380 BELL SYSTEM TECHNICAL JOURNAL 



instrumental selections, orchestral renditions, both classical and jazz. 

 Quite a number of observers were employed, some of whom were 

 present on several tests and a few on all tests. On tests made on a 

 San Franciscb-Denver-San Francisco loop involving 2,600 miles of 

 circuit, no observer was able to consistently differentiate between the 

 quality over the reference circuit and that over the program circuit. 

 On tests made on the San Francisco-Chicago-San Francisco loop 

 certain of the more experienced observers were able to differentiate 

 between the circuits somewhat more than 50 per cent of the time, but 

 this, it must be remembered, was on a direct comparison test. None 

 of the observers could tell with any assurance which was the program 

 circuit and which was the reference circuit if a few minutes were 

 allowed to elapse between switches. On the Chicago loop 264 obser- 

 vations were made on direct comparison tests on which 60 per cent of 

 the observations favored the reference circuit and 40 per cent favored 

 the program circuit. 



Included as part of the overall program, were tests to determine the 

 volume range, maximum volume obtainable and speed with which 

 the circuits could be reversed. 



On the volume range tests a source of program was obtained and 

 so regulated that it had a very narrow volume range. This was then 

 applied to the circuit with the sending end gain adjusted so that the 

 maximum volume applied at the repeater outputs was + 6. The 

 sending end gain was then gradually decreased so as to apply a 

 gradually decreasing volume to the circuit. This process was con- 

 tinued until the program volume was so weak that the line noise 

 interfered with its satisfactory reception. The amount that the send- 

 ing end gain was adjusted determined the volume range. The average 

 value for several tests was slightly in excess of 40 db. The maximum 

 volume was determined by switching a 10 db pad from the sending 

 end to the receiving end of the circuit and listening to a transmitted 

 program, noting the point at which there was a quality difference 

 between the high volume and low volume condition. It was found 

 that a slight difference could be detected when the maximum volume 

 on the high volume condition was -\- 10, thus showing the circuit was 

 capable of handling a maximum volume slightly lower than this value. 



As mentioned earlier, switching means are provided at each station 

 for reversing the direction of transmission. On the initial field tests 

 it was demonstrated that the circuits could be reversed readily and 

 at the same time maintain satisfactory overall characteristics. At the 

 present time, on receipt of proper advance notice, the circuits are 

 being reversed on commercial programs in approximately 30 seconds. 



