TRANSMISSION DATA FOR RATING TELEPHONE CIRCUITS 343 



provided that the circuit performance is the same for circuits having 

 numerically equal parameters regardless of differences in physical 

 characteristics. The number of parameters required to describe a 

 circuit is largely a matter of convenience. A small number tends to 

 make the interpolation simple, but the derivation from the measure- 

 ments complex. The converse holds for a large number. For pre- 

 paring effective transmission data for the local plant, the circuit de- 

 scription has been expressed in terms of five parameters as follows: 

 the volume loss from the transmitter of one set to the receiver of the 

 other, the sidetone volume loss at the talking end, the sidetone volume 

 loss at the listening end, the circuit noise efficiency of the station at 

 the listening end, and the distortion. An important advantage of 

 using these particular parameters is that each represents a circuit 

 characteristic generally recognized as affecting transmission perform- 

 ance. The electrical line noise at the station end of each subscriber 

 loop and the average room noise at the station are, of course, two other 

 parameters which are maintained at the reference value except when 

 computing line noise and room noise losses. 



The computation of effective ratings from repetition observations 

 and circuit measurements may be summarized as follows: The trans- 

 mission observations are preferably made on various series of circuits, 

 in each of which one parameter is varied while the others are kept 

 constant. First, a series of observations is made on circuits which are 

 identical except that the volume loss is varied by distortionless changes 

 in the trunk attenuation; preferably these should be various adjust- 

 ments of the working reference system. A second series of observa- 

 tions is made with a constant volume loss but with variations in some 

 one of the other parameters; for example, the sidetone at one end 

 of the circuit might be varied. From this series of tests in conjunction 

 with the first series, the distortionless change in volume loss, which is 

 equivalent to each change in sidetone, is determined both for trans- 

 mitting and receiving and curves of effective loss versus sidetone may 

 be established. Such curves are shown in Fig. 6. These effective 

 losses apply only for this particular volume loss, but tests with other 

 constant volume losses give essentially the same relations. The change 

 in effective transmitting efficiency is due to the fact that telephone 

 users raise their talking volume when the sidetone is reduced. The 

 change in effective receiving efficiency is due to the fact that a reduc- 

 tion in sidetone reduces the interfering effect of room noise. 



In the same way, distortion and noise are varied separately and the 

 effect of each of these changes in terms of equivalent change in volume 

 loss is determined. 



