When listeners are detecting tones of a given frequency in noise, they operate with a narrowband attentional filter which is tuned to the frequency of the attended tone. This means that tones with frequencies which match that which is being attended to, will be detected, whereas tones with frequencies outside the filter will be detected at chance levels. In the current study, attentional filters were measured on two auditory dimensions. The first experiment employed a modified version of the methods of previous studies to measure attentional filters in the frequency dimension (Greenberg & Larkin, 1968). The mean results replicated those of previous experiments, revealing an attentional filter for frequency. In the second experiment, the methods of the first experiment were used to investigate whether there is an attentional filter in the fundamental frequency (f0) dimension. To test this, sentences with a given f0 were presented in noise. Infrequently, sentences were presented with f0s which the listeners were not attending to. The changes in f0 (∆f0) tested were ± 5 Hz and ±10 Hz, with respect to the expected 220 Hz f0. The effect of ∆f0 was investigated using sentence identification scores. Mean results indicated that listeners were best at identifying the sentences with an expected f0, and sentences with ∆f0s of 10 Hz from the expected f0. Sentences with ∆f0s of 5 Hz from the expected f0, on the other hand, were more poorly detected. This could be due to the presence of an attentional filter for f0, which has a narrow bandwidth, ranging between 10 and 20 Hz. An attentional filter for f0 may have different properties than those for attentional filters for frequency. The complex nature of sentence materials might change the way that attention is allocated across the f0 distribution, so that greater changes in f0 may be enough to switch the listeners’ attention to that f0. Overall, the results did not approximate the typical attentional filter shape which was found for frequency. Therefore, the results cannot be used to demonstrate the existence of attentional filters for f0. The presence of an attentional filter for f0, however, cannot be ruled out. Further research is needed, using a greater variety of ∆f0s to confirm and further investigate the presence, and properties of an attentional filter for f0.

Two experiments were conducted. The first was a control condition and used the probe signal method similar to Greenberg and Larkin (1968) to see if a filter-like attentional mechanism was acting when listeners were presented with pure tone stimuli in the presence of background noise. The second experiment also used the probe signal method of Greenberg and Larkin (1968) to investigate the extent to which listeners direct their attention to a particular fundamental frequency (f0) when detecting complex tones masked by noise. Twenty adult listeners ranging from 23 years to 54 years with a median age of 28 years participated in both experiments. Of the 20 listeners, 8 were male and 14 were female. Both experiments used a Two Interval Forced Choice (2IFC) procedure. There were two types of trials, the target signal trial and the probe signal trial. The target frequency was presented on 71% of trials, and the probe frequencies on the remaining 29%. The results of Experiment 1 were similar to those obtained in Greenberg and Larkin’s (1968) pioneering study. The 1000 Hz target tone was detected at a significantly higher proportion than probe signals differing in frequency (p < 0.05). Detection scores were observed to be higher when probe signals had a frequency close to the 1000 Hz signal compared to when they had a frequency positioned further from the 1000 Hz target tone. Experiment 2 using complex target tones with f0 of 115 Hz (part 1) and 220 Hz (part 2) revealed a similar pattern to Experiment 1. Listener’s detection scores decreased the further the f0 of probe tones were positioned from the f0 of the target tone, revealing the shape of a band-pass filter. This pattern is consistent with the presence of an auditory attentional filter in the f0 domain for complex tones

No description available.

This thesis introduces a novel method for accurate pitch detection and speech segmentation, named Multi-feature, Autocorrelation (ACR) and Wavelet Technique (MAWT). MAWT uses feature extraction, and ACR applied on Linear Predictive Coding (LPC) residuals, with a wavelet-based refinement step. MAWT opens the way for a unique approach to modeling: although speech is divided into segments, the success of voicing decisions is not crucial. Experiments demonstrate the superiority of MAWT in pitch period detection accuracy over existing methods, and illustrate its advantages for speech segmentation. These advantages are more pronounced for gain-varying and transitional speech, and under noisy conditions.

The problem of determining the eigenvalues of a vibrational system having multiple lumped attachments has been investigated extensively. However, most of the research conducted in this field focuses on determining the natural frequencies of the combined system assuming that the characteristics of the combined vibrational system are known (forward problem). A problem of great interest from the point of view of engineering design is the ability to impose certain frequencies on the vibrational system or to avoid certain frequencies by modifying the characteristics of the vibrational system (inverse problem). In this thesis, the effects of adding lumped masses to an Euler-Bernoulli beam on its frequencies and their corresponding mode shapes are investigated for simply-supported as well as fixed-free boundary conditions. This investigation paves the way for proposing a method to impose two frequencies on a system consisting of a beam and a lumped mass by determining the magnitude of the mass as well as its position along the beam.

The problem of determining the eigenvalues of a vibrational system having multiple lumped attachments has been investigated extensively. However, most of the research conducted in this field focuses on determining the natural frequencies of the combined system assuming that the characteristics of the combined vibrational system are known (forward problem). A problem of great interest from the point of view of engineering design is the ability to impose certain frequencies on the vibrational system or to avoid certain frequencies by modifying the characteristics of the vibrational system (inverse problem). In this thesis, the effects of adding lumped masses to an Euler-Bernoulli beam on its frequencies and their corresponding mode shapes are investigated for simply-supported as well as fixed-free boundary conditions. This investigation paves the way for proposing a method to impose two frequencies on a system consisting of a beam and a lumped mass by determining the magnitude of the mass as well as its position along the beam.

The purpose of this study was to examine the effect of fundamental frequency (Fo) on stop consonant voice onset time (VOT). VOT was measured from the recordings of 56 young men reading phrases containing all 6 English voiced and voiceless stops in word-initial position across high-, medium-, and low-Fo levels. Separate analyses of variance for the voiced and voiceless stops revealed no significant main effect for Fo for the voiced stops but a significant Fo effect for the voiceless stops. Across the voiceless stops, productions at high Fos displayed significantly shorter VOTs than productions at low or mid F os. The findings indicated that researchers must take into account the Fo level at which voiceless stop VOT is measured.

Key aspects of prosody have been studied in adults for a number of years; however, less attention has been paid to the acoustic patterns of prosody in children. Thus, the purpose of the present study was to evaluate how a group of 20 pre-adolescent children use prosody to mark contrastive stress compared to a control group of adult speakers. It was of interest to investigate whether the children's use of prosody differed between boys and girls or the part of speech being emphasized. The prosodic patterns of contrastive stress were evaluated in terms of duration, fundamental frequency, and intensity change relative to a baseline production of the same sentence. In addition, a perceptual experiment was conducted to determine if listeners could reliably identify the gender of the child speakers when listening to sentence length utterances. Statistical analysis indicated that there were some differences in the duration and fundamental frequency change as a function of speaker age and the part of speech being emphasized, with relatively minor differences between genders. However it remains unclear if the acoustic differences found in this study were substantial enough to cause a salient perceptual difference. Although previous studies have identified increases in frequency, intensity, and duration as cues of contrastive stress, the present findings revealed patterns that did not consistently conform to these expectations. Limitations in the task design, individual speaker characteristics, and also the type of acoustic measure used may have contributed to these results.

Quantitative measures of laryngeal muscle tension are needed to improve assessment and track clinical progress. Although relative fundamental frequency (RFF) shows promise as an acoustic estimate of laryngeal muscle tension, it is not yet transferable to the clinic. The purpose of this work was to refine algorithmic estimation of RFF, as well as to enhance the knowledge surrounding the physiological underpinnings of RFF. The first study used a large database of voice samples collected from 227 speakers with voice disorders and 256 typical speakers to evaluate the effects of fundamental frequency estimation techniques and voice sample characteristics on algorithmic RFF estimation. By refining fundamental frequency estimation using the Auditory Sawtooth Waveform Inspired Pitch Estimator—Prime (Auditory-SWIPE′) algorithm and accounting for sample characteristics via the acoustic measure, pitch strength, algorithmic errors related to the accuracy and precision of RFF were reduced by 88.4% and 17.3%, respectively. The second study sought to characterize the physiological factors influencing acoustic outputs of RFF estimation. A group of 53 speakers with voice disorders and 69 typical speakers each produced the utterance, /ifi/, while simultaneous recordings were collected using a microphone and flexible nasendoscope. Acoustic features calculated via the microphone signal were examined in reference to the physiological initiation and termination of vocal fold vibration. The features that corresponded with these transitions were then implemented into the RFF algorithm, leading to significant improvements in the precision of the RFF algorithm to reflect the underlying physiological mechanisms for voicing offsets (p < .001, V = .60) and onsets (p < .001, V = .54) when compared to manual RFF estimation. The third study further elucidated the physiological underpinnings of RFF by examining the contribution of vocal fold abduction to RFF during intervocalic voicing offsets. Vocal fold abductory patterns were compared to RFF values in a subset of speakers from the second study, comprising young adults, older adults, and older adults with Parkinson’s disease. Abductory patterns were not significantly different among the three groups; however, vocal fold abduction was observed to play a significant role in measures of RFF at voicing offset. By improving algorithmic estimation and elucidating aspects of the underlying physiology affecting RFF, this work adds to the utility of RFF for use in conjunction with current clinical techniques to assess laryngeal muscle tension. / 2021-09-29T00:00:00Z

No description available.