Hi Southwind,
I can give an idea of why a PI may not air test as well as a buried test and it is simple. PI's are really bad about reacting to external noise from about any source. Raise the coil in the air and that coil becomes a really great antenna. Lower that same coil close to the ground and the ground helps shield the coil from much of the noise. So, in the air, I may not get nearly the depth as I can if I bury the same object.
On just the right day I can air test up to 18" on a nickel with one of my PI's, but the next day I may be lucky to only obtain an air test of maybe 12". In the field, I used to bury a nickel at about 12" for a quick test just to verify the detector was working right. In almost all cases, the signal strength from that nickel was a strong response. I could easily bury the same coin deeper but that wasn't my main objective when I would run the test.
So, can one air test poorly but do much better on buried objects. The answer is yes. Can one display a poor test even on a buried test in one location and do a whole lot better somewhere else even though the ground may be very similar. Again, the answer is yes.
If the signal coming from the PI is crystal clear with no noise, then the detector should display the approximate same depth rather the object is buried or in an air test. Keep in mind the noise level doesn't have to sound that bad to make quite a difference.
One good way to check both a buried target and an air test is to use a noise canceling coil. Unfortunately, such coils generally lack depth compared to a similar mono of the same size. Because of the nature of a noise canceling coil, I normally use an elongated housing and build the figure 8 from front to back instead of side by side. My test coil I use at home is about 16" long by 5" wide. Inside I have a coil twisted into a figure 8 with each half close to half the length of the housing. This particular coil will display two distinct target responses, one in the front and one in the back, each having the approximate depth capability of maybe a 6" coil. What makes it nice is noise is no longer a problem, so it is easier to compare depths easier if I so desired.
Now, my reason for the noise canceling coil is to check my detector inside where noise could be a big problem and not for comparing different depth capabilities. Minimizing external noise problems allows one to concentrate other things besides trying to deal with trying to minimize noise sources.
I mention this noise canceling coil as a simple means of being able to do easier comparisons. Fortunately, building this type of coil is really quite easy to do.
Reg