Search

Using only three measures of the waveform, the zero‐crossing rate, the logarithm of the root‐mean‐square (rms) energy, and the derivative of the log rms energy with respect to time [termed rate of rise (ROR)], voiceless plosives (including affricates) can be distinguished from voiceless fricatives in word‐initial, medial, and final positions. Peaks in the ROR contour are considered for significance to the plosive/fricative distinction by examining the log rms energy and zero‐crossing rate. Then, the magnitude of the first significant peak in the ROR contour is used as the primary classifier. The algorithm was tested on 1364 tokens (720 word‐initial tokens produced by four female and four male speakers; 360 word‐medial tokens produced by two males and two females; 320 word‐final tokens produced by two males and two females). Data from two male and two female speakers (360 word‐initial tokens) were used as a training set, and the remaining data were used as a test set. The overall rate of correct classification was 96.8%. Implications of this result are discussed.

The acoustic nonlinearity parameter B/A has been calculated for four linear alkanes over the temperature range from 273.15 to 473.15 K and the pressure range from 1 atm to 1000 atm by using the Lee–Kesler equation of state. The calculated results are in agreement with the experimental data in the literature. It is found that under low pressure (1 atm and 20 atm), B/A increases linearly with temperature. Under high pressure, the value of B/A increases with temperature first, however, at a slower rate of rise. After a broad maximum, the ratio B/A decreases slowly as the temperature increases. At the fixed temperature, the B/A value decreases with the number of carbon chain of alkanes. It is shown that the B/A decreases monotonically with pressure over the whole temperature and pressure ranges.

We report on temporal cues associated with contrastive palatal affricates in Serbian in minimal pairs such as veche evening (apical) and vec’e bigger (laminal). These sounds have been difficult to distinguish using articulatory features. We investigated risetime, rate of rise (ROR), closure duration, release duration and segmental duration in four vowel contexts and two word positions. The apical affricate displays shorter risetime, ROR and closure duration, with longer release duration, while the laminal affricate displays longer risetime, ROR and closure duration, with shorter release duration. While risetime differentiates fricatives from affricates (Howell and Rosen, 1983), and ROR differentiates stops from fricatives (Weigelt et al., 1990), this is the first study to demonstrate risetime and ROR distinguishing two affricates. Rate of rise distinguishes these two affricates from the palatal fricative and alveolar stop allowing for the maintenance of contrast over the three‐way distinction. Risetime co...

It was hypothesized that if microbubbles are administered as a continuous infusion, then destroying them in the myocardium and measuring their myocardial reappearance rate at steady state will provide a measure of mean myocardial microbubble velocity. Conversely, measuring their myocardial concentration will provide an assessment of microvascular cross‐sectional area (CSA). Myocardial blood flow (MBF) can then be calculated from the product of the two. Accordingly, experiments were performed in 21 dogs where MBF was altered mechanically or pharmacologically in the absence or presence of coronary stenosis. Microbubbles were delivered as a constant infusion, and two‐dimensional ultrasound was performed using different PI. The myocardial video intensity (VI) versus PI plots were fitted to an exponential function: y=A(1−e−βτ), where A is the plateau VI reflecting the microvascular CSA, and β reflects the rate of rise of VI and hence, microbubble velocity. Excellent correlations were found between flow and β, ...

In this study neural sensitivity to rate of rise was measured. The responses of 32 medial geniculate neurons to linear increments of a broadband noise from a defined baseline level were recorded extracellularly. Step rise time (3–96 ms), step amplitude (6–24 dB), and baseline level (0–73 dB SPL) were varied independently. A typical response to the step consisted of a transient burst of 1–8 spikes. Response probability increased as a function of the step’s rate of rise (amplitude/rise time), such that increases in rise time could be traded for proportional increases in step amplitude to obtain a criterion response. Increases in baseline level decreased the slope of the probability versus rate of rise function and increased the threshold rate of rise. Response probability was independent of peak level (=baseline level+step amplitude). The results show that the response of a common type of medial geniculate neuron is determined by the rate at which stimulus amplitude increases from a defined background level, indicating a possible role for these neurons in the perceptual segregation of auditory events [A. S. Bregman, P. A. Ahad, and J. Kim, 2694–2703 (1994)].

The cochlear ac potentials in response to a stimulating tone of rapidly increasing intensity undergo a rapid reduction in amplitude after reaching a certain maximum. The record seen on the cathode ray screen is indistinguishable from that reported for middle ear muscle action, yet the response described here occurs in the absence of muscles or of the entire middle ear. This effect is described as a “clamp,” but it is found to be the result of overloading in the inner ear. Within the limits of these experiments it is shown that the rate of rise of a tone makes no difference in the form of the cochlear ac response. This finding, if carried to extremes, would suggest that the inner ear behaves the same in the presence of impact noise as it does in the presence of steady‐state or slowly rising tones.

An algorithm has been developed for the purpose of voiceless stop/fricative distinction in word‐initial position using three measures of the waveform: zero crossings, energy, and derivative of the energy over time (termed rate of rise, ROR). The ROR is used as the primary classifier while energy and zero crossings are used to discard spurious, irrelevant peaks. Peaks in ROR are considered in order of magnitude for relevance to the stop/fricative distinction. The resulting algorithm was treated on 720 CVC tokens (four male speakers, four female speakers, three stops and an affricate [P,T,K,CH], five fricatives [S,SH,F,TH,H], and ten vowel contexts [IY,IH,EH,AE,AA,AH,AO,UH,UW,ER]). Data from two male and two female speakers (360 tokens) were used as a training set, and the remaining data (360 tokens) were used as a test set. The overall success rate was 96.8%. [Work supported by AFOSR.]

Unit spikes were recorded from the VIIIth nerve branch innervating the amphibian papilla of a urethane‐anesthetized toad. Tonal stimuli were applied from an earphone through a coupling tube to the surface of the saccule. The tuning curve (TC) for fast rising tone bursts rose gradually toward both low and high frequencies (HF's) from the characteristic frequency (CF) of the unit. HF continuous tones had the effect of peripheral inhibition on the response to any test tones. As the rate of rise of tone bursts was decreased, the TC rose progressively at comparatively higher frequencies, while little change was seen near the CF and under. When the slow rising (90% rise per 20 ms and slower) test tone bursts were used, the rise of TC reached a saturation point, where HF cutoff of the TC was very steep. It seems that the sharp HF cutoff arises, because the thresholds are reached after the peripheral inhibition developed during the slow rising phase of the HF test tones themselves.