• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Is Sixaxis more accurate than Wiimote?

TheSonicRetard said:
Now, we can read those charges, and use it as input data from our controllers. In order to make this data useful, we need to position accelerometers in three strategic positions. If anyone has ever studied aerodynamics before, you should be familiar with pitch, rotation, and yaw. These are the 3 kinds of acceleration we need to measure on the controller.

wii1.png


Acceleration along the Y axis.

wii3t.png


acceleration along the X axis.

wii2.png


and acceleration along the Z axis.

There's a fundamental problem here: you're confusing linear and rotational motion. The accelerometers measure linear motion, as depicted in the first two pictures. They do not measure rotational motion. They measure the angle of tilt by using the gravity vector. As such, they can measure this angle even if the remote is stationary. To say it differently: the wii remote does not measure tilt by measuring rotation.
 
JodyAnthony said:
even though i am a nintendo fanboy all the way, i still think that the sixaxis motion sensing seems much smoother, at least in my experiences.

Depends on the games you play. If you played, say, Kororinpa and then Lair, you'd swear the Wii's motion-sensing was far superior. The Wii's motion-sensing is actually very good and very sensitive when a game is calibrated correctly. Some games don't get it quite right. Same for Sixaxis I guess. There are times I've used it where it's felt clunky as hell, yet Flower has some of the best motion-sensing of any game. And yet when turning the valves in Killzone 2, I sometimes had to do the action two or three times for it to work properly.

In short: results inconclusive. :lol
 
Hero of Canton said:
Depends on the games you play. If you played, say, Kororinpa and then Lair, you'd swear the Wii's motion-sensing was far superior. The Wii's motion-sensing is actually very good and very sensitive when a game is calibrated correctly. Some games don't get it quite right. Same for Sixaxis I guess. There are times I've used it where it's felt clunky as hell, yet Flower has some of the best motion-sensing of any game. And yet when turning the valves in Killzone 2, I sometimes had to do the action two or three times for it to work properly.

In short: results inconclusive. :lol

no matter what, both systems use six axis accelerometers to determine tilting and movement. they output the exact same raw numbers. it's all up to the developer to use these numbers well.
 
leroy hacker said:
There's a fundamental problem here: you're confusing linear and rotational motion. The accelerometers measure linear motion, as depicted in the first two pictures. They do not measure rotational motion. They measure the angle of tilt by using the gravity vector. As such, they can measure this angle even if the remote is stationary. To say it differently: the wii remote does not measure tilt by measuring rotation.

Right, these aren't HOW it's measured, it's the implication. I've simplified the end result for clarity, otherwise we'd be talking physics here - calculating resultant forces of multiple forces (gravity, acceleration, etc) and quite frankly, I don't want to discuss such physics in a NeoGAF thread :lol

feel free to expand, though. The wiimote and DS3 pretty fascinating technologies, even if I don't like them much.
 

lowrider007

Licorice-flavoured booze?
soldat7 said:
I wonder if the Wii-mote could handle Flower with the same grace and precision as the DS3.

I've often wondered this, in fact I remember saying in the flower thread after completing flower I felt compelled to purchase a Wii becuase I found the motion controls so exhilarating and wanted more, but people seemed to advise me against saying that the Wii doesn't really have any games that have implemented the motion controls in the same say, or as well, is this really the case ?
 
TriangularDuck said:
no matter what, both systems use three axis accelerometers to determine tilting and movement. they output the exact same raw numbers. it's all up to the developer to use these numbers well.
Fixed for physical and mathematical truth.
 
TheSonicRetard said:
It'd be nice if people actually read the thread.

I know... quite a few explanatory posts in this thread, and they aren't even that long! And yet people do not read them.


lowrider007 said:
I've often wondered this, in fact I remember saying in the flower thread after completing flower I felt compelled to purchase a Wii becuase I found the motion controls so exhilarating and wanted more, but people seemed to advise me against saying that the Wii doesn't really have any games that have implemented the motion controls in the same say, or as well, is this really the case ?

There are plenty of Wii games that use motion sensing extremely well. ExciteBots, for instance... and Excite Truck as well, though ExciteBots really improved on the controls. And with Motion+ things are about to get a whole ton better. The sword-fighting and jet-ski games are ridiculously fun :D

And if you get a balance board, Shaun White on Wii is pretty damn awesome. Very fun snowboarding game.
 
leroy hacker said:
Fixed for physical and mathematical truth.

Yes, the other 3 axis in the name SIXAXIS refers to X, Y, and Z axis of the standard dualshock 3 controller minus the motion controls. (for those confused by the Z axis, it's the pressure sensitive triggers).
 

dyls

Member
In one of the "Iwata Asks", he says this:

Perhaps. This goes back a little bit, but before the Wii project began, I started to have doubts about what had become the current game standard of exchanging 60 signals per second. For example, if you put a camera on a car moving at high speed, and then try to reproduce the images taken by that camera at 60 frames per second, then the whole point of the car moving at high speed is lost. Now, I've also been interested in pointing devices for a while, but I've always thought that the tracking would be insufficient, and that the pointer would therefore not move as expected, if signals are only exchanged 60 times per second. Around that time, a sensor technology was released that could pick up 200 or 300 signals per second, and I felt that we should take a chance on it. I told Mr Ikeda that it would be worth a shot. This was the only thing I mentioned regarding the controller. In that sense, the union of the pointer and the sensor was very important.

Now he's talking about the pointer feature here, but it seems that if the remote is sending the pointer signals 200 to 300 times a second, it would probably be sending the tilt signals that quickly as well. I don't know if Sony went with that same rate or the normal 60 signals per second, but given how quickly the whole motion thing was put together for the PS3, it might well be the latter.
 

Haunted

Member
I always likened the Sixaxis motion control to that of the Nunchuck. The form factor is probably the deciding factor here.


It's a total non-issue though, since it all depends on the implementation anyway. Flower controls feel really really good, and Lair controls like dogshit. Both are using the same motion controller.

There are abundant examples of the same phenomenon the Wii side so I'll not even mention those. :p
 
dyls said:
In one of the "Iwata Asks", he says this:



Now he's talking about the pointer feature here, but it seems that if the remote is sending the pointer signals 200 to 300 times a second, it would probably be sending the tilt signals that quickly as well. I don't know if Sony went with that same rate or the normal 60 signals per second, but given how quickly the whole motion thing was put together for the PS3, it might well be the latter.

...he's talking about IR tracking, not accelerometer-based motion controls. That doesnt apply to the conversation at hand.
 

Apenheul

Member
The SIXAXIS is actually somewhere inbetween the Wii remote and the Nunchuck device if it comes to sensitivity, at least that's what my memory tells me from my experience with them. Wii Remote is 3.4G, SIXAXIS (I think) is about 3G. Both devices had a sample rate of 100hz. Most games would probably work on both devices, it's just that you can max-out the accelerometers in the SIXAXIS controller much easier.

I wrote a driver for the SIXAXIS that did sub-frame interpolation and (optional) lowpass filtering to smoothen results from the controller, and spline interpolation to predict accelerations that exceeded 3G (that would otherwise be clipped because the accelerometer was maxed out). Got way better results after that.
 

karas

Member
Nintendo did not cut any corners with their accelerometers in the Wiimote. They use the Analog Devices ADXL330 sensor wich is really awesome and not the cheapest in the market. An accelerometer of as good as this used to cost hundreds of dollars (and weigh a few pounds) just a few years ago.

The Nunchuck on the other hand uses an STMicroelectronics LIS3L02AL sensor, which is not quite as good but probably cheaper.

Sony never said officially which sensor they use for the Sixaxis/DS3 but it should be at least as good as the one in the Nunchuck.
 
It is a largely moot question considering motion+ is coming soon.

But the sixaxis has always kinda sucked since a single unit motion sensitive controller blows.
 

Squeak

Member
Basically tilt control is a worthless gimmick. It doesn't do anything a joystick doesn't do far better. What it desperately needs is force feedback, which the joystick at least has in it's passive form from the spring.
Motion sensing in general on the other hand is a whole other matter. It's just a pitty it should take Nintendo 3 years to come with a solution that is at least somewhat satisfactory.
They should have just have bitten the sour apple from the start and gone with a fully fledged radio or magnetic tracking solution, instead of horsing around with half arsed optical and accelerometer solutions.
 
Squeak said:
Basically tilt control is a worthless gimmick. It doesn't do anything a joystick doesn't do far better. What it desperately needs is force feedback, which the joystick at least has in it's passive form from the spring.

Play some better motion controlled games...there are many motions that would be awkward if reproduced on an analog stick or with buttons. Play Wario Ware: Smooth Moves and tell me that it would be better on a normal controller because there's absolutely no way it would be

Motion sensing in general on the other hand is a whole other matter. It's just a pitty it should take Nintendo 3 years to come with a solution that is at least somewhat satisfactory.
They should have just have bitten the sour apple from the start and gone with a fully fledged radio or magnetic tracking solution, instead of horsing around with half arsed optical and accelerometer solutions.

Well you're probably the only one who thinks the $40 remotes were expensive enough...because what you're proposing would have surely made them cost more
 
A question:

Can the sixaxis measure yaw, ie rotation parallel to the ground? If it can't the name is simply a lie since it would be missing one degree of freedom.

As far as I know, the Wii remote accelerometers cannot measure yaw. You can't do it with the gravity vector and I know of no games that use yaw. So if the sixaxis only contains an accelerometer similar to the Wii remote it should not be able to measure yaw, and hence would only be a fiveaxis.
 

KurowaSan

Member
i just went through the thread, and well, about the pointer and the distance between the controller and the sensor bar... under good light conditions where the wiimote can pick up the IR lights from the sensor, the farther away you are from the bar, the smoother it should be, as long as it can track both IR light points.

If you get closer, the distance between the two IR lights get bigger, so it 'occupies' more space in the area the wiimote can 'see', so as long as you can track both IR lights, if you get farther away, there's more room in the area the IR camera picks up for you to move the pointer.
2e1c9b5.jpg



Then again, that will also make the wiimote less accurate when tracking Z movement.
 
leroy hacker said:
A question:

Can the sixaxis measure yaw? If it can't the name is simply a lie since it would be missing one degree of freedom.

As far as I know, the Wii remote cannot measure yaw, which is rotation parallel to the ground. You can't do it with the gravity vector and I know of no games that use yaw. So if the sixaxis only contains an accelerometer similar to the Wii remote it should not be able measure yaw, and hence would only be a fiveaxis.

Well isn't that wasn't estimated using the IR sensors (at least within the bounds of the screen)?
 

mclem

Member
Neiteio said:
As a PSWii owner who's played gobs of Flower and Warhawk on the one hand and Excite Truck and Mario Kart on the other, I can honestly say I prefer the Sixaxis. It feels far more organic, like an extension of the mind. I'm honestly surprised how well it works and how tastefully it's been applied in the games I've played. I also can't fathom how there are still PS3 users out there who doubt the merits of motion control when some of the best games on the system make amazing use of it.
I suspect it's no coincidence that both of those games match nicely to the concept of a flight yoke.

The problem I have with the SIXAXIS motion controls is the fact that it's not a particularly natural setup, holding an item of that shape with two hands which you then tilt. The Wii remote's shape and one-handedness lends itself to a greater range of motions; I couldn't imagine swinging the SIXAXIS like a tennis racket or golf club, although it's quite capable of detecting such a motion.

However, when the SIXAXIS is representing something meaningful which *does* fit the way you hold the SIXAXIS, I do agree that it has the edge over the wii; and flight yokes are a natural fit for the means of holding the pad.

Neither controller quite feels right for steering wheels, but I marginally prefer the wii after having huge problems trying to steer in Motorstorm.

However, when you are going with flight yoke controls, and they're done well (go away, Lair), it's just great.
 

Ranger X

Member
The Wii remote is basically 4 axis so theorically motion recognition should be better on the Sixaxis.

In reality though, it seems easily for devs to implement the Wiimote in their game. Maybe because of the shape of the controller is also gives the impression it's better.

I suspect the Wii motion plus to add 2 axis and it will use the pointer for relativity. At this point the Wiimote will surpass the Sixaxis even in theory.

.
 

Squeak

Member
Zoramon089 said:
Play some better motion controlled games...there are many motions that would be awkward if reproduced on an analog stick or with buttons. Play Wario Ware: Smooth Moves and tell me that it would be better on a normal controller because there's absolutely no way it would be

SM doesn't just use the tilt "function", in fact I can only think of a couple of sub games using it. The rest use the accelerometers in some other way.
Nevertheless there is a huge difference from a casual "burst" game like this, to something more involved. Why do you think most large games still has support for joystick controls even if they have tilt controls as default?

Well you're probably the only one who thinks the $40 remotes were expensive enough...because what you're proposing would have surely made them cost more
If it could replace both the accelerometers and the IR-cam and the sensorbar, then it wouldn't necessarily be more expensive, and even if it was slightly more expensive it would still be so much more convincing and satisfactory to use that it would be an investment in the future.

As it is now we have to drag along with shitty accelerometer controls that in real games aren't good for much more that waggle, swipes and tuds; IR tech, that has a very limited range of motion and is somewhat unstable and finally M+ which is an addon and not even a complete solution (there will still be drifting as there is no absolute reference point, hence there will be a need to recalibrate often, IE lift the mouse, and that in turn has to be worked into all games).
 
As someone who has played a fair share of Wii and PS3 games that use motion controls. It isn't a contest the Wii Remote is far better.

Threi said:
I've said it many times and I will say it again:

I equate motion control to rumble: cool when it does it well, but not terribly necessary. Pointer controls, on the other hand, I equate to an analog stick: a near complete improvement over previous ways input and needs to be in every single controller next gen.

While I do definitely dig the motion controls on many games. The pointer is something that needs to be standard in controllers.
 

onipex

Member
Brandon F said:
Not trying to incite a flamewar, a friend asked this over the weekend and I really didn't know. Barring Motion+, is the Sixaxis more capable of tracking accurate motion movements than the accelerometers in the Wiimote?

No, the Wiimote is better at tracking accurate motion.
 
If Flower is the true justification of the sixaxis, I hope a lot more people are working on games that require slight gradual motions and fall apart whenever you have to circle back or do precise maneuvering. Sweet!
 
TheSonicRetard said:
No, I get what you mean. You mean that logically the closer you are, the more light has to hit the remote. It does. However, the wiimote doesn't measure the amount of light hit. It's binary in its implementation. It either sees the LEDs or it doesn't.

In which case it cannot detect the "size" of the dots (since size isn't the correct term, actually. it's more like the amount of light hitting the camera), but rather how close the two points are.
What I was getting at is when you stand far away the left and right led's are closer together, so when you rotate the wiimote the circle arc is smaller and more sensitive, after posting though I realised what you really meant so it didn't make much seance, the thread was moving too fast to alter my post.
 

dyls

Member
TheSonicRetard said:
...he's talking about IR tracking, not accelerometer-based motion controls. That doesnt apply to the conversation at hand.

But as I said, if the remote is communicating the pointer signals at a rate of 200+ times per second, wouldn't it stand to reason that the data from the accelerometers is being transmitted at the same rate?
 

CamHostage

Member
Neiteio said:
Did Sony really add Sixaxis at the last second? It seems too well-done to be a quick add-on. I mean, in my opinion it works better than the Wiimote, and the best part is it's put to the service of some triple-A hardcore titles that'd be a dying breed on Wii. I'd personally love to see what Sony could do with pointer technology; that's all their missing on the control front, and it's a big one.

It's often said that Sony tossed in the motion control of SixAxis at the last minute as a response to the unveiling of the Wii. I don't follow that line of cynicism (though if it had not been hinted that Nintendo was looking at stuff like tactile control, it's hard to say if Sony would have gone that route), and in fact wonder if the early PS3 banana-shaped batterang DualShock controller (which was maybe designed to be held at a variety of positions along the control bar and could be used with more versatility as a motion controller, it also looked a lot like Microsoft's SideWinder controllers, some of which had motion) was actually designed that way for the motion control function that they were secretly experimenting with.

ps3_controller.jpg


No matter whether Sony was copying or experimenting, however, the fact remains that they still hid it from developers until the E3 unveiling (even the revered Kojima team had no knowledge of it) and made no great efforts after that unveiling to make the technology easy to work with. It may not have been an afterthought as some claim, but it for sure has not ever been a priority, and that's a big reason why the technology usage has lagged way behind despite the PS3 controller having the capability to do many things Wii can (nothing with the IR of course, but Kororinpa and Mercury could both have been ported... and could have remained unsold in bargain bins for two game platforms instead of just one.)

TheSonicRetard said:
Also, to throw my own two cents in here, both the wiimote and the SIXAXIS are just as accurate as one another. There's little to no difference in the technology. The main culprit here lies within the SDK distributed. Nintendo spent millions of dollars creating a robust, easy to use SDK designed to make it easy to interpret accelerometer control. Sony, by contrast, did not, and the main reason people say they feel different (despite having near-identical technology) comes down to the software involved.

Interesting, that makes sense.
 
leroy hacker said:
A question:

Can the sixaxis measure yaw, ie rotation parallel to the ground?
Yes, the Sixaxis appears to be capable of measuring yaw. This motion is used for control in one part of Linger in Shadows. I assume this means there are two accelerometers separated across the body of the controller; I have no idea how you could do this with just one (since there's no gravity vector to help, or IR device like in Wii).
 
Durante said:
Thank you. In terms of actual motion sensing, Wiimote, nunchuck and sixaxis are more or less equivalent.
It's really too bad that Nintendo didn't update the nunchuck at the same time with Wii Motion Plus. I have yet to play a game with good nunchuck motion control, and complete dual motion sensing capabilities would have opened up so many possibilities.

Another one for the Wii 2 wish list though. That, along with better battery life.
 

Totobeni

An blind dancing ho
CamHostage said:
It's often said that Sony tossed in the motion control of SixAxis at the last minute as a response to the unveiling of the Wii. I don't follow that line of cynicism (though if it had not been hinted that Nintendo was looking at stuff like tactile control, it's hard to say if Sony would have gone that route), and in fact wonder if the early PS3 banana-shaped batterang DualShock controller (which was maybe designed to be held at a variety of positions along the control bar and could be used with more versatility as a motion controller, it also looked a lot like Microsoft's SideWinder controllers, some of which had motion) was actually designed that way for the motion control function that they were secretly experimenting with.

ps3_controller.jpg


.

I think the the grips of Original PS3 controller can actually help improve Sixaxis , if it's the same shape like the Alps controller , then it will give you the "wiimote wheel" in control feel and this can improve the player performance with sixaxis especially in flying games ( like Warhawk ) and Racing since the grip can give the player more control .

829098-alpspad_super.jpg



they should at least release limited edition of Original DS3 , I will buy one for $100
 

Windu

never heard about the cat, apparently
They both probably have the same capability in terms of tilt action. So it probably goes back to the developers and how well they can implement it.
 
leroy hacker said:
A question:

Can the sixaxis measure yaw, ie rotation parallel to the ground? If it can't the name is simply a lie since it would be missing one degree of freedom.
dnf5gp.gif


WOW! Did you see that yaw control?
 
Apenheul said:
The SIXAXIS is actually somewhere inbetween the Wii remote and the Nunchuck device if it comes to sensitivity, at least that's what my memory tells me from my experience with them. Wii Remote is 3.4G, SIXAXIS (I think) is about 3G. Both devices had a sample rate of 100hz. Most games would probably work on both devices, it's just that you can max-out the accelerometers in the SIXAXIS controller much easier.

I wrote a driver for the SIXAXIS that did sub-frame interpolation and (optional) lowpass filtering to smoothen results from the controller, and spline interpolation to predict accelerations that exceeded 3G (that would otherwise be clipped because the accelerometer was maxed out). Got way better results after that.


Ding, Ding, Ding, we have a winner! The rest of the bullshit can just go away.
 

Josh7289

Member
3 pages of this shit?

If you're not a developer with experience programming for both the Wiimote and the PS3 controller, then gtfo because you can't accurately say anything about them.
 

batbeg

Member
BDGAME said:
Look at this: http://www.youtube.com/watch?v=wy5Cc2Jj3vA&feature=related

very precise, it's like a 1:1 tilt control or close, very close.

Oh. and I believe this game is impossible to be done without tilt controls.

While I won't deny the fun of motion controls for tilt-oriented games (as they bring me closer to the classic Labyrinth experience), how are these games impossible without tilt controls? Marble Madness, Mercury Meltdown, Super Monkey Ball all did pretty damn good. Or is it something specific about Kororinpa (which, to my great shame, I have not played yet).
 

RurouniZel

Asks questions so Ezalc doesn't have to
batbeg said:
While I won't deny the fun of motion controls for tilt-oriented games (as they bring me closer to the classic Labyrinth experience), how are these games impossible without tilt controls? Marble Madness, Mercury Meltdown, Super Monkey Ball all did pretty damn good. Or is it something specific about Kororinpa (which, to my great shame, I have not played yet).

Not impossible in the literal sense. More in the "This is fucking IMPOSSIBLE!!" followed by throwing the controller in a huge fit sense.
 

HUELEN10

Member
timetokill said:
As a Wii owner who has played Flower, statements like these make me open my eyes and then roll them.
As a Wii owner who OWNS the damn game, let me agree with you with a facepalm to that guy.
 
Don't we know the actual parts used in each case? I'm almost positive I've seen specific part numbers mentioned at least for the Wiimote.

If so, somebody ought to be able to dig up data sheets which would probably be more interesting than this back-and-forth fanboy crap.
 

Neiteio

Member
HUELEN10 said:
As a Wii owner who OWNS the damn game, let me agree with you with a facepalm to that guy.
You're facepalming somebody who owns a Wii and a PS3 as well. I'm going off of first-hand experience with people who come over and are more enthralled with Flower's handling than tilt-control games on Wii (Excite, Kart and Monkey Ball in particular). It really is eye-opening for some Wii users to see their system is not alone in competent motion control, and that when well-implemented Sixaxis can meet or exceed some of the best tilt-control games on Wii. It opened my eyes; I never thought of my PS3 as a "motion" system because Sony hasn't exactly billed it that way, so I was surprised to see how well motion controls worked in a number of its games.

I know that's threatening to some, but try to think of the proliferation of quality motion control as good news for gaming in general. Again, I really hope Sony and Microsoft take the last step and add IR controls so they can be put to the service of grand-scale shooters like Killzone, Halo, BioShock, et al. Hell, I hope Sony and Microsoft adopt motion controls in general so they can be consistently put to use in quality hardcore titles.
 
I am not a fan of waggle controls.

In my experience most motion controls on the PS3 are implemented poorly and for some motions you have to hold the Sixaxis in away it was not designed to be held.

However there are a number of games that use motion controls to good effect and the holding position suits the Sixaxis.

Flower
Toy Home
Flow
Super Rub 'n' Dub.


Not only is the use of motion controls in these games well done, it as helps convey the physicallity of the gameplay in a much better way than you could with traditional controls.

In Flow & Flower the motion controls capture the organic feeling of movement.

In Super Rub 'n' Dub the motion controls simulate holding a bowl of water. The inetia physics also contribute greatly to the realisim of the motion.

In Toy Home you are controlling a wind-up toy car, the motion controls exaggerate the movements an enhance the sense that you are playing with toys, rather than a precision racing machine.

On the subject of pointer tracking, I am really happy with the improvements made to the PS3Eye. This does a very good job of tracking motion. We just need some more games.

The recent PS2 Eyetoy Hero game comes with a toy sword. This does a very accurate job of mapping the movements, and makes a lot more sense than the original Eyetoy Play series. The Eyetoy camera is fine if the object being tracked is something predetermined like the sword in Eyetoy Hero.
 
Top Bottom