ROOM NOISES IN TELEPHONE LOCA TIONS 659 



Both methods were used in the survey, because it was felt that each 

 gave information which could not be as accurately obtained from the 

 other, and also because the use of two methods enabled each one to be 

 used as a check upon apparatus defects which might occur in the other. 



In using the masking method, data were taken by two experienced 

 observers and corresponding measurements averaged. All meter 

 measurements were made by one observer. 



Results of Survey 



Noise in Business Locations. — One hundred and nine business loca- 

 tions were visited. The magnitudes of the noises measured varied 

 from that found in a doctor's quiet office to the din of a large manufac- 

 turing plant. Distribution curves for the noises measured are shown in 

 Fig. 4 for the meter method and Fig. 5 for the masking method. For 

 any point on one of these curves the corresponding per cent of all of the 

 measurements made had values equal to or greater than the indicated 

 abscissa value. 



It may be seen that with the exception of the "high" curve of Fig. 5 

 the curves for meter and masking methods are fairly similar in shape. 

 The "middle" curve has been selected to represent the masking 

 method. 



If there are excluded as extremes those noises which were so low that 

 95 per cent of all the noises measured equaled or exceeded them, and 

 those which were so high that only 5 per cent of the measurements 

 equaled or exceeded them, the spread of noise magnitudes is seen to be 

 about 40 db. The standard deviation of the measurements is about 

 12 db. 



As shown on Figs. 4 and 5 the median business room noise would 

 produce a reading of 23 db on the meter scale and a masking of 27 db 

 on the high-frequency warbler tone, 39 db on the middle-frequency 

 tone, and 31 db on the low-frequency tone. The average business room 

 noise was about 2 db higher than the median. 



Some conception of the amounts of noise represented by these figures 

 may perhaps be gained from the following. The extremely loud noise 

 measured in a local station of the New York subway while an express 

 train was passing produced a meter reading of 70 db, while the lowest 

 noises measured in the survey, in quiet residences, gave readings near 

 Odb. 



Data on noise at the business locations tested have been grouped so 

 as to show the average differences in the room noise values obtained for 

 different types of business and for different sizes of towns. It will be 

 appreciated that only a very small number of measurements were in- 



