PubMed Health⌕ Search

Biomedical subjects

R J Niederjohn

Publications and source records attributed to R J Niederjohn.

2 recordsLinked to original sources

Speech amplitude statistics and the gamma density function as a theoretical approximation.

Knowledge of the speech amplitude density provides basic data concerning the statistics of the continuous speech waveform. The objective of this paper is to present data which characterize the longtime speech amplitude density utilizing a data base of sufficient size to describe the distribution. The data base consists of approximately 45 min of recorded speech for each of the 25 speakers. Two parameters of a theoretical gamma density function are estimated from the experimental data to provide the best curve fit between the experimental data and the gamma density function. Overall results presented in this paper substantiate three main conclusions: There is a significant difference between the speech amplitude density for male and female speakers; the speech amplitude density for positive speech amplitudes differs from that for negative speech amplitudes; and the gamma probability density function is a better approximation for full-wave rectified speech than it is when considering just positive or just negative amplitudes.

Adult↗

The effects of high-pass and of low-pass filtering upon the intelligibility of speech in white noise.

This paper studied with trained, normal-hearing, young adults the effects of high-pass and also low-pass filtering on the intelligibility of speech mixed at S's ear with additive white noise at 4 S/N. Before each filter--S/N combination Ss were encouraged to adjust overall gain so as to yield seemingly optimum speech intelligibility. Gain chosen was relatively independent of S/N, being primarily a function of speech level. Results provided information on the contribution to intelligibility of various frequency ranges of the speech spectrum. Data were examined in relation also to the articulation index (AI). The intelligibility scores were in good agreement with the AI and yielded an improved understanding of the intelligibility of speech in noise.

Adult↗