Harmonisers are a tool like any other. It can be used or abused. While they are pretty good these days, there is always a degree of unnatural tone to them, and the further you get from the fundamental, the more unnatural they become. Best effects are had by processing each voice separately, which is something you can only really do on the seriously high end harmonisers/modellers/multi-FX though. Good to cop harmonies live though.
It's all down to your audience though Bob, the acoustic crowd will likely see a harmoniser as a crutch, while no-one thinks twice about Brian May using one (or two) on his guitar.
I am guessing that they need some parameters to work with - like what key you are singing in - so that they know what notes can be used.
Correct on all counts, you can also set how many harmony voices are generated and what their interval is to the fundamental.
though I suppose if your pitch was varying a lot then one "side-effect" would be that what comes out of the speakers would be magically adjusted so that it's in tune and pitch is constant.
The Autotune-type effects so beloved by the dance crowd are basically monophonic, single-voice harmonisers that replace the voice with the effected signal. That corrects the pitch and with external control can be used to change the pitch on the fly for the "Cher Effect".
It would be interesting to know the (rudiments) of the science behind it.
Put simply, it's a zero-crossing digital pitch recognition algorithm combined with a few pitch shifting algorithms. Newer, better versions have a formant algorithm to change the character of the shifted voices too.
I saw this guy called "Lonesome" Dave Ferguson at a festival. Outstanding blues harmonica player.
No shit! Dave can make my hair stand on end with one note - and I don't even
like the harmonica! I'd heard he was doing solo stuff with loopers these days.