Sony's New Motion Controller

[obonicus]

To bring it back on topic my concern is Sony's scheme just doesn't fit this model. To me, it puts a lot of requirements on the game's design. Something few developers want to mess/experiment with.

But that's my point. Sony's doesn't fit the Natal's model, and, indeed, if Natal is the next big thing they'll be too late. But Sony's does fit the Nintendo model, which is proven -- when it comes to motion controls you can be sure that no one's ignoring the 800lb gorilla in the corner that is Nintendo.

 

[Graham]

But that's my point. Sony's doesn't fit the Natal's model, and, indeed, if Natal is the next big thing they'll be too late. But Sony's does fit the Nintendo model, which is proven -- when it comes to motion controls you can be sure that no one's ignoring the 800lb gorilla in the corner that is Nintendo.

I understand, but I just don't think it's a workable strategy in their current position.
Nintendo's model was forced down developers throat. There was no choice but to invest in the system, here it's optional, risky and I'd argue expensive.
I just don't see enough developers making that bet.

 

[_phil_]

Do the buy from Microsoft of 3DV and the Z-cam patents make inviable for Sony to make a 3d cam ?. That´s the master movement i think Microsoft made here. In fact they finished the buy the same day of their conference in the morning. I could imagine Sony trying to buy it the same day in the evening...

3DV isn't the only company makind depth/3d camera on the market.

But that's my point. Sony's doesn't fit the Natal's model, and, indeed, if Natal is the next big thing they'll be too late.

Sony was playing with these 3d camera back in 2003.The ealy expansive model from 3DV apparently.
They probably don't think it's the right model for short/middle term.

 

[Shifty Geezer]

There is another option. Developers have been (bizarrely) avoiding Wii. Now they have another platform to target, one they are all familiar with by now and have engines for. Yes, just when you thought cross-platform couldn't get any more cross this gen' . Lots of Wii games could be ported, using high-level tools and engines at minimal effort to get the Wii results in HD and with a bit of shader decoration, even as PSN downloads. The possibility is there for developers to try quick cash-ins, expanding the PSMC library and encouraging adoption.

 

[Silent_Buddha]

So there will be an pverhead on PS3. However, the more stuff inside Natal, the more expensive it's going to be. If it's a little technology marvel, it could cost too much to be mainstream. PS3's advantage over that is that, though the tech is simpler and will take more assets from the PS3, it'll be cheaper and more readily adoptable.

I completely agree on the cost factor. The PS3 system should be cheaper (Even when factoring in you need both Eye Toy + PS3 Wands) to achieve something approaching what Natal is possible of doing.

And yes, I'm pretty certain certain that MS will have to use this as a loss leader. I just don't see very many ways for them to get the costs down in such a way that the system performs as they claim. Sure the camera systems (visible light + IR) may be cheap. The microphones certainly are cheap. And perhaps each individual component that will do the actual processing may be cheap. But put that all together and it probably won't be a non-trivial cost.

MS's hope/gamble here is that the system provides such a revolutionary experience that it manages to move consoles in Wii like numbers (historically speaking since we don't know if Wii momentum will continue as it has in the past).

And unlike Wii, there are more titles than just a few first party efforts that hit the top of the charts. Meaning they could just be looking at the potential of selling a whole lot more software...

One problem I see is that if the user base is expanded using a similar demographic to the Wii userbase. Will they have the same sort of problems moving traditional type games to that demographic?

But yes, overall, I absolutely agree that I don't see a way that Natal will be cheaper to manufacture than Eye Toy + PS3 Wands. But I think MS is not only more willing but certainly more capable of eating the cost if it see's an possible overall cost benefit.

As to processing requirements that's where we'll have to agree to disagree. Since we don't know many hard details on the system yet. We can't know things such as.

Is Natal going to be designed with the ability to just pick out developer specified (not MS/Natal hardcoded) key words out of multiple voice streams and properly identify each key word to the respective speaker?

Or will Natal have the ability to more closely mimic natural speech parsing and be able to (internally) transcribe everything said and then be able to output whole sentences to the X360 for use by any dev that would want to?

Likewise just how well does it track multiple people? And how many multiple people can it track. If it's tracking 48 key points per person and there are say 4 people in the camera field of view (getting crowded here ) will it still be accurately tracking 192 key points 30 times per second?

And that's just some of the considerations...

Depending on how advanced or how simplistic they go, would determine (IMO) whether the Eye Toy + PS3 wands + PS3 doing all the computations would be able to mimic it in some way.

Regards,
SB

 

[Shifty Geezer]

MS's hope/gamble here is that the system provides such a revolutionary experience that it manages to move consoles in Wii like numbers (historically speaking since we don't know if Wii momentum will continue as it has in the past).

And unlike Wii, there are more titles than just a few first party efforts that hit the top of the charts. Meaning they could just be looking at the potential of selling a whole lot more software...

I don't think games is where MS would look for revenue. Their mission was to control the content portal in teh living room. They've predicted and seen in Wii how large markets can be reached with 'casual' interfaces. Natal would make a console accessible to a wider deomgraphic than HD games machine alone. One someone buys XB360, whether for "Halo 4" or "Happy Farm Adventure", they can be sold movies and TV series, which is where, long term, the real money is (believed to be )

But yes, going after this wider market, everyone who watches TV instead of everyone who plays games, MS may be willing to push Natal as a loss-leader now they've got a handle on XB360 production costs.

 

[Silent_Buddha]

There is another option. Developers have been (bizarrely) avoiding Wii. Now they have another platform to target, one they are all familiar with by now and have engines for. Yes, just when you thought cross-platform couldn't get any more cross this gen' . Lots of Wii games could be ported, using high-level tools and engines at minimal effort to get the Wii results in HD and with a bit of shader decoration, even as PSN downloads. The possibility is there for developers to try quick cash-ins, expanding the PSMC library and encouraging adoption.

That makes me think of something else.

Considering that in recent years there's been a greater move towards cross platform titles, we may see one of two things.

1. Slow adoption for anything that can't be implemented on multiple consoles.

2. Reversing the trend and seeing greater numbers of exclusive titles.

While PS3 and Wii would share a similar "wand-like pointing" device, the graphics are vastly different. Meaning cross platform of wand-pointy games could be limited to Wii quality and effects. Likewise without a camera system (does Wii have a camera based control system?) that would limit cross platform games to only wand-pointy games.

While PS3 and X360 share similar graphics quality, their motion based schemes are for the most part quite different. But you can always have cross platform camera-y games...

So how much are many companies that are heavily invested in cross platform developement going to gamble on motion based gaming?

Regards,
SB

 

[Silent_Buddha]

I don't think games is where MS would look for revenue. Their mission was to control the content portal in teh living room. They've predicted and seen in Wii how large markets can be reached with 'casual' interfaces. Natal would make a console accessible to a wider deomgraphic than HD games machine alone. One someone buys XB360, whether for "Halo 4" or "Happy Farm Adventure", they can be sold movies and TV series, which is where, long term, the real money is (believed to be )

But yes, going after this wider market, everyone who watches TV instead of everyone who plays games, MS may be willing to push Natal as a loss-leader now they've got a handle on XB360 production costs.

Oh duh...

/me smacks forehead.

Duh, one of the more compelling reasons to get a Natal system, Minority report style UI control. And I completely forgot about their extensive movie and video services through Live.

I wish there was a way to get some numbers on just how much revenue Live and Live related sources of revenue (XBLA, Movies, Video) are bringing in.

So that right there could be the main gamble on the part of MS. Sell the Natal system at a loss hoping for Wii like sales of the console as a result and recoup losses from Live services.

Regards,
SB

 

[patsu]

There is one crucial point that is often missed, and I alluded to it earlier. Natal is a complementary technology. It can be used in conjunction with all their existing control systems, and from everything microsoft has said, it will not significantly impact performance on the console. (From what I can tell, the demo they were showing behind closed doors was *entirely* processed externally).

Sony's system is camera- plus controller-based. It can also complement existing control (e.g., the "wand" may be a "transformer" DS controller as depicted in one of the patent images). Pointing to the screen with right hand, while your left hand continues to hold the controller = Split the controller half way and point the right half at the screen.

I feel that as long as Natal cannot track individual fingers reliably, the 2 systems are similar in capability... but the physical controller will be more accurate. Note that it's not just tracking motion, it may be much more important to track sudden finger/hand impact and contact, especially for core gaming. Without a very fast "trigger" mechanism, it may be difficult or useless. Nothing in Natal addresses this issue (Jerking a la SIXAXIS seems not acceptable by players , waving like PS Eye is time consuming). May be people want to experiment with "Bang ! Bang !" for speech recognition trigger ? Hand gesture may work but has to be spot-on for all people. Miss a beat and the player may complain. This is lesson learned from Sony's EyeToy and SIXAXIS experiences, coupled with the problem you highlighted before (No tactile feedback in a pure hands free model = clumsy to use).

going after this wider market, everyone who watches TV instead of everyone who plays games, MS may be willing to push Natal as a loss-leader now they've got a handle on XB360 production costs.

After I looked at where MS is going, Natal may not be useful here. A one 3D-camera based approach is best for CE applications because those guys are very stingy. Natal is too "big" !



The third problem is resource consumption (both run-time and freaking BOM cost). If it's for casual gaming, the current system would work (i.e. user pays $40 with Sony making $$). If they want an "auto-pilot" system, Sony may need to whip out the 3D camera and rev the hardware. While they're at it, engineer it for current CPU/memory resource limitation.


The final problem is marketing. I reckon the most resource hogging part is the complex UI operation (i.e., advanced image and sound processing), not the controller tracking (see Wii). Just rev DS3 to the new controller and bundle it with every PS3. Here, Sony will need to make more investment, but the cost to user should be an added $50-60 for the new DS controller using current price as reference. For sexy use, upgrade to new camera, which may have new circuitry. Another $100 ?


And since Sony uses a physical controller, it'd be a natural next step to use a glove for finger tracking. Natal can certainly use a glove too later.

 

[Shifty Geezer]

I think to be comparable, MS will introduce a wand. If we take the magic example, the current Natal implementation suggests you call 'fireball' to case a fireball spell, where the PSMC has a button. The Natal version sounds cooler, but what if you don't want to or can't use speech? Will the player need to learn glyphs to draw for guesture instructions? There's a lot to be said for good old fashioned buttons!

 

[Asher]

Sony's system is camera- plus controller-based. It can also complement existing control (e.g., the "wand" may be a "transformer" DS controller as depicted in one of the patent images). Pointing to the screen with right hand, while your left hand continue to hold the controller = Split the controller half way and point the right half at the screen.

I feel that as long as Natal cannot track individual fingers reliably, the 2 systems are similar in capability... but the physical controller based one will be more accurate.

Again, the Natal does not preclude using a physical controller in any way. Whether this be "light bulb" based tracking like how the PS Motion Controller works or whether it's another physical controller in the hand.

People keep looking at this as some kind of mutually exclusive control, it's not. It can complement Wii and PSMC like control systems, pretty trivially.

After I looked at where MS is going, Natal may not be useful here. A one 3D-camera based approach is best for CE applications because those guys are very stingy. Natal is too "big" !

What? It's a device that goes by the TV/home theatre, it's certainly not wider than 99.99% of all TVs out there and it's certainly not too tall, either. I don't see how this is a relevant argument in any way.

 

[Silent_Buddha]

After I looked at where MS is going, Natal may not be useful here. A one 3D-camera based approach is best for CE applications because those guys are very stingy. Natal is too "big" !

That's the thing though it's perfect for CE applications because there is no size. Just like a stereo receiver you place it and forget it.

In fact for all practical purposes, it's smaller than a remote. You hold nothing, don't risk losing anything, and never have to find something that you can't remember where you put it.

Granted it's probably not going to completely replace a remote. It's probably easier to hit buttons to go to channel 82 than to say out loud "Channel 82." Er maybe not.

If priced (regardless of actual cost to manufacture, remember loss leader) in the same price range as CHEAP programmable remotes 50-100 USD then it compares favorably IF (and big IF) there's a way to demo it such that people think, wow this would be totally cool to use instead of my TV remote.

This is all with regards to the CE space and completely ignoring any possible gaming related benefits they would also be able to get.

Of course, the big limitation here is that if it can only hook up to an X360, then the CE benefits can't be the only ones. After all you can't watch regular TV through the X360 (can you? I haven't tried the media extender features much myself). Nor can you control your receiver or DVR. That would be the biggest knock against it being a CE only device.

If they also had integration with Windows 7 it could potential dominate the MCE control market. But then there's no additional revenue stream from that to offset the MS eating the cost to market it as a loss leader.

Hmmm, probably why no announcement of it working with anything other than the X360. Since they would need some way to recoup the loss (Live marketplace).

Regards,
SB

 

[patsu]

Again, the Natal does not preclude using a physical controller in any way. Whether this be "light bulb" based tracking like how the PS Motion Controller works or whether it's another physical controller in the hand.

People keep looking at this as some kind of mutually exclusive control, it's not. It can complement Wii and PSMC like control systems, pretty trivially.

Yes, then it's even more expensive than what we are looking at on the table. And the complementary use argument is not so compelling once you throw in a controller. I already acknowledge above that MS can certainly add an extra controller like the glove. Adding it just for trigger is wasteful, might as well add new capability at the same time.

What? It's a device that goes by the TV/home theatre, it's certainly not wider than 99.99% of all TVs out there and it's certainly not too tall, either. I don't see how this is a relevant argument in any way.

That's the thing though it's perfect for CE applications because there is no size. Just like a stereo receiver you place it and forget it.

Too big means too expensive. Size concern to me is secondary at the moment, although housewives will chime in later. e.g., Wii + 360 sensor bars ? If they know a smaller and cheaper alternative exists, it'd be hard to fly.

If priced (regardless of actual cost to manufacture, remember loss leader) in the same price range as CHEAP programmable remotes 50-100 USD then it compares favorably IF (and big IF) there's a way to demo it such that people think, wow this would be totally cool to use instead of my TV remote.

MS has no way to earn money back from CE. They have to make money per sales. I don't know 50-100 is sufficient for the entire sensor bar. Add the distribution and storage cost for that large unit, I doubt it.

 

[Silent_Buddha]

I think to be comparable, MS will introduce a wand. If we take the magic example, the current Natal implementation suggests you call 'fireball' to case a fireball spell, where the PSMC has a button. The Natal version sounds cooler, but what if you don't want to or can't use speech? Will the player need to learn glyphs to draw for guesture instructions? There's a lot to be said for good old fashioned buttons!

It makes me wonder if there are provisions within Natal itself for recognition of special objects. Such that if there's something built in (so that Natal can do the positional processing rather than the X360) to recognize plastic wand? Or perhaps special colored ball that could be attached to virtually anything.

Basically if there's any provisions in Natal to track something other than those 48 joints per person.

Heh, just had a funny thought. What if they sold a big red button (like the Staples Easy button ) that Natal could recognize visually? Sorry, not sure why I find that so funny. Hehe.

[edit] Additionally nothing precludes a game of that type where you have to use a fireball of using a standard X360 controller as you normally would. Then just having to raise your hand back and do a throwing motion to throw a fireball. In other words, there doesn't have to be a need to replace a controller if it makes no sense to replace it, yet at the same time have complementary motion based controls for special effects. The Analogue sticks are probably still going to be better for navigation than anything you can do with motion control for example. But now you have a whole new world of options of what you can do additionally through motion or voice control.

Regards,
SB

 

[patsu]

I think to be comparable, MS will introduce a wand. If we take the magic example, the current Natal implementation suggests you call 'fireball' to case a fireball spell, where the PSMC has a button. The Natal version sounds cooler, but what if you don't want to or can't use speech? Will the player need to learn glyphs to draw for guesture instructions? There's a lot to be said for good old fashioned buttons!

Yes, buttons are not evil. Nintendo showed us that it is also not the problem for mass adoption. Sense of purpose, complexity and required skills are the problems. You can certainly come up with a clumsy/unreliable user flow or gesture skill for a hands free controller. In short, it is not only choice of technology but also a technology usage/application issue.

 

[obonicus]

As usual doubter, remember that we haven't seen what Natal will look like as a finished package. The sensor-bar-like strip is what they hope it'll look like. The one that they're using as a prototype seems to have a smaller profile, even if it's larger overall.

One thing that occurred to me, with all these solutions: what about people with stupidly big tvs? Didn't people already have trouble with the sensor bar being too small for some ginormous televisions (leading folks to create custom sensor strips)? It seems like we'd have similar problems with people standing too far away from very large TVs, except maybe without as easy a fix.

 

[patsu]

As I understand, they need 2 cameras for triangulation. So it will be more expensive and occupy more space. If they want to use one camera, then it's like Sony's 3D camera approach. And they may really need a controller then to compensate for the loss in accuracy (without triangulation) for core gaming.

I am not doubting Natal. I like the base technology, but the fundamental approach has its assumptions and limits. The same goes for a controller-based approach like Wii and Sony. They have their pros and cons. If Natal adds a controller, then its appeal is considerably smaller because with a controller, some of its base functions has already been replaced by low tech. At which point, the user will be able to compare the response time between the controller and a hands free operation. If there is a perceived difference (e.g,, snappiness/lag, distance issue, etc), the core gamer will "see through" the initial "no controller" magic. For casual use, no controller should be more sexy and convenient.

The differentiator will be the application and marketing. That will in turn affect the pricing.

 

[liolio]

Sorry I cut your post in more fiiting way for me to answer.

Is Natal going to be designed with the ability to just pick out developer specified (not MS/Natal hardcoded) key words out of multiple voice streams and properly identify each key word to the respective speaker?

Or will Natal have the ability to more closely mimic natural speech parsing and be able to (internally) transcribe everything said and then be able to output whole sentences to the X360 for use by any dev that would want to?

I would go for the former the other option sounds a too much optimistic/unrealistic to me.

I completely agree on the cost factor. The PS3 system should be cheaper (Even when factoring in you need both Eye Toy + PS3 Wands) to achieve something approaching what Natal is possible of doing.

And yes, I'm pretty certain certain that MS will have to use this as a loss leader. I just don't see very many ways for them to get the costs down in such a way that the system performs as they claim. Sure the camera systems (visible light + IR) may be cheap. The microphones certainly are cheap. And perhaps each individual component that will do the actual processing may be cheap. But put that all together and it probably won't be a non-trivial cost.

MS's hope/gamble here is that the system provides such a revolutionary experience that it manages to move consoles in Wii like numbers (historically speaking since we don't know if Wii momentum will continue as it has in the past).

And unlike Wii, there are more titles than just a few first party efforts that hit the top of the charts. Meaning they could just be looking at the potential of selling a whole lot more software...

One problem I see is that if the user base is expanded using a similar demographic to the Wii userbase. Will they have the same sort of problems moving traditional type games to that demographic?

But yes, overall, I absolutely agree that I don't see a way that Natal will be cheaper to manufacture than Eye Toy + PS3 Wands. But I think MS is not only more willing but certainly more capable of eating the cost if it see's an possible overall cost benefit.

As to processing requirements that's where we'll have to agree to disagree. Since we don't know many hard details on the system yet. We can't know things such as.

Likewise just how well does it track multiple people? And how many multiple people can it track. If it's tracking 48 key points per person and there are say 4 people in the camera field of view (getting crowded here ) will it still be accurately tracking 192 key points 30 times per second?

And that's just some of the considerations...

Depending on how advanced or how simplistic they go, would determine (IMO) whether the Eye Toy + PS3 wands + PS3 doing all the computations would be able to mimic it in some way.

Regards,
SB

I think all this part of your post is tied together. Basically the problematic for MS is:
"The thing can sell ubberly well but to do so it will have to deliver on promise thus it will cost money to produce the proper hardware".
I think more and more that the "somewhere between 100 and 200$ cost" might have a point.
The custom chip will have some work.

So we have a custom chip that we put in the sensor itself. The chip we designed with Microsoft will be doing the majority of the processing for you, so as a game designer you can think about the sensor as a normal input device - something that's relatively free for you as a game designer.
Designers have 100 per cent of the resources of the console and this device is just another input device they can use. It's a fancy, cool, awesome device, but essentially you can just treat it from a free-to-platform perspective, because all of the magic - all of the processing - happens sensor-side.

I could see the chip being indeed pretty big, think a tiny CPU tied to a custom DSP/VPU working on some on chip scratch pad memory/edram. Firmware would be loaded on a per application/game basis, turning on the device would act mostly like a "boot on LAN". It will cost money.
I wonder if Ms plan is to make swallow the initial loss to its entertainment division (kind of their money loser) and then will make this available on 7/windows.

 

[Silent_Buddha]

As I understand, they need 2 cameras for triangulation. So it will be more expensive. If they want to use one camera, then it's like Sony's 3D camera approach. And they may really need a controller then to compensate for the loss in accuracy (without triangulation).

I am not doubting Natal. I like the base technology, but the fundamental approach has its limits. The same goes for a controller-based approach like Wii and Sony. They have their pros and cons.

The differentiator will be the application and marketing. That will in turn affect the pricing.

The Natal solution doesn't require 2 camera's for location triangulation. The visual camera will give some basic X/Y location data while the infrared camera will gives depth (and possibly additional X/Y) location data. That's internally processed into X/Y/Z and velocity/acceleration data. Er wait that's still location triangulation but not the way you think of it as in the Wii sensors for example.

So it doesn't matter how large your screen is (well I suppose a theatre sized screen and room would be problematic), the camera still see's and tracks whatever is in the field of view normally.

Regards,
SB

 

[patsu]

The Natal solution doesn't require 2 camera's for location triangulation. The visual camera will give some basic X/Y location data while the infrared camera will gives depth (and possibly additional X/Y) location data. That's internally processed into X/Y/Z and velocity/acceleration data. Er wait that's still location triangulation but not the way you think of it as in the Wii sensors for example.

That's not Natal technology. It already exists without Natal (3D camera using assorted techniques). As Graham mentioned, it has accuracy issues. They may not be able to construct a reliable 3D mesh and skeletons without 2 cameras (the real MS innovation).