First, it is theoretically considered how the fundamental mechanism of mean squaring operation of sound level or vibration level meters affects the probability distribution forms of output sound or vibration level in direct connection with the stochastic evaluation. The effect of time constant and frequency characteristic of the indicating system is reflected in the parameters of output probability distribution together with an amplitude-frequency spectrum of input noise or vibration. More explicitly, two modes of statistical approach to the above problem are given : In the first approach, some basic properties which are manifested throughout the entire random variations (e. g. , a boundary amplitude region where the output fluctuation of mean squaring circuit takes such as (- ∞, +∞), (-∞, 0), or (0, +∞)). are considered. Then, the probability distribution of output is obtained in the form of an orthonormal expansion, each coefficient of which may be expressed as function of value, variance, highorder moments, etc. reflecting the internal mechanism of indicating system. The second recti approach is in rather striking contrast to the first approach and the nonlinear operation itself is the first principal consideration in it. The response of the indicating system to a random input is analyzed in the same way as when the input signal is deterministic and then the probability distribution of random signal is taken into consideration. The theoretically derived output level distributions are able to be expressed only with informations on a time constant of the indicating system and amplitude-frequency spectrum of random input and/or indicating system and they are in good agreement with experimentally sampled values. Especially it must be noticed that an information on phase-frequency spectrum is completely unnecessary.
To study the auditory localization under the binaural listening, following two experiments are brought. One is the physical measurement which gives the transfer functions at the ear canal entrance when the sound source is located in the horizontal plane, lateral plane and the vertical median plane systematically. The other is the psychological measurement which gives the minimum audible angle to locate the sound source along each of above planes. These data show the listener's ability to discriminate the azimuth of sound source. Then the relations between the transfer function and the discrimination ability of azimuth are discussed in details. some auditory localization cues are found upon the transfer function, i. e. some peaks and deep notches. Further, it is found that pinna plays an important role in localization. So, The detailed cues generated by the pinna are surveyed more by using the artificial pinna model.