Carmack: New Rift Dev Kit In The Works, Retail Model May Run Android

Carmack: New Rift Dev Kit In The Works, Retail Model May Run Android

John Carmack

The new CTO of Oculus John Carmack hopes to ship a retail unit of the Rift sometime next year.

Speaking to Engadget, Oculus CTO John Carmack said that a new dev kit for the Oculus rift will be in developer's hands in the near future. He also revealed his vision for the retail version of the device, explaining "The way I believe it's going to play out is you will eventually have a head-mounted display that probably runs Android, as a standalone system, that has a system-on-a-chip that's basically like what you have in mobile phones."

Carmack hopes that the Rift will hit shelves sometime next year, though the team is still debating whether it will see a full retail release, a "direct-to-consumers" option, or something else entirely.

Carmack tells us that the Rift was made possible due to mobile technology's constant push forward, meaning cheaper components that are more accessible to everyday consumers. He expects mobile tech to reach 4K resolution in the not-so-distant future, making future models of the Rift much sharper visually.

Head-tracking was the most difficult component of the device to bring down to a consumer price, says Carmack. "A lot of the work at Oculus has gone into working out better position tracking," he told us. "The tracking side is something that there hasn't been as much of a push for and we're frantically working on a lot of that, because that is one of the other really large issues. But we expect that the next developer kit will have higher resolution and position tracking addressing some of these significant issues."

Are you excited for the Oculus Rift? What would it have to cost for you to pick one up? I would get one if it was somewhere around the $250 price point.

Source: Engadget

Permalink

slap a few HD cameras on it and it can beat Glass at its own game! :P

Edit: and 250 seems fair. I would pay that for the Rift if it is proven and has some good game support, dont expect me to buy it day one though.

250 sounds very optimistic. I think from a consumer standpoint, you'll be looking to shell out 350-400$. TBH, I'd rather that the device was robust and of high quality at a 400$ price-point rather than 250$ and some flimsy POS.

I'm very excited about Oculus Rift; it'll definitely change the face of gaming.

Joos:
250 sounds very optimistic. I think from a consumer standpoint, you'll be looking to shell out 350-400$. TBH, I'd rather that the device was robust and of high quality at a 400$ price-point rather than 250$ and some flimsy POS.

I'm very excited about Oculus Rift; it'll definitely change the face of gaming.

There are those of us that think that way. But as consumers have proven with the Vita, they don't care how excellent the components are that make the final product. They'd rather the cheaper, flimsier version at a lower price point.

Sad but true.

The choice of Android is a bit odd, I understand they need a heck load of per-processing on sensor data especially when going for higher accuracy but that still doesn't need a whole OS...

And the $250 mark is probably not going to happen for the next 3 years, or if you get a battered dev kit off ebay.
I am perfectly aware paying $500+ for such niche hardware is nonsense but you do need to understand that this is the price of adequate functionality, you could probably do something similar for 100 or less but the end result will be a slow unresponsive head splitting piece of poop.
It is the exact same problem as the waggle flail devices, yes at $1000+ you could get a sensor array that can track every millimeter of your movement in nanoseconds, but going 10x cheaper means the devices can barely guess what the hell is going on half the time.

$250, HD support and head tracking would be nice. But I live in Realityville, which states that reasonably priced hardware is a sin, and therefore doesn't exist until it has become obsolete.

In real money that's 154.70 Sterling. I paid that for my screen. If this can keep within the middle-of-the-road price bracket for a monitor, I suspect it will fly off the shelves faster than you can say... erm... something witty.

Something that is beginning to concern me is thier lack of haste. This $16million boost from the corporations is certainly a leap of trust from those who matter, but the downside to this appears to be it's potential revertion to a pipe dream & a prototype.

But perhaps I'm being pessimistic. This is new ground, so I don't blame them for taking the time to get it right.

In short: I feel as if they're beginning to take too long.

Mr.K.:
The choice of Android is a bit odd, I understand they need a heck load of per-processing on sensor data especially when going for higher accuracy but that still doesn't need a whole OS...

Actually, using a forked Android OS makes perfect sense for a device like this. Carmack & Co are talking about using a mobile SOC to control the device and with Android, they get an API for controlling the CPU's and motions sensors pretty much for free. It's a win win situation for everyone really. Furthermore, just because it runs Android natively doesn't mean it will look and control like a Nexus device. They can put whatever skin/UI they want on top.

Mr.K.:
And the $250 mark is probably not going to happen for the next 3 years, or if you get a battered dev kit off ebay.
I am perfectly aware paying $500+ for such niche hardware is nonsense but you do need to understand that this is the price of adequate functionality, you could probably do something similar for 100 or less but the end result will be a slow unresponsive head splitting piece of poop.

The devkit's won't be worth anything after the consumer device is released. The hardware will be hopelessly outdated. The devkit screens are what, 320x240 resolution each? Currently, they are talking 1080p screens for the first gen device and possibly 4k in the not too distant future.

EDIT: Also, hopefully it won't be niche for long. The 400$ price point (which they are aiming for according to their website) is lower than most high quality 27" screen (the 1440p ones).

Desert Punk:
Edit: and 250 seems fair. I would pay that for the Rift if it is proven and has some good game support, dont expect me to buy it day one though.

For game support look here: http://en.wikipedia.org/wiki/List_of_games_with_Oculus_Rift_support

The DevKit costs $300 as is: https://www.oculusvr.com/order/
For the consumer version they want to improve several things:
- Better resolution: At least 1920x1080, possibly 2560x1400 with Higher refresh rate (I think they've said 90-120Hz) --> could potentially increase the cost
- Better FoV and improving the optics as well as lens solution --> shouldn't really increase the cost
- Largely removing Motion Blur with low persistence display technology --> shouldn't increase the cost
- Positional tracking - 6DOF system (translational movement up/down, left/right and tilting forwards/backwards), additional to the 3DOF system employed by the DevKit (looking left/right (pitch), looking up/down (roll) and yaw). Other people are already experimenting with Workarounds for this: http://www.youtube.com/watch?v=UxSZ0fk_ubA&hd=1 --> should most definitely increase the cost
- Better/smoother software and hardware support (I think they said they are working on a solution to browse and start stuff directly from the headset and are at the moment working on an SDK for Android devices) --> shouldn't increase the cost

Although they will likely have an economy of scale at that point that they didn't have for the DevKit and that might decrease the price. I seriously wouldn't hope for anything below $300 though, potentially if they can put some feature(s) in that can really improve the experience they might even go somewhat above. They will certainly rather go for the "quality" over "really cheap and shitty" option. I don't think they would go over $400 though since they are painfully aware that the price-point plays a rather large role in their success.

Joos:

Mr.K.:
The choice of Android is a bit odd, I understand they need a heck load of per-processing on sensor data especially when going for higher accuracy but that still doesn't need a whole OS...

Actually, using a forked Android OS makes perfect sense for a device like this. Carmack & Co are talking about using a mobile SOC to control the device and with Android, they get an API for controlling the CPU's and motions sensors pretty much for free. It's a win win situation for everyone really. Furthermore, just because it runs Android natively doesn't mean it will look and control like a Nexus device. They can put whatever skin/UI they want on top.

There won't be any kind of Android SoC or internal system for rendering, that would certainly increase the price by a lot and be all kinds of stupid for people that just want to use it as a peripheral. It's just the "gaming press" doing the usual thing again, Carmack was asked where he sees the hardware in 5-10 years time: http://www.engadget.com/2013/10/18/oculus-rift-john-carmack-interview/

He answered:

The way I believe it's going to play out is you will eventually have a head-mounted display that probably runs Android, as a standalone system, that has a system-on-a-chip that's basically like what you have in mobile phones

A sheer swathe of the Gaming press took that as acknowledgement that the consumer model will apparently run Android...

See also: http://www.reddit.com/r/Games/comments/1orm64/oculus_rifts_john_carmack_says_a_new_rift_dev_kit/

Creator of the Rift here. The title of this article is very misleading. Building a mobile processor into the Rift is a long term project, we don't have any plans to do it for the first consumer product. Mobile hardware is going to get a lot more powerful in the next few years, it will allow for some big improvements in VR tech.

Having a processor onboard is helpful even when desktop computers are much more powerful. As the speed and quality of the tracking data improve, it is going to be much harder to pipe it all back to the computer for processing; bandwidth limitations alone are a bitch, especially when you have to leave enough room on the bus for things like USB audio devices, keyboards, and mice! Crunching all the numbers on an SOC that is connected directly to the motion tracking hardware would let us send only the final translation/orientation data, very small amount of data there. Doing this would also let us get away from the varying latency that most current operating systems throw into the works.

I'd definitely get it at $300, maybe higher depending on reviews and demonstrations, or if there was support for it on next-gen consoles as well as PC.

 

Reply to Thread

Log in or Register to Comment
Have an account? Login below:
With Facebook:Login With Facebook
or
Username:  
Password:  
  
Not registered? To sign up for an account with The Escapist:
Register With Facebook
Register With Facebook
or
Register for a free account here