A Voice and Ear for Telephone Measurements 



By A. H. INGLIS, C. H. G. GRAY and R. T. JENKINS 



An artificial voice and ear have been developed which are sufficiently 

 close simulations of the real voice and ear in their principal physical attrib- 

 utes to justify their use in both shop and laboratory tests of telephone 

 transmitters and receivers. 



The artificial voice and ear have certain advantages in that they can be 

 exactly specified and reproduced and can be used in determining physical 

 characteristics of instruments which are difficult or impossible to obtain 

 with real voices and ears. 



THE performance of telephone transmitters and receivers is de- 

 pendent not only upon their physical characteristics but also 

 upon the reactions of the users to these instruments. Observations of 

 the results obtained by subscribers under known conditions of practical 

 telephone use take both these factors into consideration. Tests of 

 this kind, however, are not suited to the needs of laboratory develop- 

 ment and study of instruments where data regarding their physical 

 characteristics are required. In such instances data, to be of greatest 

 value, should be taken under conditions which include those factors 

 of actual service having an important bearing on the performance of 

 the instruments. 



Among the most important of such factors are the voice and ear. 

 For making many of the laboratory tests, therefore, it has been 

 necessary to employ actual human voices and ears in order to insure 

 that all of their physical characteristics have been included in the test. 

 This procedure, however, due to uncontrollable variations in indi- 

 viduals, requires a large expenditure of time and effort to insure the 

 precision desired. Furthermore, certain instrument tests, such as 

 response-frequency characteristics, are either impossible or difficult to 

 make with the real voice or ear. 



These disadvantages, inherent in the use of the human voice and 

 ear, have been recognized for a long time. Numerous attempts have 

 been made to employ voice and ear substitutes for instrument testing, 

 and in cases where uniformity of instruments rather than their design 

 has been of primary importance, as in shop acceptance tests, such 

 substitutes have been of great value. That their use in engineering 

 design problems has not been more extensive has been due to the 

 inability to make them meet certain fundamental requirements. 



It has been the aim, therefore, in the design of the artificial vT)ice 

 and ear to be described, to overcome previous objections to the use of 



293 



