If putting the audio through a SSB mod-demod channel did result in a frequency shift, that could create some interesting effects. The Doppler effect from a passing train whistle comes to mind.RussellHltn wrote: ...
Speaking of feedback controllers, I ran across a very old Altec anti-feedback device. Looking at it, I think what it did was modulate the audio into a Single-Sideband and then demodulate it. That would cause the frequency of all sounds to shift sightly. Thus the feedback couldn't reinforce itself.
That must have been interesting to listen to. When you pushed it to where it started getting echo-y, it must have gone up or down in frequency with each "ring". It must not have worked too well as Altec made very few of them. I think it goes back to what the sound guy was saying - the mic hears and speaker. No getting around that.
In some ways, phase is even more important than frequency in terms of preventing audio feedback. Many years ago, (long before FM groups) I designed and built an audio preprocessor for my stake center cultural hall. It included a polarity reversal switch, because I had thought the two possible polarities through the electronic part of the loop could yield different profiles of the first frequencies to feed back. It turned out at least the first few frequencies to feed back were the same, regardless of the polarity setting.
What I did discover was that moving that switch from one position to the other would instantly (well, so it sounded) kill any feedback energy that had built up. I found that by periodically (or aperiodically) toggling the switch, I could contain the audible feedback to a modest volume level, even with the gain turned up quite a bit higher than would normally be possible. I toyed with the idea of an automatic polarity toggler but never built one. With a modern DSP, one could do some wild things to the phase of the signal. However, I suspect it would likely not be feasible to get more than at most a few dB gain above unity without some rather unpleasant audible effects.