• Forum server maintanace Friday night.(around 7PM Centeral time)
    Website will be off line for a short while.

    You may need to log out, log back in after we're back online.

Just thinking out loud about VDI resolution

Cherry Picker

Forum Supporter
Joined
Jan 12, 2006
Messages
13,470
Location
Dodge City KS
OK, I know some of you are probably getting tired of me bringing up VDI resolution, but I know detector manufacturers read these forums and I'm hoping to strike an idea.

While I talk about the VDI resolutions of 30, vs 50, vs 100, vs 200 here, I will mainly use the NOX 800 (50 VDI segments), and the Simplex+ (100 VDI segments) because they are two of the popular models currently.

We seem to have reached a tone-dependent period in detectors. I believe this is only because the VDI resolution, and hence accuracy, has fallen off the importance list of newer detectorists. If VDI accuracy were much better you'd demand better VDI systems from manufacturers, and they would deliver. There was a time, not so long ago, when the VDI was as important as the tone, and the combination of the two helped dig less junk.

From the responses, I've read a lot of people just don't trust the VDIs. And I don't blame them because most are almost useless on deeper targets or even shallow ones. But there are still exceptions. I use for example the NOX 800 vs the much less expensive Simplex. I have hunted mostly tot lots for the last 12 years. This has a few unique lessons that can be learned because I have cleaned out all the junk over the last 12 years, so. like with many beach hunters, I do the dig it all because there is far better odds of jewelry to junk in these sites.

Hunting these clean tot lots has also shown me how important VDI resolution can be even in very trashy parks. Once you know what trash VDIs are you know what not to recover. That is the key. The VDI & tone should be telling you what not to recover rather than what to recover. I have learned that my Simplex, with 100 target ID segments, will read a 18-21 on the pesky foil 90% of the time. If in 10 swings it ever reads above 21, even on 1 or 2 swings, it will not be foil, but something like jewelry. The same is true down in the iron range. If my Simplex reads a solid 04 on every swing, it will be a hairpin, tac, paperclip, or some other piece of iron. If the VDI on any of the swings read above 04, even on 1-2 swings a 05 or higher, 90% of the time it will be a chain, earring, or even one of those $0.25 cheapy rings.

This same concept can be used in your favorite trash filled park. Now when I had the NOX 800 with only 50 VDI segments, actually only 40 in the conductive range, that 18-21 could be a number of things from foil to gold. And that gold chain will read the same VDI as a piece of iron because the VDI resolution is so much lower. The resolution does make a difference in
VDI accuracy. It was even better with the old Whites DFX and its 191 VDI segments. The DFX has twice the VDI resolution of the Simplex, which has twice the resolution of the NOX 800.

As someone who depends more on the VDI than the tones I would love to see new models of detectors with much higher resolution.
 
What you're asking for makes sense. But I think the simple answer is that it's far easier said than done.

Increasing VDIs doesn't mean much (except from maybe a marketing perspective) unless the underlying technology can make use of it.

While an oversimplifcation, here's a rhetorical question for you to make my overall point:

You say the DFX has 4 times the VDI resolution of the Equinox 800 (and 600). Tell me, which one of those detectors is used by more hunters right now?

Ok, but you say, "imagine an Equinox 800 with a VDI of 200 instead of 50." Sure, but that's not possible right now, either b/c MIQ can't do that or Minelab chooses not to do that.
 
What you're asking for makes sense. But I think the simple answer is that it's far easier said than done.

Increasing VDIs doesn't mean much (except from maybe a marketing perspective) unless the underlying technology can make use of it.

While an oversimplifcation, here's a rhetorical question for you to make my overall point:

You say the DFX has 4 times the VDI resolution of the Equinox 800 (and 600). Tell me, which one of those detectors is used by more hunters right now?

Ok, but you say, "imagine an Equinox 800 with a VDI of 200 instead of 50." Sure, but that's not possible right now, either b/c MIQ can't do that or Minelab chooses not to do that.

I don't think people would have such a negative view of using the VDI IF it were far more reliable. To be far more reliable just takes a higher resolution as proved to me by even a $250 detector like the Simplex.

The DFX is still widely used even being 20+ year old technology. I'm sure there are a few DFX'rs on here. I know my buddy would not replace his DFX with any new detector for tot lots. And I can't blame him. I personally have seen that there isn't a detector on the market today any better at tot lots than the DFX because of the very reliable VDI system. That and it is as sensitive to small gold as any detector I have ever used.

You're probably right that the newer MIQ/SMF technology doesn't play well, yet, with a high VDI segment system. I do think if people showed more interest in a better more accurate VDI system, one of the manufacturers would start the ball rolling, and the others would soon follow.
 
I don't think people would have such a negative view of using the VDI IF it were far more reliable. To be far more reliable just takes a higher resolution as proved to me by even a $250 detector like the Simplex.

The DFX is still widely used even being 20+ year old technology. I'm sure there are a few DFX'rs on here. I know my buddy would not replace his DFX with any new detector for tot lots. And I can't blame him. I personally have seen that there isn't a detector on the market today any better at tot lots than the DFX because of the very reliable VDI system. That and it is as sensitive to small gold as any detector I have ever used.

You're probably right that the newer MIQ/SMF technology doesn't play well, yet, with a high VDI segment system. I do think if people showed more interest in a better more accurate VDI system, one of the manufacturers would start the ball rolling, and the others would soon follow.

Of course! But current tech doesn't allow for that. Just increasing the number of segements will NOT automatically make it more reliable.

For example, if you had a ruler made out of play dough and it had 1 inch increments, what good is increasing it to 1/16 inch increments? The play dough is so soft, increasing its resolution would be meaningless. Only until the ruler was made out of a harder material, like wood, plastic, metal, etc. would boosting its resolution do any good.

And yes, the Simplex has more VDI resolution, but it's not SMF. So its VDIs are less consistent (especially in mineralized soil) as MIQ's VDIs. So all those extra VDI increments aren't as useful as they might otherwise be.

You're acting like it's easily done or that manufacturers can do it, but just don't realize it. I don't think either situation applies here.
 
Of course! But current tech doesn't allow for that. Just increasing the number of segements will NOT automatically make it more reliable.

For example, if you had a ruler made out of play dough and it had 1 inch increments, what good is increasing it to 1/16 inch increments? The play dough is so soft, increasing its resolution would be meaningless. Only until the ruler was made out of a harder material, like wood, plastic, metal, etc. would boosting its resolution do any good.

And yes, the Simplex has more VDI resolution, but it's not SMF. So its VDIs are less consistent (especially in mineralized soil) as MIQ's VDIs. So all those extra VDI increments aren't as useful as they might otherwise be.

You're acting like it's easily done or that manufacturers can do it, but just don't realize it. I don't think either situation applies here.

Yes, a ruler with more segments would be more accurate as well.

In my experience, the simplex is very accurate compared to far more expensive machines. As I said I use them in tot lots where judging VDI accuracy is far easier because of less trash. A 18-21 on the Simplex will be a piece of foil 90% of the time, and I'm sure other Simplex users can weigh in. That same accuracy also applies out in that trashy park. On the NOX 800 that same 18-21 can be a far larger number of items because the VDI resolution is at least 1/2 of the Simplex.

I don't believe it will be easy, just that I think it is worth manufacturers working on. I believe the only reason they don't include it in the newer technology is that the majority of detectorists today have been swayed from caring about VDI accuracy. You seldom see anyone talk about how they wish the VDI was more accurate, so why would they spend any time trying to make it more accurate? We usually get what we ask for if there are enough voices behind the request.
 
Yes, a ruler with more segments would be more accurate as well.

In my experience, the simplex is very accurate compared to far more expensive machines. As I said I use them in tot lots where judging VDI accuracy is far easier because of less trash. A 18-21 on the Simplex will be a piece of foil 90% of the time, and I'm sure other Simplex users can weigh in. That same accuracy also applies out in that trashy park. On the NOX 800 that same 18-21 can be a far larger number of items because the VDI resolution is at least 1/2 of the Simplex.

I don't believe it will be easy, just that I think it is worth manufacturers working on. I believe the only reason they don't include it in the newer technology is that the majority of detectorists today have been swayed from caring about VDI accuracy. You seldom see anyone talk about how they wish the VDI was more accurate, so why would they spend any time trying to make it more accurate? We usually get what we ask for if there are enough voices behind the request.

Because it's so frickin' obvious that everyone wants that. The whole point of a metal detector is to accurately and precisely ID metal objects in the ground. Of course people want their machine to be as accurate and precise as possible. Some people enjoy the extra challenge (or are feeling nostalgic). But then there are older machines that can offer that reduced ability to ID targets in the ground. No manufacturer will deliberately nerf a machine unless it's to reduce it price or make it easier to use.

As for your Simplex versus Equinox 800 comparison, I'm sure Minelab could double the VDI, so that 18-21 target comes up as a 36-42 target. But would that help you or most other hunters? No, not really.
 
How deep are you on tot lots, maybe a couple of inches of sand or woodchip?
Look at it from another point, take a small low conductor coin at 8"+ in more mineralized pasture, the Nox will more likely call it correctly whereas the Simplex could quite easily call it iron, simply because the Nox can better filter the ground better vs the Simplex.
Extra ID numbers is only of use if the Detector can properly utilize them.

Yes, a ruler with more segments would be more accurate as well.

In my experience, the simplex is very accurate compared to far more expensive machines. As I said I use them in tot lots where judging VDI accuracy is far easier because of less trash. A 18-21 on the Simplex will be a piece of foil 90% of the time, and I'm sure other Simplex users can weigh in. That same accuracy also applies out in that trashy park. On the NOX 800 that same 18-21 can be a far larger number of items because the VDI resolution is at least 1/2 of the Simplex.

I don't believe it will be easy, just that I think it is worth manufacturers working on. I believe the only reason they don't include it in the newer technology is that the majority of detectorists today have been swayed from caring about VDI accuracy. You seldom see anyone talk about how they wish the VDI was more accurate, so why would they spend any time trying to make it more accurate? We usually get what we ask for if there are enough voices behind the request.
 
How deep are you on tot lots, maybe a couple of inches of sand or woodchip?
Look at it from another point, take a small low conductor coin at 8"+ in more mineralized pasture, the Nox will more likely call it correctly whereas the Simplex could quite easily call it iron, simply because the Nox can better filter the ground better vs the Simplex.
Extra ID numbers is only of use if the Detector can properly utilize them.

Good point bringing that up. Those grounds are basically like doing air tests.

When I hunted tot lots with a Vanquish 540 or AT Max, their VDIs were pretty close to the same in terms of accuracy and consistency.
 
Of course! But current tech doesn't allow for that. Just increasing the number of segements will NOT automatically make it more reliable.

For example, if you had a ruler made out of play dough and it had 1 inch increments, what good is increasing it to 1/16 inch increments? The play dough is so soft, increasing its resolution would be meaningless. Only until the ruler was made out of a harder material, like wood, plastic, metal, etc. would boosting its resolution do any good.

And yes, the Simplex has more VDI resolution, but it's not SMF. So its VDIs are less consistent (especially in mineralized soil) as MIQ's VDIs. So all those extra VDI increments aren't as useful as they might otherwise be.

You're acting like it's easily done or that manufacturers can do it, but just don't realize it. I don't think either situation applies here.

When resolution step size becomes smaller than the noise level, more resolution is useless. There are several noise component for the designer to worry about, for example: The noise created by the pre-amp; the noise created by instantaneous changes in the ground matrix as you sweep the coil, compared to the time constant of the ground filter section (think of it as ground tracking speed); inherent background noise present in the environment (e.g. rural versus urban detecting). These are but a few issues the designer has to deal with in deciding on detector resolution.
 
Because it's so frickin' obvious that everyone wants that. The whole point of a metal detector is to accurately and precisely ID metal objects in the ground. Of course people want their machine to be as accurate and precise as possible. Some people enjoy the extra challenge (or are feeling nostalgic). But then there are older machines that can offer that reduced ability to ID targets in the ground. No manufacturer will deliberately nerf a machine unless it's to reduce it price or make it easier to use.

As for your Simplex versus Equinox 800 comparison, I'm sure Minelab could double the VDI, so that 18-21 target comes up as a 36-42 target. But would that help you or most other hunters? No, not really.

I would think it obvious, but how often do we see "I don't even look at my detector, I just listen." Or "You can't trust the VDI". Both, in my opinion, are due to reduced VDI segments.

Adding more VDI segments would not change the VDI numbers from 18-21 to 36-42, a higher resolution would break 18-21 from 4 VDIDs to 8 for example of the difference in 50 vs 100 VDI segments. So what you'd have is instead of foil reading from 14-25 with 50 VDIs, you get foil reading 18-21, and leaving 14-17 and 21-25 for non-foil like gold earrings, small pendants, etc. It would make it so you can avoid the most likely trash or 18-21 and still recover the better odds of gold of the 14-17 and 22-25.

A compressed VDI, for example, may have 5 segments all in 1 VDI number. Say a VDI on your screen that says 20 may also contain the VDIs 20, 21 ,22, 23, 24, 25 on a higher resolution system. So any object in that range 20-25 would all read a VDI of 20. That is a lower resolution. A higher resolution would have VDIs 20 through 25 as single individual VDIs giving more possible identification in the foil/gold range. Sure not 100%, but as I have experienced, I'd give it a good 90%. that same 90% would also apply in trashy parks.

Yes it does make obvious sense and that is why I'm surprised VDI resolution only continues to go down and not up. I believe some of the reason is that no one expresses much concern about it. I see tons of videos on Youtube comparing tones, but in many of them, you can't even see the VDI because they don't care about it. In my case, the VDI is just as important as the tones.
 
What you're asking for makes sense. But I think the simple answer is that it's far easier said than done.

Increasing VDIs doesn't mean much (except from maybe a marketing perspective) unless the underlying technology can make use of it.

While an oversimplifcation, here's a rhetorical question for you to make my overall point:

You say the DFX has 4 times the VDI resolution of the Equinox 800 (and 600). Tell me, which one of those detectors is used by more hunters right now?

Ok, but you say, "imagine an Equinox 800 with a VDI of 200 instead of 50." Sure, but that's not possible right now, either b/c MIQ can't do that or Minelab chooses not to do that.

My thinking is that the numbers are jumpy at 40 segments, and the number of segments is smaller because it is using mutli-freq.tech and does not have high resolution. IF there were 100 or even 200 segments it would only demonstrate more, the lack of ability to be precise with this tech at this time. The number variation on each sweep would the much greater using more segments.

I was hoping that Nokta was going to solve this puzzle, but demos that I've watched don't indicate any increase in stability of segment id. over Minelab
 
My thinking is that the numbers are jumpy at 40 segments, and the number of segments is smaller because it is using mutli-freq.tech and does not have high resolution. IF there were 100 or even 200 segments it would only demonstrate more, the lack of ability to be precise with this tech at this time. The number variation on each sweep would the much greater using more segments.

That’s what I was thinking too. We’ll see how the Deus does.


Sent from my iPhone using Tapatalk
 
I would think it obvious, but how often do we see "I don't even look at my detector, I just listen." Or "You can't trust the VDI". Both, in my opinion, are due to reduced VDI segments.

Adding more VDI segments would not change the VDI numbers from 18-21 to 36-42, a higher resolution would break 18-21 from 4 VDIDs to 8 for example of the difference in 50 vs 100 VDI segments. So what you'd have is instead of foil reading from 14-25 with 50 VDIs, you get foil reading 18-21, and leaving 14-17 and 21-25 for non-foil like gold earrings, small pendants, etc. It would make it so you can avoid the most likely trash or 18-21 and still recover the better odds of gold of the 14-17 and 22-25.

A compressed VDI, for example, may have 5 segments all in 1 VDI number. Say a VDI on your screen that says 20 may also contain the VDIs 20, 21 ,22, 23, 24, 25 on a higher resolution system. So any object in that range 20-25 would all read a VDI of 20. That is a lower resolution. A higher resolution would have VDIs 20 through 25 as single individual VDIs giving more possible identification in the foil/gold range. Sure not 100%, but as I have experienced, I'd give it a good 90%. that same 90% would also apply in trashy parks.

Yes it does make obvious sense and that is why I'm surprised VDI resolution only continues to go down and not up. I believe some of the reason is that no one expresses much concern about it. I see tons of videos on Youtube comparing tones, but in many of them, you can't even see the VDI because they don't care about it. In my case, the VDI is just as important as the tones.

Just because many hunters would rather listen than look at their screens doesn't mean they don't want more accurate VDIs; you're getting the chicken-and-egg thing mixed up here. While I'm sure there are some people that love using their ear, I'm pretty sure that the majority of ear hunters would gladly stop using sound for making the "dig or don't dig" decision if the could trust their VDIs more.

You keep mentioning more VDIs, but you aren't addressing the fact that unless the technology producing the VDIs also improves, more VDIs are practically meaningless (outside from a marketing perspective).
 
My thinking is that the numbers are jumpy at 40 segments, and the number of segments is smaller because it is using mutli-freq.tech and does not have high resolution. IF there were 100 or even 200 segments it would only demonstrate more, the lack of ability to be precise with this tech at this time. The number variation on each sweep would the much greater using more segments.

I was hoping that Nokta was going to solve this puzzle, but demos that I've watched don't indicate any increase in stability of segment id. over Minelab

I agree with you there. If the MIQ scale were -19 to 80, I think we'd just see wider ranges for certain targets, ie a nickel would be anywhere from 10-16 instead of 12-14, depending on settings and hunting conditions.
 
When resolution step size becomes smaller than the noise level, more resolution is useless. There are several noise component for the designer to worry about, for example: The noise created by the pre-amp; the noise created by instantaneous changes in the ground matrix as you sweep the coil, compared to the time constant of the ground filter section (think of it as ground tracking speed); inherent background noise present in the environment (e.g. rural versus urban detecting). These are but a few issues the designer has to deal with in deciding on detector resolution.

MIQ tech has always struggled with EMI...maybe you're on to something...:yes:
 
I agree with you there. If the MIQ scale were -19 to 80, I think we'd just see wider ranges for certain targets, ie a nickel would be anywhere from 10-16 instead of 12-14, depending on settings and hunting conditions.

Here's a thought...the TID is just a digital representation of an analog signal, and as such, subject to rounding error.

Let's say the "3 bounce" 12-14 signal is generated by analog values from 12.4 (rounded down to 12) and 13.5 (rounded up to 14.) If the TID range is doubled, and chopped up into twice as many segments, for simplicity say the 12-14 range expands out to cover 24-28. The above analog signal expands to cover 24.8 to 27 which rounds the hits to 25-27, still just 3 segments. This is a "best case" scenario, but there's no harm in spreading it out more. If you still get a solid 13 (or 26), awesome, if it bounces 3-6 digits, as long as you know that's normal for your machine, all's well.

Fewer segments may make the signal SEEM "more stable" or "more repeatable" but may be reducing the overall usefulness of the underlying data the TID is representing.

It probably comes down to how repeatable/precise the machine really is, and whether marketing thinks more TID numbers will sell the machine, or more "stable" TID numbers on the display will sell the machine.
 
I agree with you there. If the MIQ scale were -19 to 80, I think we'd just see wider ranges for certain targets, ie a nickel would be anywhere from 10-16 instead of 12-14, depending on settings and hunting conditions.

I think may are not getting the right idea of the expanded VDI. I have experienced it myself, between using the NOX 800 vs the Simplex in tot lots. Think along the lines of the VDI scale is like a ruler. The more increments/VDIs on a ruler, the more accurate you can be. It would not move the VDI of a target. 6" remains 6" on the ruler whether the ruler is in 1" segments or 1/32 segments. The difference is more accurate measurements.

Yes, the tones would also change. Look at the difference between the E-Trac, 1500 VDI segments, and the CTX3030 with 50. People talk about missing the "flutty" sound of the silver magnate of the E-Trac. That is what the CTX3030 misses to do the same thing. More tones also make target ID more accurate.

I remember the day detectors first came out with nothing but a needle pointing to good/bad and we were thrilled.

As a programmer, I don't see where adding more VDIs on the new SMF technology, since it seems it is more software than hardware driven. I think that is why we haven't seen any increase in depth in the last 20 years. The hardware has had little improvement, faster processors have allowed for more complex firmware/software to allow multiple frequence technology to improve. After all, multiple frequencies have been around for a long time, it has just gotten a lot more complex.
 
What you're asking for makes sense. But I think the simple answer is that it's far easier said than done.

Increasing VDIs doesn't mean much (except from maybe a marketing perspective) unless the underlying technology can make use of it.

While an oversimplifcation, here's a rhetorical question for you to make my overall point:

You say the DFX has 4 times the VDI resolution of the Equinox 800 (and 600). Tell me, which one of those detectors is used by more hunters right now?

Ok, but you say, "imagine an Equinox 800 with a VDI of 200 instead of 50." Sure, but that's not possible right now, either b/c MIQ can't do that or Minelab chooses not to do that.

So what you are saying is White's with the DFX had more intelligence than Minelab with the Equinox. I still use the DFX and find great targets. I no longer have the Equinox 800 that I had, to many EMI problems and terrible ID numbers.
 
So what you are saying is White's with the DFX had more intelligence than Minelab with the Equinox. I still use the DFX and find great targets. I no longer have the Equinox 800 that I had, to many EMI problems and terrible ID numbers.

I certainly believe the DFX has one of the best, if not the best, VDI systems on the market. Not a depth monster, but if target ID is important it would be tough to find anything more accurate than the DFX.
 
So how many have seen VDI numbers change as you dig your hole? Plenty of times I have seen a jumpy VDI of say 17 -18 firm up to a nice solid 20 after taking out a shovel full of soil?

What about targets that give different VDI at different sites? I have one sites were pistol balls come up at a consistent 16 but on other sites they are a 19?

This is with my Nox but have seen similar variations with my Racer 2 also...

I would love to be able to hunt primarily VDI only, but I know that I will miss a lot of good targets if I do, partly because our targets are so variable...

it *might* be a more practical approach for you folks who coin shoot for modern (post 1800) coins as the targets themselves are more consistent, but for people interested in older coins and relics, I can't see it working....
 
Back
Top Bottom