? ? Audio quality on SSB is always a problem. Both transmitters and
receivers often have very limited response, IMO too limited. Why? Well
one reason is misunderstanding the research done on intellibibility and
articulation (not quit the same thing) over a period of decades. A lot
of the earliest research was done by Bell Labs and later research by
many organizations such as Armor Research, Harvard and MIT for military
purposes. One thing that has been known for a very long time is that the
human voice is not very efficient in its use of energy. It turns out
that in the male voice (there has also been research on female voices)
that most of the power is below about 250 Hz while the speech parts that
contribute to intelligibility at mostly above about 1 Khz. The power is
in the bass. So, if we have a power limited system, nearly always the
case for radio or telephone, full bandwidth is very inefficient in its
use of available power.? The simple solution is the reduce the bass and
accentuate the treble. So, a microphone with limited frequency response
and that tailored to reduce the lows and accentuate the highs, is one
way of doing this. It is also very common to use electronic filters in
the transmitter to limit the frequency response.
? ? Now, it turns out that in practice, where there is NO power limit
or noise full range speech has superior intelligibility. However, for
practical communication systems there is virtually always a limit on
power and the presence of noise. Keep in mind that where there is no
requirement for it the limiting of frequency response always damages
intelligibility. This is one of the things that is missed when reading
the research.? I have not discussed the compression of the dynamic range
but that is also commonly done. Here again there has been a tremendous
amount of research, and again, much of it mis-understood.? While it is
possible to process speech to the point where it is no longer to
identify individual voices and still maintain intelligibility it is
again high-fidelity speech which has the best intelligibility scores in
the absence of noise or power limits. So, in general the least that
works is the most desirable.
? ? AT&T discovered long ago that excessive frequency response limiting
made it difficult to recognize voices. For the telephone system some
"naturalness" was necessary to make it acceptable. So the low limit was
set at about 250 Hz and, to accomodate the limits of the telephone
system, about 2750 Hz on the high side. Going higher where possible will
improve articulation, the ability to understand speech sounds. The
limits have been more or less adopted for SSB, typically 300 to 3000 Hz.
On either end reduction will either reduce "naturalness" or reduce
articulation. Both are very common in ham equipment and I have heard
very severely limited signals on the air. Electronic filters with
extensive menus are pretty common and to quote an engineer friend
"anything adjustable is by definition out of adjustment".
? ? I am writing too much so will stop now.