Frequency Glides Research Articles

Auditory Representational Momentum: Surface Form, Direction, and Velocity Effects

In a series of four experiments, subjects were presented with inducing sequences consisting of either a series of discrete tones or a continuous frequency glide. A probe tone was then presented, and subjects judged whether the probe was the same pitch as the final pitch in the inducing sequence. When the final frequency of the inducing sequence was prolonged, subjects' memory for the final pitch was displaced forward in the direction of implied motion. When the final frequency of the inducing sequence was not prolonged, no consistent pattern of displacement was found. With descending auditory motion, faster velocities led to larger forward displacements than slower velocities, but with ascending auditory motion, velocity did not significantly affect displacement. In general, displacements were influenced by the direction and velocity of motion and by the duration of the final frequency, but not by whether the stimulus was presented in a continuous form or a discrete form. These results are consistent with Freyd's (1993) hypothesis that representational momentum does not depend upon the surface form of the stimulus and can be found when stimuli drawn from a continuous

Frequency glide discrimination in the F2 region by normal-hearing and hearing-impaired listeners

Several researchers have suggested that the dynamic formant transitions present within consonant–vowel and vowel–consonant utterances may be poorly processed by hearing-impaired (HI) listeners. To investigate HI listeners’ processing of transitional stimuli, six HI and six normal-hearing (NH) listeners were asked to discriminate between frequency glides patterned after second formant transitions in English CV syllables. The influence of glide duration, rate, and frequency extent were examined for each group in quiet and in 35-dB spectrum level broadband noise. Glide rate did not show a significant direct effect on frequency difference limens (DLs) for glide onset. Reductions in glide duration and the presence of noise each led to significant increases in DLs. The magnitudes of these effects were similar across groups. The only significant group differences were observed for gradual glides of limited frequency extent. This finding is consistent with more accurate processing of ‘‘cochlear dispersion’’ cues on the part of NH listeners [Porter etal., J. Acoust. Soc. Am. 90, 1298–1308 (1991)]. [Work supported by NIDCD.]

Frequency glide discrimination and frequency resolution as a function of age among a normal hearing population

The relationship between frequency glide discrimination and frequency resolution was investigated among a group of normal hearing individuals aged 20 to 75 years. Frequency glide discrimination for 60-ms stimuli was determined at two frequency regions (1000 and 2500 Hz) for both upward and downward trajectories; converging or diverging offset frequencies; and in quiet and background noise. Psychophysical tuning curves were measured in a forward masking paradigm at both 500 and 4000 Hz. Relative to young adults, the elderly displayed larger just noticeable differences, greater effects for frequency region, noise, and transition direction. Though there was a significant effect for age, the results indicate that hearing sensitivity was the most important variable explaining more of the variance in the data than age. Even within normal limits, sensitivity still decreased with increasing age and therefore discrimination performance and shape of the tuning curves varied as a function of age. Consequently, the effect of age is confounded with sensitivity levels despite normal audiological profiles. Altered tuning curves are present before clinical evidence of cochlear damage indicating an impaired processing of complex signals.

Temporal resolution for frequency gliding signals and its relation to the phonemic perception of /b/ and /w/

The effects of age and hearing loss on amplitude and temporal resolution were examined. The first experiment assessed the perception of the phonemic distinction between three continua of /b/ and /w/ stimuli which varied in their amplitude envelopes. The degree of hearing impairment significantly effected the importance of the type of acoustical cue with hearing impaired listeners demonstrating a greater influence of the amplitude envelope and a shift in the phonemic boundary toward longer transition durations relative to normal listeners. The second experiment examined duration discrimination in a 2AFC task using a constant stimuli method for isolated analogs of the second formant transitions. Thresholds were interpolated from the 75% correct position from probits fitted to the psychometric function. The effects of both age and hearing status were significant, however the type of hearing impairment (shallow or steeply sloping) was an important factor in temporal resolution. Additionally, among listeners with normal hearing within the frequency range being tested elderly adults showed poorer temporal resolution relative to young adults.

Dynamic frequency change among stimulus components: effects of coherence on detectability.

Experiments were performed to determine the effect of coherent dynamic frequency change among stimulus components on detection of one of those components. Detectability of a frequency glide signal, centered at 660 Hz, was measured in the presence of two additional frequency glides centered at 220 and 440 Hz. Results show that the signal is most difficult to detect when it is changing coherently with the other stimulus components, and that detectability gradually improves as the frequency change among stimulus components becomes less coherent. A control experiment indicated that detectability of the signal is influenced somewhat by the average spectral distance between the signal and the other stimulus components. In order to separate the effects of dynamic frequency change from those of altering harmonic relationships, the experiment was repeated with stimulus components not harmonically related. Results are similar in pattern, though with somewhat smaller shifts. A final experiment was performed to determine if the observed shifts in detectability might be due to changes in the maximum spectral distance between signal and masker when their frequencies are changing incoherently. Results of this experiment indicate that detectability improves when frequency change of signal and masker is not coherent, even when maximum spectral distance does not change. These data indicate the likelihood that the auditory system is able to use coherent dynamic frequency change among stimulus components.

Effects of glide slope, noise intensity, and noise duration on the extrapolation of FM glides through noise.

Listeners are quite adept at maintaining integrated perceptual events in environments that are frequently noisy. Three experiments were conducted to assess the mechanisms by which listeners maintain continuity for upward sinusoidal glides that are interrupted by a period of broadband noise. The first two experiments used stimulus complexes consisting of three parts: prenoise glide, broadband noise interval, and postnoise glide. For a given prenoise glide and noise interval, the subject's task was to adjust the onset frequency of a same-slope postnoise glide so that, together with the prenoise glide and noise, the complex sounded as "smooth and continuous as possible." The slope of the glides (1.67, 3.33, 5, and 6.67 Bark/sec) as well as the duration (50, 200, and 350 msec) and relative level of the interrupting noise (0, -6, and -12 dB S/N) were varied. For all but the shallowest glides, subjects consistently adjusted the offset portion of the glide to frequencies lower than predicted by accurate interpolation of the prenoise portion. Curiously, for the shallowest glides, subjects consistently selected postnoise glide onset-frequency values higher than predicted by accurate extrapolation of the prenoise glide. There was no effect of noise level on subjects' adjustments in the first two experiments. The third experiment used a signal detection task to measure the phenomenal experience of continuity through the noise. Frequency glides were either present or absent during the noise for stimuli like those use in the first two experiments as well as for stimuli that had no prenoise or postnoise glides. Subjects were more likely to report the presence of glides in the noise when none occurred (false positives) when noise was shorter or of greater relative level and when glides were present adjacent to the noise.

Discrimination of frequency transitions as a function of spectral and temporal psychoacoustic cues.

The present study examined discrimination of short duration frequency glides (60 ms) analogous to the second formant transition of speech. An interdependence of rate of change, frequency excursion, and duration complicates the study of frequency transitions, and to determine how these psychoacoustic cues impact upon perceptual decisions stimuli in the current studies varied as a function of offset frequency and rate of change. Three continua of computer synthesized signals were used to determine just noticeable differences (jnds) for signals described by: a constant onset frequency with diverging offset frequencies; varying onset frequencies with a constant offset frequency; and varying offset frequencies with a constant rate of change. To attenuate frequency onset information, a 20 ms of octave-band noise was appended to the beginning of each signal. Results showed that discrimination improved as a function of the available psychoacoustic information: covarying transition rate and offset frequency yielded smaller jnds than varying either offset frequency or rate of change alone. Discrimination of transitions is not a simple function of frequency excursion or rate of change and integration of rate-related and spectral properties is speculatively important in perceptual decisions regarding complex signals such as speech.

Age-related auditory discrimination of short-duration steady-state, and frequency glides signals.

Evidence indicates a developmental component to the ability of the auditory system to resolve rapid frequency changes: Both young children and elderly listeners require larger acoustic differences in order to be able to differentiate signals, and this occurs in the absence of any reduced pure-tone sensitivity. The present study examined auditory sensitivity of school-aged children, young adults, and elderly listeners in the discrimination of both frequency glides and steady-state signals corresponding to the frequency range of the second formant of speech. Frequency glides had a constant onset frequency and diverged to varying offset frequencies; steady state signals corresponded to the offset frequencies of the gliding signals. Stimuli were presented monaurally and just noticeable differences were calculated using a two-cue, two-alternative, forced-choice paradigm. Significant differences were noted as a function of age and signal type, and a shift in the psychometric functions suggesting a greater sensitivity for the young adults. Results are discussed in terms of the maturation and efficiency of auditory processes.

Neuromagnetic responses of the human auditory cortex to short frequency glides

Neuromagnetic responses were recorded to 33-ms frequency-glide tones with a 7-channel SQUID gradiometer. Ninety per cent of the tones were of rising/falling frequency and 10% otherwise similar tones but presented backwards, i.e. comparable falling/rising frequency glides. The infrequent stimuli elicited a specific response, the mismatch field (MMF), whose equivalent source suggested a generation in the supratemporal auditory cortex. Responses to rising and falling infrequent glides showed no consistent asymmetry. The MMF amplitude correlated positively with the glide magnitude. It is suggested that the generation of MMF to changes in the direction of frequency glides involves neuronal networks tuned to frequency glides.

Audiometric profile in presbycusis

A group of 40 elderly subjects were evaluated using a variety of audiological tests. Their mean age was 68 years. They had all come to the clinic because of hearing problems. The subjects were selected for this study because no cause of their hearing loss could be found other than presbycusis. They all had the typical moderately sloping pure tone audiogram configuration. The pure tone average hearing loss (PTA, average of 0.5, 1 and 2 kHz) ranged from 11 to 70 dBHL with a median of 40. The average air-bone gap was 3 dB. In addition to pure tone audiometry the following tests were performed: most comfortable loudness level and uncomfortable loudness level for pure tone; stapedius reflex thresholds; speech recognition threshold and maximum speech recognition score in quiet; distorted speech recognition; detection thresholds for frequency and intensity glides of a pure tone; slow evoked cortical potentials in response to frequency and intensity glides of a pure tone. The test results, when compared with those obtained on young subjects with normal hearing and with cochlear lesions, show a mixture of characteristics typical of cochlear, retrocochlear and central lesions.

Discrimination of frequency‐glide direction

It is often assumed that the relevant information in formant transitions includes the direction as well as the extent of frequency excursions. A study to determine the ability of listeners with normal hearing to determine the direction of linear frequency modulation of sinusoidal tones has been devised. To more closely approximate the listener's task in processing speech, music, or other environmentally important sounds, the initial frequency of each transition was selected at random from within a predefined range. Thus, for each interval of a 2Q, 2AFC listening task, the listener hears a frequency glide of the same duration and extent. In one of the two middle intervals, the direction of the transition is reversed to produce the “signal.” Since each of the four glides begins and ends on frequencies selected at random, the listener cannot rely on simple pitch differences to determine which interval contained the reversed glide. Our preliminary results will be discussed, with emphasis on the effect of the width of the “roving frequency” range on listener performance. [Work supported by a grant from AFOSR.]

Detecting single-cycle frequency modulation imposed on sinusoidal, harmonic, and inharmonic carriers

Thresholds for detecting frequency modulation (FM) of various carriers were measured in quiet and in noise. The carriers were either pure tones, harmonic complexes, or inharmonic complexes. The modulator was always a single cycle of a 12.5-Hz sinusoid. When the carrier was a complex tone, each component of that carrier was modulated by the same percentage of its starting frequency. This is the form of FM that occurs when the fundamental frequency (F0) of a complex sound is modulated. Threshold was defined in terms of this percentage frequency change. Two phases of the modulator were used, producing percepts of either upward or downward frequency glides. Experiment 1 used sinusoidal carriers with frequencies between 200 and 2400 Hz. Thresholds were constant for carrier frequencies of 1200 Hz and above, increased below 800 Hz, and were higher for the ‘‘down’’ than for the ‘‘up’’ glides, especially with low carrier frequencies. Experiment 2 measured thresholds for carriers consisting of three adjacent harmonics of complexes with F0’s of 50, 100, and 200 Hz as a function of the frequency of the center component of each complex (CF). Thresholds varied with CF in a similar way to those for the pure-tone carriers. In the frequency region 1200–2400 Hz, where thresholds did not vary with CF, they did not vary systematically with F0. Experiment 3 compared thresholds for a 20-component harmonic complex with those for an inharmonic complex. Thresholds were only slightly lower for the harmonic complex than for the inharmonic complex when presented in quiet, but were much lower when the complexes were presented in bursts of pink noise. Thresholds in noise for the completely harmonic complex were also lower than those for 20-component complexes in which only five adjacent components shared a harmonic relationship. It was concluded that a harmonic relationship between components occupying widely different frequency regions aids the detection of frequency modulation, particularly in the presence of noise.

Detection and discrimination of spectral peaks and notches at 1 and 8 kHz.

The ability of subjects to detect and discriminate spectral peaks and notches in noise stimuli was determined for center frequencies fc of 1 and 8 kHz. The signals were delivered using an insert earphone designed to produce a flat frequency response at the eardrum for frequencies up to 14 kHz. In experiment I, subjects were required to distinguish a broadband reference noise with a flat spectrum from a noise with either a peak or a notch at fc. The threshold peak height or notch depth was determined as a function of bandwidth of the peak or notch (0.125, 0.25, or 0.5 times fc). Thresholds increased with decreasing bandwidth, particularly for the notches. In experiment II, subjects were required to detect an increase in the height of a spectral peak or a decrease in the depth of a notch as a function of bandwidth. Performance was worse for notches than for peaks, particularly at narrow bandwidths. For both experiments I and II, randomizing (roving) the overall level of the stimuli had little effect at 1 kHz, but tended to impair performance at 8 kHz, particularly for notches. Experiments III-VI measured thresholds for detecting changes in center frequency of sinusoids, bands of noise, and spectral peaks or notches in a broadband background. Thresholds were lowest for the sinusoids and highest for the peaks and notches. The width of the bands, peaks, or notches had only a small effect on thresholds. For the notches at 8 kHz, thresholds for detecting glides in center frequency were lower than thresholds for detecting a difference in center frequency between two steady sounds. Randomizing the overall level of the stimuli made frequency discrimination of the sinusoids worse, but had little or no effect for the noise stimuli. In all six experiments, performance was generally worse at 8 kHz than at 1 kHz. The results are discussed in terms of their implications for the detectability of spectral cues introduced by the pinnae.

Otoneurological findings in psycho-organic syndrome caused by industrial solvent exposure.

Nine subjects with long-term (8-30 years) occupational exposure to industrial solvents and a confirmed diagnosis of psycho-organic syndrome (POS) have been studied with audiological and otoneurological test batteries. The results were compared to a matched control group of nine industrial workers not exposed to solvents and to normal data from healthy volunteers. In the clinical examination, the Romberg test identified 5/9 workers as pathologic and concurrently the stabilometry showed significantly larger sway areas in the POS-group. In the audiological test battery, the significantly pathologic tests were discrimination of interrupted speech and evoked cortical responses to frequency glides (CRA-delta-f). The saccade test disclosed abnormal findings in 5/9 workers. In the smooth pursuit test, abnormality was found at some test frequencies using pseudorandomized stimulus. The VOR-suppression test was significantly abnormal at all test frequencies. The test battery used strongly indicates CNS lesions due to industrial solvents.

Detection of linear frequency glides as a function of frequency and duration

This article compares frequency discrimination thresholds for four conditions, using a two-interval forced-choice task: (1) discrimination of a difference in center frequency for two pulsed tones (DLF); (2) discrimination of a steady tone from an up-frequency glide with the same mean frequency (UP-DL); (3) discrimination of a steady tone from a down-frequency glide with the same mean frequency (DOWN-DL); (4) discrimination of an up-frequency glide from a down-frequency glide with the same mean frequency (UPVSDOWN-DL). All glides were linear in Hz/s. Experiment 1 measured thresholds as a function of center frequency over the range 500–8000 Hz in 1-oct steps, for a fixed duration of 500 ms. The variation of glide thresholds with frequency was similar to that for the DLFs. Experiment 2 measured thresholds as a function of duration over the range 50–750 ms, for a fixed center frequency of 2000 Hz. The DLFs decreased monotonically with increasing duration. The glide thresholds decreased as duration increased from 50 to 100 ms, but did not change markedly beyond that. In both experiments, DLFs and UPVSDOWN-DLs were similar. UP-DLs and DOWN-DLs were typically two to three times greater than DLFs and UPVSDOWN-DLs, and UP-DLs tended to be larger than DOWN-DLs at center frequencies up to 2000 Hz. Experiments 3 and 4 measured thresholds for detecting glides in frequency when fixed glides in level were superimposed on all stimuli. The glides in level had only a small effect on the frequency-glide thresholds. The results are discussed in terms of their compatibility with Zwicker’s [Acustica 6, 365–381 (1956)] excitation-pattern model of modulation detection and in terms of their implications for sampling theories of glide detection.

Complex sound discrimination: Predictions of the EWAIF model

The envelope‐weighted average of instantaneous frequency (EWAIF) model of auditory perception has been successfully applied to simple, two‐component signals, simultaneously amplitude‐ and frequency‐modulated tones, and even to the complex signals used in the early “profile analysis” work. The EWAIF model predicts performance from a calculation of the envelope‐weighted average of the instantaneous frequency fluctuations inherent in almost every complex periodic sound. When the model is in error, it generally predicts better performance than our listeners achieve. A revised version of the EWAIF model incorporates a temporal processing window. The revised model was tested by comparing its predictions with listeners' performance in a frequency‐change discrimination task. The experiment requires the listener to distinguish a smooth frequency glide from a discrete, multistep transition over the same trajectory. The listeners' ability to distinguish the glide from the multistep transition, in a 2IFC task, decreased to chance as the number of steps increased. The EWAIF model performance follows that of the listeners, given the appropriate choice of temporal window parameters. [Work supported by AFOSR and NIH.]

Dispersion of waves in piano strings

A recent survey of piano acoustics literature revealed an apparent lack of attention to various aspects of dispersion in piano strings, apart from some information on its effect on inharmonicity of piano tone partials. In this article, it will be shown how group velocity of transverse waves in piano strings can be measured as a function of frequency with the aid of a short-time spectral analysis method. Examples of group velocity measurements appear to be essentially in agreement with the theoretical predictions based on a model of a flexurally stiff string. In addition, the relationship between the group and phase velocity, as a function of frequency, is also illustrated, indicating correspondence between theoretical predictions and experimental results. A short-time spectral analysis of transverse string displacement as a function of time, monitored near the bridge, has revealed a quasiperiodic succession of frequency glides. This effect is due to dispersion and is particularly prominent in the lowest bass strings. However, the same type of analysis applied to the sound-pressure signal of the corresponding radiated sound yielded somewhat different results. While frequency glides similar to those found in the string displacement spectra were partly evident in the sound-pressure spectra, strong precursive components of longitudinal string vibration origin were found to dominate the higher frequency portion of the attack transient of the radiated sound.

Detecting dynamic frequency shifts of simple and complex tones

Thresholds for the detection of frequency modulation by a single cycle of a 12‐Hz sinusoid were obtained for carriers that were simple or complex tones. Two phases of modulator were used, producing percepts of either upward or downward frequency glides during the middle of the stimulus. To ensure that subjects based detection on frequency movement, both the phase of the modulating sinusoid and the starting frequencies were randomized from presentation to presentation. For pure‐tone carriers, thresholds (= minimum detectable frequency swing/carrier frequency) were constant for carrier frequencies of 800–2400 Hz and increased below 800 Hz. Thresholds were lower for modulations producing the percept of an upward glide than for “downward” glides. This difference was much greater at low carrier frequencies. When the carrier was a three‐component harmonic complex, thresholds were almost independent of its center frequency but increased with decreasing fundamental frequency. Thresholds for harmonic and for inharmonic complexes were compared. These results have implications for the recording of voice pitch (e.g., in signal‐processing hearing aids).

Audiological and vestibulo-oculomotor findings in workers exposed to solvents and jet fuel.

Three groups of subjects with long-term (5-41 years) occupational exposure to industrial solvents have been evaluated with extensive audiological and vestibular test batteries. Group A comprised 16 subjects with a confirmed diagnosis of psycho-organic syndrome (POS), while group B consisted of 7 subjects with suspected POS. Both groups had been exposed to mixtures of aliphatic and aromatic solvents. Eight subjects with long-term exposure to jet fuel constituted group C. In the audiological test battery, discrimination of interrupted speech and evoked cortical potentials in response to frequency glides were the tests that yielded significantly abnormal results. In the vestibular test battery, considerable pathology was seen in electronystagmography, and in addition, visual suppression test and saccade test indicated CNS disturbance. In general, when a test yielded pathological results, the incidence of pathology was highest in group A and lowest in group C.

The discriminability of transition duration: Effects of the amplitude and duration of a following steady state

When listeners are asked to discriminate changes in the duration of a brief frequency glide which is followed immediately by a longer, steady‐state, pure tone, they show better performance when the duration of the glide is in the range 40–60 ms than when it is longer or shorter. This effect could result from the interaction of two psychoacoustic functions—the first, a discriminability function, which follows Weber's Law and the second, a masking function. Three experiments were undertaken to test this explanation. In experiment 1, the glide began at 400 Hz, ended at 700 Hz, and was immediately followed by a steady‐state pure tone of equivalent amplitude, also at 700 Hz. Overall stimulus duration was fixed at 200 ms, while glide and steady‐state values varied in 10‐ms intervals, as follows: 10,190; 20,180; 30,170; 40,160; 50,150; 60,140; 70,130; 80,120; 90,110. This experiment replicated previous results [D. Jamieson and E. Slawinska, J. Acoust. Soc. Am. Suppl. 1 74, S67 (1983)] in finding the discriminability (− ln η) was a maximum and bias (ln β) a minimum, when glide duration was in the interval 40–60 ms. Stimuli for experiment 2 were identical to those for experiment 1, except that the amplitude of the steady‐state pure tone was reduced 10 dB. Stimuli for experiment 3 were identical to those for experiment 1 except that the overall duration of the stimulus was fixed at 10 ms. Thus, glide and steady‐state pure‐tone components had the following durations in ms: 10,90; 20,80; 30,70; 40,60; 50,50; 60,40; 70,30; 80,20; 90,110. Results obtained in experiments 2 and 3 are consistent with the hypothesis that the effect seen in experiment 1 is due, in part, to the masking of the glide stimulus by the following steady‐state pure tone. In both cases, the point of maximum discrimination (− ln η) shifted to shorter glide durations (20–40 ms) than was found in experiment 1.