摘要:

Research concerning man's ability to speak and hear within ambient conditions of many times normal atmospheric pressure poses special difficulties in calibrating psychoacoustic equipment. In this experiment, a system for obtaining reciprocity calibrations within hyperbaric environments was employed. By means of the reciprocity system, microphone sensitivities were calculated from measures taken with a piezoelectric microphone. Frequencies were 0.1 6 kHz, and conditions of ambient pressure within a helium‐filled hyperbaric chamber were equivalent to pressures encountered from surface to 600‐ft submersion in sea water (1–19 ata absolute). The results indicate that, at frequencies below 1 kHz, increases in ambient pressure from surface to depths of 600 ft were accompanied by consistent decreases in microphone sensitivity. At 2 kHz, sensitivity did not continue to decrease for pressures equivalent to submergence from 300 to 600 ft. Resultant corrections to the calibrated microphone for increases in ambient pressure permit accurate measures to be made of stimuli for psychoacoustic experiments, and precise acoustic analyses can be made of speech produced under hyperbaric conditions.

ISSN:0001-4966    

DOI:10.1121/1.1975876

出版商:Acoustical Society of America    

年代:1971

数据来源: AIP

摘要:

A multichannel recording and analysis system is now in use for beamforming, signal processing, and noise‐field studies on large acoustic arrays. Each of the 252 data channels records and reproduces signals from dc to 20 kHz with a reproduced‐signal to system‐noise level difference of more than 40 dB and a maximum timing error between channels of less than 3 μsec. Timing signals giving 1‐μsec time resolution and digital signals with housekeeping and manually set information are also recorded. A real‐time playback system reproduces the original time and voltage relationships for all 252 channels simultaneously. A second playback system processes 40 channels at a time through a 40‐channel A‐to‐D converter and a 40‐channel 1024‐point FFT processor. This system produces digital tapes containing the signal samples the auto‐spectral‐density data and the cross‐spectral data for the 40 input channels. The data on these tapes is either plotted, further processed by a computer, or combined with other output tapes to produce the cross spectra between any pairs of the 252 input channels. Examples of current rises of this system are given.

ISSN:0001-4966    

DOI:10.1121/1.1975880

出版商:Acoustical Society of America    

年代:1971

数据来源: AIP

摘要:

Cinefluoroscopic techniques were used to observe the variations in the movement of the diaphragm during the alteration of the vibrato rate in singing. Four singers were selected for their ability to control the vibrato rate by means of abdominal pressure. All movement of the diaphragm was measured by caliper from its position at maximum inspiration before the onset of phonation. The vocal tasks were (1) nine pulsated tones sung legato on 234 Hz (male) and 468 Hz (female), (2) the same frequencies sung with normal vibrato rate, and (3) the vibrato controlled (slow, fast, slow). Observation shows the diaphragm dislodgement to vary with each subject. The average dislodgment was 4–6 mm in vocal task 1, 2–4 mm in vocal task 2, and 0–8 mm in task 3. Results suggest that the vocal vibrato is directly related to the senation of support—that it is produced by minute alterations of body pressures which are reflected in the undulation of the breath column. This variation in subglottal air pressure causes the pitch to vary as singer keeps the mass length tension and elasticity of the vocal folds constant through laryngeal controls.

ISSN:0001-4966    

DOI:10.1121/1.1975890

出版商:Acoustical Society of America    

年代:1971

数据来源: AIP

摘要:

This paper is based on an interpretive survey of the literature. Speaker recognition is defined as any decision‐making process that uses speaker‐dependent features of the speech signal. The origin and nature of these features are discussed, and three general methods of speaker recognition (aural, visual, and automatic) are described. Each method is illustrated by several experimental studies. Speaker recognition by listening appears to be the most accurate and reliable method at the present time. Speaker recognition by visual comparison of spectrograms is coming into use in criminology, but the validity of this method is still in question. Speaker recognition by machine is limited by various design shortcomings, both of concept and implementation.

ISSN:0001-4966    

DOI:10.1121/1.1975898

出版商:Acoustical Society of America    

年代:1971

数据来源: AIP

摘要:

Ambient‐noise correlation measurements have been made in a deep ocean area with a vertical hydrophone array suspended at a depth of 100 ft. Spatial correlation coefficients were obtained for vertical hydrophone spacings of from 1 to 12.5 ft over the frequency range 100–2000 Hz and for sea states 2 and 4. The measured coefficients are compared to those obtained from theoretical models of ambient noise. The results indicate that below 200 Hz the noise field is mostly horizontal and is attributed to distant sources. Above this frequency, the field can be interpreted as arising from a uniform distribution of directional noise sources at the sea surface.

ISSN:0001-4966    

DOI:10.1121/1.1975910

出版商:Acoustical Society of America    

年代:1971

数据来源: AIP

摘要:

An adaptive array beamforming technique based on an eigenvector interpretation of the conventional (i.e., time delay and sum) single‐frequency array directional response pattern is presented. In particular, the directional response pattern of aK‐element antenna array can be expressed in terms of the spectral density matrix of the spatially received data. This directional response pattern can, in turn, be expressed as a sum ofKterms, each involving a single eigenvector of the spatial spectral density matrix where each term in the sum is weighted by the corresponding eigenvalue. The array beamforming technique presented exploits the fact that positional information regarding the spatially coherent, source‐like components of the array's environment concentrates in the larger spatial eigenvalues and related eigenvectors. This “clustering” effect allows the conventional directional response pattern to be approximated by a truncation of the eigenvalue‐weighted sum. A single‐term approximation to the conventional bearing response pattern is the subject of this paper. An adaptive array processing technique is proposed which recursively obtains the single‐term truncation approximation to the bearing response pattern. The potential of this array processing technique for automatic multiple source location and tracking is examined.

ISSN:0001-4966    

DOI:10.1121/1.1975914

出版商:Acoustical Society of America    

年代:1971

数据来源: AIP