Abstract
The distribution of average power in various speech sounds as a function of frequency has been measured with an integrating audio spectrometer. This instrument divides the electrical signal into 14 frequency bands and simultaneously records the integrated square of the amplitude in each band over a measured time interval, usually 30 seconds. From these data and the known over-all calibration of the spectrometer channels one can compute the spectral distribution of average acoustic power at the location of the input microphone. The operation of the instrument and the procedure of measurement are described. Speech spectra are given which are average results for a trained crew of seven men, speaking into a condenser microphone in an anechoic chamber.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
