[Radio Wrangler]

Think of a signal entering a receiver.

What features are there in the signal to allow the receiver to be accurately tuned and for the user to tell that it's accurately tuned?

With a pilot carrier mode, there is one very obvious answer. The FDM systems I used to work on had up to 1800 SSB channels and a sprinkling of pilot tones to allow good demodulation. The pilots weren't simply the missing carriers, but just one shared for every group, supergroup and mastergroup.

With fully suppressed carrier DSB, you don't have that option. But if it's not ISB but true DSB, then you can use the knowledge that the sidebands must be mirror images around the missing carrier. Inserting an off-frequency carrier will split each tone component in the recovered audio. Tune around to minimise the splits, or to get best correlation between basebands recovered from USB and LSB sides.

With SSBSC, you lose that opportunity as well, and you're left trying to find non-sinusoidal components in the speech so you can slide the frequency with a linear offset until you get harmonic structures coming out with integer relationships.

The second harmonic of 1kHz is 2kHz but that of 1.01kHz is not 2.01kHz and the human ear hears that something sounds 'off'

Fine tuning an ssb receiver manually when you don't know the person speaking has to come down to this approach because you don't know what the speaker's natural pitch is. A lot of people do this successfully without ever knowing what they're really doing.

There was a programme to try to pack more PMR channels into limited bandspace back in the Early 70s. The Woolfson foundation was the sponsor and it involved Gosling and Macario. Their approach was to change from FM in 50kHz/25kHz channels and use SSB. One tool used was pilot-controlled companding, but the difficulty was frequency correction to get a good natural sounding voice. THe gear had to be simple enough to be affordable. A friend was doing his Phd in amongst that bunch.

David