TRANSMISSION FEATURES OF NEW TELEPHONE SETS 361 



Fundamental differences of this kind between telephone and direct 

 conversation must be taken into account in the design of a telephone 

 transmission system if satisfactory results are to be obtained. For 

 example, the talker is accustomed in a direct conversation to regulate 

 his talking volume by what he himself hears under prevailing noise 

 conditions (which incidentally are the same for the listener), by the 

 ease with which he hears the other party, and by the ease with which 

 the listener appears to hear him. By experience, under ordinary 

 conditions, the first factor mentioned, the loudness with which the 

 talker hears himself, probably comes to be the primary control on his 

 talking volume. 



These various factors also serve to regulate talking volumes in 

 conversation by telephone, but their magnitudes and the relations 

 between them differ from the condition of face-to-face air path con- 

 versation and vary from one type of telephone connection to another. 

 For example, the "sidetone" of the telephone set, being materially 

 higher than the air-path sidetone, deceives the talker, not only by 

 making him think he is talking louder than he really is, but also by 

 apparently modifying the noise conditions under which he is talking 

 in the pickup and amplification of room noise by his telephone trans- 

 mitter. Since, in addition the efficiency of the telephone circuit itself 

 may be different in the two directions of transmission, the loudness 

 heard by one party may differ more from that heard by the other than 

 in the case of air transmission. Then, too, noise conditions may be 

 and frequently are quite different at the two ends of the telephone 

 circuit. Figure 3A shows the probability of noise of various average 

 intensities at subscribers' stations as determined by several surveys 

 covering a large number of locations. On the assumption that any 

 one of the stations represented by these data may with equal prob- 

 ability call any other one. Fig. SB has been computed, showing the 

 probability of noise at the two stations of a telephone connection 

 differing by more than a certain amount. It will be noted that there 

 is about an even chance of the noise at the two ends differing by more 

 than 12 db. In view of these differences, a person's judgment of how 

 well he is heard and understood can not be as direct as in the case of 

 air transmission. 



In addition, the transmission over the commercial telephone system 

 affects the quality of the received speech more than the usual room 

 surroundings in air-path transmission. While acoustic resonance and 

 reverberation in a room do distort speech, in the extreme case to a 

 point where understanding may be difficult, such a condition is dis- 

 tinctly unusual. Equal freedom from distortion in a telephone system 



