Confidence Bar on Fisher Detectors question...

Does anyone know how Fisher derives it's confidence bar readings?
Is it just the sum total of the last few seconds of numerical sampling,
or a different calculation all together? F-75LTD is the unit I have in mind.

That's a good question to ask, I'd considered asking it myself, knowing who visits this forum.
My hunch is that it relates to the target signal strength relative to the ground signal strength/or its variation. So if for example the target signal exceeds the grounds' by some margin, it's possible to determine target info (ID value) with accuracy. If they are comparable, then a medium confidence is indicated, and if the target signal is sufficiently weak that it's giving maybe 1/10th of the grounds signal, then a low confidence ID is indicated. I don't think it refers to confidence that the target is there, just detailed ID info.
I have to admit I don't pay much attention to it, I find I work it out my own way, using the pinpoint-mode depth-gauge, target ID values/jumpiness, tones etc.

Not sure how it works but it's saying that it is confident (or not so) in what the unit is id-ing. Certainly not near 100% accurate. I make use of it.

The f-75's tone ID and numerical ID are separate from each other, I'd like to know which one the confidence meter is paired to...anyone know?

It's a common trait with the MXT---I think D. Johnson was a constant design element also. I never even look at it when I'm coin shooting. I have a MXT armed friend that lives by it and I constantly demonstrate to him the need to dig those iffy 40% readings he passes up when I dig them after using the etrac but he is incredibly stubborn. My gain. In the magnetic ground around here I have little confidence in the confidence bar.

When you get to know the machine, you get more confidence yourself to dig ever more borderline signals, like those weak "almost 100% large iron" signals that turn out to be 10 inch (30cm) deep coins. I can't imagine the confidence meter giving a very 'encouraging' reading on those.
Re: the ID on-screen or audio report, I think the answer is on-screen ID, because the conf. meter works in all-metal, too. (I think, I'm tempted to assemble the machine to verify that, now it've typed it...)
The manual mentions using the conf meter to improve your swing technique. One of the key things here is to have a smooth even-height swing, to minimise ground-signal variations caused by a varying ground-coil gap. Ground-signal variations are what causes problems detecting weak targets, not the size of the ground signal itself. Naturally, 'stronger' ground with have 'stronger' variations, too. Mineralisation/iron levels vary from inch to inch, and the surface of the ground is a bit uneven, too.



Edited 1 time(s). Last edit at 01/02/2014 11:14PM by Pimento.

Pimento is correct.

For a beginner........ the Conf bar graph may be (somewhat) helpful.
Bottom line: deep targets will never give a high Conf reading. (((This can really hurt your decision-making process))).

The Confidence Bar Graph is a "peak signal strength" indicator that is tied to the visual id number. It adds another visual display to compare the audio signal's characteristics to. It has a variety of uses:

- as a running depth meter.

- as a large deep target alert. Deep large metal like smashed cans, etc. give a Confidence Meter reading that is not in harmony with the audio signal characteristics.

- as a way to focus jumpy tid numbers. While all the TID numbers displayed for a target are useful, which number do you focus on? The Confidence Meter is useful for focusing attention on the strongest signals of the target.

- and yes, as an aid to sweep consistancy, which is one of the hardest things to learn and practice.

Hope that helps.
HH
Mike

Just dig, dig ,dig

LowBoy

TAKE A LITTLE TIME KICKBACK AND WATCH SOME OF MY DETECTING VIDEO'S BELOW ON YouTube

[www.youtube.com]

If you don’t dig it, then how are you going to know what you’re missing!
How can you have your pudding if you don’t eat your meat!

Mike Hillis Wrote:
-------------------------------------------------------
> The Confidence Bar Graph is a "peak signal
> strength" indicator that is tied to the visual id
> number. It adds another visual display to compare
> the audio signal's characteristics to. It has a
> variety of uses:
>
> - as a running depth meter.
>
> - as a large deep target alert. Deep large metal
> like smashed cans, etc. give a Confidence Meter
> reading that is not in harmony with the audio
> signal characteristics.
>
> - as a way to focus jumpy tid numbers. While all
> the TID numbers displayed for a target are useful,
> which number do you focus on? The Confidence
> Meter is useful for focusing attention on the
> strongest signals of the target.
>
> - and yes, as an aid to sweep consistancy, which
> is one of the hardest things to learn and
> practice.
>
> Hope that helps.
> HH
> Mike

Thanks Mike and others....
This is the type of answer I hoped for. I like knowing where input comes from,
and I learned about the large deep target use that I had not known....

I'm a little skeptical of Mike H's statement: "- as a running depth meter."
As the machine has no idea what the target is, and only a rough idea of the ground signal strength, making estimates of depth is a different technique. To work out on-the-fly depth, you need to filter out the ground signal as best you can (this is a normal process inside a detector) and look at the peak change above the ground-signal level. Then you ascribe it a notional depth based on pre-measured values for a 'typical coin' (US 5c ?).

Yes, there is no explanation to that statement. I get the others, but I can't make that one clear in my head.

I have never paid attention to it.
Corey

I'm not delving into how it works. I'm stating what it can be useful for.

What is so hard about using the confidence meter as a running depth meter? It is a signal strength indicator. Weak signal = deeper, Strong signal = shallower. Subject to all the variability of any depth readout (target size, detector settings, minerals, etc).

Check it out in your test bed. Or your next hunt. You can even calibrate it to your sensitivity levels for known standard objects, even factor in known variables.

HH
Mike

One obvious comment is the fact that the F75 has an on-the-fly depth meter, so it would seem reasonable to suggest that the confidence meter indicates something else.
I haven't done anywhere near enough proper measurements, yet, of ground and target signals, in a variety of different grounds, to state any results. And whilst I do detect a variety of terrain, from smooth pasture & parkland, to lumpy ploughed/plowed fields, I don't look at the confidence meter, so I have no anecdotal evidence as to what does / doesn't affect the meters' reading. Over the coming year, I may have more idea.
It might, on first consideration, seem that air-tests should give a high confidence reading, after all, there's few other signals to 'corrupt' things, just electronic circuit noise, EMI, (and perhaps clumsy sweep technique). In practice, it seems to give a rather jumpy reading. If it's effectively performing some kind of 'mathematical division', (to determine target strength relative to ground-signal strength), then in air, it would be 'dividing by zero', a recipe for unpredictable output.

Indeed so many junk targets mimic good targets is the reason for confidence bars. Sort of to give us that little extra to dig or not. Having said the above never paid much attention to confidence bars on any unit as bad ground.. depth ..masking influences can certainly play havoc with the best of them so I say when in doubt dig. Indeed probably newbies to the hobby or to a particular unit can gain by using same but my confidence comes from knowing what my unit is telling me thru experience using same.

Sorry, talking about my F5.

Been awhile since the F75 and I were buddies. Have to go back through my notes for the F75.


Mike

Been digging through my stuff looking for my field test notes on the F75. Trouble is I have some information saved on discs and some saved on thumbdrives. But I did find some of my private reports dated back to Febuary of 2007.

Regards the F75...In a nutshell, the confidence meter and the depth meter should not give conflicting reports. When they do give conflicting reports you know that something is amiss.

For example:

If you see a target give a medium or shallow depth reading, yet the confidence meter is low (1 or 2 bars), you know you have a deep, larger than a coin, target. Most of the time you will see this type of response on deep, smashed aluminum cans. You'll get a nice zinc target id in the low 70s, good audio, medium to shallow pinpoint (4" or so) but the confidence meter will show 1 bar.

Whenever the two graphs agree you can have strong confidence in the display output no matter what the number of bars being displayed. When the two graphs disagree, you then have to intrepret the results, keeping in mind that calibration is to US coin size targets.

Hope that helps.
HH
Mike

Hey Denny ..

I would suspect the confidence bar is set up with how accurate the I.D. is showing by the data collection of I.D. and the depth on say a coin sized object...plus the presence of trash in the snapshot..(by this i mean is it a hard solid hit in the pictur or is it a multiple target hit that the I.D. is showing as a conglomerate average)....And even F.E. reading skew..

Why coin sized? well thats what the whole unit is calibrated at in terms of I.D.

so say if it's showing a dime on the I.D.

the confidence meter will then be judging its response to say well its 4 inhces deep ...its in 1 bar f.e. dirt...its a clean snapshot no multiple target's. So its prett high that it is a dime or a object that resembles a dime...So it gives a high confidence reading..

but it could also decide its 8 inches deep ....its in bad F.e. dirt....and theres was an averaging of I.D. taking place for multiple target's...So now the machine is very low thinking its a dime...

and as you can guess theres all point's inbetween the two extremes I laid out in the scenario..

it just takes all the data that is already there and does GUESS work...is it accurate? well the other things like I.D. and even depth and such are not accurate most of the time ..so if yuo collect inaccurate data and report it theresa good chance the reprot is not accurate...

but anyways its nice for the cherry picker flash money type hunter.. but leave it at that..

the same info can be garnered from watching the i.d. jump around and notice the depth on pinpoint and know your ground skew condition's....but its just a selling feature to offer it.. more or less...


that's my take on it

Keith

bump

Thanks Keith;
I'm going out today, I pay closer attention to the bar, dig it, and see how well it worked...

You're right on spot Mike.
A 12" pop or beer can really fool the user, But your tip works, it just takes practice.


Mike Hillis Wrote:
-------------------------------------------------------
> I'm not delving into how it works. I'm stating
> what it can be useful for.
>
> What is so hard about using the confidence meter
> as a running depth meter? It is a signal strength
> indicator. Weak signal = deeper, Strong signal =
> shallower. Subject to all the variability of any
> depth readout (target size, detector settings,
> minerals, etc).
>
> Check it out in your test bed. Or your next hunt.
> You can even calibrate it to your sensitivity
> levels for known standard objects, even factor in
> known variables.
>
> HH
> Mike