622 



BELL SYSTEM TECHNICAL JOURNAL 



taken where the correct judgments amounted to 60 per cent. In addi- 

 tion, the frequencies where 80 per cent correct judgments were obtained 

 have been marked. The region between the 80 per cent marks is the 

 most significant — frequencies above and below would probably not be 

 heard at auditorium distances or with other instruments playing. 

 There were certain sounds that apparently extended to the highest 

 audible frequencies as judged by the ease with which the highest filter — 



AUDIBLE FREQUENCY RANCE 

 FOR MUSIC SPEECH AND NOISE 



ACTUAL TONE RANCE 



ACCOMPANYING NOISE RANGE 



TYMPANI 



BASS DRUM 



SNARE DRUM 



14" CYMBALS 



BASS VIOL 



CELLO 



PIANO 



VIOLIN 



BASS TUBA 



TROMBONE 



FRENCH HORN 



TRUMPET 



BASS SAXOPHONE 



BASSOON 



BASS CLARINET 



CLARINET 



SOPRANO SAXOPHONE- 

 OBOE 



FLUTE 



PICCOLO 



MALE SPEECH 



FEMALE SPEECH 



FOOT STEPS 



HAND CLAPPING 



KEY JINGLING 



500 1,000 



FREQUENCY IN CYCLES PER SECOND 



5,000 10,000 20,000 



Fig. 5— Audible frequency ranges. 



13,000 cycles cut-off — was detected. In these cases the lines have 

 been arbitrarily stopped at 15,000 cycles. 



With most instruments it was noticed that the actual musical sounds 

 were accompanied by other sounds such as key clicks, lip noises, 

 "buzz" of reeds, and hissing of air. The observers tried to distinguish 

 between the frequency ranges carrying the two classes of sounds, and 

 their judgments are summarized by the treatment of the lines of Fig. 5. 

 The heavy portions indicate the frequency ranges thought to convey 

 the "tone quality" of the instruments, and the short vertical lines 



