SPEECH POWER A\n ESEKGV 



635 



between tlie male ami female voices, which, however, does not affect 

 the higher frequencies enough to alter the vowel characteristics. 



Pkak Ka( tor 



The tests just described were also used to obtain the peak factors of 

 the vowels. These were lietermined by measurement of the maximum 

 Pi and Pm of each syllable and arc charted in Fig. 4. Here again there 



31, 39 



Av£itAaE Peak Factob. 



so, „ . ASX 



AvERAoe Fon 8 Male Voices 



u, yii 



AvtRAoe fo« fl Female Voices 

 i4, as, '9 1 i6, 



HidMeST Set fok any Male Voice 



60, sa ss 



Oooaooa'ua-J; 

 13 30-0 o « -;] « iq - « 

 Fig. 4 — Peak factors of vowels 



are differences between the sets for the male and female voices, the 

 former being somewhat higher, especially for the vowel e. In both 

 cases such rasping vowels as k (tap), e (ten), a (tape) have sharp waves 

 and high peak factors. Having listened attentively to all these voices 

 under test, I have become able to associate peak factors with vocal 

 qualities in the following way: the voices with the higher peak factors 

 are those which in the ordinary terminology are said to be "resonant" 



