346 BELL SYSTEM TECHNICAL JOURNAL 



same repetition rate. Presumably, other scales for expressing distor- 

 tion could be used, with other formulas, and possibly these can be 

 derived directly from repetition observations if a great enough variety 

 of circuits is covered. In any case, it appears that some such distor- 

 tion factor is needed since distortion in any complete telephone con- 

 nection is too complicated to be classified by any simple means, such 

 as specifying a cutoff frequency. 



It should be pointed out again that the computations and measure- 

 ments described provide merely a means of interpolating between 

 transmission observation results and have both the limitations and 

 advantages of an interpolation method. They cannot be used to pre- 

 dict performance of any circuit radically different from those covered 

 by repetition counts, but on the other hand, any inaccuracy in the 

 formulas used in computing the parameters is of only secondary im- 

 portance since they are used only for interpolating between observed 

 points. For limited applications to simpler circuits more direct meth- 

 ods might be satisfactory. For more complicated circuits, such as 

 present-day long toll circuits, other parameters are needed to describe 

 such characteristics as delay distortion and echoes. 



Conclusion 



The effective transmission data can be applied in practically the 

 same manner as the volume loss data which they replace. Conse- 

 quently, the effects on transmission service of distortion, noise and 

 sidetone, as well as of volume loss, can all be taken into account in 

 the design of the plant in a simple, systematic way. Such compre- 

 hensive transmission ratings are required to utilize properly the vari- 

 ous types of facilities now employed in the telephone plant, to direct 

 future developments, such as further reductions in distortion, and to 

 incorporate into the plant the new types of facilities resulting from 

 these developments. 



