LOUDNESS 



419 



This completes the data which are available on steady complex 

 tones. It is to be hoped that others will find the field of sufficient im- 

 portance to warrant obtaining additional data for improving and 

 testing the method of measuring and calculating loudness levels. 



In view of the complex nature of the problem this computation 

 method cannot be considered fully developed in all its details and as 

 more accurate data accumulates it may be necessary to change the 

 formula for h. Also at the higher levels some attention must be given 

 to phase differences between the components. However, we feel that 

 the form of the equation is fundamentally correct and the loudness 



• -HIGH PASS x-LOW PASS 



100 200 400 1000 2000 4000 100 200 400 1000 2000 4000 



Q FILTER CUTOFF FREQUENCY p 



Fig. 14 (A to D) — Loudness level reduction tests on the No. 3A audiometer tone. 



function, G, corresponds to something real in the mechanism of hearing. 

 The present values given for G may be modified slightly, but we think 

 that they will not be radically changed. 



A study of the loudness of complex sounds which are not steady, 

 such as speech and sounds of varying duration, is in progress at the 

 present time and the results will be reported in a second paper on this 

 subject. 



Appendix A. Experimental Method of Measuring the 

 Loudness Level of a Steady Sound 



A measurement of the loudness level of a sound consists of listening 

 alternately to the sound and to the 1000-cycle reference tone and 

 adjusting the latter until the two are equally loud. If the intensity 



