386 



BELL SYSTEM TECIIXICAL JOURNAL 



shows these artirulations as functions of the sensation level of the 

 speech. In Fig. 11 they are shown as functions of the width of the 

 transmiited frequency band. It should be noted that the term 

 articulation as here employed denotes only the correct interpretation 

 of unrelated speech sounds and is not a measure of voice naturalness 

 which is also an inii)c)rtant factor in the telephnnir trnnsini?sinn 

 of speech. 



loo; oioop oiooe oioov oioot onoui o 

 FiR. 10 



1 



E^E 



44H hH=^ 



^ 



OI23450l23.d50l23^50l23450IZ3^50IZ3-150l23450IJ3-150l234501J3A5 



BUinttJ r 1 o'eVya 



CO ' . . ' I ' ' ' . H— ' — ; I ' ' ■ ■ ' >/ ■ I i I t 'll \-A 





pOI23i)}Oi23il30IZ3.i:OI234J0123430i;}4SOI234)OIZ}4 90l23'd50l23ii; 



1 dowtamestib 



•^i>^ . 



lii2345Ci23^50i2345Oi23^3OI2343OI234S0l234J0l2345CI23^5OI2345 



h a'u aj zncliK s 



ARTICULATION AS A FUNCTION OF THE 

 BAND WIDTH PASSED BY HIGH OR 

 LOW PASS riLTtRS. ABSCISSA IS 



-IE cuT-orr FRtfluENCY in mlo- 

 cycles. 



Fig. U 



