[Elecraft] K3S TX Noise Gate

[Jim Brown]

On 3/18/2018 7:43 PM, Joe Subich, W4TV wrote:
> ased on the charts on his web page, Bob Heil knows *nothing* about the
> structure of human voice (speech).  There is little or no usable energy
> in speech below 200 Hz - even the lowest bass voices have little
> *speech* energy that low. 

I strongly agree with both statements.

> In addition, there is little to no energy
> between 800 Hz and 1100 Hz (it is believed by scientists that voice
> developed that way so humans could hear danger in the presence of
> voices).  Only a soprano (female) or countertenor (male) will actually
> have vocal energy when singing in the 800 - 1100 Hz range but neither
> will have speech energy in that range. 

I don't know where you get your information about this, but it is WRONG. 
The last 30 years of my professional life centered on designing sound 
systems for speech intelligibility in challenging acoustic environments, 
and I've studied it with some of the best on the planet. Yes, the 
distribution of energy in speech varies across the speech range, but the 
speech range should have the flattest practical frequency response. The 
GOOD reason for a modest boost around 3 kHz is to partially compensate 
the rolloff of the SSB transmit filters. This has been good practice for 
at least 60 years, and boost EQ is built into some of the early popular 
ham mics, of which the Shure 444 is an example.

The most critical octave bands for speech intelligibility are 1,000 Hz 
and 2,000 Hz. 500 Hz is next, followed by 4,000 Hz. I know of no serious 
practitioners who would recommend a mid-range dip except, perhaps, those 
in the broadcast "loudness wars."

73, Jim K9YC