• 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.

Sony Controller to Be Named "Move." Bye bye Gem? truly outrageous

gofreak said:
I'm not sure I follow you. If you have the camera at a known point with respect to the TV (e.g. in the centre on top of it), and you know the size of the screen (edit: and aspect ratio), and you know the position of the controller in front of the camera (and the distance/orientation for 3D pointing), you have all you need to map the controller location to a point on the TV. I think (!)

If you had a calibration process in a game, you could tell the PS3 how large the TV is (by moving the controller to each of the designated points on screen) and thus telling the PS3 where the position of the cursor is, relative to the Arc's absolute position. However, the game still wouldn't be able to tell if you are pointing the controller at that specific point on the TV (For an example where you are sitting down directly in front of your TV: If you were to hold the controller directly forward in the middle of the screen and you tried to point the controller in the upper right corner of the TV, the PS3 wouldn't recognise it since the physical 3D position would be telling the console that the controller is still pointed at the middle of the screen) making such a process useless for pointing purposes.


gofreak said:
How come? Assuming it was accurate enough, I mean.

Because it would be completely abstract. The whole idea of light gun style aiming is so that you can aim by looking down the barrel of the controller and shoot exactly where you are aiming. Using gyroscopic controls would give you no idea where you are aiming just by looking at the controller (It's like using an analog stick to move an onscreen cursor, but faster)
 

donny2112

Member
Something that just crossed my mind, but would Arc be distracting to use if playing at night with the lights out? Has that been discussed at all, and if Sony has addressed that somehow?

Just wondering. It may be something that can just be "tuned out" with enough use, though.
 

Man

Member
Not sure what you mean. It will see the sphere fine because it's a self-emitting lightsource.
You mean you would be distracted by a light source in your otherwise dark room?


Imagine it to be a torch in Siren 2...
 

Baha

Member
donny2112 said:
Something that just crossed my mind, but would Arc be distracting to use if playing at night with the lights out? Has that been discussed at all, and if Sony has addressed that somehow?

Just wondering. It may be something that can just be "tuned out" with enough use, though.

I'm assuming anything that requires camera manipulation will need light and motion-based actions will work fine in the dark as the wand glows.
 

PSGames

Junior Member
Dogenzaka said:
Yeah I don't see much potential for Natal in "hardcore" game design.
There are no buttons, therefore simple tasks are made unnecessarily complex.

You can't play Tomb Raider on a Natal. Climbing all over your living room and doing backflips just isn't fun, because none of us are Lara Croft.

On the other hand, the PS Controller uses the best of both worlds, in my opinion.

Sony's solution is more of a jack-of-all trades master of none. I don't think it's superior than the Wiimote when it comes to pointer controls. And it's definitely not as good as Natal when it comes depth and full body motion capture.
 

dofry

That's "Dr." dofry to you.
donny2112 said:
Something that just crossed my mind, but would Arc be distracting to use if playing at night with the lights out? Has that been discussed at all, and if Sony has addressed that somehow?

Just wondering. It may be something that can just be "tuned out" with enough use, though.

The less light you have, less brighter the ball needs to be in my opinion.
 

donny2112

Member
Man said:
You mean you would be distracted by a light source in your otherwise dark room?

Yes, particularly one that moves around near your field of vision. Use as a psuedo-torch in an adventure game is intriguing, though...

dofry said:
The less light you have, less brighter the ball needs to be in my opinion.

That's a good option. Has Sony said if they'll be doing this?
 

dofry

That's "Dr." dofry to you.
donny2112 said:
That's a good option. Has Sony said if they'll be doing this?

Haven't heard anything about that and I am second guessing my own statement here. The ball light is needed but how much is enough in a dark living room?
 

gofreak

GAF's Bob Woodward
Nuclear Muffin said:
If you had a calibration process in a game, you could tell the PS3 how large the TV is (by moving the controller to each of the designated points on screen) and thus telling the PS3 where the position of the cursor is, relative to the Arc's absolute position. However, the game still wouldn't be able to tell if you are pointing the controller at that specific point on the TV (You could try pointing the controller at the part of the TV you want to point at but the PS3 wouldn't recognise it since the physical 3D position would be wrong) making such a process useless for pointing purposes.




Because it would be completely abstract. The whole idea of light gun style aiming is so that you can aim by looking down the barrel of the controller and shoot exactly where you are aiming. Using gyroscopic controls would give you no idea where you are aiming just by looking at the controller (It's like using an analog stick to move an onscreen cursor, but faster)


I don't think you need to point out where the TV is or anything with the controller.

Let me put it this way... ASSUME you have good orientation data, and good x/y/z data. ASSUME you know the camera is placed on top of the TV at its centre.

With that data, you can figure out where on a (virtual) plane pressed against the TV the controller is pointing at. You have the xyz relative to the centre of the top of the TV, and its angle, so you (metaphorically) draw a line extending into infinity along the controller, and calculate the intersection point with this plane around the origin of the camera.

Then if you calibrate the system by telling it the size of the TV and its aspect ratio, it can then tell you if this intersection point is actually within the bounds of the TV its sitting on. And if so, where within those bounds.

We can place a question mark over the quality of the orientation data for now, if you like, but assuming you have good such data, all the variables you need are there.

PSGames said:
Sony's solution is more of a jack-of-all trades master of none. I don't think it's superior than the Wiimote when it comes to pointer controls. And it's definitely not as good as Natal when it comes depth and full body motion capture.

That's an exceptionally unfair characterisation.

The wiimote was built first and foremost as pointing device. Not as a motion capture device. As a motion capture device it's technically inferior even to a DS3.

The motion+ introduced 'proper' motion tracking for limited periods of time.

The Sony motion controller is built with motion tracking in mind. You can get 'proper' motion tracking without the time constraints and reset-restrictions of the M+.

I think the jury is still out on how its 'pointing' abilities compare until we have more information, but you can say right now that it is a 'master' in at least one of its domains (i.e. in motion tracking) relative to the other offerings.
 
donny2112 said:
Eh? Just curious, but is the disconnect greater or lesser than controlling a gun that you would need two hands to hold (like assault rifles and sniper rifles) with your thumb? We're still playing video games here, after all. :p

It probably sounds weird but I think controllers are far enough away from reality that I don't really compare the two. I always had the same issue with an arcade lightgun game like Time Crisis, for example, where it would feel awkward when I was using a shotgun or machine gun in the game but I was still using the handgun shaped controller. Maybe it's just me that's got this "uncanny valley of controllers" issue.
 

[Nintex]

Member
dofry said:
Haven't heard anything about that and I am second guessing my own statement here. The ball light is needed but how much is enough in a dark living room?
Wouldn't the ball of light actually be better visible in a poor lit room? It'll be the only light source that the camera needs to track.
 

Ashes

Banned
I think the gyrosopic controls, along with the other technologies should help software by giving more exact measurements of what is happening to the controller in the space provided. This data is key.
 

dofry

That's "Dr." dofry to you.
[Nintex] said:
Wouldn't the ball of light actually be better visible in a poor lit room? It'll be the only light source that the camera needs to track.

Yeah, of course. Stupid me. I'm just thinking about the amount of light needed and how much is distracting but I guess those are really not a problem and thought about before at Sony for sure.
 

Koren

Member
gofreak said:
The wiimote was built first and foremost as pointing device. Not as a motion capture device. As a motion capture device it's technically inferior even to a DS3.

The motion+ introduced 'proper' motion tracking for limited periods of time.

The Sony motion controller is built with motion tracking in mind. You can get 'proper' motion tracking without the time constraints and reset-restrictions of the M+.
I'm not sure I agree completely. How do you expect to correct drift with Arc? The camera don't have all data available. It won't measure orientation, for example.
 

gofreak

GAF's Bob Woodward
donny2112 said:
That's a good option. Has Sony said if they'll be doing this?

They haven't said, but it would be a clever thing to do.

Software could tell the controller to turn the light up or down depending on the light levels in the camera image. Would be a good way to optimise battery life too.
 

donny2112

Member
[Nintex] said:
Wouldn't the ball of light actually be better visible in a poor lit room? It'll be the only light source that the camera needs to track.

I could see a problem if the lights were suddenly turned on, but if the PSEye lost track of the wand, it could just pause the game to readjust the light level to recapture it.
 
gofreak said:
I don't think you need to point out where the TV is or anything with the controller.

Let me put it this way... ASSUME you have good orientation data, and good x/y/z data. ASSUME you know the camera is placed on top of the TV at its centre.

With that data, you can figure out where on a (virtual) plane pressed against the TV the controller is pointing at. You have the xyz relative to the centre of the top of the TV, and its angle, so you (metaphorically) draw a line extending into infinity along the controller, and calculate the intersection point with this plane around the origin of the camera.

Then if you calibrate the system by telling it the size of the TV and its aspect ratio, it can then tell you if this intersection point is actually within the bounds of the TV its sitting on. And if so, where within those bounds.

We can place a question mark over the quality of the orientation data for now, if you like, but assuming you have good such data, all the variables you need are there.

But that completely ignores the orientation of the wand. With such a setup you could move a cursor relative to its absolute position, but there would be no way for the console to actually tell if it were being pointed in the direction of where the player wanted to point the cursor.



gofreak said:
That's an exceptionally unfair characterisation.

The wiimote was built first and foremost as pointing device. Not as a motion capture device. As a motion capture device it's technically inferior even to a DS3.

The motion+ introduced 'proper' motion tracking for limited periods of time.

The Sony motion controller is built with motion tracking in mind. You can get 'proper' motion tracking without the time constraints and reset-restrictions of the M+.

I think the jury is still out on how its 'pointing' abilities compare until we have more information, but you can say right now that it is a 'master' in at least one of its domains (i.e. in motion tracking) relative to the other offerings.

Not really since the accelerometer inside the Wiimote is much more responsive than the one inside the sixaxis (300hz sampling) Plus the single axis gyroscope doesn't really do anything that can't be done with just the 3 axis accelerometer of the Wiimote.

I'm not sure I agree completely. How do you expect to correct drift with Arc? The camera don't have all data available. It won't measure orientation, for example.

Ahh, actually that's a good point. There's no way for the remote to tell the orientation of the device since the glowing ball is perfectly round. Drift might have to be corrected by software (which would mean that Wiimote + might actually be better for drift since it can be supported by the sensor bar)
 

dofry

That's "Dr." dofry to you.
electroshockwave said:
It probably sounds weird but I think controllers are far enough away from reality that I don't really compare the two. I always had the same issue with an arcade lightgun game like Time Crisis, for example, where it would feel awkward when I was using a shotgun or machine gun in the game but I was still using the handgun shaped controller. Maybe it's just me that's got this "uncanny valley of controllers" issue.

I got carried away reading your post and thinking about the arcade light gun cabinets and so on. It would be so nice to have haptic feedback on the controller. Recoil, button presses, etc but they would require more techiniques, more battery life, (compressed air?). Sylinder bolt action movement. Would be better than just rumbling on different strenghts... dreaming...drool.
 

SamBishop

Banned
Nuclear Muffin said:
But that completely ignores the orientation of the wand. With such a setup you could move a cursor relative to its absolute position, but there would be no way for the console to actually tell if it were being pointed in the direction of where the player wanted to point the cursor.

The controller is constantly culling gyroscopic data; it can tell if you angle things in a direction. In any event, I'll know for sure in a little over an hour. Catch you folks on the flip side.
 

poppabk

Cheeks Spread for Digital Only Future
Nuclear Muffin said:
Ahh, actually that's a good point. There's no way for the remote to tell the orientation of the device since the glowing ball is perfectly round. Drift might have to be corrected by software (which would mean that Wiimote + might actually be better for drift since it can be supported by the sensor bar)
There are theories that it isn't perfectly round or has more than one light source inside, or using occlusion (seems highly unlikely) will allow the ball to provide more than just co-ordinate information.
 
SamBishop said:
The controller is constantly culling gyroscopic data; it can tell if you angle things in a direction. In any event, I'll know for sure in a little over an hour. Catch you folks on the flip side.

Yeah but there's no way to correlate the orientation data from the remote with whether or not the device is being pointed in the right direction as there is no point of reference (since the camera can only detect the position of the glowing ball, not the orientation)

Hopefully we'll get some hands on impressions and someone can put these questions to rest. I reckon this setup will not be the holy grail for those who want absolute positioning and good Wii style pointer controls, but I'll be glad to be proven wrong.
 

noah111

Still Alive
I've been really busy this whole week but it's nice to see this thread boom again over the days, just got caught up, lots of stupid comments and lots of hype building blocks (you Sam B).

Regret I couldn't be here this whole time, and will probably miss the live event, but either way it's going to be awesome, it'll really help me through the day/get my mind wandering.

t-minus 1:20 :D
 

Ashes

Banned
Nuclear Muffin said:
Yeah but there's no way to correlate the orientation data from the remote with whether or not the device is being pointed in the right direction as there is no point of reference (since the camera can only detect the position of the glowing ball, not the orientation)

Hopefully we'll get some hands on impressions and someone can put these questions to rest. I reckon this setup will not be the holy grail for those who want absolute positioning and good Wii style pointer controls, but I'll be glad to be proven wrong.

Remind me again why you think the other sensors inside the arc aren't providing this data?
 

dofry

That's "Dr." dofry to you.
Nuclear Muffin said:
Yeah but there's no way to correlate the orientation data from the remote with whether or not the device is being pointed in the right direction as there is no point of reference (since the camera can only detect the position of the glowing ball, not the orientation)

Hopefully we'll get some hands on impressions and someone can put these questions to rest. I reckon this setup will not be the holy grail for those who want absolute positioning and good Wii style pointer controls, but I'll be glad to be proven wrong.

We are grasping at straws here because we don't have the necessary data on the insides of the controller, so claiming it can't do something based on ones own assumptions is just guessing. We can try to predict, not claim absolutes.
 

gofreak

GAF's Bob Woodward
Nuclear Muffin said:
But that completely ignores the orientation of the wand.

Well my post asked you to assume good orientation data...so not really... :p

Nuclear Muffin said:
Not really since the accelerometer inside the Wiimote is much more responsive than the one inside the sixaxis (300hz sampling) Plus the single axis gyroscope doesn't really do anything that can't be done with just the 3 axis accelerometer of the Wiimote.

I'm not sure about sample rates. Technically your second part isn't true...though in the end it doesn't matter much - the gyro in Sixaxis fills in yaw data which the accelerometers in the Wiimote don't provide (it defers to data from the camera to calc that in combination with data from the accelerometers).

Regardless though, my point still stands...the wiimote was not at all designed for motion tracking of any real value. If you prefer 'it wasn't any better than a DS3', and the DS3 wasn't up to much here. It was all about the pointer.

The assertion PSGames was making, that the Sony controller was trying to do everything but not beating anything at any one thing, I think, is unfair. It has been clearly designed to address certain shortcomings elsewhere. It may have some shortcomings of its own, but it certainly has its relative strengths too I think.


Koren said:
I'm not sure I agree completely. How do you expect to correct drift with Arc? The camera don't have all data available. It won't measure orientation, for example.

Nuclear Muffin said:
Ahh, actually that's a good point. There's no way for the remote to tell the orientation of the device since the glowing ball is perfectly round. Drift might have to be corrected by software (which would mean that Wiimote + might actually be better for drift since it can be supported by the sensor bar)


You don't have drift with position, it's measured absolutely with every frame. That's one half of your 6DOF tracking that's always going to be accurate regardless of the length of a motion track, and that's a step up.

As for the other half, with the orientation, again, it depends on how they're doing it and how good the data is from that. If it is depending on just gyros alone though, I believe they are not quite so temporally unstable as accelerometers... let's assume though you do need resets. With the M+ you need resets on both translation and rotation every couple of seconds, and your bound by the lowest common denominator there (i.e. the linear motion tracking). With the motion controller, only one of those would need resetting (rotation), and its probably the one that stays 'good' longer - again, if resets are needed at all. Maybe someone with knowledge of quality gyroscopes can enlighten us as to what their behaviour is like.

Re. using the cam to track orientation...I would guess they're not doing this, but there are ways to doing that with the ball, ways described in some patents they've filed. Yeah, the ball's a ball and its silhouette is invariant with orientation, but you could disperse the light/intensity falloff from the centre of the ball in patterns that would allow to do some orientation 'stuff' with the camera, and filter with the sensor data to make your best measure of the orientation. I'm guessing they're not using this though...because I'm guessing the gyro data is 'good enough'.
 
Good lord I hope they show a nice light gun game soon so we can put this to rest.

Without doubt I am certain *POINTING* has been addressed in the design to a tolerance level which is acceptable for light gun gaming and other gaming.

To say this is not accounted for is folly since we have seen on screen pointing and writing already performed.

How can it be said it won't work?
 
Mad_Ban said:
Where's the hype train gif when it's needed?
zRLtZ.gif
 

tzare

Member
why people doubt about ARC's pointing ccapabilities? wasn't it shown playing RE5 wii style? that is an evidence for me
 

Man

Member
Oh Phil... talk about taking the (seemingly) wrong path. Left SCEWWS. Joined Atari as President. Got demoted to lesser position. Never heard from again...

But he sure is tall.
 
Top Bottom