Page 2 of 2

Posted: Sat Mar 18, 2006 3:26 am
by David Levene
Anonymous wrote:Noptel, Scatt, and Rika all are capable of detecting angular (rotation about an axis) error. They are not capable of detecting translation (straight line movement along an axis) error.
JP, surely all of the systems mentioned are only capable of detecting an error, they cannot tell what caused the error. When training an event at the wrong distance however they all extrapolate the error as if it was an angular one.
Anonymous wrote:Body sway will have a large translation component and result in a small (yet easily detectable) angular error.
But couldn't that small detected angular error also have been caused by a small angular error?

Posted: Sat Mar 18, 2006 8:45 am
by Richard H
That's why it's best to combine the data from the trainer with actual observation of the shooter from another person and if possible input from the shooter. If another person is not available a video camera could be used for later analysis.

Posted: Sat Mar 18, 2006 9:56 am
by Fred Mannis
Patrick,
Thank you for posting your comments on use of the RIKA. I find the lack of comprehensive manuals on the use of these electronic trainers to be very frustrating. Despite the problems of data interpretation discussed in this thread, I think it would be helpful - for shooters as well as coaches - to have the insights that are posted here from time to time compiled into a single location. The trainer manufacturers are the obvious candidates for preparing manuals for their customers, but they seem to have no interest. Any takers?

Fred

Posted: Sat Mar 18, 2006 10:55 am
by Richard H
Fred Mannis wrote:Patrick,
Thank you for posting your comments on use of the RIKA. I find the lack of comprehensive manuals on the use of these electronic trainers to be very frustrating. Despite the problems of data interpretation discussed in this thread, I think it would be helpful - for shooters as well as coaches - to have the insights that are posted here from time to time compiled into a single location. The trainer manufacturers are the obvious candidates for preparing manuals for their customers, but they seem to have no interest. Any takers?

Fred
Isn't the rule the one who comes up with the idea does it. The best info I've found so far has been on Pilks site where it at least explains what you are looking at.

I've gotten the most working with Patrick, and looking at actual traces.

To do a good job you'd really have to work with some very good shooters who are capable of telling you what is going on and compare what they tell you to what you see on the screen.

So it's not just as easy as writing an instruction manual you'd have to involve top level pistol and rifle shooters which I think would probably be difficult. Just look at the draught of good books on target shooting the newest is probably The Ways of the Rifle (very good) but for pistol there is really nothing that is currently in print. So if there are any great coaches out there we need some new manuals.

Re: electronic trainers

Posted: Sat Mar 18, 2006 4:42 pm
by JeroenH
Patrick Haynes wrote: Here's my understanding: The Rika's emitter is placed on the pistol and you identify the location to the Rika (sides or bottom.) The Rika sensors on the receiver must plot this in relation to the emitter.
What they all do initially is register the x and y coordinate of the point of aim throughout the aiming & shooting process. From that, you get various plots and statistics.
Still, Rika is (to my knowledge) the only one also showing cant angle ('Verkantung', shown together with the heart rate I think).
Suppose you have front sight, rear sight and target perfectly aligned, and execute perfect shots with only 1 thing changing from shot to shot: the cant angle (i.e., rotating the rifle or pistol around the line of sight). Then your point of impact will vary from shot to shot. (to read more about in which direction and how far, see http://home.wanadoo.nl/jhogema/skeetn/b ... allist.htm).

I'm curious from a technical point of view how they do that.
And I'm curious if Rika users are using this feature. E.g. to see if a spirit level indeed helps to get a more consistent cant angle.


Jeroen

Posted: Sun Mar 19, 2006 10:59 am
by Ed Hall
Great link in the previous posting. I also have a much smaller article on the effects of cant (primarily with pistols) at:

http://www.starreloaders.com/edhall/12PPC12.html

The problem with the Rika is that the effects of cant are backwards! That's right, backwards. In practice, a CCW cant (from the shooter's perspective) will produce a hit to the low left if all else is consistent, but the Rika will show the hit to the right. The reason is the way the Rika picks up the signals and the relationship of the individual components.

First, the Rika works by providing two signals from the transmitter (target holder) via two LEDs protruding from the front of the base. These signals are picked up by the receiving unit most commonly placed under the barrel of the firearm. The pickup device within the receiver then translates those two signals to provide the coordinates for the resulting data stream. The relative positioning of all the components leaves room for error if any cant variations exist.

When you are in the "picture perfect world" everything is in line vertically: The transmitter/target holder is level and straight, the gun is level and straight and the receiver is directly below the bore, also level. As long as this orientation is maintained, all is well. But add cant and things change.

Let's start with everything perfectly set up and providing the solid 10.9 hits when all is vertical. Now let's tip the gun CCW (from the shooter's perspective) and see what happens when the sights are still dead in the center of the aiming area. If you notice the motion of the receiver, it has shifted to the right in reference to the transmitter. What does this tell the software? It tells the s/w that the gun is now pointed to the right of center and will show a hit to the right. This is in the wrong direction for real life cant effects.

The error described above is really only an issue if you're unaware of it and have a varying cant, especially if your cant starts out at a particular angle and then changes through the session. The results on the computer might possibly be an unexplained stringing of shots out to the side. My first recommendation when dealing with cant is to work hard to eliminate variances and then consider shot placement. A cant of a few degrees is not a real factor, IF THAT CANT IS ALWAYS THE SAME. For the Rika, that means establishing the cant during the sighting portion and then working hard to have 0.0 degrees during the shot series.

Take Care,
Ed Hall
http://www.airforceshooting.org/
http://www.starreloaders.com/edhall/

Posted: Sun Mar 19, 2006 12:27 pm
by Steve Swartz
Ed:

Two additional things to consider with the Rika cant model. First, cant is set to zero when you sight in . . . all cant numbers after that are in relation to this arbitrary zero setting (as you suggest).

Second, the computer can tell the difference between rotational and horizontal displacement. It sees the two sending points in both x and y axes; the effects of cant (tipping the top of the gun to the left ccw) and a rightward shift (swinging the receiver unit to the right) as separate movements.

So a CCW cant is picked up as a rotational displacement *to the left* (correctly) and the distance from centerline of receiver to centerline of muzzle is used to determine teh true offset. While the receiver is swung to the right when canting CCW, the software "knows" that the muzzle has actually moved left.

Hmm . . . the only way to really verify this is to use a laser-pointer type sighting device and/or a good vice with canting ability.

Steve Swartz

Posted: Sun Mar 19, 2006 4:19 pm
by Ed Hall
Hi Steve,

Thanks for the comments. I disagree with a small portion, but my Rika is packed from a trip I was just on, so it might be a little while before I set it up again. The newer s/w might intrepret it better, but the older s/w I used when studying cant on the Rika didn't interpret it well.
Two additional things to consider with the Rika cant model. First, cant is set to zero when you sight in . . . all cant numbers after that are in relation to this arbitrary zero setting (as you suggest).
Correct. After sight in, zero cant represents whatever rotational orientation existed during the sight in. I haven't experimented to see if this is actually the last sight in shot, an average, if there was a difference, or if it is based on something else.
Second, the computer can tell the difference between rotational and horizontal displacement. It sees the two sending points in both x and y axes; the effects of cant (tipping the top of the gun to the left ccw) and a rightward shift (swinging the receiver unit to the right) as separate movements.
Agree that there is enough sensor data to achieve this, but unsure if it is actually incorporated into the calculations. Again, all my testing was with a previous version of the s/w (3.61?).
So a CCW cant is picked up as a rotational displacement *to the left* (correctly) and the distance from centerline of receiver to centerline of muzzle is used to determine teh true offset. While the receiver is swung to the right when canting CCW, the software "knows" that the muzzle has actually moved left.
I'm not sure I agree with this. The rotation is about the sight line and is thereby moving the muzzle CCW with the same arc as the receiver. If we assume the muzzle to be centered between the receiver and the sight line, the described arc will have half the radius of the receiver, but will still rotate about the sight line in the same direction. Now, if we extend the bore line out to the target, it will describe an arc on the opposite side of the line of sight, for the same number of degrees, but I'm not sure trajectory is factored into the s/w.

When I did my testing a few years ago, the s/w tracked the cant correctly, but didn't register the hits correctly. A CCW cant registered as a hit to the right.

A rough test for anyone interested in providing some feedback is to sight in with a CW cant (don't overdo - 5-10 degrees is enough) and then fire a group with a proper cant orientation. See where the shots register. For my earlier testing the above setup registered hits to the right, while in actual live fire, the hits print to the left.

I will do some more testing with the current s/w when I get my system back up to see if it has a better tracking of the results, but from my testing of prior versions, it was backwards.

Take Care,
Ed Hall
http://www.airforceshooting.org/
http://www.starreloaders.com/edhall/

Posted: Sun Mar 19, 2006 6:34 pm
by Steve Swartz
Ed:

Yes, easily tested empirically and I don't know whether it reads cant in the opposite direction or not. It would just seem a somewhat foolish error, when all the data is available to read it correctly.

Here's why.

The sight line is completely irrelevant to the Rika. All the Rika knows (or cares about, frankly) is that so many millimeters above the sensor head is the muzzle (and both are in line with each other).

The position and direction of the sensor unit (vertical and horizontal displacement, and angular displacement) is read directly off the sender-receiver integration. This tells the software exaclty where the sensor unit is pointing in relation to the sending unit.

A single calculation translates the position of the sensor head (read directly) into the position of the muzzle; i.e. "take where the sensor unit is pointed and and go up X mm in the direction of cant."

It would be, well, extra work actually to get the software to translate the true offset (programmed in by the user settings) into an opposite direction.

Interesting note: the math either uses the sin (sensor directly in line with the sights-muzzle axis) or a cos (sensor 90 degrees to the side) function to make the translation. The oberved phenomenon (shots under cant recorded displaced to the right for a CCW cant) might actually be because the software was told that the sensor was mounted to the side?

Or

The software reversed the settings for in-line vs. 90 degree mounting?

Wonder if the good folks at Rika just made a straightforward math error; or if your software was set up to select 90 degree offset mounting?

Maybe I should check mine out . . . never really noticed, since my cant is generally within 1 degree . . . but this could explain some of the deviations true hole-predicted hole I have noticed on occasion!

(here we go again, demonstrating Mike's comment about Tika/Scatt/Noptel systems being more of a distraction than an aid!)

Steve

Posted: Mon Mar 20, 2006 8:45 pm
by Ed Hall
Hi Steve,

I took awhile because I wanted to unpack and verify the new s/w. Unfortunately, version 3.8T still registers hits backwards for cant error. I didn't do any math, but it also appears to be more pronounced than it should be. IOW, a CCW cant of 10 degrees seems to move the hit further than it should from center, as well as in the wrong direction. I would like some others to verify my findings please...

I haven't found any settings (missing something?) for use of the receiver anywhere but under the muzzle. I have tried it on top of a scope, but it read the normal motion mirror-imaged. The reason I referred to the line of sight is that this is the axis for the cant. The shooter would still be centered visually with sights in aiming area.

As to being a distraction, I actually seldom use the Rika any more. When I started out I found that the Rika taught me how to better "see" my dry fire. Once I learned that, I used the Rika less. I do still use it to study how pure my trigger is though, as I mentioned on TT elsewhere.

I hope someone else can verify my tests above.

Take Care,
Ed Hall
http://www.airforceshooting.org/
http://www.starreloaders.com/edhall/