Slashdot Mirror
The Inside Story of Microsoft's 'Project Natal'
Lanxon writes "Wired has published a lengthy behind-the-scenes feature documenting the inception, development and technological struggles of Microsoft's Project Natal, now known as Kinect. The feature is the result of conversations the magazine had with a number of key developers and researchers behind the project, and unprecedented access to Microsoft Research in a number of countries, over the course of three years."
Actually most projects in Microsoft Research never come close to making any money, they are just meant to be feeder projects for future products. The fact that this one came to market so quickly is really unique.
Most people forget how much money some of the big companies like Microsoft and IBM put into pure research efforts (i.e. no immediate plan for a ROI)
Exactly, just about every single console has tried to do what Kinect is trying to do.
The Dreamcast had the Dreameye
The PS2 had the EyeToy
The PS3 has the PlayStation Eye
Even the 360 has the Xbox Live Vision
All of them failed to change gaming as we know it. Yeah, the EyeToy had a few games that made use of it, yeah, they were fun in a weird way but they were just tech demos.
We've been trying the camera + console combo for a decade and none of them have really changed gaming... at all. Perhaps Kinect will change that, but chances are, its just hype. I mean, has anyone ever -used- Kinect except for a few people? Plus, attempts to "revitalize" a console midway through its lifespan have historically failed, the only "success" would be the expansion pack for the N64 the rest (SegaCD, Sega 32x, etc.) have all spectacularly failed.
Taxation is legalized theft, no more, no less.
They should get Carol Vorderman to do the advert. "Now you can consolidate all your super sucky controllers into one, easy, Mega sucky interface"
The Eye Toy for PS2 is actually one of the most successful console accessories ever, selling 10.5 million units worldwide despite a slow start. Apparently it was actually novel and useful enough for its time to generate grassroot interest. The failure you are looking for is The Eye for PS3.
My son will be 6 years old tomorrow. This is the first thing I'm getting him for Christmas. We don't own a Wii, and this looks more fun anyway. I'm just one person of course, but I'm sure there are plenty of people out there in the same situation as myself.
“Since the dawn of time, humanity’s long journey has led us to countless discoveries, Yet with each leap forward for civilisation, more people have been left behind. But our quest has taken us to a completely new horizon. History is about to be rewritten. This time human beings will be at the centre -- and the machines will be the ones that adapt. After five million years of evolution, is it possible that the future of humanity is humanity itself?”
That the article repeats verbatim such a quote from Microsoft's presentation without even a slight nod to the gross self-aggrandization clued me to the fact that the whole piece is yet another corporate advertisement disguised as news.
I mean, in addition to the whole story starting with an emergency meeting on mid 2007 about the need to "reimagine a new direction for the Xbox" yet failing to point out it was all due to the runaway success of the Wii. It actually sticks out like a sore thumb to see these VPs panicking about something that the article refuses to acknowledge exists.
Most people forget how much money some of the big companies like Microsoft and IBM put into pure research efforts (i.e. no immediate plan for a ROI)
Forget? Is there anything to really forget? Back in the day, IBM did quite a bit of pure research and now, occasionally you'll see something about them writing "IBM" with atoms or making a computer that's really fast but nothing like the research they did.
BTW, nothing will ever beat AT&T Bell Labs - especially for the pure research they did back when they existed - all that's gone now.
Basically, a guy like William Shockley wouldn't have a chance in this day and age.
RIP America
July 4, 1776 - September 11, 2001
Make your son a favor, and give him something to play outside instead of a gadget.
Maybe you can get the Kinect thing for yourself, as well...
3DV Systems, an Israeli company, so it seems: http://www.engadget.com/2009/06/03/microsofts-project-natal-roots-revealed-3dv-systems-zcam/
Probably find more info with a search on "3DV Systems" on the GOOG.
Yes. A friend (who works at MS) has a demo unit. This blows away the Nintendo Balance Board, because you can actually JUMP! It's better than the Sony Move, because you don't need a stupid looking controller to play it. It's very accurate at reading your motions in a 3d space. Microsoft has a winner on their hands with the hardware. If they can get some good titles behind this, it's going to be awesome.
Why does Kinect seem familiar? Microsoft saying "It will change the industry" and "Big players are developing it" and "You'll wonder how you ever did anything without it!" and "Everything will change after this". Oh yeah, this is what they said with Windows Mobile 6.
The more reports I get back the more it seems like just with Windows Mobile 6 that marketing is over promising features. That isn't to say Kinect is "snake oil" because of a lot of it does seem to work but that it is rough...."rougher" than they want to let on. So we'll get a big advertising blitz, Microsoft will declare it was a huge success, and then summary die because it is expensive and never quite work as smooth as they advertised while the competition runs wild. All of this is like WinMo6.
Anyhow, there is some merit to the tech but it feels like it is going in the wrong direction. Its like the belief one can effectively replace a keyboard with voice recognition. VR is useful in itself but not as a keyboard replacement! As an HMI issue, gesture controls found on Wii "work" because the interface is simplified not because of waggle. Replacing waggle sticks with Kinect without doing the requisite "simplification" is going to be a disaster. I wish they would abandon schemes like "replacing the controller for games" that are more smoke and mirrors than practical execution. Go with practical stuff like if my console notices I put the controller down to answer the phone, door, kids jump in my lap, or whatever, pause the game. If the console notice I'm no long in front of my TV for an extended period of time, go into hibernate mode. Stuff like this is more useful than trying to figure out how build a fighting game by waving my arms and kicking with my legs.
But in any event, $150us is too much for all of it. If it was built in at the start that would be one thing but it is too late now.
Microsoft was caught faking their demos for their Eye Toy style motion controls for the past two E3s.
And first had reports constantly talk of horrible lag between arm movements and response.
You can't play games while sitting down.
You need to have a relatively large area completely clear in whatever room you are using it in.
There are lighting issues that cause the system to have trouble recognizing motion.
There are clothing issues that cause the system to have trouble recognizing motion.
It doesn't recognize fine motions like finger actions.
It only supports two players.
It should be obvious why Microsoft was forced to fake their demos and so far there has been almost zero real world demos outside of carefully controlled showings with people Microsoft is certain are going to hype the product.
Not that any of that really matters. There isn't any game even remotely interesting or fun that Microsoft has show for the system. When Nintendo showed off the Wii they were confident enough to let everyone at the very first E3 where it was shown play it and had games that both gamers and non-gamers really wanted to play.
Most likely the only people who will end up buying this shoddy tech from Microsoft are the same ones who jumped on the dead HD-DVD format.
I hate to sound cynical, but this will only be the 3rd Wiimote killer since the 6-axis took it's aim. With all of Kinect's extra features over the de facto, it should fit really nicely next to all of the iPod-killers of yor.
The Kinect is the Firewire of the Firewire-USB war. Overly expensive and unsupported compared to it's contemporary cousin even if it may have better technology. It only takes a (USB2-style) revision to make the things Kinect has moot and the populace moves on...
Once you start despising the jerks, you become one.
Yesterday I watched a TV show on German TV in which they used the Kinect system. It was broadcast on a major channel watched by millions. Two players had to play against each other in a quite unspectacular game. It was embarrassing, both due to the boring gameplay and the technical weaknesses of the system. During the first game, suddenly the split-screen switched to a single screen such that only one player could see his avatar. Nobody knew why and the TV show host quickly reacted and said "well, now you had a training run". After the second run one of the players complained that his moves were not recognized at all. He kept jumping all the time but the cameras obviously did not capture these jumps.
To be fair, I think things like Move and Kinect are really just ramping up for the next generation of console equipment that will have this stuff as standard. I doubt they're counting on doing anymore than breaking even, if that for now. Getting developers onboard and used to the tech so that they could really push it next gen is probably a good thing too- look how many 3rd party developers really struggled to take advantage of the Wii, most stuff that's been churned out since release has been utter crap, and it's taken a while to get some good 3rd party stuff out there.
There was a story some weeks ago about how Natal could previously even read sign language, and detect finger gestures, but to make the equipment fit in the $150 price range they switched to a lower resolution IR camera. I'd imagine they'll put the higher resolution back in for the next gen console so FPS players can issue commands with hand gestures and that sort of thing, which they could've done this time if they'd made Natal prohibitively more expensive. I guess the technology is too new and expensive to really push it to it's limits right now, but by the time the XBox 720 or whatever comes out it may not be.
I'm also not sure at least in the case of Natal (I don't know about Move) that the R&D will be wasted even if it flops in gaming. I'd imagine hands free interfaces are something Microsoft is hoping to capitalise on elsewhere in the future. It's like things like multi-touch and gesture recognition, it's not new, but it's really come into it's own in recent years finding it's way on mobile devices and becoming a must have feature. It may be that Natal wiill find it's niche in for example TVs, to provide hands free control of them or something like that.
I was in Best Buy last weekend and they were demonstrating the Kinect. My 9 year old daughter, who doesn't like to play console based video games, started playing one of the demos. It was sort of like Breakout, but you kicked or punched the balls instead of using a remote. I had a hard time getting her to leave.
Sorry. Much falseness here.
First of all, the on-board processor was dropped. Kinect places its entire processing load on the 360 (putting a 10-15 % load on the CPU).
Second, the "regular camera" in it runs at 640x480 30Hz, which is bad enough, but the infrared projector runs at 320x240, which is abysmal and the reason the device cannot detect things like finger movements.
The device is not ambitious at all; it's old tech. We've been doing depthmaps from two image sources for about as long as we've had cameras. The fact that Microsoft has made several cuts in hardware and functionality and is still pricing the device at $150 is disheartening, although not surprising given their pricing on things like proprietary hard drives.
Basically, it's overpriced junk that could be not just matched but beaten on any computer or on the PS3 by two webcams, two microphones, and a 5k program (total cost: $60). I would applaud them for pushing development of userspace control, but what's far more likely to happen is this thing is going to bomb hard enough to scare investors away for the next thirty years. They are actually hurting the market.
Here, here, here, and here.
Actually, Kinect only allows skeletal recognition for its own Avatars -- that functionality is not available to games developers. Both consoles are doing an equal amount of skeletal recognition in the SDK exposed to developers -- none.
Because, quite simply, everyone else knows it's not worth it. In fact, Sony was offered the Kinect technology and chose not to use it, specifically because they knew its marketability is limited.
Sony has already tried the no-controller camera-driven games with the EyeToy, which bombed. There have been all kinds of toy programs using webcams, which are all forgotten. Adding a depth camera does not fundamentally change the interaction -- in fact, it barely affects it at all. In userspace, Kinect is EyeToy, is doomed. You cannot play engaging games without a controller.
If they cannot sell two cameras and a toy motor for less than $150 I'll eat my hat. Do you really believe, for instance, that they "can't sell" a 250GB hard drive for less than $129.99 when normal 250GB hard drives can be had for less than $50 (a third of the price, ultimately)?
If I put a box of trash up on eBay for only a 10% markup over what it all cost me, how many people do you think will buy it? The market doesn't care what you spent. It cares what the price is and what they get out of it. $150 is going to sail right over everyone's head. You may bookmark this post and refer back to it when the Kinect line is officially discontinued (I will give it say Summer of 2012, after Microsoft loses a lot of money, which is what their games division has been doing all along.)