开云体育

People must understand that the power distribution in speech is not very efficient. Thus speech processors. They do several things but all are to the end of modifying power distribution to make it more efficient in power limited transmission systems. Note that as a rule of thumb the ratio of peak to average in male speech is about 10db. If a transmitter is set to pass peaks without distortion its average output will be very low. A processor reduces the ratio of peak to average in various ways and also modifies the spectral distribution in the speech to accentuate the parts that contribute to intelligibility and articulation. An ALC is really a sort of peak limiter to reduce drive when it will drive the amplifier into distortion. Most ALC systems are fairly simple so an external processor is often more effective. With an external processor the ALC can be almost inactive but will provide a safety factor.
The idea is to increase average power to overcome noise. This also applies to the modification of the speech spectrum. Since low frequencies contribute the most to speech power but relatively little to articulation it is common to limit low frequency response and enhance high frequency response since the highs contribute most to it. However, the effect of combined spectrum shaping and amplitude limiting is not simple so some care must be applied in deciding what exactly to do.
There is a perfectly enormous amount of literature on speech processing (and music processing) in the technical literature. Bell Labs has research going back a century, worth doing some reading.

--
Richard Knoppow
Los Angeles
WB6KBL
SKCC 19998