I think TJN63 is correct....Here is a old post from several detectorists from a defunct website that may help you...actually it will help you alot
Q. User gets alot of +95 readings
A. Hi Edward,
I stumbled into your post so I thought I would add my opinion if you don't mind.
Now, from a technical standpoint, a coin can read higher than normal as the result of the ground conditions. The analysis is basically done using math means that adds the ground signal to the target signal. There is a tendency for some objects to read upscale on a target ID detector in highly mineralized ground. As the target signal becomes weaker and weaker, the ground signal begins to dominate, thus the higher reading. George Payne wrote about this, I quote "In addition, if the target is deep it will more than likely tend to read higher on the VDI meter in mineralized ground".
Terra Diggers Personal Note: The reason that I don't get +95 readings when bench testing is that the +95 phenomena (where target is too deep for VDI to get accurate reading) needs ground mineralization to occur. Air testing has no mineralization present.
Now, if there are few "hotrocks" in your area, I would recommend you turn on the +95.
Years ago, when I was using another Whites unit, I found that accepting the highest 3 or 4 highest negative numbers (-96 through -92) to positive would also increase the number of deeper coin finds without causing me to dig much trash. So, this is something else to think about. I used to call this sort of cheating but after finding out just how the mineralization distortion occurs, I call it distortion compensating.
User: Jim M.
On both the XLT and DFX extremely deep non-ferrous tagets will id, both audibly and visually, as +95. Thus, if you want to find the really deep goodies, the astute XLT or DFX user will ALWAYS set +95 to ACCEPT at OLDER sites. If one REJECTS +95, they will ALSO reject the really deep targets. But here is another "get more depth" hint" which will also get more depth: program your machine to also ACCEPT -95. Yeah, -95. And, then watch the VDI meter for faint, "iffy" signals which give BOTH a +95 AND -95 "double bar" graph readout.
To make a long story short, I got 8 such signals in short order. But signal number 8 wasn't so "iffy," so I dug it. It turned out to be a very old token. Whoa! Little Jimmy then re-traced his steps, and EACH (and I mean each!) of those other double-bar signals produced coins dating from 1898 to 1917.
User Dave
+95's can be very deep targets also, and GOOD ones at that. A way to tell is if the reading goes below 95.5 ,and, or, you get the slightest hint of a pinpoint. When I get a +95, a I really investigate it with many swings and checking for a pinpoint, if it goes below +95.5, even just to +95.4 or +95.3....I Dig. Been using this method for many, many years, and DEEP targets have varied from 2 foot deep Buckles. and many coins at depths of the 16 inch range. This is one of my Best kept secrets.
When dealing with HotRocks in high iron mineralization here in Va, the Pinpoint mode is my main tool. When pinpointing, hotrocks tend to not give a pinpoint signal at all, or the signal smears and moves around with no definite point.
Trash Tips
A deep GOOD target when pinpointed will sound much softer than the discrimination mode. Trash will sound weak or borderline in the discrimination mode and LOUD in the pin point mode as a general rule.
Reg
As a new detectorist, please learn how to extract coins without damaging the ground