This paper, describes a program time regulator by sampling method using a revolving head assembly, mainly its principle, mechanism for experiment, problems of head scanning, and the effects by the following factors-sampling width, time constant at sampling boundary, regulating rate, etc. The results clarified by this experiment are as follows: 1)To eliminate echo distortion, reproducing output of each magnetic head must be switched by an electronic circuit. 2)In speech, most suitable sampling width was 30 ms. 3)To Resolve the click and the level fluctuation at the sampling boundary, it was affective to set the switching time constant at 2〜3 ms. 4)In the case of large rate of regulation such as about 100% expansion or 50% compression, too the intelligibility was little destroyed.
抄録全体を表示