A Closer Look At The AimTrak (with video goodness)

[PsychoAU]

So what was the final verdict on what the best settings were in regard to off-screen reloading?  I have only tried Lethal Enforcers and Area 51 so far, but Area 51 works fine for off-screen reloading, but LE does not.

I have it set for the trigger to be gamepad button 1 for both on and off screen and the second button is gamepad 2 for both on and off screen.

Is there a good general setting that will work for most games?  I think we should compile a list of which games need which settings to work.  I know that some behave differently than others.

[Ginsu Victim]

Off-screen in LE works. I had my trigger set to the same button for the on-screen and off-screen. Worked fine.

[DJ_Izumi]

Are you sure the LASER was accurate in this test?

The laser is mounted BELOW the optical sensor, so for the laser point to hit the center where the optical sensor is aiming, the laser has to be angled upward to hit the same point.  If you stepped farther back, instead of hitting the same point it would hit higher.  If closer, it would hit lower.  The laser itself would need to be readjusted every time you chance stance.  With light guns with laser sights, I've found that matching the laser to 'iron sight' of the gun, this match even changes as you reposition, like crouch or step back.

But all this testing at different angles seems silly to me anyway.  I find that even a Guncon will become inaccurate if you shoot from a different position than that you calibrated it at.  You should be standing or sitting in your playing position and calibrating for that position.

[RandyT]

Are you sure the LASER was accurate in this test?

The laser is mounted BELOW the optical sensor, so for the laser point to hit the center where the optical sensor is aiming, the laser has to be angled upward to hit the same point.  If you stepped farther back, instead of hitting the same point it would hit higher.  If closer, it would hit lower.  The laser itself would need to be readjusted every time you chance stance.  With light guns with laser sights, I've found that matching the laser to 'iron sight' of the gun, this match even changes as you reposition, like crouch or step back.

There are a couple of problems with what you are saying here.  The first is that even if the laser was mounted so that it was emanating 1 inch below the "iron sight", plus or minus 1 foot would yield an error of plus or minus 1/4".  And the error would be very uniform and in a very specific direction based on the angle created by the intersection of the laser and the sight.  This is not what is happening in the video.

The other thing is that this type of system is 100% relative to the IR markers, not specifically the "iron sights".  In this case, the laser becomes the sight.  So as long as the laser beam is parallel to the PCB, the system should perform in exactly the same manner as if one were using the hard sight which is aligned to the barrel (provided that the PCB is mounted parallel to the barrel as well.)  The calibration routines should make the necessary corrective offsets.

But all this testing at different angles seems silly to me anyway.

The reason to look at this makes sense, as folks are concerned about friends or family members of different heights being able to use the gun without recalibrating.  For me, it's not really important, as I will likely be the only one to use it.

I find that even a Guncon will become inaccurate if you shoot from a different position than that you calibrated it at.  You should be standing or sitting in your playing position and calibrating for that position.

This is true, but to a smaller degree over the same amount of deviation.  

RandyT

[huygens]

btw, has anyone tried the wide angle lense from a door peephole trick yet?

I was thinking about this too. I think a wide angle lens would mess up accuracy if the firmware wasn't changed. A wide angle lens distorts the camera image. You'd have to account for this distortion to get accurate postional information. It seems like it would make it possible to stand closer to the screen though.

[Ragtag83]

Long time reader first time poster...always wanted to say that. First off, you guys are great and this is such a wonderful forum. Thanks for sharing all the great info I've been able to take away. Keeps me entertained into the wee hours of the night. Second off, I've got a question regarding the extra pins for extra buttons on the AimTrak. I am wiring up GunCon 2s and can get the trigger and aux working right off the bat but the P1-P5 connections are being a pain. I got P3-P5 to work on my first gun and none of them on the second gun.  I have checked an rechecked my wiring and it's starting to drive me a little nuts. Anyone else having any issues with them? Any info or insight would be appreciated 

[Lilwolf]

I never thought about it, but this is what aljupy did on the first page (but with a different gun, and a different wide angle lense).  But since we have to hack these into our own gun, it seemed perfect for it. 

I've also been looking around for the eye toy wide angle lense.  It seems like they only came out in with one game (fitness game?).  And they are going for 30+ bucks each on ebay.  If it was perfect, I would consider getting one to try...  But was hoping for another source.

Anyone know of a good source for wide angle lenses that might work?

btw, has anyone tried the wide angle lense from a door peephole trick yet?

I was thinking about this too. I think a wide angle lens would mess up accuracy if the firmware wasn't changed. A wide angle lens distorts the camera image. You'd have to account for this distortion to get accurate postional information. It seems like it would make it possible to stand closer to the screen though.

[ammitz]

Good review Randy 

[shmokes]

there are people so good with a gun that they can have someone throw a quarter in the air and they will shoot it out of the sky. 

No there aren't.

[RandyT]

there are people so good with a gun that they can have someone throw a quarter in the air and they will shoot it out of the sky. 

No there aren't.

Uh-Huh.

[Ginsu Victim]

here's a guy doing it using a paintball gun. (frame by frame you can see this isn't a fake -- but, as with most videos online, it could be)

[bkenobi]

Annie Oakley did.

http://coins.about.com/od/famousrarecoinprofiles/qt/annie_oakley.htm

[ammitz]

And some of these guys do it

http://www.firearmstalk.com/forums/f48/shooting-coins-2273/

[PsychoAU]

There was a special on the History channel a few months ago that showed a guy shooting a bullet through the eye of a small, metal washer. 

But the big question is, how do I get my Aimtrak that accurate?

[shmokes]

http://rstvideo.com/trailer/lucky-luke/

This is as close as you'll get.  If you spend all day tossing a coin in the air and shooting at it you will eventually sometimes hit it.  But that is, of course, not what we're talking about.

[ammitz]

This guy knows how to do it, Tom knapp, he's the one from the history channel.

http://www.tomknapp.net/about/sharpShooters.php

[Ragtag83]

I got more pins workings, yay. FYI to those that may struggle as I did, the config utility is a pain.  I am not sure if my computer has issues with it or what, but it keeps crashing and doesn't always like to send the new button setup though it tells me that it has. A few restarts of my computer and the config utility, and a bit of stubbornness on my part seems to have done the trick...

[brandon]

http://rstvideo.com/trailer/lucky-luke/

This is as close as you'll get.  If you spend all day tossing a coin in the air and shooting at it you will eventually sometimes hit it.  But that is, of course, not what we're talking about.

"Tidernas snabbaste cowboy, Lucky Luke, får i uppdrag att rensa upp bland bovar och banditer i Daisy Town. "  ?  

umm.. my thought exactly... 

[ammitz]

"Tidernas snabbaste cowboy, Lucky Luke, får i uppdrag att rensa upp bland bovar och banditer i Daisy Town. "  ?  

umm.. my thought exactly... 

In case you don't understand, it's swedish 

[brandon]

"Tidernas snabbaste cowboy, Lucky Luke, får i uppdrag att rensa upp bland bovar och banditer i Daisy Town. "  ? 

umm.. my thought exactly... 

In case you don't understand, it's swedish 

yeah.. my Swedish is a little rusty (ie. non existent)  

I'd be eager to learn though...



Swedish Bikini Team

[ammitz]

You would rather want to learn danish instead (very similiar) , take a look a some danish girls from a commercial....

[MaximRecoil]

there are people so good with a gun that they can have someone throw a quarter in the air and they will shoot it out of the sky. 

No there aren't.

http://rstvideo.com/trailer/lucky-luke/

This is as close as you'll get.  If you spend all day tossing a coin in the air and shooting at it you will eventually sometimes hit it.  But that is, of course, not what we're talking about.

Ed McGivern (probably the best who ever lived) could shoot dimes out of the air, among other amazing feats, with a revolver. Walter Groff could shoot dimes out of the air as well:



It is obviously easier to hit quarters or especially half-dollars.

[PsychoAU]

It is obviously easier to hit quarters or especially half-dollars.

Or women.

[brandon]

It is obviously easier to hit quarters or especially half-dollars.

Or women.

yeah.. just don't lead em as much

[Turnarcades]

Finally, I've shot the apple off the mans head, shot the coin, shot the cowboy hat and the UFO with one shot (each) every time 

Silly as it sounds, this is the only thing that is important to me in a light gun as Point Blank is the major reason I want a light gun - there is no finer gun game for fun, pace and challenge.

As Randy said, some develop a natural feel for a gun as an extension of their arm. Gun nuts us Brits ain't, but in the arcades I could hammer through Point Blank hardly missing anything, justby instinctively aiming in the right place without looking through a sight. This translated over to the PS1 version too when I had my G-Cons originally back in the day, and hitting those UFO's on 'VERY HARD' level was instinctive and accurate nearly every time. I'm hoping I can do this with the AimTrak at some point so I get that feeling back.

Calibrating by shooting off-screen was something I found worked on the LCD Crapguns, but though this helped accuracy towards all edges of the screen, the problem with these guns for me was always IR interference and poor tracking. This meant grouping shots towards a screen edge (or even in the centre) was great for a few shots, but then it would randomly jump across the screen. For example in Point Blank on the stages where you got 6 shots and a point-graded target (like the police firing range stages), I'd keep my gun dead centre but two of the six shots would register somewhere on the other side of the screen - very annoying.

For anyone testing gun accuracy, I believe Point Blank is actually the best game to test it on, as the 'cardboard cut-out tin can alley' style play has no kind of bias system (grouping shots based on first shot), excellent collision detection and the perfect mix of shooting tests (pure accuracy on some stages, rapid fire on others and rythym grouping on others). I'm still 50/50 on this product so even if someone can find a silly way to calibrate far from as instructed, but it can be saved and works considtently, I'm in for definite.

[brandon]

Point Blank is one of THE lightgun games that all other games should be judged by!  Much more challenging and fun that most rail shooters.  My brother in law and I practically destroyed a brand new guncon1 playing Point Blank 3.  I had to replace the trigger switch because we wore it out.

[bigster]

Thanks for the videos.  The third one is really helpful.  I currently have two LCD Topguns but I cannot ever get them to stay calibrate for more than 1 game, and doing so takes forever.  I am considering switching to aimtrack. 
My concern looking at the video is the height differences.  I am 6'5 and most of the people who play the arcade with me are shorter 5'10 and under.  Will each person have to recalibrate every use?  Like if we are playing buck hunter (PC edition of course) 4 players with one gun, that would require recalibration after each turn....

[Neverending Project]

Thanks for the videos.  The third one is really helpful.  I currently have two LCD Topguns but I cannot ever get them to stay calibrate for more than 1 game, and doing so takes forever.  I am considering switching to aimtrack. 
My concern looking at the video is the height differences.  I am 6'5 and most of the people who play the arcade with me are shorter 5'10 and under.  Will each person have to recalibrate every use?  Like if we are playing buck hunter (PC edition of course) 4 players with one gun, that would require recalibration after each turn....

From my initial assessment it does appear that you will want to recalibrate every time the gun moves more than a foot or so from the point it was initially calibrated. That of course depends on the size of monitor, the player, and style of play, but for the most part it looks like the "aim" will be off when the height changes significantly.

I believe Andy is working on a new firmware revision which will have updated calibration techniques. And I think this update will find a compromise between changing gun height and distance after calibration. I hope to get some new videos done when/if he updates the firmware.

[smalltownguy]

I believe Andy is working on a new firmware revision which will have updated calibration techniques. And I think this update will find a compromise between changing gun height and distance after calibration. I hope to get some new videos done when/if he updates the firmware.

Here is what I propose:

• Offscreen calibration, for sure (looks like Andy's already implementing this)
• A 'stacked' calibration delay. Meaning:
• Hold down the trigger for a longer period of time to initiate the 'full calibration'
• Hold down the trigger for a lesser period of time to do a 'mini' calibration, which will tweak the main calibration settings.

The idea here is that the main calibration will put in place all of the adjustments to dial in the gun for edges, tracking compensation, etc for the screen in use. But the 'mini' calibration would allow for a quick tweak when a new user grabs the gun. Maybe the mini would entail shooting at a few strategically placed cursors to help the gun adjust for height, etc.

Example: I calibrate my AimTrak on my own using the 'main' calibration, before I have any friends over. Then I mark the distance the calibration was completed at (tape on the floor, visual reference, whatever). Later, when all my friends are playing, each one of them does the 'mini' calibration in between turns.

This might not work on some games, depending on the length of time between player turns. Any thoughts?

[drventure]

I'm probably WAY oversimplifying here so if I am, flame away 

But, once you've calibrated the gun at a specific distance, could it not, more or less recalibrate automatically just by comparing the old distance between the farthest left and right LED's that it sees and adjusting the scale accordingly?

Maybe that's what it already does internally, but I'm just guessing.

Ages ago, I worked on a GIS app that allowed users to pick points on a scanned map with no geolocation, and the same point on a digital map and create a mathematical "profile" of the scanned map that would allow the system to accurately determine the LAT LON of any point on it via just a few known points.

Seems like this problem is very similar.

I'm watching the product progression closely!

[bkenobi]

If a second LED bar were used on the bottom of the screen, would these problems be reduced?  What if the LED's were located at the 4 corners of the screen?  I don't know how the algorithms work explicitly, but the goal should be to have the gun work from any location as accurate as possible IMO.  If more LED's are needed, add them!

[syph007]

I suspect its a tradeoff between number of led, and distance from screen.  Remember the board is already aimed upwards.  If there was a bar at the bottom as well, the eye would have to aim forward, and we'd have to stand back alot further.

I'm happy with the current setup.

[bkenobi]

Hmmm, that's probably true.  Good point.

[swamprat96]

Yes ok but- modern arcade guns use sensors all around the screen - and you can stand close. So whats different?

[RandyT]

If there was a bar at the bottom as well, the eye would have to aim forward, and we'd have to stand back alot further.

That's not true.  The camera doesn't always have to have a second set of markers in view simply because they are there.

The logic would go something like:

See both: Average position
See top only: aim is below
See bottom only: aim is above.

Obviously it's more involved at code time, but that would be the starting point.  This, BTW, is why you can get closer in the arcades.

RandyT

[syph007]

If there was a bar at the bottom as well, the eye would have to aim forward, and we'd have to stand back alot further.

That's not true.  The camera doesn't always have to have a second set of markers in view simply because they are there.

The logic would go something like:

See both: Average position
See top only: aim is below
See bottom only: aim is above.

Obviously it's more involved at code time, but that would be the starting point.  This, BTW, is why you can get closer in the arcades.

RandyT

Yes but if it can only see top or bottom for part of the time, then how during that time is it anymore accurate than the current setup.

Edit: I suppose it woudl be somewhere in the middle of a 1 bar system and a 2 bar system (that the camera could always see).

[RandyT]

If there was a bar at the bottom as well, the eye would have to aim forward, and we'd have to stand back alot further.

Yes but if it can only see top or bottom for part of the time, then how during that time is it anymore accurate than the current setup.

Accuracy wasn't the issue you brought up.  You were concerned with distance.

[AndyWarne]

Yes ok but- modern arcade guns use sensors all around the screen - and you can stand close. So whats different?

The Sega system which uses LEDs around the screen works on an entirely different principle. The LEDs are controlled by the game board. Its not a very accurate system, at least on the CRT Naomi cabs. Maybe better on later LCD.

[brandon]

If there was a bar at the bottom as well, the eye would have to aim forward, and we'd have to stand back alot further.

That's not true.  The camera doesn't always have to have a second set of markers in view simply because they are there.

The logic would go something like:

See both: Average position
See top only: aim is below
See bottom only: aim is above.

Obviously it's more involved at code time, but that would be the starting point.  This, BTW, is why you can get closer in the arcades.

RandyT

Do you think that perhaps the LEDs in arcades also blink at different frequencies so that if the camera can only see one it know which one it is based on timing?  Sort of like a lightgun and CRT?  or maybe some of the "lightguns" in arcades are ultrasonic sort of like the Power Glove...

[AndyWarne]

Do you think that perhaps the LEDs in arcades also blink at different frequencies so that if the camera can only see one it know which one it is based on timing?  Sort of like a lightgun and CRT?  or maybe some of the "lightguns" in arcades are ultrasonic sort of like the Power Glove...

Yes. If larger numbers of LEDs were used, they would need to be controlled by the gun logic so the gun would know which LED(s) it is looking at. This means the simplicity of the design is lost because a controller board would be needed and/or a PC application to control everything, plus more wiring, driver circuitry etc.