Support NeoGAF

[evilromero]

As far at setup goes, in the interview Merrick said it doesn't require any type of calibration for your display other than just putting the sensor bar near it. He also said it works with any display type.

Let's hope that includes projectors. And let's also hope it's as painless as he states. It better be just placing a sensor on your tv. But even then, how will the sensor attach to your tv? How will it LOOK? That's important. People don't want an ugly eyetoy sitting ontop of their tv or center channel speaker.

 

[GaimeGuy]

Let's hope that includes projectors. And let's also hope it's as painless as he states. It better be just placing a sensor on your tv. But even then, how will the sensor attach to your tv? How will it LOOK? That's important. People don't want an ugly eyetoy sitting ontop of their tv or center channel speaker.

oh give me a break. If they can have an ugly console type thing or an ugly dvd player or a big fucking tv taking up space, they can have a little sensor.

and WHY wouldn't it work on projectors? the SENSOR is what's used to locate the controller. That's why they designed it to use a sensor.

 

[evilromero]

oh give me a break. If they can have an ugly console type thing or an ugly dvd player or a big fucking tv taking up space, they can have a little sensor.

and WHY wouldn't it work on projectors? the SENSOR is what's used to locate the controller. That's why they designed it to use a sensor.

I'm sorry but there is an aesthetic principle for living room design now. Several tvs take advantage of modern aesthetics, as do some home electronics. People aren't forced to hide their toys anymore. Now, the Rev is a fine piece of hardware (visually). I don't think there is any fear of that being seen as geeky.

If Thom from the Fab 5 can make Televisions mix and match with living room furntiture, then we've reached a new age.

 

[Robobandit]

The 1up guys mentioned doing loops with the plane demo, so I would think it is capable of registering arcs..

DEMO: PILOT WANGS
Manipulate a biplane through the air, trying to fly through rings scattered around the Isle Delfino hub world of Super Mario Sunshine.
IMPRESSIONS: This was about all the different ways the Revolution can detect tilting the controller. It was as if the controller was the airplane itself - as long as your movements weren't too sudden, the on-screen action would mimic your movements with very little lag time. After about a minute I was pulling dramatic dives and loop-de-loops, bullseye-ing plenty of rings.

http://www.1up.com/do/newsStory?cId=3143782

or maybe it is late and I'm just reading it wrong. I think there is definitely more to this controller than we've been told..

 

[heidern]

It's hard to tell if he ever actually worked on hardware. It looks like he did a lot of work on APIs and the like, but it's hard to tell from a piece like that. Interesting, though

Hey you only said technical guy first. Besides, someone that's had years of experience in things like APIs will have at least an appreciation for hardware. And seeing as he is in the know regarding the system I would assume knows what he's talking about.

AFAIK, only the "acceleration" of up+down is detected... not how far its moved or its x,y coordinates. I figure you'd need to have additional "sensor bars" on the floor below, cieling above, and in the back of the room to get actual 3D coordinates of the remote.

Tell me, do you need more than two eyes to see things in 3 dimensions?

Yes, but all with the remote at an arcing angle? That's a hell of alot of info to "crunch" in real-time.

It's not all that much really. There's probably a sensor near the top of the remote. By triangulating this with the 2 main sensors you can get the x,y,z co-ordinates. These are in the form of 3 floating point numbers. 3 floating point numbers are the equivalent of one corner of a triangle. It's not much info for systems that process millions of triangles.

My theory is that there are two sensors on the remote, one at each end. The x,y,z co-ordinates of both of these are calculated, so we have two points. What do you get when you join two points in a 3D space? A line(Vector) in a 3D space. This line represents the rotation of the device. But remember, this line is calculated by the game. All the hardware needs to do is get the two points. So getting the co-ordinates is easy. Getting the rotation is the equivalent of animating the forearm of a character, hardly a problem for systems that can animate thousands of characters with many limbs.

Anyway, from the IGN faq:

The sensors read the pointer's every move in real-time space. They can detect up, down, left and right motion, and also translate forward and backward depth. The controller's sensors also recognize twisting, rotating and tilting movements. In short, any motion made by arms and wrists can be translated to Revolution games.

Another question is... do the seperate controllers have any idea how apart they are from each other?

The controllers have no knowledge of anything. All they do is transmit info to the Revolution to interpret. The Revolution actually does everything. Since the Revolution has access to all the controllers and all their exact positions it can take into account how far apart they are.

 

[xabre]

My theory is that there are two sensors on the remote, one at each end. The x,y,z co-ordinates of both of these are calculated, so we have two points. What do you get when you join two points in a 3D space? A line(Vector) in a 3D space. This line represents the rotation of the device. But remember, this line is calculated by the game. All the hardware needs to do is get the two points. So getting the co-ordinates is easy. Getting the rotation is the equivalent of animating the forearm of a character, hardly a problem for systems that can animate thousands of characters with many limbs.

I agree, the device simply determines the x, y and z co-ordinates through a method of triangulating the distance of the remote in respect to the two sensors. So at the remotes initial position when starting up a game, sensor A determines its distance from the remote, sensor B determines its distance from the remote and the initial position can be given co-ors of 0 for each the x, y and z axis. Any movement away from the initial position changes the distance of the remote in respect to the two sensors and the console is able to determine the new co-ordinates and thus determine its position in 3D space. I have no idea how the maths work, but the principle is straight forward and simple.

Also, in regards to representing rotation I also agree, however, in order for the device to detect both pitch and yaw you would need a plane, not a line, which means four sensors in the remote unit.

As for how the pointer aspect of the device works, I have no idea.

 

[Aurora]

This was posted from firefly in the 3DRealms forums. And from someone that has tried the controller.




http://forums.3drealms.com/ubbthrea...Board=othergames&Number=949214&page=0&fpart=2

Sounds fantastic

It just seems to be getting better and better.

 

[JoshuaJSlone]

AFAIK, only the "acceleration" of up+down is detected... not how far its moved or its x,y coordinates.

I don't know exactly how this controller is working, but if the system could always tell the vertical speed/acceleration, it'd be a simple calculation to keep track of height relative to the starting point.

Also, in regards to representing rotation I also agree, however, in order for the device to detect both pitch and yaw you would need a plane, not a line, which means four sensors in the remote unit.

Only 3 points would be necessary for a plane.

 

[wonderfuldays]

Oh jebus christ.... This is the last thing I wanted to hear.

I said several times in other threads that I can't imagine the Rev Con being better than the $99 Gyration mouse and keyboard combo we bought a year ago and found absoultely horrible in just about everything.

Now you are telling me that it's exact same tech? YEESH!!

Interest in the Rev Controller -1 googleplex.

You are such a funny guy Shog.
So you are trying to telling us that you are using that kind of mouse now? :lol

 

[KINGMOKU]

You are such a funny guy Shog.
So you are trying to telling us that you are using that kind of mouse now? :lol

From what I can tell of his post's, he seems like the type of guy that has hairy arms and just smashes his keyboard with his fist's.

"Shog smash stupid rev remote!" "Shog hate change" "RAAAAAAAHHHHHHHH!!!!"

 

[Gahiggidy]

I don't know exactly how this controller is working, but if the system could always tell the vertical speed/acceleration, it'd be a simple calculation to keep track of height relative to the starting point.

....

errr... but wouldn't the game quickly lose track of where the starting point is? Part of the reason we'll need to place the sensor-bar under the tv is so that the z-axis has an origin point. Don't see where a corresponding y & x-axis origin points are available to give true 3D coordinates.

 

[SpoonyBard]

So no-one understood what I was trying to say on the first page? I know that PEACE guy had no idea...

 

[PkunkFury]

errr... but wouldn't the game quickly lose track of where the starting point is? Part of the reason we'll need to place the sensor-bar under the tv is so that the z-axis has an origin point. Don't see where a corresponding y & x-axis origin points are available to give true 3D coordinates.

HOw is the sensor bar only providing a z origin point and preventing reading x and y data. I'm really confused where you are coming from here. You know the sensor bar contatins two sensors right? Those 2 points can then by triangulated with the remote location to find the position of the wand in space

 

[wonderfuldays]

First.. I called this 4 years ago when Nintendo "heavily invested" in Gyration.
http://www.gyration.com/intl/uk/pr-nintendo_uk.htm


Here's the tech description...
http://www.gyration.com/intl/uk/technology_uk.htm


Here's a vid of the GyroRemote... With almost all the same features...
http://www.gyration.com/files/demos/GyroRemote_web.wmv

Thanks for the great information.

 

[SpoonyBard]

HOw is the sensor bar only providing a z origin point and preventing reading x and y data. I'm really confused where you are coming from here. You know the sensor bar contatins two sensors right? Those 2 points can then by triangulated with the remote location to find the position of the wand in space

But only in 2D space. For 3D location you need four sensors.

 

[DDayton]

I've been assuming that you would calibrate the unit for vertical and horizontal motion in each game, as that could differ in each setting. If we are right in understanding that the Revolution can track the controller in 3D space, then the calibration wouldn't be for the system to register movement... rather, it would be for the user to define their "zero position". The Rev should be able to get an "absolute" status and location check on the controller at any time (in full 3D space), as well as get info from the controller on things like rotation and such... but as the player's relative location may change from game session to session (depending on whether you sit down, stand up, etc.), the game needs to know what to register as the "normal" location for each session.

Right?

 

[mariosbrother]

How about this...

3 sensors. 1 in the system itself and the 2 that sit by the screen (whatever the screen may be... tube, plasma, LCD, projection). The controller itself has two transmitters in the top and bottom. The 3 sensors are used together to triangulate the exact 3-D positions of the two controller transmitters, thus providing a 3-D vector representing the direction the controller is facing. The 2 sensors at the screen are used for determining the screen position relative to the controller, and hence, what part of the screen the controller is or is not pointing towards. The mysterious top of the controller could even include a sensor that helps in detecting this. That takes care of 3-D movement in 3-D space. The rotations? Sensors internal to the controller that actually detect this sort of movement. Multiple controllers? So what? No more sensors needed than the 3 to get positioning.

The controller itself has rumble. The attachments very well could - and each attachment will most definitely be powered by the main remote unit. Nintendo knows well enough that people aren't going to want a thousand different batteries all over the place. They'll most likely include or at least offer a rechargeable battery pak that will be used to power the remote along with any attachments.

My only real curiousity is the detection of exactly where on the screen the remote is pointing to. I would think that you'd have to place a sensor in one upper corner of the screen and the opposing lower corner, so to detect a vector thats pointing above the 2-D plane created by the screen. I'll have to think about this one more....

 

[SpoonyBard]

I've been assuming that you would calibrate the unit for vertical and horizontal motion in each game, as that could differ in each setting. If we are right in understanding that the Revolution can track the controller in 3D space, then the calibration wouldn't be for the system to register movement... rather, it would be for the user to define their "zero position". The Rev should be able to get an "absolute" status and location check on the controller at any time (in full 3D space), as well as get info from the controller on things like rotation and such... but as the player's relative location may change from game session to session (depending on whether you sit down, stand up, etc.), the game needs to know what to register as the "normal" location for each session.

Right?

Yeah, and I think that's what the home button does.

 

[argon]

Arstechnica has an interesting article speculating on the Rev remote technology:

The remote control thingie (which I'll call the "wand," I guess) sends out simultaneous IR and ultrasound blasts. The Revolution detects the time delay between when the ultrasound hits the right sensor and when it hits the left sensor. This gives the position of the wand relative to the vertical centerline of the TV screen. (Imagine a plane slicing the TV in half vertically and extending out perpendicular to the surface of the screen. The time delay gives the distance of the wand to the left or right of that plane.)

The ultrasound signal travels only about 300 m/s, but the IR signal travels much faster—so much so that the delay between when it hits one sensor and when it hits the other is negligible. Since the IR and the ultrasound blasts are emitted simultaneously, the time difference between when the IR hits the sensors and when the ultrasound hits can be used to measure the distance of the wand from the TV screen.

Finally, the wand itself has something in it that can detect the angle of the wand relative to the plane of the floor. (Well, relative to the direction of gravity really, but presumably the floor is nice and level.) This could be something like the sudden motion sensor chip in PowerBooks and other laptops, for example. But some of those chips seem to "re-center" themselves if you hold the laptop very still, so maybe the technology is slightly different.

although I don't know how bluetooth would fit into this..

 

[BorkBork]

Arstechnica has an interesting article speculating on the Rev remote technology:

OMFG Revolution is a BAT!

Bats > Owls!

 

[Xrenity]

It is bluetooth actually, confirmed.

 

[Gahiggidy]

HOw is the sensor bar only providing a z origin point and preventing reading x and y data. I'm really confused where you are coming from here. You know the sensor bar contatins two sensors right? Those 2 points can then by triangulated with the remote location to find the position of the wand in space

Well, I'm assuming the sensor bar isn't triangulating but simply caculating the "ping-times" to gauge the distance from tv set.

 

[DrGAKMAN]

But only in 2D space. For 3D location you need four sensors.

I think there's two sensor in the remote-controller (one near the front end and one near the back end). Don't quote me on that, but it just makes sense.

Each of the 4 sensors (two in the bar, 2 in the remote-controller) one communicates to one another determining speed, distance, dirrection, motion to & from each other. Just the two sensors inside the remote-controller (always being the same distance apart of course, but) can tell what angle/possition the remote-controller is in by if one is being lowered/elevated differently than the other or if one is being rotated around the other. Adding the communication with the two sensors inside the sensor bar and the distance from them to the front sensor in the remote-controller and thus it can tell how it's being moved (up/down, left/right, towards/away). Futhermore, adding in the triangulation of the back end sensor in the remote-controller and BAM the tracking can sense any angle as well...so any kind of swing (as shown in the video), twist or tilt can be detected on any of the aformentioned planes (up/down, left/right, towards/away).

I imagine Bluetooth handles the communcation, but the logic from each of the four sensors is like sonar. The front end sensor may also use some form of IR to determine more precision in aiming and the back end sensor would contain some sort of gyration detection in direct relation to the front end sensor for determining tilting, rotation & angulation. Or Nintendo could've put that window on the front of the controller to throw reverse engineers off...in fact, them not explaining the technology plus expressing their concern over the competition stealing their ideas in the past pretty much tells me Nintendo won't tell how it works, nor will they want us to know.

 

[SpoonyBard]

Well IR is obviously unusable for anything, but using ultrasound instead of RF (Bluetooth) for position finding is pretty easy to believe. Because the speed of sound is so much lower than the speed of light, calculating the delta times between the sensors and the controller would be much easier and cheaper. Bluetooth (or rather, Nintendo's own version of Bluetooth) would be only used for communication between the controller and the console.

edit:
If the controller has one ultrasound transmitter, I'd think that you'd need four receivers. If the controller has two transmitters, three receivers would probably be enough... I'm not sure though, it's hard to think these things without visual aid. And for that I'd need a 3D modelling program or something

And two transmitters would drain the battery more.

 

[PkunkFury]

Gyration mice are only half of the story here. They do not track position of any sort, only rotation.

The closest device I can think of to what Nintendo might be emulating for the rev controller is the P5 data glove by Essential Reality, shown at the following link (if someone can host the image, please do):

http://www.vrealities.com/P5.html

The P5 glove comes with a sensor bar, much like the revolution will. The glove can be tracked in 3D space in all 3 directions of position, and all 3 directions of rotation (6 degrees of freedom in total).

The best part is the price!! notice it is only 60 dollars on the site I linked, and I remember seeing it for around $40.00 regularly when we picked one up at work. That is 40 dollars for the sensor bar, and a glove with flex technology enabling the tracking of bendable fingers! One could assume the revolution remote is far easier to manufacture than a glove would be, however wireless/battery and rumble may drag the price back up. All in all I think this tech could be priced per controller, at about the same price as the X-Box 360 controller. And if only one sensor bar is needed for controllers, then the price of extra controllers could be significantly less!!


Tracking:

The P5 tracks rotation and position of the glove using a sensor bar. Now, I've seen enough of the inside of the bar to know it only has two receptors, but the triangulation in 3D crux is a good point, I'm not quite sure how the glove is tracking it but it works. It is possible the receptors themselves are handling one degree of freedom on their own, and the reason for two is to triangulate the other degrees. Maybe each receptor has three detection points that can somehow work together?

The glove itself has little LED lights around its casing (I think 7 in total) which are probably what the sensor bar is tracking. You don't really see the lights light up, but the bulbs are visible on the glove (the orange points in the photo) so I guess they are just off of the visible spectrum. My theory is that these lights blink on and off in rapid succession, and the computer knows which light is on at each time. The receptors then determine the position of each currently visible light and use that data to identify the orientation of the glove. All 7 lights don’t need to be seen to gather the glove information, only 3 or maybe 2 with some good prediction algorithms. That is why the lights are spread out all over the glove, to be sure some are visible at all angles. As with any optical based system, occlusion is a problem.

Does the rev actually use this technology?? I don’t know. I don’t see LEDs on the casing of the rev controller, but the remote is a far less complex shape. Maybe just having those bars on the front and back of the remote are generating the same effect?? I’m just assuming the rev uses similar technology since it achieves the same end result, and is easy to afford.

My guess is that the rev only needs one detection point to be visible at a time (hence the one “LED” in the front of the controller). This is because the LED/tracking system is only generated position data, unlike the P5’s position and rotation. Rotation detection requires that the location of multiple tracking points be compared in relation to one another to determine an orientation. I’m guessing rotation data is sent to the rev from an internal gyro, similar to the one in WarioWare Twisted, and the gyro information is combined with the position information for a full 6 DOF!!!
Awesome!! As a side note, the P5s biggest problem I think is the rotation, I will try it again at work on Monday. I imagine Nintendo will have improved a bit on the tech anyway by the time the rev releases, so expect the remote to work Pretty Damn Well

Edit: thought the expansion slot on the back was another window (oops)

 

[PkunkFury]

Well IR is abviously unusable for anything, but using ultrasound instead of RF (Bluetooth) for position finding is pretty easy to believe. Because the speed of sound is so much lower than the speed of light, calculating the delta times between the sensors and the controller would be much easier and cheaper. Bluetooth (or rather, Nintendo's own version of Bluetooth) would be only used for communication between the controller and the console.

Yes, I asked about the tracking technology being used before and someone replied bluetooth, but I don't think the press release meant blue tooth was tracking, rather that bluetooth was communicating the data. My initial thought was that the rev used radio frequency, like an intersense tracker, but now my P5 theory has me wonder if some form of light tracking is involved. I wish someone at TGS had asked more hardware related questions about this stuff

 

[SpoonyBard]

That P5 thing is pretty interesting, I think Rev controller could be using something very similar. Those LEDs are most likely normal IR-LEDs you have in all your remote controllers...

With IR you need constant visual connection to the reveiver unit, so I don't think it's viable for normal living room use. Using ultrasound would be much better. You need either multible transmitters or multible receivers for the system to be accurate, whatever system is used. Using multible IR-LEDs, like in P5, would of course be the cheapest option.

 

[PkunkFury]

Here is a link to Intersense’s consumer level VR wand, which also does everything Nintendo’s controller will do, but in an entirely different way than the P5 glove accomplishes things:

http://www.intersense.com/products/prec/is900/PCTracker.pdf

The Intersense tracker determines position using ultrasonic transmissions which are sent to the wand device from tracking strips. The time in which the wand receives them is calculated to determine the distance between the wand and the strips, and this data is triangulated to determine the wand’s position. No triangulation mystery here as every time I’ve used one of these we have had three trackers (or more) set up around the screen. You’ll notice that even in the link I posted, the diagram shows 3 trackers being hooked up to the processor.

Orientation is determined by internal gyros, and I believe some accelerometers are also built in to help deal with tracking and prediction when the tracker gets lost.

As far as the revolution remote goes, I think it will be closer to the P5 than the Intersense tech, but I hope I’m wrong. Intersense tech is much better and includes many more fail safes for when the positioning signal gets occluded. If Nintendo can borrow anything from the Intersense system, I hope it is the internal gyro for dealing with rotation, as tracking rotation optically seems to be far to risky, and occlusion could be frequent. Let’s hope they don’t cheap out on this

I’m guessing Nintendo isn’t using the ultrasound tracking intersense uses, simply because of the Infra-red looking windows at the front of the controller. I can’t find a price on this consumer version of the Intersense tracker, but I’m dead certain it is a heap more than any of us would like to pay for a game controller.

 

[Particle Physicist]

...

 

[SpoonyBard]

The Intersense system seems to be bit overkill for just gaming purposes. It doesn't have to be THAT accurate.

 

[PkunkFury]

The Intersense system seems to be bit overkill for just gaming purposes. It doesn't have to be THAT accurate.

Yeah, all of my experience with it has been using VR caves or walls where the accuracy is a bit more crucial over a wider range. But still, the press won't be so positive if Nintendo puts out a controller that doesn't work as accurately as current control methods. I only posted about it because it is another device that accomplishes exactly what the revolution controller will be doing, and because I'm hoping Nintendo is using the same internal gyro idea for roatation. Since WarioWare had an internal gyro at no extra cost to the game cart, I should hope Nintendo can afford to include one in the controller as well.

 

[Shao]

I'm not as techy as you guys but I find this interesting. btw the control doesn't have a "window" at the back, only on top. The back is where the attachments fit in. oh and attachments will feed off the remote's batteries hence the cable - and Nintendo may consider using rechargeable batteries as well as a charging dock. This is promising news, hopefully the batteries can be good judging from the DS/GBA ones. What do we do when they run out and we want to keep playing though??

I'm not sure this necessarily works like a spatial mouse. i.e. I can take my PC mouse, face a different direction from my screen but pushing forward with my mouse will always equate to up on the screen. This shouldn't be an option with the Rev.

I think the Rev will always map the controller's position in relation to where the screen is. While full motion capture is possible lets not forget that the main area of focus is breaking down the barrier between TV and gamer. This is just my theory, but it should relate to stuff like "centering the mouse" in 3D space - an idea I feel is too techy for non-gamers.

In the demos the people had to stand on a red "X" on the floor, so position and distance had to be configured in some way I guess, but that doesn't explain for height differences or different ways of playing/holding/aiming etc.

Also it seems people can change and shift position freely. In the demo video we have people changing play position, this guy who has supposedly playtested it explains how he started by holding his hand in the air but eventually decided to rest it on his lap.

Just curious but what kind of calibration does Warioware Twisted use?

I would imagine off the top of my head that the sensors are used with the wand's pointer ability, you can point at the screen from anywhere and it will know. The other degrees of movement are done by gyros. Using the two systems this way, wouldn't you be able to play sitting down, pause, then go to the other side of the room and continue playing as if nothing changed?

 

[KonVex]

I think the IR is only used to switch the console on/off.
If this would be done using the regular bluetooth signal, you could switch off all systems in your neighbourhood at the same time!
Also you couldn't switch the system on, because the controller hasn't linked up with the system yet.

 

[Yoschi]

Hello there

I only read thru half of that thread, so don't kill me if some of that info has already been posted pls
I just think this thread needs more technical descriptions:

How a Lightgun works:
http://entertainment.howstuffworks.com/question273.htm

how a classic gyro works:
http://www.accs.net/users/cefpearson/gyro.htm
http://en.wikipedia.org/wiki/Gyroscope

and video's of a classic gyroscope and inertia(explanation):
http://www.gyroscopes.org/1974lecture.asp

gyrobot & iPod remote & iBOT Movie:
http://www.barello.net/Robots/gyrobot/index.htm
http://homepages.cae.wisc.edu/~tyser/ipod_remote.htm
http://www.dynopower.freeserve.co.uk/homepages/movies/Ibot.mpg

CR03 against vibration and shock:
http://www.spp.co.jp/sssj/qanda-e.html#nazegaibu
and how it works:
http://www.siliconsensing.com/documents/Info/Why Choose The Silicon Ring Gyroscope.pdf


some gyro's or accelerometers for those who make project's themselves:
http://www.acroname.com/robotics/parts/R40-PG2033.html
http://www.memsic.com/memsic/products/product.asp?prodid=40
http://www.analog.com/en/cat/0,2878,764,00.html

other interesting info-link's:
http://csdl2.computer.org/comp//proceedings/ismar/2003/2006/00/20060268.pdf
http://www.baesystems.com/engineering/examples/vsg.htm


i am currently experimenting myself with accelerometers to build something similar to the gyration mouse. so i did some search a while ago to figure out wich would be the best sensor to use (and cheap too). The accelerometers i am currently experimenting with are the one's from memsic (the MXD2125GL). For those who want to know, they measure the angle of tilt, NOT the velocity of moving in a direction(I thaugt so first)

I hope I don't went too off-topic with this...

 

[Frankfurter]

Someone posted this in the Shack thread: " "How does it compare to a mouse?" From what I experienced, it seemed to be more precise than a mouse, but it's also much faster because it requires only a much smaller movement of the hand to achieve the desired effect. You just instantly point the controller at any part of the screen and bam!, that's where you're looking.

Woah! That sounds really great. More precise than a mouse... nearly unbelievable.

 

[Yoschi]

forgot to say, i can make a video for you to see how an accelerometer would work, with the help of an oscilloscope.

ya just have to ask

 

[Shogmaster]

You are such a funny guy Shog.
So you are trying to telling us that you are using that kind of mouse now? :lol

The mouse is not being used anymore, no sireebob. It sucks. And we had that for about a year. Stopped using it about 3 months ago.

From what I can tell of his post's, he seems like the type of guy that has hairy arms and just smashes his keyboard with his fist's.

"Shog smash stupid rev remote!" "Shog hate change" "RAAAAAAAHHHHHHHH!!!!"

Yeah dufus, I hate change so much that I used the mouse when you didn't even know such things existed. *rolleyes*

How about the simple fact that I'm trying to relay to you expereience I've already had with the same technology? It's not a magic bullet you guys are imagining. It's got a lot of limitations and won't work the way you guys think. It's relatively a simple device as far as how it controls the cursor. Nintendo can design and code software around it to maximise it's usefulness in games, but it won't be doing all those magical things you guys are imagining right now.

But the real problem I had with the gyration mouse was that it was too imprecise to replace a real one in a desktop environment. It's simple as that.

 

[nightez]

Am I the only one that thinks programming this thing is going to be a nightmare?

 

[Frankfurter]

Why?

 

[jgkspsx]

Am I the only one that thinks programming this thing is going to be a nightmare?

Yes. If Nintendo doesn't do a half-assed job in signal conditioning, it's all just a bunch of analog values to be read. That's half the beauty of the idea -- from the programming side, nothing much changes. That's why the Prime 2 conversion was so fast.

 

[DrGAKMAN]

I think the IR is only used to switch the console on/off.
If this would be done using the regular bluetooth signal, you could switch off all systems in your neighbourhood at the same time!
Also you couldn't switch the system on, because the controller hasn't linked up with the system yet.

That makes alot of sence...but where's the IR receiver on the console? It's hidden on my VCR's & DVD player, but they also have display windows they can hide it behind. I'm thinking if any sort of IR is used, it's for what you say (powering on & off the system) and maybe for adding precision to aim/pointing dirrectly to the screen (an invisable laser point that bounces back from the screen to determine exact possitioning).

 

[heidern]

Before I did get it wrong, you do need 3 points to calculate 3 degrees of rotation. So my theory is that there are 3 sensors in an equilateral triangle formation in the controller. Maybe an inch or so apart. Sensors can calculate the difference in joints on a finger so I imagine them being real close won't make accuracy a problem. The position of these is triangulated and sent to the Revolution. So the game receives 3 x,y,z co-ordinates whether they be (0,0,0) or (-100,4,27) or even (1,009,345, -263,856, 2). That's it, the game only recieves 9 numbers.

I think the Rev will always map the controller's position in relation to where the screen is. While full motion capture is possible lets not forget that the main area of focus is breaking down the barrier between TV and gamer. This is just my theory, but it should relate to stuff like "centering the mouse" in 3D space - an idea I feel is too techy for non-gamers.

I think that this won't be hardware dependent. As per my theory, the game only receives positional values for the game to interpret. If developers want to implement centering they can, if they want a fixed start they can as well. The way I think it will work is that acceleration is the key not just the amount of movement. In an fps, if you quickly flick it left the character will sharp turn left. If you slowly flick it left the character will slowly turn left. If you carry on turning it left slowly it won't affect the speed much. Assuming this control is great, aiming at the centre of the screen just requires a quick flick. So to recentre, all you do is slowly move to your new centre and the a quick flick to the centre from there.

The problem with having the screen as the centre is that your position of use is limited. Say you want to play with the remote facing left being the centre and just flicking from there or maybe the most comfortable centre in terms of hand position is at a slight angle, not smack bang in the centre of the screen? That would require more adjustment trying to find the centre. I imagine it will be more intuitive than that. As long as the player is facing the screen, I think the natural adjustment will enable any player to recentre without any difficulty.

Am I the only one that thinks programming this thing is going to be a nightmare?

Nightmare isn't the right word. You have to assume that this controller is better or at the very least a different but worthwhile experience. Otherwise the whole thing's entirely pointless anyway. But assuming that this has value, and taking the assumption that game programmers want to make good games, what I'd say is that programming this is going to be a time eater to some extent.

Game programmers have been programming in 3D for a decade now. To them this is no more difficult than any animation or physics or collision detection work they have done before. It is not a new mystery. It requires the skills they have used before like vector mathematics and mapping analogue interface to analogue game movements. Taking the assumption that this adds value to the game, game programmers will want to work with this thing. They'll probably be falling over each other for the opportunity. So again, noightmare isn't the word. It's a challenge, but a good one. Pressing a button to swing a bat is trivial, interpreting a remote full swing is no doubt a challenge, but its one programmers will relish as a way to make a game better.

The other side is publishers. To them, the cost of this is time. Time = money. In this case programmers time. However it is a matter of weeks, maybe months. You're looking at a cost of $1000s, maybe $10,000s in more extreme cases. However that isn't much in a world where million dollar budgets are the norm. So I imagine the tradeoff of higher sales will very much make it worthwhile. EA for example could slack and see Rev Madden launch at GC style 80K. Or they could invest $50,000 into this and get opening sales of 280,000. The risk to reward payoff is very high with this so I think if publishers greenlight a Revolution game in the first place, greenlighting some funds to develop for the controller is a no brainer. Also if you have sequels, you can split the intial cost between them.

 

[Phranky]

Uh... Did it occur to any of you that perhaps Nintendo has already created some ready to use tools to implement the main feature of their console? I very much doubt Nintendo are going to talk about reducing the time and cost it takes to develop a game and then go right ahead and make it difficult to program for.

 

[AB 101]

Oh jebus christ.... This is the last thing I wanted to hear.

I said several times in other threads that I can't imagine the Rev Con being better than the $99 Gyration mouse and keyboard combo we bought a year ago and found absoultely horrible in just about everything.

Now you are telling me that it's exact same tech? YEESH!!

Interest in the Rev Controller -1 googleplex.

Holy shit.

I agree with the Shog man on this one.

 

[heidern]

Uh... Did it occur to any of you that perhaps Nintendo has already created some ready to use tools to implement the main feature of their console? I very much doubt Nintendo are going to talk about reducing the time and cost it takes to develop a game and then go right ahead and make it difficult to program for.

Perhaps. But I reckon most developers will ignore the tools and just develop their own for the specific needs of their game. Really, one programmer could come up with all sorts of stuff in a couple of weeks. I don't think it's too much of an issue really.

 

[Shao]

I think that this won't be hardware dependent. As per my theory, the game only receives positional values for the game to interpret. If developers want to implement centering they can, if they want a fixed start they can as well. The way I think it will work is that acceleration is the key not just the amount of movement. In an fps, if you quickly flick it left the character will sharp turn left. If you slowly flick it left the character will slowly turn left. If you carry on turning it left slowly it won't affect the speed much. Assuming this control is great, aiming at the centre of the screen just requires a quick flick. So to recentre, all you do is slowly move to your new centre and the a quick flick to the centre from there.

The problem with having the screen as the centre is that your position of use is limited. Say you want to play with the remote facing left being the centre and just flicking from there or maybe the most comfortable centre in terms of hand position is at a slight angle, not smack bang in the centre of the screen? That would require more adjustment trying to find the centre. I imagine it will be more intuitive than that. As long as the player is facing the screen, I think the natural adjustment will enable any player to recentre without any difficulty.

Some good stuff, I quite like your idea. Only thing I don't like is it works alot like a mouse in that if I was to hold it upside down and try to aim at the top of the screen - it would aim at the bottom. If I want to hold the stick sideways like a fancy gun slinger, I can't. I don't want it to be like a spatial mouse at all which is kind of what Shog is mindlessly complaining about.

I don't know the ins and outs, but I was just contemplating the uses of the controller that we've seen. In a game where pointing at the screen is not relevant for example, movement would be purely based on gyros (or whatever else inside the remote). Perhaps a game which has the remote act as a plane control stick and you hold it upright, or a fishing rod where you manipulate the on-screen graphics. These don't require the game to directly target anything on-screen, or even point at the screen.

As opposed to a lightgun game where the sensors have to know exactly what angle you are playing from and where you are pointing at, in games like time crisis and FPS, gyros might not have much use at all, only where you are aiming at would matter - you could suggest that the technologies could work independently.

Particularly if you were to move your position, having to reset the stick would not be user-friendly. I want to be able to hit a bullseye on screen even if I 'm walking around my room. This is shown in the demos. Would this be achievable with your set up?

 

[mariosbrother]

Am I the only one that thinks programming this thing is going to be a nightmare?

It won't be a nightmare. Nintendo will provide a robust API I'm sure, so the programmer is only going to have to say "let me know when the controller has been rotated so I know to change some stuff on screen."

It's like programming for wireless networks versus wired networks. Programmers don't say, okay, pulse these bits on this frequency channel and if that's not working, try this other channel, blah blah blah. No. They open a socket connection as they would with any other type of networking connection and all the underlying details are masked.

I'm sure all of the tracking and rotations and other controls are handled in hardware and Nintendo will provide various ways to get at changes in this data.

 

[mariosbrother]

Only thing I don't like is it works alot like a mouse in that if I was to hold it upside down and try to aim at the top of the screen - it would aim at the bottom.

I doubt this will be the case.

 

[heidern]

Some good stuff, I quite like your idea. Only thing I don't like is it works alot like a mouse in that if I was to hold it upside down and try to aim at the top of the screen - it would aim at the bottom. If I want to hold the stick sideways like a fancy gun slinger, I can't. I don't want it to be like a spatial mouse at all which is kind of what Shog is mindlessly complaining about.

No you don't get it. You can do this both ways. By giving the developer the positions of sensor 1, 2 and 3 the developer can interpret those any way he wants. If the controller is turned upside down the developer can account for that and invert the processing. Or not. The choice is there to work this any way you want. Because the interpretation is done in software you have total flexibility. And because this is the equivalent of processing one limb the hit on the processor is pretty much negligible.

 

[John Harker]

Way too much in this thread, so I'll just ask.

Can it/do you think it can sense strength of motion?

For example, if your playing tennis, and you want a gentle hit, you swing softly and that will translate to a soft hit, vs you swing full swing across the entire length of your body, and that translate to a more powerful stroke, or will it just strigger a preset swing that the character is programmed to do?

What I'm saying is, such as in the those videos Nintendo released, will it matter if you just flick your wrist in the direction you want to swing/move across the length of the TV, or will it actually make a difference it you load back your arm, and do a full-strenght swing across the length of the room?

Am I getting my point across clearly? I'm not sure.

Also, it def. has tilt sensors/gyros in addition to the bluetooth external sensors and the internt receiever chip, right?
Anyone know exactly what else is inside it?

Thanks