390 



BELL SYSTEM TECHNICAL JOURNAL 



level, are given. The intensity levels, ^\, of the free wave which 

 sounded equally loud are given in the third row. In the fourth row 

 the values of the constant C, the calibration we are seeking, are given. 

 The voltage level added to this constant gives the equivalent free 

 wave intensity level. In the fifth, sixth and seventh rows, similar 

 values are given which were determined at the threshold level. In the 

 bottom row the differences in the constants determined at the two 

 levels are given. The fact that the difference is no larger than the 

 probable error is very significant. It means that throughout this 

 wide range there is a linear relationship between the equivalent field 

 intensity levels, /3, and the voltage levels, 20 log V, so that the 

 formula (13) 



,3 = 20 log V + C 



can be applied to our receivers with considerable confidence. 



500 1000 



FREQUENCY IN CYCLES PER SECOND 



I. 1 — F'ield calibration of loudness balance receivers. 



L = 60 db.) 



5000 10000 20000 



(Calibration made at 



The constant C determined at the high level was determined with 

 greater accuracy than at the threshold. For this reason only the 

 values for the higher level were used for the calibration curve. Also 

 in these tests only four receivers were used while in the loudness tests 

 eight receivers were used. The difference between the efficiency of 

 the former four and the latter eight receivers was determined by 

 measurements on an artificial ear. The figures given in Table II 

 were corrected by this difference. The resulting calibration curve is 

 that given in Fig. 1. It should be pointed out here that such a calibra- 

 tion curve on a single individual would show considerable deviations 

 from this average curve. These deviations are real, that is, they are 

 due to the sizes and shapes of the ear canals. 



^ The ordinates represent the intensity level in db of a free wave in air which, 

 when listened to with both ears in the standard manner, is as loud as a tone of the 

 same frequency heard from the two head receivers used in the tests when an e.m.f. of 

 one volt is applied to the receiver terminals. 



