A Closer Look At The AimTrak (with video goodness)

[Loafmeister]

I don't have a problem with people being a bit picky and wanting a gun to work as they see fit. Let's face it, we all have our preferences for joysticks, trackball, spinners, etc., and all our reasons for prefering our babies are valid, because the decision is based on what's important to us.  So why should this be different because its a light gun?

However, I do take objection with using real world examples to validate whether this is a good product.  I have a friend who was a pretty good shot for hunting.  Amazingly, he was also AWESOME playing those old target games, you know the ones that used some robotics for animals, targets that moved/lighted up to indicate to shoot it (not a video game, I can't remember what these were called).  Anyway, he was great at them.  One day, he tries either time crisis or Virtua Cop (can't remember) and sucked totally.  Something about it being presented on a 2d plain just messed him up.  Maybe because it's a totally different skillset that's required?  I don't know.

Anyway my point is, I think it's cool to see a comment that states "hey compared to my guncon on my PS2, this gun isn't very accurate" or "I've got a problem when I rotate". Those are good points and need to be brought up so that either tweaks can be done at his/her end to fix the issues, or to allow Andy to improve the product via a patch.  No one is asking for detractors to do any "glad handing" towards Andy.  However, since we're not the friggin' engineers of this product, what is the point of bring this into the debate?  It's cool to comment on what others have said but the words used hint at a position of knowledge as far as THIS SOLUTION is concerned and this isn't the case.  If one hasn't even tried it yet, that shouldn't be the stand taken, IMHO anyway.

Randy, I don't doubt the quality of your products, but hammering at engineering points, considering you haven't even tried it yet may reflect poorly on yourself. Again, not saying to shut up but withholding the judgmental tone until you'd tried it yourself might be a good idea.

It's all down to expectations of course.  All I want is to have the gun be as accurate as possible but I understand it may not be perfect. When I used to use my Amiga Lightgun (still have it ), it was weak on corners a bit but I adjusted.  When I had a super scope for my SNES, it wasn't always bang on, but I adjusted.

Relevant to this thread, with my LCD Topgun's, I had to learn that for the guns to be accurate, I had to calibrate them differently from the top of the screen compared to the bottom (top left corner, aim at the corner, top right corner, aim about half an inch to the right. Bottom right corner, aim about half an inch lower and 1/4 of an inch to the right and finally for the bottom left corner, 1 inch lower and half an inch to the right.  Once this was done, I found the guns to be very accurate with my setup, with the laser pointer being "close enough" on just about everything. No it wasn't 100% accurate, but it seemed pretty similar to the accuracy I got by playing Time Crisis in the arcades and that's just fine for me.

I should be getting my own aimtrak modded guns later this week. Because it uses a similar technology, I'm fully expecting to have to tweak the calibration similar as I did with the LCD Topguns' and that's a freakin ok for me.  I'll post my informed comments on the quality of the gun then.

[Loafmeister]

how accurate are the setups you have? Are they consistent with shots?  I am not trying to be rude I want someones honest opinion on the setups they have and if they are dead on all the time?  Also when you compare them to the AimTrak please do so to other arcade/video products past and presents.  No real guns they kind of kill the fun after the first shot.

The GunCon2's (and 1's) are the "gold standard" for me.  They are consistently the most accurate, which is why I chose to build a dedicated shooting setup around them.  Like all raster based guns, they have a little jitter at the extreme edges (right where the tube meets the bezel), but overall are very consistent and accurate.  If the Aimtrak can be tweaked to be as good as these, even if it means "trick" calibration, you can bet I will be sounding the "buy" bell.

RandyT

That's good to hear and fair.

[Ginsu Victim]

Loafmeister:
If the Aimtrak calibration works out for you with the adjustments you mentioned, be sure to let us know exactly where you're aiming. I'd love to try it out.

[Loafmeister]

Ginsu: Will do. All I can say is it was trial and error for the LCD Top Guns but it definitely helped to find the sweet spot (distance was also part of the sweet spot - unfortunately for the LCD Topgun's, I just find I have to be a little too far). If I get it to work, I'll post a vid!

[Neverending Project]

I was just thinking about the typical use that these guns will get on my machine, and it will most likely require recalibration - not necessarily between every game, but in between game sessions. If I am playing, I will stand in the middle of the screen. If I am playing two-player, I will definitely be off to one side by a couple of feet. I also imagine if my kids play, they will not be standing on a ladder.

So this brings me to my point. If calibration is mostly dependent on the current position of the gun relative to the sensor, maybe we can request that the AimTrak is able to store several calibration setups. Maybe they can be called with a special button press. Then once we hit the sweet spot for calibration, it is saved in spot 1. We hit the sweet spot for the left side and it can be stored in sport 2, etc. Then when you switch positions you just call up the saved calibration, rather than recalibrate.

I'm not sure it is really saving that much work, but just thinking out loud.

[Ginsu Victim]

Ginsu: Will do. All I can say is it was trial and error for the LCD Top Guns but it definitely helped to find the sweet spot (distance was also part of the sweet spot - unfortunately for the LCD Topgun's, I just find I have to be a little too far). If I get it to work, I'll post a vid!

Luckily distance isn't a problem. My gun is about 2 1/2 ft away and I sit on a stool. With a 19" monitor, I can't get too far away anyway.

[TheShanMan]

I was just thinking about the typical use that these guns will get on my machine, and it will most likely require recalibration - not necessarily between every game, but in between game sessions. If I am playing, I will stand in the middle of the screen. If I am playing two-player, I will definitely be off to one side by a couple of feet. I also imagine if my kids play, they will not be standing on a ladder.

So this brings me to my point. If calibration is mostly dependent on the current position of the gun relative to the sensor, maybe we can request that the AimTrak is able to store several calibration setups. Maybe they can be called with a special button press. Then once we hit the sweet spot for calibration, it is saved in spot 1. We hit the sweet spot for the left side and it can be stored in sport 2, etc. Then when you switch positions you just call up the saved calibration, rather than recalibrate.

I'm not sure it is really saving that much work, but just thinking out loud.

I share your concern about 1 vs. 2 players and kid vs. adult. Seems like a pretty good idea for a solution if there is an inherent inaccuracy that comes from holding the gun in different locations, which can't be overcome with firmware updates or whatever.

[RandyT]

However, I do take objection with using real world examples to validate whether this is a good product.  

Unfortunately, (fortunately?) this type of control is different than a joystick, or a spinner, or a pushbutton.  This one actually has a real-world analog.  So while you may not think it's appropriate to compare the two, it is, to the same extent that a comparison between a real car steering wheel and one used in a video game is appropriate.  The steering wheel example is a good one, because even though many wheels only roughly approximate a real car steering wheel in that they provide the user with the ability to steer a simulated device to the extent allowed by the game, no-one really expects to be thrown to one side of the room, or smell burning rubber when it's turned too hard .  Likewise, a gun control should allow one to accurately "shoot" something within the software's ability to sense that something is being "hit" by that shot.  If the software is limited by the control, then the control falls short of what the software expects, and user skill has little to do with it, unless the control also unnaturally limits the users interaction with that control.

If one hasn't even tried it yet, that shouldn't be the stand taken, IMHO anyway.

Are you new around here?  Look at any of the new products I announce and find me one where this hasn't happened en masse.  It can't be ok because it only happens to me (can it?)

Randy, I don't doubt the quality of your products, but hammering at engineering points, considering you haven't even tried it yet may reflect poorly on yourself. Again, not saying to shut up but withholding the judgmental tone until you'd tried it yourself might be a good idea.

The videos are pretty clear, and to be honest, unless you understand the well-known and widely used principle of these systems, I'm not sure why you are finding fit to pass judgment on my comments.  Andy stated that the current calibration method did not account for overscan, and I offered a very simple and effective way to help calibrate for it.  If he has a different idea, then that's fine too.  It was an honest attempt to help make the one I end up with better as well.  My statements about lens distortion come from over 7 years of optical fabrication experience, and experience with the injection molded lenses used with these little mass-produced camera modules.  But if that's not the cause, then that's great too, as it means it's probably an easier situation to correct.  But I threw it out there in case it wasn't considered.

It's all down to expectations of course.  All I want is to have the gun be as accurate as possible but I understand it may not be perfect. When I used to use my Amiga Lightgun (still have it ), it was weak on corners a bit but I adjusted.  When I had a super scope for my SNES, it wasn't always bang on, but I adjusted.

If that's as high as your expectations are, then you have nothing to worry about.  They were poor compared to any of the newer stuff, and in that light, the games helped to cover that by being less demanding accuracy-wise than more recent titles.

Relevant to this thread, with my LCD Topgun's, I had to learn that for the guns to be accurate, I had to calibrate them differently from the top of the screen compared to the bottom (top left corner, aim at the corner, top right corner, aim about half an inch to the right.

While this was indeed a pain, it was the distance that made the LCDTopGuns a non-starter for me.  The gun tip needed to be back 10' from a 36" monitor to be just within operating parameters.  And even with all that, the accuracy wasn't incredible, just passable.  It's also a bit finicky about how one moves the gun.  I was able to get through HODIII with it, but it seemed like a chore.  Not at all an "arcade experience".  My expectations are admittedly higher for this one, due in no small part to the claims made by the manufacturer.

RandyT

[aljupy]

Hello:
I´m a big fan of lightguns. And.... like RandyT  I have lots of Guncon1, guncon2, Zappers, Dreamcast, Xbox lightguns..etc..etc...

While this was indeed a pain, it was the distance that made the LCDTopGuns a non-starter for me.  The gun tip needed to be back 10' from a 36" monitor to be just within operating parameters.

That is solved with a Wide angle lens

And even with all that, the accuracy wasn't incredible, just passable.  It's also a bit finicky about how one moves the gun.  I was able to get through HODIII with it, but it seemed like a chore.  Not at all an "arcade experience".  My expectations are admittedly higher for this one, due in no small part to the claims made by the manufacturer.

I agree that. Guncon2 is Excellent for CRT. But Topgun beats it, and it is for All screens. I´m very sorry for people that got a defective model.

See my vids

[RayB]

Useless bickering for a "1.0" product. Even Microsoft never gets it right on the first iteration. Cut Ultimarc some slack and wait for the 1.1 fix. (I'm pretty confident a solution exists like what I suggested above, but my post probably flew over the heads of most).

[Loafmeister]

aljupy: I've seen your vids. I was never able to get the eyetoy wide angle lens but looks like it worked great for you!

[aljupy]

aljupy: I've seen your vids. I was never able to get the eyetoy wide angle lens but looks like it worked great for you!

Many Thanks.
Most wide angle lens should work for it. They are very cheap.

I do not want to do OffTopic. I wish the best for AIMTRACK lightgun, and I will love to see that lightguns for Pc are a growing market.
Also I wish that people that get dissapointed with TOPGUN gets a nice experience now.

I just posted that vids because I read that RandyT has a Topgun inbox thinking that is not accurate like a guncon2

-Bet Regards

[RandyT]

I just posted that vids because I read that RandyT has a Topgun inbox thinking that is not accurate like a guncon2

I think a lot of that had to do with the distance and the space between the bars.  Being able to get that close to them gets you a much brighter reference.

The accuracy you are showing is very impressive, but your video brought back to me something that I had forgotten about.....lag.  It's not horrible,  but it was something I noticed while playing.  The AimTrak definitely looks better in that department, and the GunCons are pretty much "real time" in contrast.

Still, a very cool mod you did with the lens.  There was a lot of speculation as to whether that would work and it's neat to see that someone succeeded with it.  Makes me want to find one and re-visit the hardware.

RandyT

[BamBam]

Useless bickering for a "1.0" product. Even Microsoft never gets it right on the first iteration. Cut Ultimarc some slack and wait for the 1.1 fix. (I'm pretty confident a solution exists like what I suggested above, but my post probably flew over the heads of most).

I second that.  It was getting a bit nasty. 
Patience folks.  Now everyone back to their room for a timeout 

[TheShanMan]

Useless bickering for a "1.0" product. Even Microsoft never gets it right on the first iteration. Cut Ultimarc some slack and wait for the 1.1 fix. (I'm pretty confident a solution exists like what I suggested above, but my post probably flew over the heads of most).

Third'ed. Randy, I think it's fine for you to try to help Andy and your motives may be 100% pure, but diving into the engineering implementation details adds no value to this thread because that's only relevant to you and Andy. I think it would be better to take up that discussion with him via PM. I know it sucks, but as a competitor you need to tread very lightly when it comes to questioning/critiquing a competitor's product, even if you don't have a similar product and have no plans for one. Right or wrong, it's just way too easy for you to come across as a jerk, as evidenced by some of the reactions you've gotten.

Seems like this isn't the first time this issue has come up (I could be thinking of something else though). Hopefully you can find a good balance on that tightrope you're on. I hope you take this as an attempt to be helpful because I am not angry or trying to jump on you.

[Ginsu Victim]

Seems like this isn't the first time this issue has come up (I could be thinking of something else though).

Spinner argument. It got ugly. If I recall, Saint had to lock it while he was gone for the weekend.

[RandyT]

Useless bickering for a "1.0" product. Even Microsoft never gets it right on the first iteration. Cut Ultimarc some slack and wait for the 1.1 fix. (I'm pretty confident a solution exists like what I suggested above, but my post probably flew over the heads of most).

Third'ed. Randy, I think it's fine for you to try to help Andy and your motives may be 100% pure, but diving into the engineering implementation details adds no value to this thread because that's only relevant to you and Andy. I think it would be better to take up that discussion with him via PM. I know it sucks, but as a competitor you need to tread very lightly when it comes to questioning/critiquing a competitor's product, even if you don't have a similar product and have no plans for one. Right or wrong, it's just way too easy for you to come across as a jerk, as evidenced by some of the reactions you've gotten.

Well, just to keep the record straight, the solution offered up by RayB (the one he refers to in the quote you quoted) is the very same one I offered before he made that post.  He calls it a "multiplier", whereas I refer to it as a "factor table".  It makes no sense to me that anyone would see RayB as being "helpful" and me as being "derogatory" unless that individual has already "taken a side".  And if that's truly the case, there's really no reason for me to be concerned about what they think about what I say.  There is a difference, though; I have already given Andy my money.

I think it would be better to take up that discussion with him via PM.

This isn't a joint development.  I am interacting here as any other member would, and is, as evidenced by RayB's earlier comment.  

I don't take offense at your comments, and others shouldn't take offense to mine.  One of two things will ultimately happen as a result of them.  Either the product will improve to a level beyond what is shown in the videos, or prospective buyers won't have overly high expectations from the product.  Both outcomes benefit everyone in the community.

RandyT

[bkenobi]

I disagree.  I think it's good to hear different opinions about the hardware.  Randy doesn't have a competing product and is asking the same questions that I'm interested in having answers for.  Maybe there's only 2 of use, but before I spend my money on a pair of AimTrak's, I want to know that they will outperform my GunCon2 setup.  Ehhh, whatever.  

[Hoopz]

One of two things will ultimately happen as a result of them.  Either the product will improve to a level beyond what is shown in the videos, or prospective buyers won't have overly high expectations from the product.  Both outcomes benefit everyone in the community.

RandyT

Two outcomes based off the assumption that the videos are the end all, be all of how it operates.  It's one user's experience and may not be the same for everyone/anyone else.  I'm not saying it's perfect as I haven't had a chance to fiddle with mine yet.

[AndyWarne]

 I offered a very simple and effective way to help calibrate for it.  If he has a different idea, then that's fine too.  

I think you will find that you repeated my idea, ie to position the cursor a known distance from the edge and have the gun aim at this rather than the edge of the screen. Maybe read the posts again...

Have you used a reasonable mobile phone camera recently? I dont think you will find any noticeable picture distortion.

[AndyWarne]

I posted this before in my mini review, but maybe I should mention it again.  I noticed the non linear x tracking as well, but I noticed it only happened when i moved the gun by rotating my body.  If i kept the gun as the pivot point, the cursor and sites always lined up.   I would expect you'd need some kind of tweak to compensate for the gun rotating about a fixed point (the player) a fixed distance from that point.   Possibly a modified calibration routine could take this into account.

It would actually be possible to completely compensate for this. But, this would introduce an error when moving physically back or towards the screen. So the present firmware (which is not the one being shipped as I want to add some more tweaks) gives a happy medium between side to side and front to back body movement.

[RandyT]

Two outcomes based off the assumption that the videos are the end all, be all of how it operates.  It's one user's experience and may not be the same for everyone/anyone else.  I'm not saying it's perfect as I haven't had a chance to fiddle with mine yet.

I'm not basing everything on just the videos.   Most are saying that the difficulties are similar.  Whether or not they are concerned about it seems to be the variable.  FWIW, I'm hoping my experience is different as well, and if it is, I will be the first to tell the OP to send their unit back for a good one, or try it on an LCD screen (which he should do anyway, if he has one available to him).

I think you will find that you repeated my idea, ie to position the cursor a known distance from the edge and have the gun aim at this rather than the edge of the screen. Maybe read the posts again...

Actually, I did read your post.  Mine was in response to it.  You were concerned about the static cursor image being hard to see or being off screen.  Moving it in a box formation 10% in (which should accommodate any overscan situation) should address both of those concerns, without a negative impact on accuracy.  A quickly moving mouse cursor forming a small box is also a better target to aim at for calibration purposes.

Have you used a reasonable mobile phone camera recently? I dont think you will find any noticeable picture distortion.

As I said, I put it out for consideration.  In a mobile phone application, a little edge distortion is no big deal.  In an application like this, where you might be using a small section of the CCD at the very periphery of  the lens, not for simple "artistic imaging" but for coordinate use, and it could be.  But some scrutiny of the raw data will flesh that out.  If that's not the problem, be happy

RandyT

[RandyT]

Anyone want to hear my impressions of this thing, or am I going to be accused of being a jerk?

[HaRuMaN]

Anyone want to hear my impressions of this thing, or am I going to be accused of being a jerk?

dew it...

[Loafmeister]

Got my 2 aimtrack modified guncons today, looks pretty sweet! The following are my INITIAL thoughts on this, some issues I've come up with, without having put much time in troubleshooting them. I suspect these issues will work themselves out.

#1 - poor calibration:
Issue: the gun was not very accurate and I was having issues trying to calibrate it. For example, it was having difficulty registering the trigger for the top right corner.

Solution: I was concerned that my 19" LCD at about a 70 degree tilt was the culprit. First, I turned off all external lights (in case of a leak heat source) but that didn't help.  I then proceeded to check out the Ultimarc website and remembered that the LED pins need to be pointing at the gun.  As my monitor is not at a 90degree angle, I needed to bend the LED's towards the gun (as documented).  So I did so, bent them pretty much as far as they could go, tried it out and ...  SUCCESS!  I was able to do the calibration and right from the get-go, I'm super impressed with the accuracy.  Can I say it's perfect?  No, but I’m not getting the half-inch error rate that was mentioned before and that’s with the stock firmware (I think it was mentioned there was improvement in this area?)

#2 - buttons not responsive:
I tried Operation Wolf, it tracks near perfect, but the trigger sometimes won't respond, sometimes the bottom button (under the handle) will work as the grenade button (whatever it is) most of the time it won't. I had trouble getting OpWolf to recognize that button in the menu (had to hold it a while).

Solution: None yet but I believe I might know the cause

#3 - the aimtrack config utility says the gun is not connected (but the mouse cursor is moving, controlled by the gun!).  I also get a similar message when trying to use the firmware utility.

Solution: None yet but I believe this is tied to #2.

What I believe is happening is the gun I am using is plugged into a USB extension cable, which is plugged into a USB hub, which itself, is plugged into a USB cable extension, which itself is then plugged into one of the ports. Because of the type of USB hub I’m using, I HAVE to use a USB extension cable, but maybe the one I’m using is too long. I’ll experiment and will try and plug in the USB hub directly into the USB port (just letting it hang there for a bit won’t cause damage).
 
Anyway, I've already got ANOTHER USB hub inside the cab because of the use of U360’s, the USB trackball and the USB Spinner so maybe I'm pushing it a bit with a second hub, though it's never been an issue before when I was using the

Whatever the case, I'm pretty sure the guns are fine, I'll also try the other gun. I think someone mentioned an issue with the LCD Topgun driver, so I will also remove that, just in case its interfering.

Whatever the case, initial impressions are very good, tracking wise I was accuracy on par with the LCD TopGun’s (which I got good success) but at half the distance!  It’s true though, you really have to keep the gun within a certain distance from where you calibrated but this is no surprise and is a non-issue for me.

I'll have to hunt down some arcade gun holders that I can attach to the side of my CP so that they could be holstered when I don't use them.  Anyone know if there are some that will fit well with the guncon’s?

More comments coming…

edit:  Just thinking about accuracy and the "half-inch" mentioned before. It's true that I'm using a 19" monitor so if there are some accuracy issues that come up, it's possible they will be lessened because of the smaller sized monitor I am using.  Whatever the case, works for me so far.

[Loafmeister]

Anyone want to hear my impressions of this thing, or am I going to be accused of being a jerk?

That seems to be a loaded question .  Assuming you have any issues, if you bring them up and say "what can be done to fix this?", how can anyone take that as being a jerk?

[Neverending Project]

Anyone want to hear my impressions of this thing, or am I going to be accused of being a jerk?

You're wife thinks you're a jerk, and you still give her your opinions, right? 

For the record, I am KIDDING! I know I don't you well... I am just joking on a comment you made a while ago. You can't always trust that the sarcasm font is rendered on everyone's computer.

[smalltownguy]

Anyone want to hear my impressions of this thing, or am I going to be accused of being a jerk?

Unlike some other predisposed mindsets, I for one would like to hear your impressions, Randy. It would be nice to hear from another enthusiast. I don't care if you sell stuff. I want to hear what you think.

[TPB]

I think it would be better to take up that discussion with him via PM.

If you're conducting a Beta testing program, then sure, keep the communication confidential and discreet.

But that's not the case here.  Andy is not seeking Beta testers.

The AimTrak is on sale to the general public.  As such, it's open to scrutiny.  People are making purchase decisions about whether they part with their hard-earned cash.  They deserve to know the truth about the AimTrak, warts and all.

When problems were found with the LCD TopGun (the distance problem, and the lack of MAME drivers when initially released), no-one curtailed their criticism, despite the TopGun also being a 1.0 product at the time.  And that's fair enough.  It was on sale to the general public, so it was open to scrutiny.

The same principle applies here.

Of course, Ultimarc have an impeccable reputation for excellence, so there's little doubt the problems will be ironed out by Andy over the course of time, whether through firmware updates or the next version of the product.

In the meantime, the version that's currently on sale, is open to scrutiny.

Andy wouldn't want it any other way.  I'm sure he's happy to accept feedback presented in a constructive manner.  I doubt he'd expect to be afforded "special treatment" in any way.

[Loafmeister]

Hmmm, I'm going to have to read up on this.  With the configuration applet, I noticed sometimes the device id keeps rotating through each of the device numbers, like it's confused, one can also see the green bar flip to red, than to green, etc.  This has happened with both guns, both were not plugged in at the same time too.

[RandyT]

That seems to be a loaded question .  Assuming you have any issues, if you bring them up and say "what can be done to fix this?", how can anyone take that as being a jerk?

If there's a way, it will be found.  Trust me.

I just spent 3 hours with it, and did some comparisons with Guncons on a PS2.

First of all, no software tweaks or firmware upgrade.  My machine is a 98SE based unit and the supplied software requires .NET to run.  I've never been able to get it to take on this machine for some reason, so everything is "out of the box" as far as settings are concerned.

My machine has a 27" tube and a piece of dark tinted plexi (invisible to IR) for the front window.  If I put the LEDs on top, it would be behind the bezel, so I used a chunk of 3M automotive double sticky and stuck it on the screen just above the active part.  Advice: if you have a large screen, get the LED's as close to the active part as possible.  I then took a look from the side of the machine and bent the LED's so they were straight out toward the front.

Like everyone else, I started out by holding the unit in my hand.  It moved the cursor.  So far so good.   Next I taped a laser pointer to the board, and tried to calibrate.  Crazy problems.  Basically the same type of issues shown in the video.  I was able to get close a couple of times, but there always seemed to be a significant error in some part of the screen.  Advice: Don't do this.  It's way too easy to tilt the unit when it's in your hand, and tilting is not something you want to do while calibrating.

So I gave up playing and committed myself to at least getting the board mounted in the GunCon2 shell (bad ebay seller experience, made good by ebay...ended up being free ).  Whittled away a little chunk of plastic inside the barrel and hot melted the board into place, making sure it was as square and level to the barrel as possible.  Initially I was concerned that the camera board was not square to the rest of the unit, but it appears to be intentional.  I used the main board as my reference.

Immediately, I saw an improvement.  The gun grip provides a much better reference to keep things in a consistent orientation from calibration through to use.  But even so, I was still seeing problems.  It seemed like no matter what I did, I couldn't get a good calibration at 3' from the monitor face, even though the camera could technically, "see" the lights.  

I was getting a little frustrated so I decided to go out and play with the GunCon2 on the PS2.  The simplicity of setting up these can really be appreciated after what I was dealing with, but you can't use a GunCon2 on an LCD, and the LCDTopGuns are no walk in the part to set up and use either, so that's no knock specifically against the AimTrak.  My GunCon2s are on a 37" monitor, so keep that in mind when I talk about deviation.  What sounds large on this screen would be quite small on more average sized screens.  OK, one shot in the center of the screen from about 5' back and the the cursor was following the sights from between dead-on in the center to about 1/4" off (remember the scale) toward the corners.  Extreme edges with the GunCon2's are pretty much out of the question.  The top is pretty good, the bottom is rough (can only accurately get within 3/4" of the frame), and the sides are so-so, with one being good right to the edge and the other giving up about 3/8" from the edge.  Tracking is silky smooth with the GunCons, and lag is minimal.  Compared to the AimTrak, lag is about the same, but cursor movement appears to be smoother (on my machines.)

Ok, back to playing with the AimTrak. Two more attempts at calibration and then decided to drop back a foot.  There was an unquestionable improvement.  Advice: Don't push the distance thing.  Move back to the furthest point you will be happy with, and start there.  So I "trick" calibrated about 3 more times, using the best skillz I acquired from the LCDTopGun, and watched the cursor track my sights to within about 1/4" (again, remember the 27" screen size) from 4' away.  The bottom is still a little sketchy, but no worse than with the GunCon2's on the PS2.  Advice: Once you find a spot where you can get good calibration, remember it, mark it, do whatever you need to do to make sure you play from that point the next time.

Finally happy that things were looking good, I decided to try T2.  The cursor seemed choppy, so I disabled the mouse support and left lightgun enabled.  Still choppy.  Rebooted the system and tried again.  This time, the cursor was running smooth and I was able to pop robots without paying attention to the cursor, even though the trigger switch was on a wire in my left hand and the gun case wasn't even screwed together.  I think I might have actually smiled at that point in time.  The wet blanket came 5 minutes into the game when I forgot to let off on the fire button before the 5 seconds was up and it went into calibration mode.  Advice to Andy:  Default the hardware to go into calibration mode when the trigger is held off-screen for 10 seconds.  Much less likely to happen by mistake.

I'll close this terribly long post with some observations:

Just having the light from the LED's strike the CCD does not seem to be enough for good calibration to be possible.  There seems to be a "sweet spot" that is based on the distance between the LED's and the center of the screen, the distance between LED's and camera, and the elevation difference.  It's my opinion that the OP's videos were showing an attempt to use the gun outside of this sweet spot, either too close or at a poor elevation relationship between camera and LEDs.

If GunCon2 performance on the PC is anything like GunCon2 performance on the PS2, and one is using a tube based monitor, I can't come up with a compelling reason to switch to an AimTrak. There are more limitations inherent in the technology it uses, and the accuracy is about the same when well calibrated.

The calibration routines as they are currently need help.  Pointing at a location where screen meets bezel is never as good as pointing at an on-screen target.  It also leaves tube users at a disadvantage, as some overscan is almost always present.  And because tube screens are not nearly as geometrically perfect as LCD or plasma screens, a 3 point calibration may not be enough with this type of technology. "Trick calibration" is a mostly unintuitive process and shouldn't be required of the user.  This was also a shortcoming of the LCDTopGun.

The accuracy shown in the LCDTopGun videos in this thread probably aren't going to be achieved by the AimTrak.  But the AimTrak has less lag, and when properly calibrated, enough accuracy to be a better solution in actual use.

I wish the software didn't need .NET to run.  Probably not a big deal for most.  Was for me.

Conclusion:

If you have an non-tube display that isn't huge, don't mind being back maybe just a hair farther than you would otherwise like to be, and have a specific shooting stance you will always use, then you would be crazy not to give this a shot.  Within it's limitations, it seems to work well.  You just need to understand what the limitations are, and patiently adjust.

If you have a tube display (especially a larger one) and you want to get closer to it, you might want to look at the GunCon2 and PC drivers first.  Even if the distance doesn't bother you (it's pretty acceptable, even with a 27") you may still want to hold off a bit until the calibration routines take tubes into consideration a little better.  If you are used to "trick calibration" techniques from using LCDTopGuns, then it might not be a big deal for you.  If not, you might end up really frustrated, really quickly.

Personally, I'm happy I purchased it and am looking forward to revisiting some of the old titles with the new toy.

RandyT

[TheShanMan]

Seems like a very fair review, Randy. It's much easier to put aside the fact that you compete with Andy on other products once you've actually used the device and have clearly tried to be fair in your review rather than just focusing on the negatives. Let's hope feedback like this will lead to an improved product from Andy - I'm sure it will!

I'm glad I'm way too busy right now to buy. There are obviously some kinks and/or unoptimized user experience to be worked out. Hopefully when I'm ready to buy it'll be a much smoother setup process.

[AndyWarne]

A very fair review.

I will think about the "sweet spot" issue, as I am not totally sure what this is about. The config utility may be useful here because it displays the actual X-Y location and if this is not making sense then there is something not right in the positioning. Being too close is going to be a problem but the isue is how close is too close?
One possibility is to make a change which introduces a complete cut-off when too close rather than attempting to work but not accurately.
Changing the calibration to start only when aiming off-screen is a good idea and I will make that change.
Also will change to optionally calibrate using pointer at a known distance from the edge.

[severdhed]

thanks randy, nice review.  i personally am very excited about this product, and im sure that will all of the feedback that is coming in, that Andy will get these things working great.  my hope is that the bugs are worked out and ready to go by the time my birthday rolls around in February

[Cananas]

One possibility is to make a change which introduces a complete cut-off when too close rather than attempting to work but not accurately.

Yes. I think it's better a cut-off when too close, like LCDTopGun do.

[Ginsu Victim]

Very nice write-up, Randy. You made some great points and I'm glad to see Andy agrees (I sure as hell liked your ideas). I knew you would be fair and you were.

[bkenobi]

Thanks Randy!  I am currently using GunCon2's with the PC driver (as I've mentioned before), so I'm looking for an improvement in setup and capability.  Your review is EXACTLY what I needed to know.  Right now, it looks like I will be staying with my current configuration.  HOWEVER, as soon as some of these calibration issues are dealt with (or my CRT TV dies), I'll be looking to upgrade.

[Loafmeister]

Great review Randy, lots of good info there too and a great recommendation for the off-screen calibration.

[vartech]

That seems to be a loaded question .  Assuming you have any issues, if you bring them up and say "what can be done to fix this?", how can anyone take that as being a jerk?

If you have a tube display (especially a larger one) and you want to get closer to it, you might want to look at the GunCon2 and PC drivers first.  Even if the distance doesn't bother you (it's pretty acceptable, even with a 27") you may still want to hold off a bit until the calibration routines take tubes into consideration a little better.  If you are used to "trick calibration" techniques from using LCDTopGuns, then it might not be a big deal for you.  If not, you might end up really frustrated, really quickly.

Personally, I'm happy I purchased it and am looking forward to revisiting some of the old titles with the new toy.

RandyT

How would I connect a GunCon 2 to an arcade monitor using a arcadevga?

[Lilwolf]

btw, has anyone tried the wide angle lense from a door peephole trick yet?  I hear they are about 5-10 bucks and will come with a decent end that could be added to a gun.  (I hear you want the 160degree, not the 200).  But I'm not sure if will help with the distance... or if the aimtrack already has a wide angle lense on the end)