The Future of Kinect
microsoft.com
microsoft.com
Damage control regarding the Kinect is everywhere now, including this article. I just saw that the game Destiny is jettisoning Kinect support and because of that, is now playable at a higher resolution. MS is scrambling to make the "Kinect reserve" configurable instead of static. I believe the June update for the Xbox has this functionality, but how many devs will just go, "Who cares" and set it zero just like Bungie did with Destiny? Most I assume, especially if their games play at 720p on Xbox w/ Kinect but 1080p on PS4.
I really think the Kinect is more or less dead and have blogged about it in the past:
http://nothingjustworks.com/microsoft-just-gave-up-on-the-ki...
FWIW, I have the new Xbox with the Kinect and even as a Kinect fan I'm underwhelmed. I also had to disable the voice turn on feature because when its on, the xbox doesn't fully turn off, so a little fan runs 24/7 on the xbox power supply and it makes just enough noise to be annoying when watching TV and movies. In the age of silent tablets and passively cooled devices, a non-stop hum from a fan on a media center device is inexcusable. Couldn't they passively cooled the power supply?
I'm sad to write that "just dance kids edition" equals tears during menu navigation. Works fine during gameplay, its just the wrong UI for anything involving menus or detail work or ... pretty much anything other than dancing.
The fundamental insight of the kinect "problem" is its a transparent interface that people hate about 10% of the time. That sounds great, people love it 90% of the time, right? No, because its transparent, it disappears 90% of the time and all that remains is the boiling cauldron of hate.
As a real world insight into how much people love interfacing with electronics using large motor movements vs pushing buttons on a box, look at how remote controls have completely disappeared from the marketplace and everyone loves changing channels by getting up off the couch and walking around, and its such a great hit of exercise too. Oh wait the marketplace has gone the opposite direction. Yeah that too.
The other part I don't like is Microsoft pays an entire army of astroturfers who have poorly written scripts. Astroturfers should at least show some respect, by not having poorly written scripts. If your astroturfing script looks like a parody of customer support videos or a parody of the wikipedia list of logical fallacies, you're just doin it wrong. Even fanboi trolls are better than bad astroturf scripts.
Also, a "cheap camcorder" is nothing compared to this. The Kinect is a 3D sensor. You would need 2 "cheap camcorders," stereo vision algorithms/hardware to process it in real-time, and even then the quality typically wouldn't be as good as the Kinect.
Tangentially, it's also disheartening (at least for the time being) that Apple recently bought Primesense - the maker of the original Kinect. Primesense was selling a better 3D sensor for developers/researchers. Apple shut it down as soon as they bought them.
https://github.com/OpenKinect/libfreenect2
We're always looking for more contributors!
The Kinect v1 was also a pain to get working on x64 Windows, so hopefully things will be a bit easier this time around
The reason why the software works so well is because it is a lot more than a colour video camera. It captures depth images too.
'Just give me the (high level) data on any OS' seems reasonable. With, of course, access to lower level stuff if you want to create your own depth algo for example.
You want data and the various extremely advanced interpretations of it, something that is not trivial to build on your own. So there needs to exist at least one more processing layer before you can consume it for most applications. http://en.wikipedia.org/wiki/OpenNI is a good step, but until those other layers exist reliably in a stable cross-platform way, "just the data" is nearly useless.
I think MSFT is doing it right by really working on it, as we all are aware they have had their share of flops.
Putting out a great product specially like kinect is pretty exciting. Oculus is cool but its FB buyout puts it in a strange place and it was years out anyway from general adoption. The kinect is much simpler and cheaper and has already been out on the market which, gives it some validation. Plus, I am not sure how many people will go through the trouble of putting on goggles in wide range adoption. That being said I think simpler devices like the kinect have more mass market appeal, and thus the adoption of devices like kinect will ultimately be instrumental in the success of more complex interfaces like Oculus, etc...
http://www.spartanpr.com/xbox-one-kinect-u-turn-the-pr-impac...
If you're comparing sales figures from XboxOne (w/ Kinect) versus PS4 (w/o Move), I think that adds too many variables to be object right now.
Let's see how this thing goes. Good thing is there is more option now. If you don't care about Kinect, you're not forced to buy it anymore. It doesn't change the fact it's an incredible piece of technology.
http://news.xbox.com/2014/05/xbox-delivering-more-choices
First, beginning on June 9th, in all markets where Xbox One is sold, we will offer Xbox One starting at $399. This is a new console option that does not include Kinect.
Maybe it's a lack of compelling software, but most "regular" folk that I know use their Kinect as a Dance Central machine. Kind of how most people buy a wii for Wii sports and nothing else.
Observationally "regular folks" play games on their phones, tablets, wii, not xbox, vs "real gamers" have $3000 gamer PCs with gamer mice and multiple $750 graphics cards and overclocked CPUs and water cooling. I guess someone could aim for the non-lucrative market in between...
Why would you expect them to buy a completely different design just because they liked something else made by the same company? It's a fact that regular people made the Wii extremely successful, do you really expect more?
This got me thinking a lot but I'm not sure things like Kinect will eventually help devices like the Oculus Rift. Case: my girlfriend loves the Kinect i.e. how it will turn on my TV/Xbox with voice commands, sign her in via face recognition and let her navigate to Netflix easily not even having to get out of bed or move. My girlfriend is very non-technical and doesn't care about gadgets like most of us on here do. I couldn't see my girlfriend in a million years wearing a Oculus headset and I'd never hear the end of it if she saw me wearing one...
And this is an insightful comment in that the tide seems to be turning anti-kinect, and once the "value is socially constructed by others, not internally" crowd decides as a group that the kinect is embarrassing, its toast in the marketplace unless it can downscale quickly and smoothly which most products can't do.
Occulus can survive while its in growth phase and fashion oriented people don't like it. That's OK, its a big world and there's space for all. But if the kinect org and support system and ecosystem is size "X" and "not being cool anymore" results in a mandatory nearly instant decrease in size by a factor of perhaps 3, that could kill the entire product by having trouble down scaling.
With narrow casting I'm not even sure what "cool people" watch anymore but all you need is some comedians making fun of people using Kinect or maybe some parody commercials during the superbowl or during reality TV or whatever, and that product is simply toast. Watch for that cultural phenomena, and then watch for product discontinued notices six months later.
All that product needs is a sequel of that "40 year old virgin" movie where sony pays for product placement for the main character to use a kinect, or maybe a stoner/teen comedy movie with a kinect in it, and its all over.
However if you were a teen in the 90's i can tell you that girls playing video games was as rare as a gypsy with a mortgage.
Now girls play video games but it was unheard of only a decade ago. Things change, although it takes time but don't be surprised if your son's gf someday will be comfortable with the Oculus.
The future of kinect is now going to be carried out on PC, can't wait to see what people cook up with the v2.
Kinect + Leap Motion + Oculus + Control VR gloves. This set up could deliver some serious immersion, at least in the aspect of accurately bringing the user's hands into a virtual space.
But this shows once again the conflictual nature of Microsoft's two strategies with the Xbox One. It's like the gaming part is more of an afterthought, and what they really want is for Xbox to become everyone's streaming box - a $400 streaming box.
This notion of "one OS everywhere" is what we'll get us a Windows 8 smartwatch with 2GB of RAM and a Celeron processor - because who wouldn't want that?!
For now. But they also have a perfectly capable $179 streaming box that blows e.g. the Apple TV out of the water performance-wise and has a large number of popular game titles.
The XBox One is apparently popular enough to sell a couple of million devices (roughly in the same ballpark as most other streaming boxes) and packs enough punch to gently go from $400 to the 360's price point.
What I wonder about is what's in it for Microsoft. First of all, the music + movie + some apps market is much larger than the hard core gamer's market. And they make it awfully hard for people to buy into their ecosystem. A $50 streaming box would make that Windows Phone (which cannot stream to Apple TV or Chromecast) much more interesting for a lot of people than having to buy a game console for streaming.
I've always thought of video games for Microsoft as a vanity project in the same way family sedan car companies build supercars. It provides a high-performance focal point for development that trickles down useful technologies everywhere else in the manufacturing process.
DirectX, for example, probably wouldn't have ever happened if Microsoft just cared about making office productivity software.
Isn't that 30% made to take care of the social features, like "sharing a clip of the last frag I just achieved" in a FPS or something like that ? If you want to have these advanced features, you need a number of tasks to run in the background and I would not be surprised by the 30% CPU time.
Since the CPU is low performance, the OS could easily eat 50% of its total performance, leaving scant little for games.
If you believe the rumours Dave Cutler worked on the hypervisor of the Xbox.
Meanwhile, the OS on an Xbox One eats 3Gb of RAM, even when games are running: http://arstechnica.com/gaming/2013/07/report-os-overhead-tak...
PS4 has a similar memory overhead for the OS.
Modern consoles are nothing like the old-school ones where once a game was running it had nigh on total control of the underlying hardware.
They're already on clinical trials.
Kinect is a really cool enabling technology, I've seen several interesting applications, and I'm sure there are a lot more.
I guess Rift is going to do the same.
I don't know if Microsoft is exploring VR applications of the Kinect, but they haven't really done anything with VR at all. Sony has Morpheus, but Microsoft doesn't have anything.
They'll probably just miss out on VR like they missed out on mobile, and then spend billions trying and failing to catch up 4 years from now...
Nice to see microsoft have a non-sucky website now.
M$ is blowing it with Kinect One again. They are over promising, and not delivering. We bought instant fast paced gameplay bullshit the first time with X360. This is how first Kinect games actually worked:
http://www.collegehumor.com/video/6348385/mega-difficult-xbo...
or games that advertised kinect, but used only microphone lol.
And here we go again with Microsoft showing fast movements and twitch games while technology is simply not there yet. Slow down, build from the grounds up, slow paced uses first until you can reach 1-2 frames of delay instead of alienating whole user base by delivering laggy shit.