Will we ever get this type of metal detecting technology?

maxxkatt · Sep 28, 2018

In the Army I was a 32G20 or Fixed Cryptographic Equipment Repairman and worked on crypto computers that were use in fixed locations like army headquarters all over the world.

But all my experience was analog and digital electronics theory and field work. Not audio processing. This was the domain of the US Navy submarine fleet. Of course being in the Army I had no access to any information about Navy electronic systems. Everything below is idle speculation on my part. Too hot to detect in Atlanta these days.

Ok, what follows is what I have gleaned from the Internet since I have no training or experience in audio signal processing.

One of the tools the US Navy uses is passive sonar digital analysis of propeller noise and other noise characteristics of every ship on the ocean. This database of audio characteristic is stored in their onboard computer systems and the computers compare the signals heard with those in the database and come up with a match to ID the ship. I have read that it can identify individual ships from a long distance away with out the enemy even knowing we are lurking and identifying those ships.

So can this technology be applied to metal detecting? We process those audio signals with our ears and the better metal detectorists are better at this than others. I know the technology exists, but can it be applied in a practical and economic manner to make it feasible?

So the questions are…

1. do our metal detectors have the computing power to analyze these complex audio signals?
2. can there be developed a database of known audio responses of all the junk and good targets and stored in the metal detectors ram?
3. Can this be done in economic manner that keeps the resulting metal detector under $2,000?
4. Would you buy such a machine for $2,000?

If this technology could be developed and implemented in a sub $2,000 machine I would suspect it would be the ultimate cherry picker.

I also think it could be used with a high quality digital color display even better than the CTX 3030 and show you a cross section of what is actually below your coil and using icons indicating size of coin, type of coin (eg silver, clad, etc), relic type eg shoe buckle or ox knob and type of junk, nail, bottle cap, can slaw and relative size of each object.

Ok, maybe not now, but in the future. The microprocessors are always getting smaller, faster, cheaper and the same goes with the memory.
Any thoughts from anyone and the real technical guys? I know there are some on these forums with tons of technical experience in electronics.

Jason in Enid · Sep 28, 2018

I think you are believing peoples own hype a little too much. Claims that people can distinguish targets based on the sound are delusional. Its possible to separate by gross characteristics but nothing more.

Metal detecting hit its theoretical wall back in the 80s. The only development since then has been multi-freq which is extremely complicated and difficult to make work, which is why so few do it. There is far more information gleaned from the analysis of the EM return signal than can ever be determined from sound.

Porsche914 · Sep 28, 2018

http://noktadetectors.com/invenio-metal-detector.asp

A bit more than $2k

georgeinsc · Sep 28, 2018

Maxx you must remember that those machines that we used on the boats cost many millions of dollars. First was the millions spent on research and development and then the cost of the machine itself was also very expensive.

Passive Sonar operates strictly on picking up the sound of a ship or piece of machinery.

Currently the detectors we use emit a sound based on the conductivity of the target.

It would require an new technology not based on conductivity of the target to be able to get a more definitive sound out of a detector.

I do not think that the astronomical cost of research and development would ever be offset by the sales of a metal detector used in a hobby.

The Russians have always kept a "fishing vessel" outside of every port where our submarines operate recording the sound signature of each ship that enters or leaves the port.

lytle78 · Sep 28, 2018

The “signature library” approach works because the “signal” which is captured, analyzed and stored is relatively invariate - a given noise sources signature doesn’t change much.

The “signature” of a buried metallic object is subject to vastly more variables. The detector is not a passive system, but an active one. Almost all detectors have lots of user controlled variables which can have an effect on the returned signal. In addition, ground mineralization, moisture level, the presence of signals returning from “non target” metallic (and mineral) objects near the target all modify the signal input to the detector.

Foragist · Sep 28, 2018

Metal detecting technology, I believe, can and will be improved. Perhaps a major breakthrough is coming, but not how you are suggesting...exactly. Our machines do not analyze sounds. They analyze electromagnetic responses of metals in proximity to the coil's electromagnetic field(s). The detector's magical computing ability is to translate these EM responses into visual and audio signals for the human user to further analyze. So I think some goals for improvement should be A) Better EM field emitting B) Better EM response analyzing C) Better digital/visual translation to the human user, and D) Better audio translation to the human user. Then of course, they could always improve durability, waterproof capabilities, ergonomic designs, backlighting, etc. etc.
Just wait and see...or start working on it. There will be improvements.

That Nokta technology that Porsche914 linked is a huge step in the right direction, but I think even that is just the beginning.

flyguy784 · Sep 28, 2018

Lots of good info here. I think Jason said it best. Your signal analysis issue would be in the returned signal from the coil, that's where all the variables are. It'd be easy to "assign" a given tone to a given return. The problem is, there's so many variables in what an objects return signal would be. Based on everything that's been mentioned. Soil type, depth, orientation of the target, moisture all alter this return signal to a degree.
Listening to a sub and having a library of those captured sounds, then comparing the signature of an unknown to the rest in the library, by computer, is another technology.
Just way too many variables in the targets beneath our coils.

Tom_in_CA · Sep 28, 2018

You've gotten excellent answers so far. Another thing I would add, is that the medium that we have to cut through, is FAR different than what you're talking about. Because that technology is seeing through water. We are seeing through solid ground.

The same question comes up about fish sonar: It is now so refined, that they can even tell the type and size of fish, at great distances. Hence why oh why is md'ing limited to just a foot (on coin size) and a few foot (on Toaster oven size) ?

The reason why fish sonar can see so far, is that water present zero impediment to the send and receive signal. Contrast to ground, which is ALREADY a solid object. So too do I think the same comparison exists to what you're saying. Not to mention the fact that ANY "signature" you could derive from a gold ring: I guarantee you that I can wad up a piece of foil or fold up a snippet of aluminum, to give that exact same signature.

Tom_in_CA · Sep 28, 2018

Porsche914 said:
http://noktadetectors.com/invenio-metal-detector.asp

A bit more than $2k

That is over-hyped nonsense. In the text , it talks about "To know what type of a metal the detected target is ". Yet when you scroll further down, to look at sample screen pix, you see that it's doing nothing more, than categorizing the metals via conductivity. Which all metal detectors ALREADY do. And as we all know, the devil is in the details: That many metals share the same conductive ranges, Doh!

And the shape showing feature of that machine is going to be useless for most all of the hunting we typically do. The problem is: The pixel size is too large (about 1" squares) to be of any use. Thus all the objects we typically look for (coins, rings, etc...) will all be ...... doh ... 1 pixel.

Hence I don't see much use for that machine, except to lull the buyer, who thinks he's going to get magical TV-like images of objects in the ground, and tell gold apart from aluminum, etc.... It won't do that.

Will we ever get this type of metal detecting technology?

maxxkatt

Jason in Enid

Elite Member

Porsche914

Elite Member

georgeinsc

Elite Member

lytle78

Elite Member

Foragist

flyguy784

Elite Member

Tom_in_CA

Elite Member

Tom_in_CA

Elite Member

Similar threads