116 BELL SYSTEM TECHNICAL JOURNAL 



The difficulties encountered as a result of "singing" form one of 

 the most troublesome problems connected with the actual operation 

 of these systems. When a portion of the sound emitted by the pro- 

 jectors reaches the transmitter with sufficient intensity, that its re- 

 production is as great as the originally emitted sound from the pro- 

 jectors, and with such phase relation that it tends to aid the original 

 sound, the system will emit a continuous note. Moreover, when the 

 portion of the sound from the projectors which reaches the trans- 

 mitter is not sufficient to cause a continuous note, it may be sufficient 

 to cause considerable distortion of the speech or music. In excessively 

 reverberant halls these conditions are often fulfilled when the actual 

 amplification is so small that the people at the distant points are 

 scarcely able to hear the speaker. In all cases in our experience the 

 difficulty has been sufficiently overcome by properly placing the 

 transmitter with respect to the projectors. The situation is very 

 much helped by the presence of the audience, which adds considerably 

 to the acoustic damping of the room. 



It will be seen, therefore, that the acoustic conditions of the space 

 in which loud speakers are used are of considerable importance. 



Characteristics of the System 



The first requirement of the system itself is that it shall reproduce 

 speech or music faithfully. A system is said to do this, or in other 

 words, its quality is called perfect, when the reproduced sound contains 

 all of the frequencies, but no others, contained in the original sound 

 striking the "pick-up" mechanism, and when these frequencies have 

 the same relative intensities that they had in the original. 



An imperfect or distorting system is one which fails to fulfill this 

 requirement. There are two main types of distortion which had to be 

 considered; the first being the unequal amplification of the system 

 for the various frequencies constituting the sound and the second 

 being the introduction of frequencies not present in the original 

 sound. For simplicity of discussion, this last class will be divided 

 in three parts, namely: the effect of transients, the effect of asym- 

 metric distortion and the effect of disturbing noises. 



The effect of transients has already been mentioned in connection 

 with acoustics and they, of course, produce the same type of dis- 

 tortion whether they occur in the acoustic or the electrical system. 

 Transients occur whenever the sound changes either in pitch or in- 

 tensity, and are introduced at the beginning and ending of each 

 speech sound. This modification of the characteristics of the speech 



