Auto will for sure stabilize the ID when you find it jumpy at maxed "stable" sensitivity on targets. I've been playing with that a lot on undug stuff, switching from max stable sensitivity to Auto, and I'm finding Auto is much easier to quickly stabilize the ID without the need to wiggle and work the target in manual maxed sensitivity. If auto shows marked improvement in target ID stability on a known "good" target such as a coin then I think it's a sure indication that I'm running manual too hot. I think particularly in rough ground (hot rocks, minerals, RF noise, and/or iron) Auto may in fact find coins that manual wouldn't because the machine is doing it's best to reset the threshold and keep things from overloading. I think it's worth working tough spots like this in auto and then going back over it in manual or vise versa to see what the results are. If anybody does that please post a thread. It would be very interesting to read. I'll try to get around to that myself some time this summer and throw up a report.
Until I peg down exactly how depth is affected at maxed stable sensitivity versus something a good bit lower and also compared to Auto I'm going to have serious doubts about which is giving me best depth, at least on the 15x12 SEF. Besides doing what I can with my test garden I plan to do some tests in the field. Kind'a tweaked that I'm getting a lot of noise from the overhead power lines where I put the garden. Should have thought that one threw, but the SEF has been so stable for me that I haven't really had to think about that much at sites. I think maybe noise band 1 might have eliminated the noise even at max manual sensitivity but only played with that for a minute before the rain. If I can't get manual at least to something halfway to max then that limits my comparisons of various settings and I may have to move the garden to a quieter area.
I think my next and most logical step is going to be to start carrying a silver dime with me at all times so when I get to a new site I can compare the various settings. I'll find a clean spot and cover it back up with soil and then start playing with auto, max, halfway, etc on sensitivity to see which gives the best response. I'll probably stick the dime at say 9 to 11" deep, because 9" is the bare minimum I'm shooting at in depth to go deeper than my prior machines. If I find any setting is giving me a good hit on it then I'll go another 2 or 3" deeper and play with them all again until I find out which setting is still hitting the target good and which ones are no longer able to register it. That's probably the only way I'm going to know for sure what adjustments to sensitivity are working best at which sites. Yea, a freshly buried coin isn't the same as something that is undisturbed in the ground but the results should probably be the same as far as which is penetrating the ground best. I'll expect slightly better depth on undug targets but this should be very informative. Might have to start making this routine for myself, particularly if I find that with each new site max, auto, or a lower sensitivity changes as to which is giving best depth and target ID/sound.
I'll report back with my findings as I go. The only machine that seemed to lose depth with max stability (threshold, ID, etc) for me over the years was the Explorer. It made adjustments tricky because everything looked fine but it was costing me depth or making targets vanish. I remember one day in which a friend was hunting with his Explorer in auto or at least a much lower manual and/or semi auto. We were taking turns with his machine and I kept maxing it out when I was using it. We pretty much covered the same general area (high trash and minerals) and he I think got 2 or 3 silvers from memory that were not very deep, like maybe 2 to 4" at the most. I kept getting skunked and that was the first time I realized something was going on with the sensitivity on it.
After all the above jazz I'll probably then compare what sensitivity setting I find to be the deepest at a site to where the pump or sweep method of setting sensitivity has it set. I'd like to see something like the pump method end up being the same sensitivity setting as the above so I don't have to go about sticking a coin in the ground at every new spot. I have noticed when I've played with the pump method that the threshold will null approaching the ground at a sensitivity setting that seems fine doing the sweep method, so that tells me that the pump method might be better as it's indicating that the sensitivity is still set too high.