There are two persistent myths that need busting.
Myth 1 - Running the sensitivity too high in bad ground puts too much power to the ground overloading the machine.
Myth 2 - Lowering the sensitivity in bad ground reduces power to the ground thereby improving depth by not overloading the machine.
If you connect the Explorer transmit winding to an oscilloscope and adjust your sensitivity up and down you will notice that the transmist signal strength does not vary. While building custom coils for my Explorer I have had my sens at 1 all the way to 28 and the transmit signal is always the same.
Therefore the Explorer transmits at 100% power regardless of user settings. Therefore one cannot overpower the ground or put less power to the ground via user settings.
That said what really happens when you adjust sensitivity? Only the Minelab engineers know for sure, we can only guess.
I have had deep targets with my sens at 28 manual that gave a not great but decent signal. I then began lowering my sens and as I did the signal became weaker and weaker until at one point the signal was no longer there and I had a steady threshold when sweeping the target. The same is often true for switching to semi-auto sens. Often the signal vanished when switching to semi-auto.
This is why some camps say the higher the sens the better, but wait! The trouble is this does not hold true 100% of the time. Sometimes when I have had a deep target with my sens at 28 lowering my sens resulted in the signal improving, crikey!
This suggests that things are not as simple as we would like them to be. It suggests that targets (this includes the ground) do not increase and decrease in signal strength at the same rate when you raise or lower the sens. For example lets say you have a deep coin with a nearby rusty nail and some semi-bad ground. Lets say with the sens at 22 the coin is the stronger of the three targets (coin, nail, ground). As you increase the sens the ground or nail may overtake the coin and drown it out.
What aboug gain? The only thing I'm pretty sure of is that gain is processed after sensitivity. Therefore if you adjust your sensitivity lower to a point that the target is no longer there, boosting your gain to 10 won't amplify a target that is no longer there.
But again things are not quite that simple. There are two general camps, those who run a high gain and lower sens (sometimes semi-auto) and those who run lower gain and higher sens. Both will make finds the others cannot. Which method produces more finds in a given area depends on the soil and site conditions.
The guys running lower sens and high gain WILL make some deep finds the high sens camp will miss. This is because higher sens results in more faint high pitched signals from the ground. Was that last faint high pitched chirp a deep silver coin or one of the half dozen high pitched ground chirps you get per swing from running your sens hot? You can't see the forest for the trees.
On the other hand there are clearly deep signals that simple vanish with the sens at lower settings so the high gain camp will miss those.
At the end of the day there is no one approach that is better than the other 100% of the time. In fact if a given site has stopped producing using one approach it makes sense to adjust your settings and try another, you may make a few more finds.
This can be true for discrimination also. I often run all metal but with rusty crowncaps notched out. I have dug many silver coins that were hiding underneath more shallow rusty crowncaps. The odd thing I noticed is this, with the crowncaps notched out the machine locked onto the silver pretty good. But when I switched to true all metal with no discrimination (I do that as a rule when pinpointing targets) the machine instead locked onto the rusty crowncap solid and pulled the cursor down to the rusty crowncap area. Had I not been running with the crowncaps notched out I would have walked over that coin in all metal.
But again its not 100%, sometimes the oposite is true, I'll get a signal on the coin in all metal but with the crowncaps notched out I get only a null. Arrrgh!
Charles