Roger (K8RI) writes:

For voice communications you want the frequency response narrow.
Probably from a low of 200 or 300 HZ to a high or 2500 or 3,000 Hz.
This is where the majority of the intelligence is located.

You need up to 7 KHz or so for sibilants and some other phonological
features which are occasionally phonemic. This high-end response
makes it possible to distinguish between 'f' and 's' in
communications.

--
Transpose mxsmanic and gmail to reach me by e-mail.