Loudness balances were collected from three subjects, employing 57 combinations of 13 frequencies extending from 20 to 5000 cps. Determinations were made from levels close to threshold to the highest levels tolerable by the subjects, with 5-dB intervals. Stimuli were applied by an earphone through individually moulded ear inserts, and were measured in decibels sound-pressure level a few millimeters from the eardrum. For each subject, a symmetric and transitive subset of the data was employed for constructing a set of matching functions and a set of equal-loudness contours. Matching functions generally were linear, or showed a double inflection (in log-log coordinates). Equal-loudness contours generally agree with previous determinations, but show a steeper rise towards low frequencies; below 300 cps, and at medium levels, the contours approach straight lines with a slope of -40 dB/decade. Obtained matching functions are compared to three sets of theoretical functions, favoring the matching functions derived from loudness theories developed by Zwicker and by Zwislocki. stood, the distributions of parameter values obtained for a larger number of individuals may serve as a basis for a standardization of the "normal" equal-loudness relations.Apart from philosophical implications, these two different approaches may lead to significant differences in the final results. Assume, for instance, that the matching function between two frequencies at various intensities for any given subject is $ shaped, but that the positions of the inflections (and, possibly, their directions) vary between subjects. If the raw data from such experiments are pooled over a large number of subjects, it is quite conceivable that the various inflections will largely cancel out, and the investigator will be led to the conclusion that such a matching function is linear. While such a conclusion may well serve a purpose for practical applications, it does not further the understanding of the basic mechanisms of hearing.In addition to concentrating on the individual equalloudness relations, the present study attempts to remedy some other apparent shortcomings of previous investigations. First, owing to various instrumentation problems, it is possible that some of the earlier studies suffer from inaccuracies in the measurement of the intensity of the acoustic stimulus, particularly at the lowest and highest frequencies. In the present study, the stimuli were measured a few millimeters from the eardrum, thus assuring an accurate and highly reproducible specification of the input to the auditory system. Further, 778 Volume 42 Number 4 1967 Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 152.2.176.242 On: Mon, 15 Dec 2014 11:56:03 EQUAL-SENSATION CONTOURS FOR LOUDNESSowing to the particular method of stimulus application employed (to be described later), it was possible to extend the frequency range downwards to 20 cps and at the same time keep distortion of the stimulus below 1% at al...
The proposed model consists of a chain of three energy reservoirs through which energy from an infinite supply is fed to a modulator which in turn drives a firing mechanism. The modulator consists of a variable permeability p that depends on instantaneous basilar displacement a through: p = (a - AO)2 for a greater than AO, and p = O for a less than or equal to AO, where AO is a constant. The firing mechanism consists of a Poisson generator (process) whose average output rate is proportional to energy flow through the modulator, and a classical "leaky integrator" neuronal model driven by the Poisson generator. The model, containing three fixed and eight free parameters, was examined with respect to statistical properties of spontaneous activity, relation between overall firing rate and level of stimulation, various adaptation and recovery phenomena within time ranges from a few milliseconds to several seconds, period histograms for one- and two-tone (phase locked) stimulation, suppression of responses to one tone by subthreshold levels of another (phase locked) tone, and neural masking. Model behavior in general was satisfactory. Deficiencies in single-cycle histograms at medium and high levels, and insufficient onset peaking in PST histograms, were attributed to the malfunctioning of one particular segment of the model, and a possible remedy was suggested.
For each of four subjects, a set of equal-loudness contours constructed from obtained loudness balances was compared to contours of contra- or ipsilateral sinusoidal stimulations that activate the acoustic reflex to the same degree, as indicated by a constant change in acoustic impedance at the eardrum. Two subjects produced “normal” equal-loudness contours, showing close coincidence with the respective equal-reflex contours; the remaining two subjects produced “abnormal” equal-loudness contours, exhibiting significant departures from the respective equal-reflex contours. Results indicate no systematic difference in the sensitivity of the acoustic reflex to ipsi- and to contralateral stimulation. Middle-ear attenuations calculated from the magnitude of the acoustic impedance at the eardrum, together with estimates of the reflex-generated attenuations, were employed to transform the sound-pressure levels (SPL's) of the equal-reflex contours to relative values of volume velocity of the cochlear fluid. The transformed contours support the hypothesis that the acoustic reflex is activated to the same degree by sinusoidal stimulations producing the same total number of neural impulses within about 200 msec. For some subjects, equality of loudness between two sinusoidal stimulations appears to be attained under this same condition.
A model is proposed containing three parallel systems; system 1 is a single channel consisting of an inhibitory circuit being fed by a fixed permeability responsible for spontaneous activity in parallel with a displacement-driven permeability, while systems 2 and 3 each consist of a number of identical permeability-volume-permeability channels with on/off switches activated at levels specific for each channel. The (continuous) output of each system drives a Poisson generator whose output "events" are delivered to a firing mechanism in the form of a "leaky integrator" delivering actual "firings" in simulated real time. The model quite accurately reproduces (1) stimulation and recovery phases of PST histograms obtained at six levels of stimulation for six auditory fibers (including "dead" period after stimulation), (2) single-exponential functions of recovery from adaptation for both onset and steady-state part of probe tone, (3) increments in response as independent of time of application of stimulus increment, (4) a specified hazard function for firings, (5) interval histograms of driven activity, (6) loss of phase locking with increasing frequency of stimulation, and (7) extended dynamic range for onset of response. The model fails in reproducing specific characteristics for response to decrements, and for response to AM.
Acoustic impedance at the eardrum was measured on three subjects as a function of frequency (20 cps to about 1000 cps) and intensity (minimum 65 dB SPL; maximum 150 dB SPL, at eardrum) of a sustained ipsilateral sinusoidal stimulation by means of an inversely driven Zwislocki acoustic bridge. At lower intensity levels, acoustic impedance was independent of level, resistance and reactance varying with frequency according to R =af -1/2 and X=bf -a/4, respectively, from 20 cps to about 300 cps; above 300 cps, resistance remained essentially constant, and negative reactance decreased rapidly. At higher levels, one subject showed variations in impedance attributable to the acoustic reflex. At levels above 120 dB SPL, all subjects showed a rapid decrease of both impedance components at all frequencies. Middle-ear attentuations calculated from the magnitude of the acoustic impedance, together with previously obtained equal-loudness contours for the participating subjects, support the hypothesis that equality of loudness between two tones is attained when both generate the same total number of neural impulses within about 200 msec. The decrease in impedance at the highest levels is attributed to subharmonic radiation from the eardrum and/or a shift in the axis of rotation of the ossicles.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.