PLANNING AND OVER-ALL PERFORMANCE 



19 



< 



15 



'JJ 



m 

 (J 



UJ 

 Q 



LU 

 > 





if) 

 in 

 O 



100 



200 



300 400 500 600 800 1000 

 FREQUENCY IN CYCLES PER SECOND 



2000 



3000 4000 



Fig. 3 — C.C.I.F. objectives for frequency characteristic of voice channel. 



No specific objectives were agreed upon for the frequency characteris- 

 tics of the 12-channel groups as such, but there was an expectation that 

 ±2db could be achieved except for frequencies adjacent to the filters 

 in the split group. 



For program channels, the C.C.I.F. recommendations were also 

 adopted in respect of the two-band (6.4 kc) and three-band (10 kc) 

 arrangements. To meet the requirements of these channels and of teleg- 

 raphy, an overall frequency stability objective of ±2 cycles was adopted. 



Noise and Crosstalk 



Noise objectives were established to be reasonably consistent both 

 with Bell System and C.C.I.F.* objectives for circuits of transatlantic 

 length. 



The objective for the rms circuit noise at a zero level point in the 



* The methods specified by these two bodies for the assessment of circuit noise 

 differ in three respects, the units employed, the frequency weighting employed, 

 and the fraction of the busy hour for which the specified noise may occur. The 

 meters concerned are the Bell System 2B noise meter (FIA weighting network) 

 reading in dba and the C.C.I.F. psophometer (1951 weighting network) reading 

 in millivolts across 600 ohms. The relationship between readings on the two meters 

 is discussed in a later paper and it will suffice here to note that, for white noise 

 dbm (COIF) = dba (Bell) - 84. 



