Hi Kevin,
Your concept of pyrites not being detectable is, for basically true. It takes an extremely sensitive detector to detect them and even then, the chances are what is being detected is another form of iron oxide commonly associated with pyrite.
Now, VLF's are much more susceptible to these various oxides and as such can generate quite a strong signal depending upon the oxide involved. Fortunately, on a good VLF, the ground balance is wide enough that normally one can adjust it and cause the oxide signal to change from a positive to negative response or visa versa. If this happens then the dominant reason for detection is because of the oxide content and not any metallic content.
Keep in mind this does not mean the rock doesn't contain valuable ores because it can. It just means the signal is controlled by the dominate factor and that is the oxide material. The fact is, iron oxides and signals from them are, according to one of Candy's patents, in the order of 40 times stronger or more than signals from a non ferrous object of similar size. As such, any key iron oxide can clearly dominate the response.
Now, on a VLF one can sort of determine what is the dominate factor by simply adjusting the ground balance. However, on a PI, that isn't quite the case.
On a PI, magnetite or commonly known as black sand does not generate a response if the detector has certain circuitry to balance out the earth's magnetic field. On some PI's this balancing act isn't perfect so there can or will be signals from certain black sand type rocks.
Now, there is also something called magnetic viscosity commonly discussed when discussing maghemite (another iron oxide). This magnetic viscosity is the term given to the response generated by maghemite when energized by a strong magnetic field. The result is a residual decay that sort of mimics a signal from a metalic object. Now, this signal has absolutely nothing to do with detecting tiny iron particles but has to do with the iron oxide creating a decay signal because of the magnetic viscosity concept.
More can be learned about magnetic viscosity and its effects on metal detectors by simply googling the association. Unfortunately, most discussions are based upon the mathematical reasoning and the concept is usually not discussed in simple terms.
Now, here is something interesting that is often interpreted incorrectly. Many detectors will generate a response in areas where a fire has occurred. Now on a VLF, this response may be because of a piece of burnt wood that has transformed into a slightly conductive object, or it could be because some of the iron oxide was transformed by the heat into maghemite. In either case, both will generate a positive response. In the case of the maghemite the signal is more dependent upon the difference in ground signals.
Now, on a PI, the response from the burnt wood isn't detected. The signal is too weak and too short for that to happen. However, the ground response because of the transformation of the maghemite is detected and can be a reasonably strong signal.
This topic is getting a little deep so I will stop here since one doesn't really need to know the technical aspects of why something is happening. However, it doesn't hurt to know at least some info.
Reg