摘要:

The acoustic volume‐velocity wave produced by the human vocal cords frequently approximates a triangular shape. A theoretical examination of the frequency‐domain properties of the triangular wave is consequently of interest in speech analysis. The occurrence of spectral zeros in the glottal wave is of particular importance. When they fall proximate to formants, they can have appreciable influence upon the quality of voiced sounds. The present paper derives equations for the complex frequencies of the zeros. Arbitrary values of wave asymmetry (leg ratio) are treated. In a number of cases, for which the asymmetry can be expressed as a ratio of small whole numbers, exact solutions are obtained. In the less simple cases, solutions are approximated by using a digital computer. Detailed nomograms are presented for the complex‐frequency loci of the zeros as a function of waveshape.

ISSN:0001-4966    

DOI:10.1121/1.1936991

出版商:Acoustical Society of America    

年代:1962

数据来源: AIP

摘要:

Descriptions of vocal‐tract configuration commonly neglect the pharynx. Side‐view sketches of the “supraglottal cavities” traditionally offer little distinction of pharyngeal outline from vowel to vowel, or from sound to sound of whatever nature. Actually, the pharyngeal cavities assume shape and sizes highly characteristic and predictable for given sounds—especially vowel sounds. When viewed cineradiographically, in sagittal‐plane projection, the mesopharynx and hypopharynx appear to be maximally constricted during the [a] vowel, maximally extended during the [u]vowel, and similarly extended but with a decidedly different shape—during the [i] vowel, for example. Articulatory adjustments in anticipation of an associated vocalic nucleus are also predictable and characteristic in terms of the particular approaching vowel, especially, but exclusively, in the same “syllable.” Thus, the pharyngeal configurations for the respective consonantal portions of [la]and [lu], for example, are radically different. Static x‐ray mews are at best only partially descriptive, but cineradiography provides the dynamics prerequisite for reliable estimates, and correlated sound‐spectrographic analysis confirms the time‐axis particulars. Lateral x‐ray projections never truly define the vocal tract, and anteroposterior views reveal the consistent asymmetries of cross‐sectional areas and disclose postulated “midlines” as illusionary. Preadjustments of the epipharynx, too, are out of phase with classically described articulatory gestures. (A short cineradiographic film and individual extracted frames correlated with sound, sound spectrograms, over‐all intensity, oscillograms, and frame marking are offered in demonstration of statements and criticisms above.)

ISSN:0001-4966    

DOI:10.1121/1.1936992

出版商:Acoustical Society of America    

年代:1962

数据来源: AIP

摘要:

A computer program has been developed which attempts to identify plosive consonants. Input to the computer is sampled, quantized spectral data. Several measurements are made on the burst portion of the sound. The identification of final plosives is made from these. Initial plosives use the burst information combined with formant transition cues as described in our other paper (G10).

ISSN:0001-4966    

DOI:10.1121/1.1936999

出版商:Acoustical Society of America    

年代:1962

数据来源: AIP

摘要:

The method of measurement depends on cross‐correlating a local replica with pseudonoise signals received from a distant transmitter. Partial correlations with the energy coming over distinct paths, and, thus, occurring at different parts of the pseudonoise cycle, can be identified on a fascimile type of display chart. Furthermore, if both the replica and incoming signals are heterodyned with a stable oscillator before being cross‐correlated, the carrier in the noiselike transmission can be displayed clearly on the charts, thus allowing its phase stability to be measured over long periods of time. Sample charts obtained by cross‐correlating pseudonoise signals mixed with electrical noise are shown. The method has been applied to signals received simultaneously on 2 hydrophones. This allows the stability of single‐path transmission to be studied. Some examples of results obtained by this technique in long‐range acoustic transmission are displayed. [This work was supported by the Office of Naval Research.]

ISSN:0001-4966    

DOI:10.1121/1.1937007

出版商:Acoustical Society of America    

年代:1962

数据来源: AIP

摘要:

A program of low‐frequency underwater sound‐transmission tests using a fixed source at Eleuthera, Bahamas, and fixed hydrophones at Bermuda has been conducted for about a year. The signals received at Bermuda have a characteristic pattern of amplitude variation, with both short‐period and long‐period components. The short‐period component has a spectrum of variation rates which is strongest around 0.12 cps. The long‐period component varies at a rate of a few cycles per hour. The work reported here is concerned mainly with the short‐period instability. One‐hour samples of the signal received at Bermuda have been selected to include as wide a range of sea states and wind velocities as possible and to be distributed through the seasons. For each sample. the spectrum of short‐period variation rates has been computed and the distribution of received levels tabulated. The characteristics of the samples have been examined for correlation with sea conditions at the transmitting and receiving ends of the transmission path. [This work was supported by the Office of Naval Research.]

ISSN:0001-4966    

DOI:10.1121/1.1937009

出版商:Acoustical Society of America    

年代:1962

数据来源: AIP

摘要:

When an analog signal is sampled at a periodic rate, the spectrum of the resultant digital signal will be periodic in frequency with a period equal to the sampling rate. This is commonly referred to as the modulation of folding property of periodic sampling. The experiments discussed in this paper illustrate the use of the folding property of periodic sampling as an aid to efficient digital spectral analyses of narrow‐band sonar signals. In sampling an analog signal, the normal procedure is to select a sampling frequency equal to twice the highest frequency present. However, for a narrow‐band signal, this procedure can result in inefficient analysis. An alternate sampling procedure, which has been discussed in the literature, is to use the signal bandwidth as the criteria for selection of the sampling frequency. Using the signal bandwidth and the folding properties of sampling as the criteria for selection of the sampling rate results in a significant reduction, for narrow‐band signals, in sampling frequency, and, thus, analysis time. The experiments discussed illustrate the use of this technique for a narrow‐band signal and for the more complex spectrum introduced by infinitely clipping the input narrow‐band signal.

ISSN:0001-4966    

DOI:10.1121/1.1937016

出版商:Acoustical Society of America    

年代:1962

数据来源: AIP

摘要:

In‐flight experiments to measure turbulent boundary‐layer pressure field along the fuselage of a large transport airplane at 10 000‐, 20 000‐, and 30 000‐ft altitudes for flight speeds of from 0.4–0.67 Mach are reported. The data include power spectra and mean‐square‐pressure of the turbulent boundary layer, and power spectra of the boundary‐layer‐induced panel vibration. These data are compared with published information on the boundary‐layer pressure fluctuation of a rigid wall.

ISSN:0001-4966    

DOI:10.1121/1.1937020

出版商:Acoustical Society of America    

年代:1962

数据来源: AIP

摘要:

Phased sequences of spark‐schlieren photographs have revealed several interesting aspects of the periodic‐flow phenomena of the Hartmann whistle, supplementing quantitative acoustical and fluid‐flow measurements. The periodic debouching of the fluid occurs in the form of a powerful vortex ring, the normal shockwave near the mouth of the resonant cavity oscillating impulsively through a large amplitude. A second normal shockwave periodically traverses the high‐amplitude oscillatory “standing‐wave” motion between the external normal shockwave and the end of the cavity. As a result, a periodic, impulsively starting, high‐temperature air jet will be emitted from any small aperture at the end of the cavity. There is also evidence of feedback action of the sound field from the nozzle to cavity. It has been discovered that an unstable oscillatory condition also occurs over a limited range of subsonic jet speeds for certain locations of the cavity with respect to the nozzle; this range is greatly increased by the presence of a wake‐producing trip or of a local reflecting surface. [This work was supported by the Office of Naval Research, under contract.]

ISSN:0001-4966    

DOI:10.1121/1.1937023

出版商:Acoustical Society of America    

年代:1962

数据来源: AIP

摘要:

A normal‐mode solution of the wave equation can be written asφ=e1ωl ∫0∞ J0(Kτ) F (Z,K)dK, whereF(Z,K) is a solution of(d2F/dz2) + [ω2/c2−K2] F=0. The specificF(Z,K) can be determined for a deep‐source shallow receiver by using the WKB approximation with the usual point source and surface‐boundary conditions. This solution can then be written as an infinite series of integrals, each of which is identifiable either as a real or a complex ray. This solution is not applicable in the immediate vicinity of the ray‐turning (horizontal) points. The first integral (the upcoming ray) is evaluated by the saddle‐point method. Where real rays exist, the usual ray equations are obtained, except for a slight correction to the intensity equation. It can be shown that this solution can be continued around the ray‐turning point to give an evaluation of the sound in the shadow zone. This solution will yield results similar to Pekeris' for his simple shadow case, and is similar in some respects to the work of Keller and Seckler. It justifies the use of ray optics after a caustic. It also justifies caustic‐intensity calculations based on ray optics. More important, it shows that the signal form is unchanged, if it is propagated by real rays, and is modified if it is propagated as a complex ray.

ISSN:0001-4966    

DOI:10.1121/1.1937032

出版商:Acoustical Society of America    

年代:1962

数据来源: AIP

摘要:

Increasing use of higher‐frequency underwater acoustic devices has prompted thein situmeasurement of sea‐water attenuation and sea‐bottom backscattering of acoustic pulses in the frequency range 150 to 2100 kc. Although the measured attenuation coefficients usually agreed closely with the viscous‐relaxation theory of absorption, marked increases in attenuation were noticed on several occasions. The backscattered signal amplitude was found to depend on bottom type and frequency, as well as on grazing angle and pulse length. The normal “reflection” coefficient is closely inversely proportional to the frequency, but can be interpreted as a scattering coefficient, if the scattered level is assumed proportional to the beam cross section, which in these measurements was itself inversely proportional to the frequency. Theory indicates that the backscattering coefficient is a function of surface roughness of the bottom, rather than of particle size for the wavelengths used here. However, surface roughness is dependent on particle size.

ISSN:0001-4966    

DOI:10.1121/1.1937039

出版商:Acoustical Society of America    

年代:1962

数据来源: AIP