Topic: FPS programmers lazy?

[dubnobasswithmyheadman]

OK, by now we're all familiar with the bounding box control scheme (Red Steel, etc). And we all know what the biggest problem with it is. Move the controller too far, you lose sight of the sensor bar, you lose control of your character.

My question is this: why not use the tilt sensor to control movement instead? You could turn outside of the sensor bar's range with no problems.

I realize that it would need calibration. In other words, the game needs to know which tilt position is neutral (aiming at the screen). That's where the pointer comes in. The game already KNOWS when you're pointing at the screen, and it could use this opportunity to recalibrate the tilt sensor.

A combined tilt/pointer approach could be the magic formula. Without the tilt sensor, we get bounding box problems. Without the pointer, the tilt sensor would be lacking calibration.

So, we should use them together to make up for their individual shortcomings.

What do you guys think? Could it work?

[Requiem]

Could it work?

no

[BlkPaladin]

No that would open it up other problems because you would need to hold the controller completely still to not move. Other problems would be that it would need to processed before it could be used at suitable data where as a device such as the analog stick doesn't need to be processed as much before hand. Supposedly by what I have read on IGN that Metroid seem to have come up with a really good working alternative to the bounding box problems.  

[D_MaN87]

Some may disagree, but I loved the way aiming was implemented into COD3.  I had no problem using it, and it was the closest thing I have used to a mouse other then a mouse itself.  I hope MP3's control scheme is great, as IGN says, but I saw no problem with COD's.

[Crimm]

Well I think it COULD work.  I'll explain.  The Remote already uses triangulation to determine position, so having it use math to do this isn't that crazy.

It would need a starting point, so you would have to calibrate it first.  Every time the remote detects the bar it would need to recalibrate.  That means that if it can see the bar it is using the data from calculating angles.
When it can't detect the bar it simply calculates its position based on distance formula.



X: Final distance traveled.
Xo: Distance already traveled.
Vo: Initial velocity (how fast it was going before you started this last movement)\
a: Acceleration (it would have to take an average over split seconds, at least that's how a lot of sensors do it)
t: Time.  How long you moved it for.

From that you could get Position by:

Px=Pxi+dx
Py=Pyi+dy

Px is the final position on an x-axis, Py is the same on the y-axis.
Pxi is where it started on the x-axis, and Py is the same on the y-axis.
dx and dy are the distances traveled on each axis.

This wouldn't be a real simple thing to do, it would require the use of memory (perhaps the remote's internal memory). It could be an interesting thing for some of the people who have toyed with the remotes to try.

[Kairon]

But the wiimote sensors don't have the sensitivity to detect acceleration the likes of which even humans can do. We can accelerate faster than the wiimote can detect, which is why 1-to-1 sword play through this method and true 1-to-1 anything is highly improbably as well. The data would be innacurate and there would, of course, be a calculation lag on anything like it.

~Carmine "Cai" M. Red
Kairon@aol.com

[MarioAllStar]

Lazy programmers?

Well I've never heard of such a thing!

[dubnobasswithmyheadman]

BlkPaladin: I'm sure they could implement some stabilization algorithm like "mouse smoothing" in PC games. Also, i don't think the input would need to be processed. Tilt control wouldn't be any different than tilting an analog stick.

Crimm: No need to determine position of the controller in my scenario. Just how much the controller has tilted from the neutral position. The neutral position can be determined by the sensor bar.

Kairon: No need for detecting acceleration in my scenario either. I bet the tilt sensor can keep up to speed with human movement. Look at Kororinpa for an example. I've heard it's very responsive.

[Kairon]

Quote
Originally posted by: dubnobasswithmyheadman
Kairon: No need for detecting acceleration in my scenario either. I bet the tilt sensor can keep up to speed with human movement. Look at Kororinpa for an example. I've heard it's very responsive.

...oh snap. You mean, use the nunchuck tilting to determine when to tilt your view up or down, rotate right or left? ... SWEET!

Player's will have to get used to seperating their aim and "facing" actions into two seperate hands though...

~Carmine "Cai" M. Red
Kairon@aol.com

[Mario]

<Ubisoft> We tried it, IT'S NO GOOD.  

[ShyGuy]

hmm, the name breaks tables yet there is no avatar.... THIS IS A CRISIS OF LAW.

On the subject, I liked Elebits handling of FPS on the Wii.

[NeoThunder]

I have to say my problem with Red Steel was the fact the pointer had to go ALL the way to the edge of the screen before you turned, and when you turned...it was one speed.  When you play a FPS you turn different speeds a lot, if i were to describe it I would say it was like turning first.....pointing and shooting second.

Call of Duty 3 I thought had a very good method.  Basically if you had the pointer in the middle of the tv....you didn't turn.  Then the further out from the center you got you started to turn, and the more from the center, the faster you turned.  The only flaw I found was it was hard to keep from moving as the "center" was a small spot.

I think two improvements that need to be made is somehow fix it so i'm not spinning out of control when the remote doesn't see the sensor bar anymore, and probley make the "center" a little bigger so the screen isn't always moving so much, but still has an analog turning.

[zakkiel]

Crimm: what does position have to do with dub's suggestion?

Dub: Haven't you played the minigames that use tilting to steer? They're fun and all, but the last thing I would base an FPS on. Even analog sticks are more precise.

[couchmonkey]

I actually find tilt steering just as good or better than the analog sticks...it's not more precise, but it gives me precision without having to perform such fine manipulations with my thumb.

But pointing is definitely more precise and I'd definitely go with that for FPS.  They just need to get the turning right.  It'll come in time...just as graphics get fine-tuned on traditional consoles, the controls will be fine-tuned on Wii.

[Ceric]

Rayman has a lot of different control setups in it.  But by far the one I hate most on a regular basis is turn-key style.  It worked ok for Super Monkey Ball but in Rayman they had you flying a bat with it and it felt terrible.  You have to hol teh Wiimote like a pencil to get any sort of accuracy out of it.  

Though personally, I pay for an on-rail shooter like the one in RRR.  Though I would expect better graphics, more locals, and variety.

[Stimutacs Addict]

....and perhaps branching paths and the option to backtrack at will... so you follow set paths but have choices at every corner... plus add the ability to take cover at will (instead of most on-rails shooters where certain locations allow you to duck for cover, you could press the Dpad/c/z button to choose to peek around a ledge/wall/corner)

..but the real fun in fps is owning your roomates.. so in that sense, I NEED a solid FPS (im borrowing Red Steel from a friend next week so i'll probably have different opinions of how it should all go down).

[ryancoke]

I think one of the upcoming FPS' should try an alternate control sceme where you control the movement of the player with the stick on the nunchuck and use the wiimote strictly for aiming within the screen.  Remember Goldeneye? The analog stick was used to control all movement except for strafing and when you used the R (or L, it's been a long time) for precise aiming.  I think it would be a good option for players who don't like the "bounding box" control style.

[Ceric]

I'll agree with branching paths but no user driven backtracking.  If you missed it you'll have to play through the level again.  It  could include Orignal Doom/Faceball style Deathmatch play.  Online and local.

[trip1eX]

Monkey Wars has the best fps controls so far and it's 1 of 50 mini-games in Super Monkey Ball.

[dubnobasswithmyheadman]

Quote
Originally posted by: couchmonkey
I actually find tilt steering just as good or better than the analog sticks...it's not more precise, but it gives me precision without having to perform such fine manipulations with my thumb.

But pointing is definitely more precise and I'd definitely go with that for FPS.  They just need to get the turning right.  It'll come in time...just as graphics get fine-tuned on traditional consoles, the controls will be fine-tuned on Wii.

What I had in mind (when i made the original post) was actually a hybrid approach. Use the pointer as long as the remote is pointed at the screen, but keep track of the tilt at the same time. Then if the pointer goes out of view, use the tilt as the primary control, to keep the player at least somewhat in control.

I'm sure you could optimize the transition from off-screen to on-screen by using some sort of average between the two methods. When you're near the edge of the screen, the crosshair position could be an average between the pointer position and the tilt.

Finally, when the pointer is near the center of the screen, the game should recalibrate the current tilt value as the neautral position. This is just in case the tilt can become skewed over time.

My entire motivation was to eliminate the wackyness that happens when you lose line of sight to the sensor bar. It's the most immersion-killing part of the Wii in my opinion. Just when you really get into a game, that's when you start making more exaggerated movements... and then you get nailed with this bug. So that's why i'm trying to think of solutions for it...