STANDARD VOLUME INDICATOR AND REFERENCE LEVEL 105 



types have been employed. For the majority of the tests, the sources 

 of program were high quaUty recordings, convenient because of the 

 ease and exactness with which the programs could be repeated. For 

 some of the tests, however, actual speakers and musical instruments 

 were employed with direct microphone pickup. 



A number of the types of volume indicators in common use were 

 represented in these tests. Since the 700A Volume Indicator was 

 common to all of the tests, it has been chosen to represent the r-m-s 

 type of volume indicator in the data presented below. The peak- 

 reading type was represented by the especially constructed experi- 

 mental instrument, whose fundamental circuit is shown in Fig. 2. 

 The resistances controlling the rates of charge and discharge of the 

 condenser were adjustable, permitting a range of characteristics to be 

 obtained. The adjustments for which the data referred to below were 

 obtained, correspond to a rate of charge of the condenser such that 

 impulses of single frequency applied to the input for 0.025 second would 

 give a reading within 2 db of the reading obtained with a sustained 

 wave of the same amplitude. The rate of discharge of the condenser 

 was about 19 db per second. These rates are generally similar in 

 magnitude to those specified by the International Consultative Com- 

 mittee on Telephone Transmission (the C. C. I. F.) for broadcast 

 service, and by the Federal Communications Commission for modula- 

 tion monitors. 



The d.-c. amplifier and d.-c. milliammeter which indicates the charge 

 on the condenser included features, not shown in the simplified sketch, 

 which made the response logarithmic. The instrument had a sub- 

 stantially uniform decibel scale covering a range of 50 db. 



The data from four different series of tests, made at different times, 

 were collected in one body, and distribution curves were plotted show- 

 ing the relative frequency of occurrence among the data of the different 

 levels at which incipient overload was detected. Curves for tests on 

 a Western Electric 94B Amplifier, which is an amplifier designed with 

 negative feedback and therefore having a relatively sharp cutoff, 

 similar to a radio transmitter, are illustrated in Fig. 4. It will be 

 noted that the curve obtained with the r-m-s volume indicator has a 

 slightly greater spread than that for the peak-reading volume indicator. 

 Twelve different observers took part in these tests, and 13 samples of 

 program were employed, including male and female speech, dance 

 music, piano, violin and brass band selections. 



The data may more readily be interpreted when plotted in the form 

 of cumulative distribution curves, obtained by integrating the above 

 distribution curves. Cumulative curves for the data just referred to 



