 |
|
|
| |
Feature: Improving Your Motion Control Games
by Staff [Console/PC, Mobile Phone, Mobile Console]
|
|
| |
|
July 27, 2010
|
| |
Wii MotionPlus developer AiLive writes, in Gamasutra's latest feature, on how to improve the accuracy of motion control input in games for Wii, Move, and Kinect.
The feature, written by four of AiLive's staff -- including Rob Kay (formerly of Harmonix) and company co-founder and former SCEE R&D member Ian Wright -- details techniques for separating different types of control input and also for making the game more comprehensible to your users, as below:
"Any game that needs to recognize the kinds of moves players are performing faces the same requirement of segmenting continuous motion into discrete categories. The game needs to identify when a motion has begun and when it has completed. There are two primary ways to do this. Which is best ultimately depends on your game design.
You can ask the player to segment their own motion. Normally this takes the form of holding down a button while the player performs a move.
The advantage of this approach is ambiguity is entirely removed in the sense that the player is responsible for telling the game when they start and stop a move. The disadvantage is that pressing a button to signal intent complicates the control scheme and may seem artificial.
The second approach is to get the game to segment the player's motion. In this approach the player simply concentrates on moving. A good approach, suitable for many games, is to tell the player when to perform a motion within a time window (e.g., "Ready, Go!") that naturally arises during gameplay and is communicated to the player. For example, Namco's We Cheer 2 takes this approach."
The full feature, which goes into much greater depth -- and includes some of AiLive's predictions for the future of motion control -- is live now on Gamasutra.
|
| |
|
|
(which processes on board)?
It moves from saying that you can only detect good wiimote rotation with gyros - to noting that Move uses a camera for much improved tracking.
Wellll, the wii uses a camera too. and it can leverage the 2 IR blobs in the sensor bar to figure out some depth and rotation. (er,... at least I thought I was seeing the camera leveraged in this way for motion control in the first red steel, and several of the wario-ware minigames.???) perhaps I was wrong?
*Assuming I'm right for a moment: this frustrates me all to hell. Because you're implying that the old wiimote is just crappy accelerometers. In my experience, the old wiimote is STILL BETTER than the upcoming HD competition, because it still offers the most precise "light gun" experience. I'm probably just being overly sensitive, but I've been waiting to see how PS3 and Xbox will gloss over their failings in this arena. So omitting it entirely in the article made me nervous about the writer's goals.
Would very much like to know if developers can leverage wii's imager alongside the accelerometers and gyro data. Or if you omitted it because you can't actually use it.
* in related news, I had no idea the Move had a magnetometer.
Is this accurate?!?!?! Arent' those expensive as all hell? Don't they measure the magnitude and direction of a magnetic field? (so what's generating this? PS3 eye? wha?)
Very curious if this detail is a slip up, or the scent of an amazing ace up Sony's sleeve. ???
* I've read that wiimote's black and white imager captures and process at around 100 fps.
- I thought I read the the PS3eye could go up to 120 fps if it halved resolution (since it's basically streaming web cam images to the PS3, and not processing them in the Eye - USB can only send so many big color images so fast). So i thought it was good rule of thumb that if you see video of yourself displayed in game, you are running at the higher resolution/slower speed. (which then needs to be uncompressed, processed, recognized - which might explain some of the lag complaints).
I'd guess standard Move setup uses faster frame rate (and doesn't show you video of yourself), and the device uses a big ball so it can show up in the lower resolution. But I dunno. would love to read this definitively, somewhere. ...Or are developers only allowed to access data after it's been processed? So nobody outside Sony knows for sure?
- Annnnd at what speed is Kinect running? Is it also sending jpeg compressed images over usb? how fast? (or is it sending more than one camera feed over usb? eek?). Again, are developers only allowed to access data after it's been processed?
I think theses would be much more interesting information tidbits to share, rather than weird confusing vague shit like "In other words, Kinect gains more general tracking capabilities at the cost of more expensive inference."
* In related complainery-fu, this struck me as the most clumsy confusing puzzle-of-writing in the whole piece: "In general, robust motion recognition requires less information than robust motion tracking. That being so, many devices that cannot support tracking nonetheless support recognition."
I just found myself taken out of the article, wondering if this sentence had be rewritten endlessly by committee until it no longer made sense.
Basically, i think it is trying to say:
'You don't need much motion data to trigger our recognition tricks. Because of this, devices that suck at accurate tracking can still offer some usable recognition tricks/triggers.'
yeah? i'm not sure. weird?
(i mention this point purely with the hope that one of the 4 authors could turn to the others and say "you see? You &#$%^@ guys skrewed the article!" Since i used to find my text rewritten by engineers and higher-ups)
@driver 01z: Yes, this is the problem we face as left handed people. It could be easily avoided if they made left handed models, a Wiimote with an analog stick, a Nunchuck with an A button and they force developers to include a lefty switch mode. As it stands I cannot play games like Metroid Prime 3, Punch-Out, or even Mario Galaxy properly. It's very sad for me being left out. Wii Motion Plus makes this even worse in games like Zelda Skyward Sword unless they offer you a lefty mode as 1 to 1 movements are obviously not going to work for lefties.
@ Eric
The IR pointer function of the Wiimote allows it significantly superior and close to absolute rotational determination within the reading window compared to Move, while Move has the advantage in absolute translational determination. Referenced rotational sensing helps the Wiimote perform well in shooters. Move would require the user to either wave the wand around a great deal (not ideal for shooters) or supplement data with its MEMS sensors, which as the article explores, are subject to accumulatable error - depending on the performance of the magnetometer at least.
Something I'm interested in: Obviously the article makes reference to motion sensors in general and a little about specific platform idiosyncrasies, but I would have liked to learn more about hybrid information gathering - the wiimote can establish its absolute position but only intermittently (when it's pointed at the 'sensor bar'), but can you/how can you optimally marry that into information from the relativistic sensors?
Prior controllers are not a problem for people of either handedness because you're just using your fingers and thumbs. Are you telling me that you type on a keyboard with one hand? Is it harder for you to type with the fingers on your left hand? Probably not, little finger movements to press keys and buttons shouldn't effect people of either handedness.
The Wiimote is entirely different and requires arm and hand movement, which does effect handedness, and coordination. As a lefty the natural thing is to pickup the Wiimote in my left hand. You obviously pick it up with your right hand or else you'd be complaining about the same things I am. The Wiimote lacks an analog stick, which is used for movement in most games, all of which have previously been controlled with your left thumb (and are still controlled with your left thumb if you are right handed). I'm now forced with either controlling movement with my right thumb, which goes against every game I've ever played and is unnatural to me, or I'm forced to hold the Wiimote pointer in my right hand which is painfully unnatural feeling and highly inaccurate being that I'm left handed.
Write your name with your offhand and tell me how it feels? Throw a baseball with your offhand and tell me how it feels? You don't see the problem with the Wiimote/Nunchuck because you don't have to deal with it. Play a game the way I do and suddenly you'll understand. And this is only in games that even ALLOW you to switch which hands you hold the controllers in. Games like Punch-Out! have no such option and force me to swing with the opposite hand that is shown on screen. I.e. for left handed punches I have to punch with my right hand! How irritatingly confusing! In No More Heroes (one of my all time favorite games!), I'm forced to do the opposite of what is shown on screen to perform supplexes because the game assumes I'm right handed.
Not to mention Wiimotion Plus games that don't have an option to switch handedness. I'm horribly afraid that I will not be able to play the next version of my favorite game, Zelda, because of the use of Wiimotion Plus. The perfect example of how the Wii is designed for Right Handed people is Zelda. Link who was left handed in every game prior to the Wii, is now right handed because most people are right handed and how awkward would it be if you're swinging your right hand while Link is swinging his left hand? If the new Zelda offers no such option to switch to lefty mode... I'll have to use Backhanded Slashes to control Link making Normal Slashes, do you see the problem now?