Main > Lightguns

Wingun, lightgun, guncon 2 troubles aplenty

<< < (8/11) > >>

felix pontifex:
when this stuff finally gets done with I think I am going to learn enough Flash in order to make some simple light gun games. That is after I give myself a vacation from learning anything new and frustrating. As I keep learning more about light gun games I am finding that there are many more ports to consoles of arcade games than I initially assumed, which makes me feel better about my guncons.

When it comes to the IR camera delay, do you find that it hampers gameplay for those games? Do you know if aimtraks have the same delay? I'm curious mainly because I am having some wishful thinking about my penguin pistol. Maybe the camera they use will be higher quality and have less delay. I'm sure you will shoot me down with your advanced technical knowledge.

By the way, what company made SINGE and when?

Howard_Casto:
I don't own an aimtrack so I can't really tell you if it's better or not, but all IR guns inheriantly have a delay.  A true lightgun literally uses the monitor's timing to determine position, so reaction is instant.  For an ir gun, position has to be calculated, either on a chip within the gun itself or via software.  Both wiimotes and the aimtrack do a little bit of both.  That's where your delay comes from. 

That being said, the delay isn't that big of a deal, and a lot of it is a visual delay.  Yeah the cursor has a time keeping up with your movements, but when you stop (which you will to shoot) it catches up. 


I seriously doubt that those wiimote guns will be more accurate.  The reason they came out is due to the wiimote's short throw on the B trigger, making gun shells feel mushy.  Don't get me wrong, that's what I would use on a cab as well for the wiimote route, but that's because of their form, not function.

IR guns biggest drawback is accuracy.  A good lightgun is accurate within one pixel, a level of accuracy you'll never get with an ir gun.  IR guns, even well calibrated ones, usually have about an half an inch of slop.  Luckily the targets on most gun games are fairly huge.  Try duck hunt if you have doubts.  Strangely enough that game has the smallest, most moving targets of any lightgun game, especially the skeet shoot levels.  Most people with IR guns tend to leave the cross hair on, to cope with inaccuracies.  So if you want to "shoot from the hip"  with the cursor turned off, it might not be the setup for you.

The IR guns biggest strength is their adaptability.  When you change your monitor setup they'll still work, ect....  Also one of the reasons they are so popular for mame is the positional gun games.  The majority of mame's gun games are actually positional gun games (giant joystick, and not a real gun).  You won't be playing games like terminator 2 or operation wolf on a real lightgun... they just aren't compatable.  IMHO these are the games that are the most fun, much more so than the traditional lightgun games.


SINGE isn't a single game, it's an Offshoot of the Daphne emulator.  It was originally intended as a means of making fan games, but it devolved into the "porting" of all the laserdisc games that daphne doesn't emulate.  This includes the American Laser Games, like maddog mcree and ect.  You buy the digital leisure copies of the games and SINGE rips them into more arcade accurate versions, playable through singe.    These games are slowly going out of print, so you can pick them up on ebay/amazon for a few bucks a piece.  If you don't like these games then don't waste your time, but if you DO I highly reccomend using singe, because the engine that the digital leisure ports run on is just awful.  Half of them won't even play on windows 7 they are so broken.  Btw... these are some of the few games you might have trouble with the ir guns.  The bulk of the stages are easy, even by lightgun stages, but for the bonus quick draw scenes, you pretty much have to have the crosshairs turned on.  They are still quite playable though.

Howard_Casto:
I played around with house of the dead III a little more today.  I'm pretty sure that game is just broken now that I've tested a bit more.  I'm using a no-cd crack, because... well you don't want to have to put in a cd on a mame cab and I think it may have screwed up the configuration a little.

I tried a  few of my real joysticks and had mixed results.  My matcatz xbox 360 controllers worked just fine, which is odd considering they are xinput devices and the games were released in 2005.  As expected, the rumble feature doesn't work with the pad (xinput doesn't support true force-feedback).  So next I tried a xbox classic controller, expecting it to fully work as it's a true directX joystick.  The joystick wouldn't respond at all, even though in the game's cfg menu I could use it to remap buttons.  Oddly enough force feedback does work?

I thought of a different way to deal with the pc games.  I downloaded the now ancient troubleshooter app that used to fix pc lightgun games for the old act-labs guns.  I was hoping there was an ini file or something, and sure enough there was.  It has the memory addresses of all the cursor postions to those games. I'm not big on making memory hacking apps, but it might be that the troubleshooter could be modified to handle more than just the act-labs guns.

felix pontifex:
I had another question about calibrating a wii remote through a script: If you were to make a script the asked not for just the calibration of the screen's corners, but of multiple places on the screen (such as corners, the center, and other points on the screen like the top, bottom, sides, and other spots in between the middle and the sides, making a grid if you were to connect the dots) would it possibly be able to tighten up the accuracy of the gun at all? I recognize that it would not make pixel perfect aiming per-se, but if the user (me) were willing to perform a longer more detailed calibration would that data be useful in minimizing the "slop" presented by the limits of the hardware? Or is good calibration simply a matter of programming?

Howard_Casto:
It's hard to say honestly.  If I were a bit better at math I could probably give you a better answer, but the truth is I really don't know. 

For the method I use you really only need two points for calibration, the upper-left corner and the lower-right one.  Those two points will give you the minimum and maximum values on the screen.  Most calibration routines go a bit further and ask you to shoot the center of the screen.  This is mostly to reduce human error.  You might be a bit off at hitting the corners, but it's pretty darn hard to screw up hitting the Center.  So you'll take your max values and divide them by two.... if they are greatly different from the center point, you can average out your min and max values with the center point to even things out a little more. 

I've also found in my tests that just finding the center point helps a lot.  I wrote a test script that allows you to press the down arrow and center the cursor.  The idea is you point the wiimote at the center of the screen and press down to zero it in.  It works amazingly well.  But inconsistencies in the wii cursor report keeps it from being perfect.  You see glovepie averages the cursor positions into the range of 0 to 1 so it can be multiplied by the screen resolution or what have you easily.  The only problem is a wiimotes REAL range can go below zero and above one.... in other words we don't know the exact range.  So a three point calibration system, like the one used in terminator 2 would give us the min and max values of your own personal monitor as well as one point of reference for averaging. 

My guess on your suggestion?  More averaging, in a perfect world, would improve accuracy, but in reality every time the user makes a bad shot, it would throw the whole table off, and if you give a whole grid of values it's going to be a huge pain in the butt to calibrate and people won't bother lining up the shot.  Also remember that glovepie can't save values, so you are going to have to calibrate every time the script is ran, unless you read the offset values while the script is running and hardcode them manually. 

I think three, carefully aimed shots, is the best solution.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

Go to full version