602 BELL SYSTEM TECHNICAL JOURNAL 



From the transmission standpoint, tape or automatic sending, i.e., 

 sending from a tape perforated in accordance with a telegraphic code, 

 is merely a matter of increasing speed and accuracy of the characters 

 transmitted.^ Automatic tape recording of radio signals, such as by 

 the syphon recorder or similar device, removes the advantage obtained 

 from the tone character of the signal, so useful in ear reception, and 

 substitutes for this tonal character the less acute ability of the eye to 

 distinguish between signals and noise on the tape record.* In sacrific- 

 ing this ability to receive with greater accuracy in the presence of con- 

 siderable noise there is, however, a gain in the speed of receiving radio 

 signals. The tape record, which is in permanent form, makes it possi- 

 ble for several operators simultaneously to transcribe different parts of 

 the received message at speeds much slower than the transmitting 

 speed. 



Printing telegraphy goes one step further in removing the human 

 element from the process of receiving and substituting a mechanism 

 which must be impelled to a definite act by each current element re- 

 ceived. The printing mechanism inevitably records what it receives 

 without using any judgment factor in the process other than the me- 

 chanical application of such fixed criteria as have been put into it by 

 the designer. Unless the transmitted signal is received with such in- 

 tensity and character as to be the controlling signal at the receiving end, 

 errors will usually result. The use of printing telegraph equipment on 

 radio circuits,* therefore, makes the signal-to-noise ratio necessary for 

 the receiving of satisfactory copy greater than would be required for 

 either aural reception or tape signal recording. 



It is of considerable interest to compare the approximate minimum 

 values of signal-to-noise ratio required for satisfactory * transmission of 

 intelligence by single side band long-wave radio using the customary 

 double side band carrier telegraph. This has been done in the table on 

 the following page. 



When automatic means for recording the signals are applied at the 

 receiving end of a radio circuit it is desirable that considerable uni- 

 formity exist in the output level of the receiving equipment. This is 

 even more important when printing equipment is used. Such a condi- 



* Obviously, "satisfactory" cannot have a definite quantitative meaning which is 

 applicable to all modes of communication under all variations in the observed types 

 of received noise. For example, "crashy static" would probably not be as serious in 

 receiving by ear as it would be in receiving by other means. Then too, there is the 

 personal judgment factor in determining just what constitutes "satisfactory" com- 

 municaton. The table is set up on a relative basis using quantitative values of sig- 

 nal-to-noise ratio which appear to represent the worst condition under which com- 

 munication could be eft'ected with only an occasional error. Of course, communica- 

 tion can be continued under much worse conditions, but with an increase in the 

 number of errors. 



