摘要:

Professional and amateur bass/baritone singers with experience both as soloists and choir singers heard their own voice mixed with a piano accompaniment while they performed a specially composed solo selection. Then the piano accompaniment was replaced by a recording of a choir made in the bass section during rehearsal and the subjects were asked to sing the bass part. In each case the subject's voices were recorded, and comparisons were made between similar vowels sung at the same pitch and sound pressure level. The output spectra are noticeably greater in the region of the singer's formant [J. Sundberg, J. Acoust. Soc. Am.55,838 (1977)] in the solo singing mode. However, differences in the source spectra, obtained by inverse filtering of the formants, are not consistent among the singer subjects.

ISSN:0001-4966    

DOI:10.1121/1.2021852

出版商:Acoustical Society of America    

年代:1984

数据来源: AIP

摘要:

In the course of measuring the real‐ear attenuation at threshold (REAT) of experimenter‐inserted E‐A‐R™ foam earplugs on 100 subjects, it was noted that attenuation was significantly correlated with hearing level (for normal listeners, HTL<20 dB) at test frequencies from 2–8 kHz. The relationship was most robust at 6 and 8 kHz. For hearing levels greater than 20 dB, attenuation appeared independent of hearing level. A hypothesis was developed to explain the relationship for the normal listeners, based upon the fact that the high‐frequency attenuation of the plug was nearly bone‐conduction‐limited. The hypothesis suggested that the attenuation of a foam earplug modified to give substantially lower protection would not exhibit the same relationship. Data for such a plug were collected for 50 subjects, and indeed demonstrated no correlation between attenuation and hearing level. Implications of the experiment with regard to (a) hearing level requirements for hearing protector attenuation test subjects, (b) utilization of hearing impaired listeners to measure REAT at suprathreshold (with respect to normal listeners) sound levels, and (c) constancy of hearing protector attenuation as a function of sound level, are discussed.

ISSN:0001-4966    

DOI:10.1121/1.2021856

出版商:Acoustical Society of America    

年代:1984

数据来源: AIP

摘要:

The audiometric test data of a population of 265 employees working in one facility were studied to determine group mean hearing threshold changes. The data were available covering periods of seven to 13 years. During this time period, the population was exposed to a workplace noise environment which resulted in time‐weighted average (OSHA) exposures of 86 to 103 dBA. Personal hearing protection (insert type protectors) was used by mill employees throughout the time period. Essentially no change in the mean threshold hearing level was found by the analysis. Certain other hearing conservation program effectiveness indicators proposed by others were also used to evaluate this population. The results of applying these indicators generally follow the trends shown by others for effective hearing conservation programs.

ISSN:0001-4966    

DOI:10.1121/1.2021860

出版商:Acoustical Society of America    

年代:1984

数据来源: AIP

摘要:

The wave vector filter (WVF) concept offers a promising new diagnostic tool for analyzing both structure‐borne and fluidborne noise fields. For instance, as with intensity measurements, WVF measurements allow one to distinguish between radiated and nearfield, or incompressible, noise components. In this paper some of the issues that arise in implementing the WVF concept, e.g., spatial nonstationarity, are analyzed in the context of a harmonically line driven rib stiffened, submerged, infinite plate. The plate is assumed to be thin and the ribs are modeled as locally reacting. First the response of this spatially infinite system is determined in both the spatial and wavenumber domains for a few representative geometries. And this is interpreted as the output from a “perfect” WVF. By response is meant plating strain or acceleration or near‐ or farfield pressure. Next these fields are analytically sampled over a finite domain thus introducing the issues of spatial nonstationarity and other finite filter effects. Results are compared for few and many ribs and various structural loss factors. The effects of structural resonances are discussed.

ISSN:0001-4966    

DOI:10.1121/1.2021863

出版商:Acoustical Society of America    

年代:1984

数据来源: AIP

摘要:

The excess air pressure from a nuclear explosion may be of the order of 28 000 kg per square meter (40 pounds per square inch). This may be compared to normal roof loading, for instance 200 kg per square meter (100 pounds per square foot). This ratio happens to be around 140, but it could be much higher closer to a blast. The Department of Defense has sponsored graduate‐level courses on “hardened” construction, in which the author participated. Advice was given against such construction as glass‐front office buildings. An alternate construction has been envisioned by the writer, in which the excess pressure on the building would be reduced by pressure‐relief ports (a patent on this has expired). It is hoped that a combination of hardened frame and yieldable front would provide an economical compromise for semi‐hardened construction. This might have application to tornado resistance as well as nuclear attack.

ISSN:0001-4966    

DOI:10.1121/1.2021870

出版商:Acoustical Society of America    

年代:1984

数据来源: AIP

摘要:

This talk will describe a network‐based system for speaker‐independent, isolated‐digit (one‐nine, oh,andzero) recognition and will discuss the results of an extensive series of system tuning and evaluation experiments. The digits are modeled by pronunciation networks whose ares represent classes of acoustic‐phonetic segments. Each are is associated with amatcherfor rating an input speech interval as an example of the corresponding segment class. The matchers are based on vector quantization of LPC spectra. Recognition involves finding minimum quantization distortion paths through the networks by dynamic programming. The system has been tested using nearly 6000 tokens of speech by 250 talkers, including a subset of a large database developed by Texas Instruments [G. Leonard, Proc. 1984 IEEE ICASSP]. The best recognizer configurations achieved accuracies of 97–99%. Performance over 21 geographically defined talker groups included in the TI database will be discussed.

ISSN:0001-4966    

DOI:10.1121/1.2021873

出版商:Acoustical Society of America    

年代:1984

数据来源: AIP

摘要:

Recent research (Huttenlocher and Zue, ICASSP 84) indicates that segmental information for isolated words provides strong constraints for lexical access. Furthermore, the results suggest that the lexical constraints provided by segments around stressed syllables are stronger than that around unstressed syllables. The present study focuses on two related issues. First, we investigate the amount of lexical constraint provided by stress information alone. Second, we implement a system that derives the stress information from the acoustic signal. In order to determine the lexical constraints provided by stress information, the polysyllabic words in the Merriam pocket dictionary are mapped into their corresponding stress patterns. The results indicate that, from stress information alone, the largest class size constitutes 28% of the lexicon. An overall expected class size of 15% illustrates the constraining power of the stress information. In order to exploit these findings, we develop a system that determines the stress pattern of isolated words and performs subsequent lexical access. The system initially segments the speech signal into broad phonetic classes. From these segments, syllable nuclei are determined. Next, known acoustic correlates of stress, such as duration, energy, and fundamental frequency, are extracted for each syllable. The stress pattern is established through a relative comparison of the syllable feature vectors. Finally, lexical access based on the derived stress pattern provides a list of word candidates. Phonological rules are incorporated to account for variations in the number of syllables from the lexical base forms. The system is evaluated on a database of 1500 isolated words, spoken by eight speakers, with varying degrees of difficulty for syllabification. Preliminary evaluation suggests that the majority of error can be attributed to initial segmentation and syllabification. Less than 5% is due to the stress algorithm. [Work supported by the Office of Naval Research under contract N00014‐82‐K‐0727 and by the System Development Foundation.]

ISSN:0001-4966    

DOI:10.1121/1.2021878

出版商:Acoustical Society of America    

年代:1984

数据来源: AIP

摘要:

Four scala tympani implant subjects, two with multiple‐electrode prostheses and two with single‐electrode implants, were tested with a single‐channel wearable sound processor after 0–8 months' use. A comparison between pre‐ and post‐implant scores on the Minimal Auditory Capabilities (MAC) Test showed little difference, except that all subjects improved on the postoperative lipreading test when electrical stimulation was added (“aided” condition). Improvement was noted in aided over unaided performance on videotaped tests, including a medial consonant test, two‐digit numerals test, and an open set sentence test. Subjects showed a two‐ to fourfold increase in speech tracking rates in the aided over the unaided condition. These baseline observations indicate that subjects quickly begin to experience improvement in face‐to‐face communication skills even during the first months of using a single‐channel cochlear prosthesis. [Work supported by NIH.]

ISSN:0001-4966    

DOI:10.1121/1.2021884

出版商:Acoustical Society of America    

年代:1984

数据来源: AIP

摘要:

Initial assessments will be presented describing the performance characteristics, effectiveness of speech feature transmission, and user acceptance of a wearable, 16‐channel electrotactile aid. Normal‐hearing and deaf subjects, wearing the device, received a battery of subtests evaluating the discriminability of male‐female voices, number of syllables, syllable stress,F1 movement,F2 movement, voicing, manner, and place of articulation. Factors affecting the design of experiments to test the efficacy of the device will be discussed, along with strategies for developing a comprehensive training curriculum for auditory awareness, speech reception in conjunction with lipreading, and speech production. [Work supported by Technology for the Handicapped, Inc., Tacticon Corporation, and SBIR/NIH.]

ISSN:0001-4966    

DOI:10.1121/1.2021887

出版商:Acoustical Society of America    

年代:1984

数据来源: AIP

摘要:

An alternative to the classic reverberation‐time measurements for determination of sound absorption coefficients (ASTM C423) is the steady‐state method which measures the energy average sound pressure level of a reference source (of known sound power) with and without a sample in a reverberation room. In addition, sound power measurements in reverberant rooms require the use of a reference source to determine the room corrections. To successfully use these techniques, sound power must be known with great accuracy. With the advent of sophisticated sound intensity instrumentation, it becomes very easy to determine sound power in almost any type of space. In this paper, the authors report the use of sound intensity to test several reference sound sources in various environments, including reverberant, hemi‐anechoic, and semi‐diffuse, to determine the changes in measured sound output. In addition, the reference sources' emissions are compared with manufacturers' specifications based on techniques similar to S12.5, and intensity techniques. The differences among results for the sources/spaces/techniques will be discussed.

ISSN:0001-4966    

DOI:10.1121/1.2021895

出版商:Acoustical Society of America    

年代:1984

数据来源: AIP