Precision quartz oscillators have three main sources of noise contributing to frequency fluctuations: thermal noise in the oscillator, additive noise contributed by auiliary circuitry such as AGC, etc., and fluctuations in the quartz frequency itself as well as in the reactive elements associated with the crystal, leading to an I-1 type of power spectral density in frequency fluctuations. Masers are influenced by the first two types of noise, and probably also by the third. The influence of these sources of noise on frequency fluctuation vs. averaging time measurements is discussed. The f-'-spectral density leads to results that depend on the length of time over which the measurements are made. An analysis of the effects of finite observation time is given. The characteristics of both passive and active atomic standards using a servo-controlled oscillator are discussed. The choice of servo time constant influences the frequency fluctuations observed as a function of averaging time and should be chosen for best performance with a given quartz oscillator and atomic reference. The conventional methods of handling random signals, i.e., variances, autocorrelation, and spectral densities, are applied to the special case of frequency and phase fluctuations in oscillators, in order to obtain meaningful criteria for specifying oscillator frequency stability. The interrelations between these specifications are developed in the course of the paper.

No abstract

Physical measurements of the transfer function from a free−field sound source to a microphone in the subject’s ear canal indicate that there are two independent localization cues generated by the pinna. For sound sources in the vertical median plane, there is a systematic change in the frequency response as a function of elevation angle, and a disparity between the left−ear and right−ear responses which also changes systematically with elevation angle. Independent psychophysical measurements indicate that these pinna cues are detectable by subjects, and both are used by subjects in vertical localization tasks. Subject Classification: 65.62, 65.75.

A mathematical model based on statistical decision theory has been devised to represent the human auditory localization task. The known localization cues have been represented as Gaussian random variables, so that their interaction in a given experiment can be analyzed (and predicted) along the lines of classical detection/estimation theory. We have applied this technique to most of the horizontal and vertical localization experiments reported in the literature during the past ten years, encompassing over 200 subjects and 20000 trials. Using a nonlinear regression program we have been able to estimate the standard deviations of four of the auditory localization cues, allowing objective comparison of their relative accuracy. The resulting model provides a relatively good fit to the published results on 40 localization experiments. Subject Classification: [43]65.62, [43]65.35, [43]65.58. LIST OF SYMBOLS (C•S) •' m M N P,(m) decision criteria speaker sp•cing number of responses in error by m speakers span-dependent component of (• index on sound sources index on cues index on response error number of experiments number of trials in a given experiment number of sound sources probability of response in error by m speakers

Time-domain analysis of firing-rate data from over 200 fibers from the auditory nerve of cat has been used to estimate the formants of the synthetic-syllable stimuli. Distinct groups of fibers are identified based on intervals between peaks in the fiber firing rates. The large extent of some of these groups--over an octave in terms of characteristic frequency--and the lack of short intervals in the longer-interval groups suggest that the behavior of the nonlinear cochlear filters for these signals is effectively wideband with steep high-frequency cutoffs. The measured intervals within each group are very similar, and correspond to the period of the formant that dominates the group's response. These intervals are used to estimate the dynamic speech formants. The overall formant estimates are better than those of the previous spectral analyses of the neural data, and the details of lower-formant dynamics are tracked more precisely. The direct temporal representation of the formant in contrasted with the diffuse spectral representation, the dependence of spectral peaks on nonformant parameters, and the distortion of the spectrum by rectification. It is concluded that a time-domain analysis of the responses to complex stimuli can be an important addition to frequency-domain analysis for neural data, cochlear models, and machine processing of speech.