RATING TR-iNSAIISSION PERFORMANCE 123 



telephone circuit, such as the replacement of one receiver by another, 

 is measured in terms of the change required in the reference trunk 

 to give a loudness balance for the second condition. In this way, 

 measurements are also obtained of the effect on the loudness repro- 

 duction ratio of the various parts of telephone circuits. When the 

 circuits used commercially consisted of apparatus and lines similar to 

 those in the Standard Cable Reference System and the major con- 

 trollable factor was the loudness reproduction ratio, such measurements 

 constituted reasonably adequate means for indicating the comparative 

 functioning of circuits and apparatus. 



The noise on a telephone circuit may be measured in various ways. 

 The method which has been most generally used is that of comparing 

 it with the controllable output of a fixed source of a complex wave shape 

 and adjusting this output until it and the line noise are judged to have 

 equal interfering effects. 



With the availability of circuits and apparatus having widely differ- 

 ent distortion effects, the volume ratings became insufficient for indi- 

 cating the relative performances of commercial circuits. The earliest 

 method used in rating distortion effects was one in which observers 

 listening to transmission over the circuits, gave judgments as to their 

 relative merits. By so comparing various kinds and amounts of distor- 

 tion, two at a time, relative ratings can be established for placing them 

 in order of merit. This procedure was particularly useful in the early 

 days in working out the designs of transmitters and receivers, especially 

 from the standpoint of the location in the frequency range of their 

 points of maximum response. While such a judgment method has the 

 shortcoming of not providing quantitative ratings it has been found 

 that experienced observers can in general obtain results which are 

 relatively consistent with the results of more definite measuring 

 methods. Such judgment comparisons of distortion effects are fre- 

 quently used, particularly in exploratory work, and are still more or less 

 necessarily relied upon in setting limitations on circuit properties which 

 primarily affect the naturalness of reproduction. 



To provide for the need of a method for measuring the relation be- 

 tween the reproduced and impressed sounds from the standpoint of 

 effects of different kinds of distortion, use has been made of the articula- 

 tion testing method.^' In this method, which has been widely used in 

 recent years, lists of syllables, usually meaningless monosyllables, are 

 called over the circuits to be rated and the percentage of syllables cor- 

 rectly understood is taken as a measure of the circuit performance. 



3 "Articulation Testing Methods," Fletcher and Steinberg, Bell Sys. Tech. Jour., 

 Oct., 1929. 



