Speaker identification utilizing noncontemporary speech.

The noncontemporariness of speech is important to both of the two general approaches to speaker identification. Ear-witness identification is one of them; in that instance, the time at which the identification is made is noncontemporary. A substantial amount of research has been carried out on this relationship and it now is well established that an auditor's memory for a voice decays sharply over time. It is the second approach to speaker identification which is of present interest. In this case, samples of a speaker's utterances are obtained at different points in time. For example, a threat call will be recorded and then sometime later (often very much later), a suspect' s exemplar recording will be obtained. In this instance, it is the speech samples that are noncontemporary and they are the materials that are subjected to some form of speaker identification. Prevailing opinion is that noncontemporary speech itself poses just as difficult a challenge to the identification process as does the listener's memory decay in earwitness identification. Accordingly, series of aural-perceptual speaker identification projects were carried out on noncontemporary speech: first, two with latencies of 4 and 8 weeks followed by 4 and 32 weeks plus two more with the pairs separated by 6 and 20 years. Mean correct noncontemporary identification initially dropped to 75-80% at week 4 and this general level was sustained for up to six years. It was only after 20 years had elapsed that a significant drop (to 33%) was noted. It can be concluded that a listener's competency in identifying noncontemporary speech samples will show only modest decay over rather substantial periods of time and, hence, this factor should have only a minimal negative effect on the speaker identification process.

Production of intoxication states by actors--acoustic and temporal characteristics.

This paper is the second of a series; the first has been published (J Forensic Sci, 1998;43:1153-62). The goal in the initial pair of experiments was to determine if speakers (actors) could effectively mimic the speech of intoxicated individuals and also volitionally reduce the degradation to their speech that resulted from severe inebriation. To this end, two highly controlled experiments involving 12 actor-speakers were carried out. It was found that, even when sober, nearly all of them were judged drunker (when pretending) than when they actually were severely intoxicated. In the second experiment, they tried to sound sober when highly intoxicated; here most were judged less inebriated than they were. The goal of this second paper is to identify some of the speech characteristics that allowed the subjects to achieve the cited illusions. The focus here is on four paralinguistic factors: fundamental frequency (F0), speaking rate, vocal intensity, and nonfluency level. For the simulation of intoxication study, it was found that F0 was raised along with increased intoxication but raised even more when this state was feigned. A slowing of speaking rate was associated with increasing intoxication, but this shift also was greater when the speaker simulated intoxication. The most striking contrast was found for the nonfluencies; they were doubled for actual intoxication, but quadrupled when intoxication was simulated. On the other hand, the shifts exhibited by the subjects when they attempted to sound sober were not as clear cut. Indeed, no systematic relationships were found here for either F0 or vocal intensity. Both speaking rate and the number of nonfluencies shifted appropriately, but these changes were not statistically significant. In sum, discernable suprasegmental relationships occurred for both studies (but especially the first); further, it is predicted that useful cues also will be found embedded in the segmentals (the sounds of speech).

Effects of ethanol intoxication on speech suprasegmentals.

The effects of ingesting ethanol have been shown to be somewhat variable in humans. To date, there appear to be but few universals. Yet, the question often arises: is it possible to determine if a person is intoxicated by observing them in some manner? A closely related question is: can speech be used for this purpose and, if so, can the degree of intoxication be determined? One of the many issues associated with these questions involves the relationships between a person's paralinguistic characteristics and the presence and level of inebriation. To this end, young, healthy speakers of both sexes were carefully selected and sorted into roughly equal groups of light, moderate, and heavy drinkers. They were asked to produce four types of utterances during a learning phase, when sober and at four strictly controlled levels of intoxication (three ascending and one descending). The primary motor speech measures employed were speaking fundamental frequency, speech intensity, speaking rate and nonfluencies. Several statistically significant changes were found for increasing intoxication; the primary ones included rises in F0, in task duration and for nonfluencies. Minor gender differences were found but they lacked statistical significance. So did the small differences among the drinking category subgroups and the subject groupings related to levels of perceived intoxication. Finally, although it may be concluded that certain changes in speech suprasegmentals will occur as a function of increasing intoxication, these patterns cannot be viewed as universal since a few subjects (about 20%) exhibited no (or negative) changes.

Perceptual confusions of high-pitched sung vowels.

Questions exist as to the intelligibility of vowels sung at extremely high fundamental frequencies and, especially, when the fundamental frequency (F0) produced is above the region where the first vowel formant (F1) would normally occur. Can such vowels be correctly identified and, if so, does context provide the necessary information or are acoustical elements also operative? To this end, 18 professional singers (5 males and 13 females) were recorded when singing 3 isolated vowels at high and low pitches at both loud and soft levels. Aural-perceptual studies employing four types of auditors were carried out to determine the identity of these vowels, and the nature of the confusions with other vowels. Subsequent acoustical analysis focused on the actual fundamental frequencies sung plus those defining the first 2 vowel formants. It was found that F0 change had a profound effect on vowel perception; one of the more important observations was that the target tended to shift toward vowels with an F1 just above the sung frequency.

Production of intoxication states by actors: perception by lay listeners.

The effects of ingesting ethanol have been shown to be somewhat variable in humans; there appear to be but few universals. Yet, questions about intoxication often are asked by law enforcement personnel (especially relative to DUI), clinicians and various individuals in social settings. A key question: Is it possible to determine if a person is intoxicated by observing them in some manner? A closely associated one: Can speech be used for that purpose? Two of the many issues related to the second of these questions involve the possibility that (1) speakers, especially actors, can effectively mimic the speech of intoxicated individuals, and (2) they may be able to volitionally reduce any speech degradation which results from intoxication. The approach used to test these two questions tasked auditors to determine if these simulations were possible. To this end, young, healthy actors chosen on the basis of a large number of selection criteria were asked to produce several types of controlled utterances (1) during a learning phase, (2) when sober, (3) at three simulated levels of intoxication (mildly, legally and severely drunk), (4) during actual, and parallel, levels of intoxication, and (5) at the highest intoxication level attained but when attempting to sound completely sober. Two aural-perceptual studies were conducted; both involved counterbalanced ABX procedures where each subject was paired with him/herself. Listeners were normally hearing university students drawn from undergraduate phonetics and linguistics courses. In the first study, they rated the actors as being more intoxicated--when they actually were sober but simulating drunkenness--88% more often than when they actually were intoxicated. In the second study, they were judged as sounding less inebriated when attempting to sound sober (than they actually were) 61% of the time. These relationships would appear to impact a number of situations; one of special importance would be the detection of intoxication in motorists.

Effects of three parameters on speaking fundamental frequency.

Speaking fundamental frequency levels and usage (SFF, F0) are of interest to many investigators who study human speech and voice. Substantial research in the area has been carried out; common foci include SFF as related to infant cry, age, gender, adolescent voice change, language, race, voice pathology, and so on. Yet there still are a number of relationships which are not well understood and three of them will be addressed in this project. They involve the long-held notions that (1) a secular trend exists with SFF being lowered over time, (2) the use of university students in research of this type will create bias because they are physically different from average individuals, and (3) SFF can vary systematically for different types of speech (especially for oral reading and extemporaneous speaking). Experiments assessing these questions were carried out, but only certain of the postulates were supported. That is, while some evidence of a secular trend was found, it appeared inconsequential during the past quarter of this century; second, although university students were found to be slightly larger than a cohort approaching the average population, only minor vocal differences were found. Finally, it was observed that, in general, oral reading resulted in higher mean SFF's than those for spontaneous speech. However, this difference was not robust and, due to reversals, the resulting metric did not appear to be of good predictive value for individual speakers.

Longitudinal research on adolescent voice change in males.

Many of the characteristics and processes associated with adolescence are reasonably well understood; others are not. Still unanswered are questions about adolescent voice change. For example, when does the average child start the process; what changes result and what is their extent; when is it complete? Further, how closely do shifts in vocal level parallel, or even predict, the adolescent process? This project was carried out in order to obtain data which would complement available information on the subject and provide additional specificity about voice change and pubescence--at least for boys. To that end, a longitudinal study was conducted in which 48 males were tracked over a 5-year period. Voice measurements were made--including speaking fundamental frequency (SFF) and phonational frequency range (PFR)--as were others relating to physical size (height, weight, and six body dimensions). It was found that (1) stable adolescent voice change (AVC) measurements could be obtained, (2) maturation patterns appeared predictable on the basis of AVC changes, and (3) both the onset and duration of voice change extended over a longer period of time than had been previously thought.

That golden voice--talent or training?

This article is an updating and expansion of a lecture presented at the Thirteenth Symposium on the Care of the Professional Voice, New York, June 1984; a somewhat limited version of the manuscript was published in the Transcripts of that symposium. Basically, an attempt will be made by an observer from another field to lay the groundwork for the better understanding of the "voice teacher." Accordingly, some of the dialogue to follow may be a little informal; it certainly is interpretive. The latter portion of the discussion will involve suggestions/illustrations that may be useful in further enhancing this somewhat underrated field and bonding it to others. While the comments found below reflect only the opinions of the author, it is fair to state that most of them have been articulated by others.

Speaking fundamental frequency patterns of Japanese women.

The purpose of this research was to obtain information on the speaking fundamental frequency (SFF) mean levels, variability and patterns of Japanese females and contrast them with those of Caucasian speakers of American English. To this end, 56 young women--32 Japanese and 24 Americans--read standard passages in either or both the Japanese and English languages. It was found that the Japanese exhibited higher fundamental frequencies than did the Americans for all speaking conditions, and this contrast was statistically significant. Additionally--and unlike the American speakers--most Japanese women exhibited bimodal SFF distribution patterns. It is judged that the observed differences in level and distribution result primarily from differences in the structure of the two languages.

Listening preferences for voice types as a function of age.

This experiment was designed primarily to generate information about the preferences of older listeners for various classes of voices. Speech samples for that purpose were elicited from 80 speakers, who provided the desired stimuli (sentences) under frequency and intensity control. Specifically, there were eight cells in the design, each represented by 10 speakers (5 male and 5 female); all combinations of low, medium, and high speaker fundamental frequency (SFF) were combined with soft, middle and loud vocal intensity (VI) productions--except for the low-SFF/high-VI combination, which proved impossible to obtain. Listeners were four groups of 20 individuals equally divided as to sex. The two older of these groups, designated as the experimental subjects, were: older adults (60-70 years of age) and the elderly (80-90 years). The two younger groups served as controls; they included young adults (20-30 years of age) and middle-aged adults (aged 40-50 years). Listeners rated each speech sample on a 5-point preference scale varying from "like very much" to "dislike very much." The results suggest that most listeners prefer medium intensity voices. Other preference tendencies were toward low-pitched voices and (slightly) toward male speakers; but these trends were not as marked as the first. Most importantly, there were no systematic differences in voice type preferences between or among the older and younger groups--or between male and female listeners.

Voice stress evaluators and lie detection.

Proponents of vocal stress analysis systems argue that they are able to detect spoken deception by analysis of "stress" in the voice signal. Presumably, they do so by examining traces made by laryngeal microtremors which, they claim exist in the voice, are associated with stress, and ultimately are associated with lying. However, most research that seeks to identify the relationships between microtremors and laryngeal function has produced negative results, and data on the ability of voice analyzers to detect stress from speech--or to identify spoken deception--have been negative or "mixed" in nature. Since perspectives based on available results leave a number of questions unanswered, a series of experiments has been undertaken. The first was focused on the basic acoustic/temporal correlates of stress in voice (the subject of an earlier report), the second on examination of stress by commercial voice analyzers, and the third on the detection of relatively high-risk lies by this same type of voice analysis procedure. It was found that correct stress/nonstress identifications occurred only at chance levels; the lie/nonlie identification scores were quite similar with professional "examiners" performing at about the same level of accuracy as other auditors. The following review is divided into two parts: a history of the controversy and a presentation of the two cited experiments.

An acoustic approach to diver navigation.

Three experiments were carried out to assess the capability of divers to localize acoustic signals underwater and to navigate by them. In the first experiment, divers attempted to determine the correct distance to several underwater transducers projecting acoustic stimuli on a horizontal plane. The second experiment consisted of two related studies where the diver/subjects attempted to discover which of many possible stimuli would produce the most robust perception of underwater "sound movement," or the Underwater Auditory Phi Phenomenon (UAPP). A third experiment consisted of navigational swims by divers; the acoustic stimuli utilized were based on those identified in prior experiments as the most preferred. The results demonstrated that divers are able to discriminate among signals emanating from acoustic sources at various distances underwater and to do so at levels well above chance. Second, divers judged 500-Hz square waves to be the signal which best facilitated an acceptable UAPP; thermal noise and 1-kHz square waves followed in that order. However, these differences were only slight and, in practice, divers maintained that the noise signal was the most useful. Third, it was found that divers apparently can effectively navigate by means of auditory signals alone--at least within certain limits. Finally, a significant decrease in the discrimination abilities of divers for frequencies above 6 kHz suggests that intensity cues may not be as robust as time-of-arrival information with respect to underwater sound localization.

Criminal identification comparison: aural versus visual identifications resulting from a simulated crime.

This research was undertaken to investigate and contrast visual and aural/perceptual identifications of a previously unknown individual from a set of photographic and tape-recorded exemplars following a simulated crime. All participants were volunteers for the "criminal," the victim of an "assault," and all suspects drawn from a Reserve Officer Training Corps (ROTC) class; the "witnesses" were 61 students in a law class. The student/witnesses were divided into four groups. Group A made three identifications (serially) one day, one week, and two weeks after the crime took place. Group B saw the photographs and heard the tapes only once a week after the incident, and Group C only after two weeks had elapsed. Witnesses in a fourth group (D) followed the same schedule as did Group A; however, they were presented with foils similar in appearance and speech to the criminal. Group D was presented both the foil and criminal in the photographic lineup at the final judging session. The results demonstrated that visual identification can be quite accurate although not consistently or predictably so. By contrast, aural/perceptual identifications were relatively poor. No strong trends for latencies were observed, either for repeated trials or for procedures involving different initiation latencies; nor did confidence levels appear to be related to accuracy of judgment. Finally, when a similar looking foil was included in the identification task, there was a weak trend for the foil to be chosen more often in subsequent trials. These results support the position that eyewitness--and especially earwitness--testimony should be viewed by judges and juries with greater caution than has been the case in the past; by appropriate instructions, juries should be given assistance in interpreting and in assigning appropriate weight to this kind of testimony.

Speech intelligibility at high helium-oxygen pressures.

Word-list intelligibility scores of unprocessed speech (mean of 4 subjects) were recorded in helium-oxygen atmospheres at stable pressures equivalent to 1600, 1400, 1200, 1000, 860, 690, 560, 392, and 200 fsw daring Predictive Studies IV-1975 by wide-bandwidth condenser microphones (frequency responses not degraded by increased gas density). Intelligibility scores were substantially lower in helium-oxygen a 200 fsw than in air at l ATA, but there was little difference between 200 fsw and 1600 fsw. A previously documented prominent decrease in intelligibility of speech between 200 or 600 fsw because of helium and pressure was probably due to degradation of microphone frequency response by high gas density.