摘要:

Hearing-impaired and normal-hearing listeners labeled synthetic consonant-vowel stimuli (/bɑ/, /dɑ/, /gɑ/, /be/, /de/, /ge/) presented at moderate and high signal levels. First formant(F1)regions were synthesized at normal and at attenuated levels to test whetherF1attenuation might reduce upward spread of masking, making information contained in higher formant regions more available. Performance was tested in quiet and in broadband noise sufficient to mask initial release bursts. Although complete removal ofF1consistently reduced performance,F1attenuation of up to 18 dB led to increased labeling accuracy, particularly in the /ɑ/ vowel context. Benefit associated withF1attenuation was more consistently seen for hearing-impaired than for normal-hearing listeners and, in particular, for listeners with steep increases in audiometric thresholds between the first and second formant regions of the test stimuli. The availability of initial bursts as a source of place cues during testing in quiet did not reduce the benefit associated withF1attenuation.

ISSN:0001-4966    

DOI:10.1121/1.419303

出版商:Acoustical Society of America    

年代:1997

数据来源: AIP

摘要:

A targeting-and-extracting procedure of speech enhancement for hearing aids in the presence of background noise, especially competing speech, is proposed. The procedure is composed of two steps: targeting by a fixed (or deterministic) beamforming array, followed by a post-targeting extracting step. Emphasis is placed on the extracting step, which performs noise cancellation based on the acoustic difference between the desired speech and interfering speech. Either comb filtering or attenuation is applied to the signal in accordance with the current voiced/unvoiced/silence state of the desired signal. The comb filter design is based on the fundamental pitch frequency of the desired speech. Algorithms for deciding the voiced/unvoiced/silence state and determining the fundamental frequency are developed. The performance of the system is evaluated through computer simulation. The simulation results indicate significant noise cancellation and intelligibility improvement.

ISSN:0001-4966    

DOI:10.1121/1.418518

出版商:Acoustical Society of America    

年代:1997

数据来源: AIP

摘要:

Utterance verificationrepresents a key technology in the design of a user-friendly speech recognition system. One essential element when designing such a system is the ability to maintain a uniform performance over a wide range of acoustic conditions. An acoustic mismatch between training and testing conditions often results in an undesirable performance degradation. This paper addresses the issue ofrobustnessin utterance verification of a speech recognition system. Two techniques, namely signal bias removal (SBR) and on-line adaptation, are studied. The SBR algorithm is used to deal with global mismatch conditions caused by handset and channel differences. The on-line adaptation algorithm is used to adjust verification threshold at runtime for achieving a desirable trade-off between false rejection and false alarm in new test conditions. Various on-line adaptation schemes are investigated. We show that both supervised or unsupervised adaptation can effectively adjust the verification threshold to achieve a desirable performance trade-off irrespective of the initial setting of the threshold. We report on connected digit recognition/verification results formatchedandmismatchedtraining and testing conditions. At a 5% digit string rejection rate, the proposed robust utterance verification system gives a reduction in string error rate between 32% and 35% over the conventional system, while still correctly rejects over 99.9% of nonvocabulary utterances.

ISSN:0001-4966    

DOI:10.1121/1.418519

出版商:Acoustical Society of America    

年代:1997

数据来源: AIP

摘要:

The attack of most bowed notes shows an initial part before Helmholtz triggering occurs (thepre-Helmholtz transient), during which the stick-slip interaction promotes frequencies other than that of the string’s fundamental. Depending on the particular combination of bowing parameters, this state is characterized either by periods that areprolonged, or by a division of the period into two or more parts,multiple flyback. An onset with perfectly periodic motion (Helmholtz triggering) directly from the very start is also possible. A sample of violin tones representing these three classes of attacks, and with different duration of the pre-Helmholtz transient, has been collected by the use of a computer-controlled bowing machine. The tones were evaluated by 20 advanced string students and professionals in a listening test, judging the acceptance and quality of the attacks. The maximum acceptable duration of the pre-Helmholtz transient was estimated to 50 ms (⩽10 nominal periods, open G string, violin) for attacks with prolonged periods, and 90 ms (⩽18 periods) for multiple-flyback attacks. These values refer to a neutral start in a neutral context, such as when practicing a scale. A playing test, in which the performances of two professional violinists were analyzed, confirmed these results, and showed that the same limits apply to a larger group of bowing styles as well.

ISSN:0001-4966    

DOI:10.1121/1.418520

出版商:Acoustical Society of America    

年代:1997

数据来源: AIP

摘要:

Data on the internal acoustic pressure signals in an experimental recorderlike flue organ pipe are presented. A dimensionless representation appears to be a powerful basis for the analysis of these data. The dimensionless amplitude of the fundamental is, for a given geometry, a function of the Strouhal number only. For the first hydrodynamic mode of the jet, this amplitude is independent of the acoustic mode involved. The dimensionless amplitude of the second harmonic displays two different behaviors, depending on whether the jet is laminar or turbulent. This specific Strouhal number dependency implies a strong influence of the distanceWfrom the flue exit to the labium on the timbre of recorderlike instruments. In recorders the ratioW/h, wherehis the height of the flue exit, is adjusted by craftsmen to a value close to 4. This specific ratio ensures an optimal harmonic content to noise ratio in the produced sound. Increasing the distanceWyields a more powerful sound but at the expense of additional turbulence noise. For these large ratiosW/h, transitions between different hydrodynamic modes of the jet are observed at low blowing pressures. This phenomenon is avoided by the choice of a recorderlike geometry.

ISSN:0001-4966    

DOI:10.1121/1.418521

出版商:Acoustical Society of America    

年代:1997

数据来源: AIP

摘要:

A simple one-dimensional representation of recorderlike instruments, that can be used for sound synthesis by physical modeling of flutelike instruments, is presented. This model combines the effects on the sound production by the instrument of the jet oscillations, vortex shedding at the edge of the labium, and turbulence in the mouth of the instrument. The jet oscillation model used is a modification of the semi-empirical model by Fletcher [J. Acoust. Soc. Am.60, 926–936 (1976)]. The steady-state drive of the acoustical oscillations in the pipe by the jet motion is represented by a pressure jump in the mouth of the instrument. Vortex shedding at the edge of the labium during steady-state operation is taken into account by the use of a free-jet model. The combined effects of this nonlinearity and the jet-drive model enable one to correctly predict the steady-state amplitude of the fundamental. The turbulence noise source is represented by an additional pressure jump across the mouth of the instrument having an amplitude scaling with the square of the jet velocity. This simple model appears to correctly predict the noise level and its spectrum, both with and without oscillations of the jet. The transient response of the model is triggered by the initial volume injection into the mouth of the instrument and is dependent on the steepness of the driving pressure rise.

ISSN:0001-4966    

DOI:10.1121/1.419481

出版商:Acoustical Society of America    

年代:1997

数据来源: AIP

摘要:

The inertial cavitation and associated acoustic emission generated during electrohydraulic shock wave lithotripsy were studied using high-speed photography and acoustic pressure measurements. The dynamics of cavitation bubble clusters, induced in vitro by an experimental laboratory lithotripter, were recorded using a high-speed rotating drum camera at 20 000 frames/s. The acoustic emission, generated by the rapid initial expansion and subsequent violent collapse of the cavitation bubbles, was measured simultaneously using a 1-MHz focused hydrophone, The expansion duration of the cavitation bubble cluster was found to correlate closely with the time delay between the first two groups of pressure spikes in the acoustic emission signal. This correlation provides an essential physical basis to assess the inertial cavitation produced by a clinical Dornier HM-3 shock wave lithotripter, both in water and in renal parenchyma of a swine model. In the clinical output voltage range (16–24 kV), the expansion duration of the primary cavitation bubble cluster generated by the HM-3 lithotripter in water increases from 158 to 254 μs, whereas the corresponding values in renal parenchyma are much smaller and remain almost unchanged (from 71 to 72 μs). In contrast, subsequent oscillation of the bubble following its primary collapse is significantly prolonged (from 158–235 μs in water to 1364–1373 μs in renal parenchyma). These distinctive differences between lithotripsy-induced inertial cavitation in vitro and that in vivo are presumably due to the constraining effect of renal tissue on bubble expansion.

ISSN:0001-4966    

DOI:10.1121/1.418522

出版商:Acoustical Society of America    

年代:1997

数据来源: AIP

摘要:

The acoustic features of 216 baboongruntswere investigated through analysis of field-recorded calls produced by identified females in known contexts. Analyses addressed two distinct questions: whether the acoustic features of these tonal sounds could be characterized using a source-filter approach and whether the acoustic features of grunts varied by individual caller and social context. Converging evidence indicated that grunts were produced through a combination of periodic laryngeal vibration and a stable vocal tract filter. Their acoustic properties closely resembled those of prototypical human vowel sounds. In general, variation in the acoustic features of the grunts was more strongly related to caller identity than to the social contexts of calling. However, two acoustic parameters, second formant frequency and overall spectral tilt, did vary consistently depending on whether the caller was interacting with an infant or participating in a group move. Nonetheless, in accordance with the general view that identity cueing is a compelling function in animal communication, it can be concluded that much of the observed variability in grunt acoustics is likely to be related to this aspect of signaling. Further, cues related to vocal tract filtering appear particularly likely to play an important role in identifying individual calling animals.

ISSN:0001-4966    

DOI:10.1121/1.418523

出版商:Acoustical Society of America    

年代:1997

数据来源: AIP

摘要:

The information echolocating bats receive is a combination of the properties of the sound they emit and the sound they receive at the eardrum. Convolving the emission and the external ear transfer functions produces the full spectral information contained in the echolocation combination. Spatially dependent changes in the magnitude spectra of the emission, external ear transfer functions, and the echolocation combination ofEptesicus fuscuscould provide localization information to the bat. Principal component analysis was used to reduce the dimensionality of these complex spectral data sets. The first eight principal component weights were normalized, rotated, and used as the input to a backpropagation network model which examined the relative directionality of the emission, ear, and the echolocation combination. The model was able to localize more accurately when provided with the directional information of the echolocation combination compared to either the emission or ear information alone.

ISSN:0001-4966    

DOI:10.1121/1.418524

出版商:Acoustical Society of America    

年代:1997

数据来源: AIP

摘要:

The Acoustic Thermometry of Ocean Climate (ATOC) program of Scripps Institution of Oceanography and the Applied Physics Laboratory, University of Washington, will broadcast a low-frequency 75-Hz phase modulated acoustic signal over ocean basins in order to study ocean temperatures on a global scale and examine the effects of global warming. One of the major concerns is the possible effect of the ATOC signal on marine life, especially on dolphins and whales. In order to address this issue, the hearing sensitivity of a false killer whale (Pseudorca crassidens) and a Risso’s dolphin (Grampus griseus) to the ATOC sound was measured behaviorally. A staircase procedure with the signal levels being changed in 1-dB steps was used to measure the animals’ threshold to the actual ATOC coded signal. The results indicate that small odontocetes such as thePseudorcaandGrampusswimming directly above the ATOC source will not hear the signal unless they dive to a depth of approximately 400 m. A sound propagation analysis suggests that the sound-pressure level at ranges greater than 0.5 km will be less than 130 dB for depths down to about 500 m. Several species of baleen whales produce sounds much greater than 170–180 dB. With the ATOC source on the axis of the deep sound channel (greater than 800 m), the ATOC signal will probably have minimal physical and physiological effects on cetaceans.

ISSN:0001-4966    

DOI:10.1121/1.419304

出版商:Acoustical Society of America    

年代:1997

数据来源: AIP