel quality for each series step. Results will be discussed in terms of whether the spectral integration effects within the 3.5 bark decline as the frequency separation between the resonance components increases. [Work supported by NIDCD R01DC00679-01A1.]

Prosodic influences on consonant production in Dutch: Effects of prosodic boundaries, phrasal accent and lexical stress

Prosodic influences on phonetic realizations of four Dutch consonants (/t d s z/) were examined. Sentences were constructed containing these consonants in word-initial position; the factors lexical stress, phrasal accent and prosodic boundary were manipulated between sentences. Eleven Dutch speakers read these sentences aloud. The patterns found in acoustic measurements of these utterances (e.g., voice onset time (VOT), consonant duration, voicing during closure, spectral center of gravity, burst energy) indicate that the low-level phonetic implementation of all four consonants is modulated by prosodic structure. Boundary effects on domain-initial segments were observed in stressed and unstressed syllables, extending previous findings which have been on stressed syllables alone. Three aspects of the data are highlighted. First, shorter VOTs were found for /t/ in prosodically stronger locations (stressed, accented and domain-initial), as opposed to longer VOTs in these positions in English. This suggests that prosodically driven phonetic realization is bounded by language-specific constraints on how phonetic features are specified with phonetic content: Shortened VOT in Dutch reflects enhancement of the phonetic feature {−spread glottis}, while lengthened VOT in English reflects enhancement of {+spread glottis}. Prosodic strengthening therefore appears to operate primarily at the phonetic level, such that prosodically driven enhancement of phonological contrast is determined by phonetic implementation of these (language-specific) phonetic features. Second, an accent effect was observed in stressed and unstressed syllables, and was independent of prosodic boundary size. The domain of accentuation in Dutch is thus larger than the foot. Third, within a prosodic category consisting of those utterances with a boundary tone but no pause, tokens with syntactically defined Phonological Phrase boundaries could be differentiated from the other tokens. This syntactic influence on prosodic phrasing implies the existence of an intermediate-level phrase in the prosodic hierarchy of Dutch.

Characterizing the sound quality of air-conditioning noise*1

The aim of the psychoacoustic study presented here was to characterize listeners' preferences for a set of sounds produced by different brands and models of indoor air-conditioning units. In addition, some synthetic sounds, created by interpolation between recorded sound samples, were integrated into the set. The multidimensional perceptual space and the corresponding physical space representative of the sound set were determined with multidimensional scaling (MDS). Then the preferences for different classes of listeners were related to the physical space. The best spatial model yielded by the MDS had three common dimensions and specificities. The three dimensions are correlated with the ratio of the noisy part of the spectrum to the harmonic part (NHR), with the spectral center of gravity (SC) and with loudness ( N ). Two classes of listeners can be distinguished in terms of preference. For one, preference varied primarily with loudness, whereas for the other it varied more with SC and NHR. However, for one class the preference grew with the parameter NHR, while it decreased for the other class. The results replicate under different laboratory conditions and indicate the usefulness of this sound quality assessment approach for characterizing appliance noises.

Characterizing the sound quality of air-conditioning noise

The aim of the psychoacoustic study presented here was to characterize listeners' preferences for a set of sounds produced by different brands and models of indoor air-conditioning units. In addition, some synthetic sounds, created by interpolation between recorded sound samples, were integrated into the set. The multidimensional perceptual space and the corresponding physical space representative of the sound set were determined with multidimensional scaling (MDS). Then the preferences for different classes of listeners were related to the physical space. The best spatial model yielded by the MDS had three common dimensions and specificities. The three dimensions are correlated with the ratio of the noisy part of the spectrum to the harmonic part (NHR), with the spectral center of gravity (SC) and with loudness ( N). Two classes of listeners can be distinguished in terms of preference. For one, preference varied primarily with loudness, whereas for the other it varied more with SC and NHR. However, for one class the preference grew with the parameter NHR, while it decreased for the other class. The results replicate under different laboratory conditions and indicate the usefulness of this sound quality assessment approach for characterizing appliance noises.

Influence of syllable-coda voicing on the acoustic properties of syllable-onset /l/ in English

Properties of syllable onset /l/ that depend on the voicing of the syllable coda were measured for four speakers, representing different nonrhotic British English accents that differ in their phonetic realization of onset /l/ and in their system of phonological contrast involving onset /l/ and /r/. Onset /l/ was longer before voiced than voiceless codas for all four speakers, and darker for two of them as measured by lower F 2 frequency, and for these two and one other as measured by spectral center of gravity (COG). There were no coda-dependent differences in f 0 in the /l/, and F 1 frequency differed only for the fourth speaker. The vowel was also longer for all four speakers when the coda was voiced (as expected), while F 1 was lower and F 2 normally higher. One speaker provided data with fricative or affricate onsets: fricated segments were longer before voiced codas, but no coda-dependent COG differences were found. At least when the onset includes /l/, phonological voicing of the coda seems to be reflected in complex acoustic–phonetic properties distributed across the whole syllable, some properties being localized, others not. We describe these properties as variations in a bright–somber dimension. In most accents, when the coda is voiceless, the syllable is relatively bright: small proportions of periodic energy which is relatively high frequency at the syllable edges, and a high proportion of silence or aperiodic energy. When the coda is voiced, the syllable is relatively somber: a high proportion of periodic energy which is relatively low frequency at the syllable edges, and relatively small amounts of silence and aperiodic energy. Other accents use other combinations, dependent on the phonetic and phonological properties of liquids in the particular accent. The association of onset darkness and coda voicing does not seem to be ascribable to anticipatory coarticulation of features essential to voicing itself; this observation provides support for nonsegmental models of speech perception in which fine phonetic detail is mapped directly to linguistic structure without reference to phoneme-sized segments.

Electrophysiological correlates of musical timbre perception

Timbre perception has been studied by deriving a multidimensional space of the perceptual attributes from listeners’ behavioral responses. The neural bases of timbre space were sought. First, a psychophysical timbre dissimilarity experiment was conducted. A three-dimensional space of 16 synthetic sounds equalized for fundamental frequency, loudness, and perceived duration was designed. Sounds varied in attack time, spectral center of gravity, and energy ratio of odd/even harmonics. Multidimensional scaling revealed a three-dimensional perceptual space with linear or exponential relations between perceptual and physical dimensions. Second, in an electrophysiological experiment, the mismatch negativity (MMN) component of event-related potentials was recorded. The MMN is elicited by infrequently presented sounds differing in one or more dimensions from more frequent ones. Although elicited without the focus of attention, it correlates with the subjects’ behavioral responses, revealing the neural bases of preattentive discrimination. Eight sounds were chosen within the perceptual space. Changes along individual and combined dimensions elicited an MMN response. MMN latency varied depending on the dimension changed. In addition, preliminary analyses tend to show an additivity of the MMN waves for some pairs of dimensions. These results shed light on the neural processes underlying the perceptual representation of multidimensional sounds. [Work supported by the CTI program of the CNRS.]

Dynamic center-of-gravity effects in consonant–vowel transitions

Lublinskaja has reported that dynamic changes in the spectral center-of-gravity (COG) of one formant in selected Russian vowels led to changes in the identification of the vowel. Movement of the COG was effected by simultaneous amplitude modulation of two formants placed at the end points of the desired frequency transition [Iyer et al., J. Acoust. Soc. Am. 109, 2294 (2001)]. The present study explores whether such COG effects will extend into the processing of consonant–vowel transitions (CVs). A [dɑ]–[gɑ] continuum was generated using a Klatt synthesizer with a 50-ms CV transition and a 200-ms vowel. The onset of F3 varied from 2018–2818 Hz in 80-Hz step. The F3 of the vowel remained constant at 2527 Hz. Three different stimulus sets differed in terms of whether the F3 transition was a: (1) formant, (2) FM tone, or (3) virtual frequency (VF) glide. In the VF stimuli, the frequencies of two tones matched the onset or offset frequencies of the F3 transition and modulation of their amplitudes produced a gliding spectral COG. Preliminary results indicate that listener performance is not strongly affected by changing the means by which spectral changes to F3 are made. [Research supported by a grant from OSU’s College of Social and Behavioral Sciences to L. Feth and an INRS award from NIH to R. Fox.]

An EPG and acoustic study of fricative production and spectra

In order to test the correspondence between variability in fricative production and variability in the fricative spectrum, the spectral center of gravity (COG) is compared with the EPG center of gravity measured along the horizontal dimension. The CV coarticulatory effects in particular are examined. Data are taken from CV tokens produced by four female speakers of Australian English. Tokens consisted of the coronal fricatives /T s S D z Z/ in seven monophthong vowel contexts. Rounded vowels are excluded so as to minimize spectral changes due to lip-rounding. The consistency between the articulatory and the acoustic data is particularly striking for the palato-alveolar fricatives, although results are also highly consistent for the other two fricative places of articulation examined. However, it is not always clear that the variability in fricative production is conditioned by the vowel context. Overall, the sibilant fricatives seem particularly resistant to coarticulation and to variability in production, while the nonsibilant dental shows a good deal of variability, including some conditioning by the vowel context. These results have implications for the hyper- and hypo-theory of speech production. It is suggested that sibilant fricatives do not lend themselves to the articulatory imprecision, which characterizes other perceptually salient, and typologically common, speech sounds.

Effect of vocal effort on spectral properties of vowels.

The effects of variations in vocal effort corresponding to common conversation situations on spectral properties of vowels were investigated. A database in which three degrees of vocal effort were suggested to the speakers by varying the distance to their interlocutor in three steps (close--0.4 m, normal--1.5 m, and far--6 m) was recorded. The speech materials consisted of isolated French vowels, uttered by ten naive speakers in a quiet furnished room. Manual measurements of fundamental frequency F0, frequencies, and amplitudes of the first three formants (F1, F2, F3, A1, A2, and A3), and on total amplitude were carried out. The speech materials were perceptually validated in three respects: identity of the vowel, gender of the speaker, and vocal effort. Results indicated that the speech materials were appropriate for the study. Acoustic analysis showed that F0 and F1 were highly correlated with vocal effort and varied at rates close to 5 Hz/dB for F0 and 3.5 Hz/dB for F1. Statistically F2 and F3 did not vary significantly with vocal effort. Formant amplitudes A1, A2, and A3 increased significantly; The amplitudes in the high-frequency range increased more than those in the lower part of the spectrum, revealing a change in spectral tilt. On the average, when the overall amplitude is increased by 10 dB, A1, A2, and A3 are increased by 11, 12.4, and 13 dB, respectively. Using "auditory" dimensions, such as the F1-F0 difference, and a "spectral center of gravity" between adjacent formants for representing vowel features did not reveal a better constancy of these parameters with respect to the variations of vocal effort and speaker. Thus a global view is evoked, in which all of the aspects of the signal should be processed simultaneously.

Phonological boundaries and the spectral center of gravity

A critical limit of 3–3.5 Bark has been reported for the ‘‘spectral center of gravity’’ effect between the second and third formants (F2 and F3) and is assumed to correspond to a perceptually natural boundary [A. K. Syrdal and H. S. Gopal, J. Acoust. Soc. Am. 79, 1086–1100 (1986)]. This hypothesis was tested for both the [+/−back] and [+/−coronal] distinctions in English vowels. Subjects identified two sets of three-formant synthetic vowels which varied orthogonally in F2 and F3 and ranged between /ᴜ/–/ɪ/ ([+/−back]) or /ɝ/–/ᴜ/ ([+/−coronal]). For the /ᴜ/–/ɪ/ distinction, a boundary shift was observed solely as a function of F2, ruling out an invariant F3–F2 boundary. For the /ɝ/–/ᴜ/ series, both F2 and F3 influenced boundary location. There was a relatively stable F3–F2 boundary for the mean identification responses in this case, but it occurred at less than the predicted 3–3.5 Bark difference. Follow-up results will also be presented. [Work supported by NIDCD.]

Latency of MEG M100 response indexes first formant frequency

Magnetoencephalography recordings from the auditory cortex of subjects listening to synthetic vowels show a close correlation between the timing of the evoked M100 response and the first formant frequency (F1). These results are consistent with evoked magnetic field latencies elicited by tone stimuli, which show 100 to 300-Hz tones associated with latencies up to 30 ms longer than 500 to 3000-Hz tones. In experiment 1, three-formant vowels /u,a,i/ were presented at two fundamental frequencies (F0=100 Hz, 200 Hz). The M100 latency was a function of the vowel identity and not F0: M100 was significantly shorter for /a/ than /u/, consistent with the spectral center of gravity being at a higher frequency for /a/. In experiment 2, single-formant vowels (/u/:F1=330 Hz, /a/:F1=720 Hz) were covaried with two F0 values. M100 latencies were shorter for /a/ (high F1) than for /u/ (low F1), at both fundamentals. In experiment 3, subjects listened to pure-tone complexes with frequencies and amplitudes matching the F0 and F1 energy peaks of the stimuli in experiment 2. M100 latencies showed the same pattern: latency covaried with the energy peak corresponding to F1, suggesting that the sensitivity to the energy in the F1 range is not speech-specific. Finally, an experiment involving two-formant, amplitude-varying vowels will be presented.

The ‘‘center-of-gravity’’ effect in dynamic tones

Earlier work [e.g., Xu et al., J. Acoust. Soc. Am. 101, 3149(A) (1997); Chistovich, ibid. 77, 798–805 (1985)] has demonstrated the spectral center-of-gravity (CoG) effect. The studies have largely concentrated on steady-state signals. Recently, Lublinskaya (1996) has investigated the CoG effect in dynamic synthetic speech. In the study presented here, the CoG effect in dynamic two-tone complexes was investigated. The frequencies of the two tones were held constant while their amplitudes were varied linearly: the amplitude of the lower frequency tone was decreased while that of the higher frequency tone was increased. This causes a variation in the spectral CoG, although the locations of the frequency content remain constant with time. Subjects matched the slope of a linear glide (target) having the same center frequency as the two-tone complex (standard) using a double-staircase procedure. The experiment was carried out for various center frequencies (500–2000 Hz) and frequency separations (0.5–8 ERBs). The center frequency was also roved between trials. Matching experiments were also carried out with another linear glide as the standard. The results of these experiments will be presented and the center-of-gravity effect in these dynamic two-tone complexes will be discussed. [Work supported by AFOSR and OSU-GSARA.]

Dipolar relaxations in glycerol: a dynamic fluorescence study of 4-[2'-(dimethylamino)-6'-naphthoyl]cyclohexanecarboxylic acid (DANCA)

Author(s): Bismuto, E; Jameson, DM; Gratton, E | Abstract: Solvent dipolar reorientations following excitation of DANCA in glycerol at various temperatures were investigated by using multifrequency-phase fluorometry. The time evolution of the emission spectrum was determined by performing lifetime measurements at a number of wavelengths across the emission band and then reconstructing the spectrum at selected times after excitation. At low temperatures, a large displacement of the time-resolved spectrum was apparent. The spectral center of gravity was observed to move faster at short times and then to exhibit a long decay. The spectral width was also observed to increase at short times and then to decrease at much longer times. Several models have been considered to account for the major features of our experimental results. A two-state model agreed well with the data only in the high-temperature regime; at lower temperatures, the data were not consistent with the characteristic features of the two-state model. In the high-temperature regime, we obtained a reorientation rate for the glycerol molecules using the two-state model. Other models which consider a continuous translation of the emission spectra as a function of time also failed to account for the experimental results. Instead, a good qualitative agreement was found with the model proposed by Weber in which the charge separation of the excited fluorophore is followed by reorientation of the solvent molecules around the individual monopoles. © 1987, American Chemical Society. All rights reserved.

Aspects of a model of the auditory representation of american english vowels

This paper describes some of the implications of Ludmilla Chistovich's “spectral center of gravity” (SCG) effect for a model of the auditory representation of American English vowels. Chistovich's work on defining a critical distance for the SCG effect is closely related to two of the most fundamental problems in speech communication research: the relation between acoustic attributes and phonemic features and the problem of invariance despite large acoustic differences between speakers. Firstly, experimental findings relating to the SCG effect are reviewed. Secondly, a model incorporating these perceptual effects is described, and thirdly, three aspects of the model are discussed: (1) the resulting feature analysis, (2) normalization, and (3) acoustic variability as represented in the model and its relation to Steven's quantal theory of speech production.