It took him only three hours, but here we are: Microsoft’s Kinect motion controller thing for the Xbox360 has been hacked, and a preliminary open source driver has been made available for Linux. According to the hacker in question, it was surprisingly simple.
The hack and driver comes courtesy of Hector ‘marcan’ Martin, and has been dubbed Libfreenect, and has been released under version 2 or 3 of the GPL. Both the depth and ordinary camera work, and even though a lot of work still needs to be done, it’s impressive it only took a few hours to get where we are now.
One of the major problems right now is the audio chip. “The audio chip (the Marvell) requires firmware and more init and does a TON of stuff including the crypto authentication to prove that it is an original Kinect and not a clone,” marcan writes, “Who knows what this thing does to the incoming audio. This should be interesting to look at.”
He has the video to prove it all.
You can get the code from its git.
I love it when a plan comes together!
should microsoft care if it is hacked and used with other systems other than 360?
It shouldnt matter what you do with the device, you paid for it, they got their $$$ unless they subsidize it hoping to get their money back through sold kinect games, are they subsidizing the device?
As far as I know, this time they’re making a profit on the device itself.
I really doubt this, particularly with the huge marketing campain for this costing a truckload. But then again I think the whole console experience has been very costly overall for Microsoft.
Got to hand it to them though, it takes balls to happily enter such a cutthroat market so heavily dominated by the Japanese. And they’ve certainly carved out a good chunk of it, the question is at what cost?
yes, they should. if manufacturing costs are the same as starter console costs then they are selling it at loss, where they count on selling games to cover them. in that case their only defense is limited to manufacturing shortages.
i know i’m buying one tomorrow and i don’t own neither xbox nor windows machine. i just see places where i could apply it in my software.
Yes, it must. If somebody was thinking about buying 360 *just because* it had an unique technology as Kinetic, potentially things may have now changed, and 360 lose its appeal.
I.e. it’s the whole ‘360 gamesystem’ that is losing value.
That said, I’m very glad for the hack. Well done! 😛
We should thank Microsoft for this device.
It makes my Linux Machine looks even better.
(BTW, is it true they paid 500Million$ only for ads?)
I’m not sure which I’d rather use with my computer, a kinnect or a Playstation Move. Sure, waving your arma about like an idiot is cool (?), but I’d hate, for example, if Firefox interpreted scratching my nose as an instruction to open my web history… Particularly if I have company…
. With the Move, at least, everything you do with it would have to be deliberate, to the point some actions require buttons. The Kinnect’s biggest selling point is also it’s biggest disadvantwge, in my opinion.
It could be cool tough for some powerpoint presentation
Good job and congratulations to this hardware hacker.
d(>w<)b
Does he get $2000 then?
Not bad for 3 hours work if so.
$3000, apparently.
http://www.adafruit.com/blog/2010/11/10/we-have-a-winner-open-kinec…
Plus an additional $2000 goes to the EFF.
What I’m still waiting for is any indications that this can be used to make a DIY Surface table even easier to make. Anyone know if this is the case? More particularly I’d love to know if it would be possible to use this as an input for a large interactive Whiteboard type application with the user being able to “write” on the display using their hands.
Are there any possibilities this opens the door to that aren’t related to robotics?
–bornagainpenguin
Hi,
As far as I know the kinect device is like a web camera on steroids – a video feed with the addition of depth information and an array of microphones to allow the position of sounds to be determined (rather than just one microphone).
The new Linux driver makes this information available to software, which is a very good start, but…
There’s a massive amount of software (image recognition, voice recognition, and something to combine the results of both into gestures) that is still “yet to be implemented”; and someone will need to invent some sort of API (and libraries) that’s suitable for handling the wide variety of gestures, etc; and only after all that is done will application developers be able to start looking at adding support for the new gestures, etc to their applications.
Basically what I’m saying is: It’s one small step for man (but there’s still a long way to go before it’s a giant leap for mankind).
– Brendan
Sure, if you’re doing this in order to recreate Kinect games. But that’s not why people want this. It is useful for many more things than just people-tracking.
Cheap motion capture, here we come!
Just think about it, 2 kinects opposed to each other to get full body realtime motion capture for animation..
Of course, i guess we have to wait for somebody to make a body tracking library as good as microsoft’s one.. but it’s definitely a matter of time, and the kind of stuff people in academia like to do..
I thought about that but I don’t think two Kinects would work if they are looking at the same thing.
It’s active IR time of flight. Think about it.