E-Targets for F-Class

[GSS]

Is this thread shut down? Someone (not me) wants to post something but gets the error "The content you want to post is not allowed"

???

It's a genuine and regular post with nothing contentious. I would have thought...

That's funny
Seeing how you have already seen it and it's been previewed by you maybe you will have to post it for them.

Is it a friend / ally from over yonder?

 

[MichaelJ]

Yeah, I'm new here... I'm the one trying to post some thoughts -- hopefully just to shed some light on where I got years ago with E target stuff.

 

[GSS]

Yeah, I'm new here... I'm the one trying to post some thoughts -- hopefully just to shed some light on where I got years ago with E target stuff.

Looks like you are from Australia and you just joined yesterday... Interesting
Geoff bringing in troops / reinforcements.. ?
;-)

 

[GeoffR]

That's funny
Seeing how you have already seen it and it's been previewed by you maybe you will have to post it for them.

Is it a friend / ally from over yonder?

George, it's my son. He has something I think/hope will be constructive to say and hopefully take some heat out of this debate about how shot points of impact are calculated. At least I hope so!

Geoff.

 

[GSS]

George, it's my son. He has something I think/hope will be constructive to say and hopefully take some heat out of this debate about how shot points of impact are calculated. At least I hope so!

Geoff.

Thank You, the relation disclosure is appreciated.
I though the heat in the debate had died down but constructive discussion is good, I guess.

I would imagine;
Some would likely better / rather spend their time cranking out product to catch up on order backlog but the debate will always be there for some as well.

 

[GeoffR]

Thank You, the relation disclosure is appreciated.
I though the heat in the debate had died down but constructive discussion is good, I guess.

I would imagine;
Some would likely better / rather spend their time cranking out product to catch up on order backlog but the debate will always be there for some as well.

Whatever.

I have had no input to what he says. He is his own man. He's just trying to figure out what it is in his text that seems to make the filter barf. I presume it is some kind of filter.

 

[MichaelJ]

Hi all. Yep, I'm GeoffR's son. I tried to post before, but I've decided to just dot-point the fun stuff.

- I'm disappointed at how toxic this debate seems to look. I may be wrong... But it's in everyone's interest for users to challenge target manufacturers to get it right. Verifying results is a great start; I've heard of cases on various targets where the results are anomalous, and while we won't eliminate that, we can certainly minimize it. I will point out that I think GeoffR and adamjmac are probably arguing towards different goalposts, which isn't very helpful.

- I developed a solution for taking 3+ time measurements and returning a shot location to the user. It seems to do the job quite well, at least from what I've been hearing.
- The math is quite flexible. If you have a sensor field that makes a perfect square, and you receive a shot making timer values all 0, it'll place the shot dead center without resorting to special cases. Likewise for the horizontal and vertical axes where you have two 0's and two equal numbers from the timers.
- In principle, you can place a number of sensors anywhere on the target plane and, as long as you know where they are in a consistent frame of reference, you'll get a meaningful shot location.

- I've never evaluated open-target designs before, so I can't claim that the solution I've developed translates well to the physics challenges they face. Closed-target designs appear to dodge the need to resolve the properties of the shockwave cone, though, which is nice.
- There's information out there indicating that the open-target design faces a more complex problem to solve, which would necessitate a different design.
- The Mach angles of the shock waves for different shots will likely alter the timer measurements (at least on open-target designs) because the wavefronts that hit the sensors originated in front of the target, not in the target plane. The Mach angle, of course, is a function of projectile velocity.
- It therefore follows that you could use different projectiles, loading configurations, calibres, etc. to hit different parts of the target but produce similar timer measurements, as in adamjmac's example above. Using only the timer data in that situation wouldn't be acceptable.

Finally, "multilateration" is really just an approach to the problem of locating something in space. It's not the E-target system, and honestly, you could probably use a different method anyway and get a good result. I chose it because though it took some study at the time to understand it, it was very effective at simplifying the input requirements. It doesn't require angles, but I can be creative: I'd be interested in seeing if the Mach angle improves the solution. However, I'd assert that System Design is paramount - you can't use multilateration if your design doesn't let you.

 

[Toejoe12]

He's just trying to figure out what it is in his text that seems to make the filter barf.

I think the forum just implemented "multilateration"

 

[Down South]

When I started this thread I figured it would last a while but damn, this thing is like the energizer bunny.......

 

[adamjmac]

I've never evaluated open-target designs before, so I can't claim that the solution I've developed translates well to the physics challenges they face.

This is the point. Geoff has claimed some fundamental insight about open sensor targets, not closed. Of course you can solve a closed system with 3 sensors and no velocity. 4 sensors gives you a measure of accuracy. Everything he said is correct, applied to a closed system. There's no denying that. I respect your work developing an algorithm for closed sensor systems, but it absolutely does not translate to open.

Open sensor targets are a 6 degree of freedom problem, requiring 6 sensors minimum, without making assumptions. Geoff is attacking my personal credibility on the basis of a principle that to someone familiar with the open sensor problem is akin to a perpetual motion machine.

Proof by counter example. Timings of 0,0,0,0 does not necessary mean the shot was fired in the center. The shot could have landed to the left with a right to left yaw angle.

I think we can all let this go now. I will not be writing any further on this subject.

 

[spclark]

I though the heat in the debate had died down but constructive discussion is good, I guess.

I certainly hope so [italics mine above] otherwise this Forum is doomed to follow others that are no longer worth bothering with.

 

[sdean]

I certainly hope so [italics mine above] otherwise this Forum is doomed to follow others that are no longer worth bothering with.

You can say the same thing only so many ways. It then gets old and boring. This has been the same discussion by the same people for two years. I don't think a closed system has been sold here since then

 

[GSS]

I think we can all let this go now. I will not be writing any further on this subject.

You are a smart man Adam,
You sold over 1000 units in your 1st year and you have customers lining up to buy your product.
Probably much better use of time satisfying customer orders / backlog.

Huge Thanks for delivering a great product at an affordable cost.
G-

 

[GeoffR]

This is the point. Geoff has claimed some fundamental insight about open sensor targets, not closed. Of course you can solve a closed system with 3 sensors and no velocity. 4 sensors gives you a measure of accuracy. Everything he said is correct, applied to a closed system. There's no denying that. I respect your work developing an algorithm for closed sensor systems, but it absolutely does not translate to open.

Open sensor targets are a 6 degree of freedom problem, requiring 6 sensors minimum, without making assumptions.

Geoff is attacking my personal credibility on the basis of a principle that to someone familiar with the open sensor problem is akin to a perpetual motion machine.

Yes. And for that I apologise. I think we were coming at it from different angles and talking at cross purposes.

Proof by counter example. Timings of 0,0,0,0 does not necessary mean the shot was fired in the center. The shot could have landed to the left with a right to left yaw angle.

I think we can all let this go now. I will not be writing any further on this subject.

Yes, I think we can wrap it up.

 

[243Mendoza]

The short (and possibly glib) answer is because it doesn't [require Vt]. It is entirely time based - the times being the TDOA's (Time Differences of Arrival) from which we create isochronic curves (hyperbolas) and exploit their properties. A four sensor system produces twelve usable isochrones from which we derive four triangles (triplets), the apexes (intersections) of which should coincide (and generally do). The amplitude of the resulting shape (polygon) indicates the accumulated errors (and area of uncertainty). Ideally, all intersections should be at the same point (no amplitude meaning zero errors). There are always errors but this number provides us with a indication of the current quality of measurements by that particular target.

We require speed of sound in order to convert time to distance.

To determine the speed of sound we require a temperature input. Temperature is pretty easy to reliably determine and calibrate.

So that's all we need: four (or more) TDOA's (one per sensor, with at least one being 0 - zero) and a temperature.

We (my son and I) are thinking about posting something a bit more comprehensive if indications are that anyone is really interested.

 

[243Mendoza]

Using the speed of sound is a common error. The shock wave is NOT a Mach wave and travels at a different speed and that includes it orthogonal components. The Mach wave is generated by a infinitesimal disturbance which does not include the shock wave.

 

[MichaelJ]

You can say the same thing only so many ways. It then gets old and boring. This has been the same discussion by the same people for two years. I don't think a closed system has been sold here since then

Indeed. Closed-target systems traditionally have come as a holistic package of stuff, too. My understanding is SMT and ShotMarker only provide some key components to attach to existing targets. On face value, it's difficult to compete on cost, and the flexibility to choose the other components that suit you is understandably attractive.

Also I should caveat my above response with the fact that I haven't really given any of this much thought since about 2012 or so.

Obviously since then the community knowledge of the physics and mathematics required for acoustic E-targets to work has matured somewhat. The more important discussion is about standards, calibration methods, and independent result verification methods, which should highlight the practical issues with the different kinds of E-targets and hold manufacturers to account on the accuracy of their systems.

I understand that conversation is ongoing in the background, and hopefully that helps us push towards having targets that are actually fit for competition purpose. Targets that shooters can be reasonably confident in.

For example, say you observe that your group size is smaller at 1200 yds than at 1000 yds. Can you compare the system results to physical shot positions on a target and verify that, despite it probably not making sense, the results reflect reality? If not, the manufacturer should want to know.

If it really is a quality product, it'll stand up to constant and thorough scrutiny. Manufacturers should welcome that scrutiny.

 

[Crow]

...hopefully that helps us push towards having targets that are actually fit for competition purpose. Targets that shooters can be reasonably confident in.

Great post. Especially liked the section above...

To me, the standard that e-targets have to meet is: On average, are they as accurate as manual pit service? Determining just how high that bar really is will be an interesting exercise.

 

[Doug Beach]

Since we’re at 27 pages...
Caveat: I like e-targets, at least for practice and club matches. I’ve never shot a “big” match on e’s.
I’ve played this game for 24 years, this month. I’m a High Master, Nat. Record holder, blah, blah, blah.
I don’t shoot F-class, only sling, so ignore me if you must.
Manual targets have lots of faults, as has been discussed ad nauseam. These faults, however, are not compounded by distance, or terminal velocity. I would guess that a typical human puller locates and scores shots with more accuracy than an e-target (they are, after all, putting the spotter in the actual bullet hole).
E targets, on the other hand, don’t gossip with the target next door, don’t cheat, don’t care who’s winning, don’t adjust their speed, or accuracy, for emotional reasons, etc.
If you think that you have full control over that last 1/4” at 600 + yards, you have something to teach me, for sure. This game is full of luck, and e-targets remove one kind of luck, and insert a different kind.
Go shoot. If you’re really good, and even slightly lucky, you’ll win some matches. With your massive haul from winning, and a few bucks, you can buy yourself a beer or three.

 

[Bindi2]

I would guess that a typical human puller locates and scores shots with more accuracy than an e-target (they are, after all, putting the spotter in the actual bullet hole).
E targets, on the other hand, don’t gossip with the target next door, don’t cheat, don’t care who’s winning, don’t adjust their speed, or accuracy, for emotional reasons, etc.

If the target is not repeatable blame the maintenance crew. The ET can only work accurately when the human factor does its job. I will back the ET.