Today's hot video: Wiimote on PC acceleration reading
by Ross Miller 
{ Dec 1st 2006 at 12:25AM }
We're going to be honest with you: three out of the top five most viewed clips on YouTube today are gameplay videos of Tales of Destiny, but without knowing the Japanese language we have a hard time enjoying it. So we move on to the next top video (notice how we didn't say "hottest" in the headline), which is yet another Wii-related clip.The video is an early documentation from WiiLi.org, a group that is trying to figure out all the technical workings of the Wii remote, with the ultimate goal of creating a "live Linux CD that will allow the Wii to boot into a fully fleged Linux Desktop without the need of a modchip, or any hardware modification." You may not understand (or care to understand) their purpose, but let's be honest: their logo, a penguin dressed as Mario, is just plain adorable!
The gurus are able to detect the accelerations (and it appears somewhat its orientation) of the Wiimote with their computer, but it is unclear to what extent they understand the wavelengths presented. Some might find it yawn-inducing, but it's a promising start.
Reader Comments (Page 1 of 1)
T-Man @ Dec 1st 2006 12:42AM
they aren't "wavelengths" they are graphical representations of the data being generated by the wiimote.
not having watched the video I would guess that it is a real time charting of the Gs being experienced by an individual accelerometer or the value on an axis being generated by a gyro.
Esch @ Dec 1st 2006 1:02AM
I don't understand how any Wii owner could not be excited by this development -- I find it odd that you mention that this is boring and useless quite a few times in your post.
As far as the video goes, this'd be great for teaching elementary physics.
PTT @ Dec 1st 2006 1:20AM
Thankees for changing it, I've noticed the change. It was just... I don't know, one of those nerd things. It was kind of like how some people write "Rediculous," and there's always the one guy that pops up and says "...it's RIDICULOUS! Sorry, I'm hard on spelling errors." ...?
Anyway, thanks for listening to random commenter-person, and indulging ^^
Al @ Dec 1st 2006 1:24AM
Nice! It looks like they're making progress. I wish them luck, some really cool software could result from this.
Chris @ Dec 1st 2006 2:00AM
@#1
" they aren't "wavelengths" they are graphical representations of the data being generated by the wiimote."
This graphical repesntation is a wave pattern with wavelentgths moron. All joystiq is saying is that it's unclear what they are saying it means.
mocax @ Dec 1st 2006 2:07AM
I wonder if they can transmit the wiimote sounds to a bluetooth handsfree set
T-Man @ Dec 1st 2006 2:46AM
"This graphical repesntation is a wave pattern with wavelentgths moron. All joystiq is saying is that it's unclear what they are saying it means"
a wavelength is a measurement. a distance. like 2mm or 5ft or 6 microns. You can't see a distance, you can see the two objects that define this distance (in the case of a wave two peaks). But you can't see distance. This graphic is a real time plotting of a line graph, data point from wiimote vs time. Thats all, the only waves involved are the RF waves transmitting the data from the wiimote.(that wavelength is 122.4mm iirc)
Sam Lee @ Dec 1st 2006 4:21AM
Wow, maybe we can develop games on PC that uses Wiimote
Or maybe sooner or later, someone will hack the PS3 to use Wiimote
reppy @ Dec 1st 2006 6:14AM
How long until someone makes a hentai game for use with the Wii-mote?
Lou D @ Dec 1st 2006 7:03AM
Once they realize that the Wii-mote reports sensor bar co-ordinates on a 1024x768 plane, we will have a nice handheld mouse for our PC's to...as soon as you mount 2 IR sources 1.5" appart at the top or bottom of your monitor...and if your resolution in only that high...
ConstyXIV @ Dec 1st 2006 8:13AM
@8:
Does the PS3 run Linux?
Does the PS3 have Bluetooth support in Linux?
If you answered "yes" to both those questions, it's not too far off (though perverse and utterly wrong).
Skwurl @ Dec 1st 2006 8:32AM
Guys, it's called a tripod.
Either that or don't drink 20 cups of coffee before using a hand held video camera.
Good discovery though.
webrunner @ Dec 1st 2006 4:41PM
Honestly, guys:
The wavelength is the distance between two identical points on a wave.
The frequency is how fast they come over time.
The signal on a wave can be encoded by changing amplitudes on a wave.
I don't think this qualifies as a wave unless you move the remote semi-rhythmically.
It's true that most waves are just a function of amplitude over time, but they do need to be fairly regular (have up parts and down parts in a consistent fashion) for it to work.
Moogle @ Dec 1st 2006 8:58AM
Hmm, I wrote a huge post bitching about the apparent (but not confirmed) lack of rotation, making the wiimote a 3 axis device, and technically inferior to the PS3 controller if it *really* is 6 axis.
The truth, I realized, is that even with no rotation sensors, 3 axis of accelerometers gives you 5 total axis. That is really enough to do pretty much anything *except* rotation about the axis of gravity. That screwed me up a little in excite truck at first, since you have to turn the wheel in the vertical plane, not horizontal.
When I first heard of the wiimote and no one had yet figured out the IR/camera thing, what I first thought of was a project of one of my favorite professors:
http://www.gvu.gatech.edu/~jarek/videos/Twister1.mov
http://www.gvu.gatech.edu/~jarek/videos/Bender1.mov
http://www.gvu.gatech.edu/~jarek/videos/Camera1.mov
The device they're using has small wired sensors that precisely read a magnetic field. The system costs tens of thousands of dollars from what I heard.
Unfortunately the wiimote will never be that good, but it can still, I think, do almost everything in the first two videos well enough, except twisting the plane vertically. (as long as you point it at the screen frequently to reorient it)
If it really came down to a game that really required it, you could always clip a small box onto the back of the wiimote to add whatever sensor you wanted.
As evidence of the lack of rotation, for example in Wii Sports golf, you can swing simply by pointing the wiimote downards, letting it fall, and jerking it back up, simulating the centrifugal force the game expects of a circular swing. Maybe they're ignoring the rotation data if it *is* there, but it definitely isn't necessary. I wouldn't be surprised if Nintendo decided that rotation sensor didn't add anything to gameplay and didn't put any in.
The only big thing I can think that rotation could add is the ability of FPS's like red steel or metroid to have better left/right turning control even after the pointer has left the screen. It would make it easy to turn around quickly if the player is jumpy. Maybe. Then again, if you assume the rotation point is the player's wrist, you can do this using only accelerometers anyway.
But even then, I'm still a little miffed that they chose to save $5* per remote by leaving it out. Sure, it adds up to hundreds of millions of dollars, but they should cater to my unreasonable and unnecessary demands, dammit!
(* the cheapest I can find +-3g 2 axis accelerometers is about $10 each in packs of 1000, they could get it cheaper but still need more circuitry on the wiimote to accomodate it)
ConstyXIV @ Dec 1st 2006 9:12AM
@13:
Umm, the Wiimote does do rotation. Go to your local Gamestop, go to the Wii kiosk, hit the home button, and twist the remote around. Watch the cursor.
Paul P. @ Dec 1st 2006 9:18AM
That's not the type of rotation he's talking about.
RUSKULL @ Dec 1st 2006 2:43PM
To #14 The pointer is a separate technology.
All in all, this video actually really puts into perspective just how complex that little white controller really is.
Moogle @ Dec 1st 2006 10:49AM
It can extrapolate that from the direction of gravity or/and from the angle of the dots on the sensor bar. (which is what I meant by 5 axis from 3 accelerometers) Go play excite truck and try to steer by moving your hands forward and back instead of up and down. It just doesn't work.
I'm not saying I know for a fact that it doesn't tell 3 axis of rotation, I've just found no evidence yet that reassures me that it is there.
boywundr @ Dec 1st 2006 11:52AM
Is it me or is this dude ROCKING side to side in the beginning of this video? Between his voice and his "movements", he SCREAMS Rain Man. Yeah... Wopner at 5 o'clock.
Anyway, this is interesting, but way over my head and understanding.
Dave @ Dec 1st 2006 11:23AM
Moogle, if you're referring to depth, Wii Boxing uses depth. If that's not what you mean, can you give an example to each dimensional axis in which you are referring to, including the ones that it does have?
Thad @ Dec 4th 2006 1:33PM
"You may not understand (or care to understand) their purpose, but let's be honest: their logo, a penguin dressed as Mario, is just plain adorable!"
Well thanks, Joystiq, I really appreciate that condescension. I'm just a dumb gamer, spare me the fancy computer talk and show me something cute!
Paul @ Dec 1st 2006 1:50PM
adolph, your statement is ridiculous. hackers doesn't mean evil doing, but reverse engineering a closed process.
read the whole fucking article before posting, they aren't hacking the servers, they just want wiimote support on linux for a Wii Linux live-cd.
All Your Lost Socks @ Dec 1st 2006 2:04PM
Um, adolph, this has nothing to DO with WiiConnect24. They are just working with the Wiimote as a control device, just like one would hack a controller, keyboard, or mouse for extra functionality. What you refer to is cracking, which is destructive, not hacking, which is meant to be constructive. Besides, the post you quoted said nothing that you bashed it for.
This is indeed geeky, really cool, not useless. Good to see some intelligent detective work going on with this thing!
Majortom @ Dec 1st 2006 4:02PM
Too bad they will need a disc with a wii barcode on the udnerside close to the middle.
That is one of their protections.
crono141 @ Dec 1st 2006 3:29PM
Can't wait for the Wiimouse thats PC compatable. I know their ultimate goal is to get linux on CD to work with Wii, but I think that producing an app that would turn your wiimote into a mouse pointer/gyromouse would be infinitely more useful.
kiswa @ Dec 1st 2006 3:44PM
Could someone fill me in on where these other axes came from? Last time I did any three dimensional (3D) rendering, there were only 3 axes (X, Y, and Z) used to define space.
Am I missing something here?
Jason @ Dec 1st 2006 5:53PM
Moogle, you are absolutely right and I'm a bit miffed about the whole thing myself. All they needed was ONE gyroscope in addition to all the accelerometers. They could have used the gyroscope they used in WarioWare Twisted. There simply is no way to measure rotation on a plane that is parallel to the earth with the current setup. This is precisely why the controls for all the FPS games are a bit jacked up right now.
I can't believe this... the controller was so close to perfection.
All Your Lost Socks @ Dec 1st 2006 7:15PM
Moogle, read this - when you get down to the bit about the cows, your question is answered (I think):
http://wii.ign.com/articles/748/748607p3.html
Moogle @ Dec 1st 2006 8:39PM
AYLS - sound exactly like excite truck except for the jumping.
kiswa - Rotation. Basically if the wiimote is pointing in only one direction, moving it up, down, left, right, fore, and back can be recorded as 3 float values. + or - G force, in effect. Whichever axis happens to be facing vertically will normally read +1 G's. If you are allowed to spin the remote in each plane perpendicular to each axis, you need 3 more float values to describe the position/motion of the wiimote in space. You can determine 2 of those three rotational axis by using gravity. But the other axis would still be kinda useful. In some cases you can probably fake it by assuming the player is turning with his wrist or elbow, if you don't need the side-to-side movement for anything else.
Jason - They actually needed 3 gyroscopes, or 3 additional accelerometers. What if you're holding the wiimote on it side and rotating it about the vertical axis, or on its end pointing up or down?
Jason @ Dec 2nd 2006 12:42AM
Moogle, I wouldn't be thinking of covering every possible situation, just the most obvious one: camera panning. Right now that's pretty much impossible! The more sensors the merrier, but that one gyroscope combined with the existing accelerometers would be more than enough for most applications.
I guess, with the single 3-axis accelerometer they have, they could use the side-to-side forces to pan the camera (assuming it's precise enough), but any tilt to the left or right would be instantly interpreted as a camera pan, making it likely that the camera would go completely nuts!
If they had a second 3-axis accelerometer toward the base of the Wiimote they could throw out data where the side-to-side forces were equal on both sensors (like when you twist/tilt the controller lengthwise like a log) and only accept those where the further sensor measured more side-to-side force than the closer one. That *might* work, but it's a whole lotta effort just to do what one single-axis gyroscope could already do for you.