But it seems you are mixing apples and oranges.

Not really, because the ability to select sidebands while in sync mode
is a
big plus in mitigating against ACI. Of course, one can do the same
with just SSB,
but going through the trouble of having a sync detector and not being
able
to select sidebands seems perverse nowadays.

Sync detection means a locally generated "carrier" is present, so
if the signal fades the lack of a strong carrier is not a factor.
It does nothing to prevent fading (which I bring up because someone
recently said something along those lines here) it merely helps when
the signal fades.

There is so much loose talk about this, but I think the above is a good
way of putting
it. First of all, how could one *prevent* fading if the channel is
fixed? Sync does not prevent fading,
but as long as the detector is able to reasonably track the carrier
phase in the fading medium, it will
provide better performance than an envelope (amplitude) detector, so in
this sense sync
detection does mitigate against fading.

Selectable sideband really has nothing to do with synchronous
detection, other than that using the phasing method it's relatively
cheap to implement compared to an expensive IF filter.

True, this is why I think it's perverse when most of the expense is
already in
designing a PLL-type sync not to go the extra step in being able to
select sidebands.
It's a question of implementation.

I'm not even sure where we've veered off to. I thought
the previous comment was something like synchronous detection
wasn't all that important. I'd say that's true, given that
people lived without it till it became a feature in relatively
recent years. Someone listening to broadcast radio (am or
shortwave) that are relatively strong may be the ones to
benefit the most, because you can get deep fades where
the sidebands are still nice and strong. Signals
that you have to strain to hear, it's far less likely to
be useful, because they are already below a minimum strength.
You'd want to pull in other techniques at that point, and that
includes the narrow IF filter that has good slopes.

The opposite is true. When you have very high SNR conditions,
an envelope detector performs almost as well as a sync detector, but it
is
in the low-to-medium SNR region where you will see the performance
improvement, especially in terms of signal intelligibility. This can
be verified
by anyone with a sync detector radio. Try listening to a strong signal
in the AM band with and without sync: there is almost no perceived
change in
signal quality. Try the same on shortwave, you'll see more of an
improvement in sync mode.

RK