Totally happy for them to wait until they've nailed it 100%.
What's another year after this long waiting for VR?
Virtual reality – the technology behind strapping on a headset and finding yourself in another world – has been "the next big thing" for decades. Now a Southern California startup, Oculus VR, says that the wait is over. Well, almost. Oculus Rift VR system prototype You can't buy the Oculus Rift yet, but when you can it will …
Why the focus on gaming ? There's a traditional route used to bootstrap new tech into the mainstream consumer market.
It worked for VCR's, the Innanet and, I suspect, a fair portion of (one)hand held tablets. It could be made to work for VR headsets as well.
Team up with a few Valley Impresarios and they could be (ahem) 'arriving' sooner than they think.
Nope. It depends on how rapidly (and precisely) the display reacts.
I was once in a lecture by Fred Brooks where he show diapositives of the VR headset (more like a gigantic shoebox enclosing your head) he was working on at the time. "It's just not fast enough, my wife gets sick in it" was the message. And back then, getting shaded polygons on-screen was a major feat.
It's pretty obvious that this device is a gamechanger. Have a look at the reaction of real people testing it:
The main problem right now is the low resolution. Another possible problem is that it will not support Microsoft XBox One or Playstation 4, according to Oculus Innovative because their graphics hardware is too weak.
There are competitors in the works too. Sony already have 3D glasses out (aimed at designers), and are working on a cheapified version for their new console, but it's for some reason apparently not as impressive, not sure why.
Anyway, I'd say it's a pretty safe bet that the Oculus or something much like it is the future of computer gaming.
No mistaking the use of Unreal 3 in the Medieval Rollercoaster...Whose graphics hardware is too weak: The 'their' was ambiguous...?
Agree it looks fun. But the VR that's tied to a walking machine (forget the name) is my personal favourite wishlist gadget. Exercise and fun combined!
> The 'their' was ambiguous...?
Yes. I wish one could edit posts. It's the console graphics hardware that's too weak: to avoid motion sickness you want to run the game at 60 frames per second or more, and the consoles are simply not built to do that, certainly not at 2K resolution (the minimum expected resolution of release-version Oculus Rift).
There are several "walking machine" type devices, I think the one with the biggest buzz is the Omni.
I have a pair of VR920s.
they are 640x480 each eye, motion sensors.
they are rubbish.
Short of making the stuff cheaper by using one single big old 720p LCD panel rather than micro panels I really fail to see what is so 'game changing' about oculus ? crap resolution only a bit better than the vr920s (though as they have far larger Fov the perception is resolution in vastly lower.
Haven't downvoted you, but...
The consumer OR will be at least 1920x1080 = 2073600 pixels, so over three times the pixels of the VR920s.
Head tracking will be 6dof rather than 3dof. And MUCH more accurate.
FOV will be at least 90° (>110° diagonal)rather than the VR920s 32° (like looking through a cardboard tube?)
Latency will be MUCH lower: well under 20ms.
Cheap and robust, developer excitement and support.
So not much better than the VR920 really...
It's pretty obvious from even casual use that most phone LCDs have appalling response characteristics and terrible motion smearing and use in a visor configuration would accentuate it. The experimental DN3D VR system we built back in 1995 had expensive hi quality panels and didn't have obvious motion blur.
Looks like cutting costs with phone LCDs was a mistake, let's hope they don't jump for cheap pentile OLED, scaled to full FOV the patterning will be maddening to some users. It annoys me enough even when I can't see the sub-pixels clearly. A price hike for Occulus looks likely.
I've got a devkit and have chatted a bit with the Oculus guys, and I assure you that they're the real deal. The hardware is great, even with the resolution handicap. Yeah, right now it's got terrible screen door - but honestly, after about 10 seconds, you don't see it, because it's *SO FREAKING COOL*.
Honestly, I don't notice much of a motion smearing issue; the way it's described makes it sound like one of those old passive TFTs, but I don't think it's even occurred to me to think about it during the time I've used mine.
I'm sure that whatever glasses Sony has come up with put maximum priority on trying too hard to look sexy on a model who will put them on and stand in front of a fan, not with little details like FOV, latency, and so forth.
One really cool thing about the Rift is the way they provide distortion-free output - they just invert the distortion using the video hardware. Duh; easy as pie for a 3D card to do; expensive as hell to do with glass. And they even introduce chromatic aberration into the *output image* to correct for the known characteristics of the lenses, which is why you'll see fringing on the screenshots if you look closely enough.
It's not earth-shattering development, but it IS the kind of off-the-beaten-path, let's-just-solve-the-problem thinking that the Oculus is saturated with: Using a headband with a top piece instead of trying to prop the damn thing on your ears just because "everyone knows it's VR *glasses*" is another one. Form followed function, and that's an industry first that was requisite for them to nail all the other ones.
Which they have. If the Oculus isn't the future of gaming, gaming hasn't got much of a future.
;invert the distortion using the video hardware. Duh; easy as pie for a 3D card to do; expensive as hell to do with glass'..
........................"invert the distortion"....
........................"expensive as hell to do with glass"
Can you explain / expand on that please?
> a minimum dual 1080p displays. Avoiding motion sickness requires at least 60 fps.
The developer version is just 1280x800 / 2, easily manageable by any halfway decent gaming machine. The release version will have higher res, but I doubt it'll be as high as 3840 × 2160 / 2, I'm not even sure if there exist 7" screens with 4K resolution.
> rules out consoles for the next 7-10 years
It rules out the Xbox One and the Playstation 4 (and yes, those are not due to be replaced for 7 - 10 years). However, the Steambox machines from Valve are a different matter: they have sufficiently powerful hardware and it is already known that they will support Oculus.
"Avoiding motion sickness requires at least 60 fps."
Not in my experience. Latency is the biggie, and even now the Rift is excellent in that regard; I've also played a few tests and homebrews on it that dropped far below 60fps, and didn't suffer a hint of motion sickness as a result. Granted, anecdote is not proof, but I work in this area and have experience with this kind of stuff, so I don't think it's as critical as you might think. I'm sure to some extent it depends on the game type, too; bright scenes are more framerate-critical than dim ones, certain kinds of panning are more obviously bad than other motion, etc.
"The developer version is just 1280x800 / 2, easily manageable by any halfway decent gaming machine."
Yes, but it's generally a much better result to run at 1920x1200 with AA and other settings cranked and then have the panel downsample; you'd think that on-panel downsampling wouldn't be as good as natively-produced output with good quality settings, but apparently there's something else going on. It could be that some of the techniques used to render poly edges and textures don't work as well if the user is seeing them rendered on a surface with such a large FOV - ie, they don't intend for the the stuff to subtend such a large angle in your vision. Or there's some interaction with the distortion the Rift applies?
Not sure. Either way, the Rift *is* going to want some pretty serious hardware, since with the final version it will likely be advisable to either drive it at even higher resolution and downsample, or use MSFAA (or whatever the hell it's called), or some combination of those things. And due to the huge FOV, getting things perfect at a pixel level becomes really important, so fudging on AA and AF starts to hurt a lot more than it does when you're targeting a console player with his 40" TV five feet away.
"I'm sure there will be porn (there always is), but personally I'm not sure low-res rendered porn would be a big seller, even in believable 3D."
Rule 34... there already is. Actually, rule 34 would mean there'd be porn *about* the Rift, so... ahem, well, let's never mind that.
Yeah, there already is - check riftenabled.com, which has a rather eclectic gathering of homebrew, indie, and pro Rift-supported stuff. I haven't tried the things actually labeled 'porn' - there are only two possible results there, and I'd like to avoid both of them.
I did try a little demo thing that shows a laser-scanned, leotard-wearing model, and I can tell you that the low-res thing itself isn't necessarily a dealbreaker, because the feeling of "thereness" is so intense. If you maneuver yourself close to the model, you actually get the feeling you're invading someone's personal space - as one commenter put it, "I felt like I was going to get slapped". This isn't hyperbole, or, "You rationally think that you'd be invading someone's space", but the actual hardwired, "I shouldn't be doing this" feeling.
If the experience is powerful enough to trigger that, with a low-res prototype and a single static model in the first generation of test applications made by one guy, there are going to be some interesting offerings once people with some horsepower start in on this. While I doubt that *live rendered* porn is going to work out for the mainstream - the uncanny valley is plenty uncanny enough in 2D - I could see people doing the equivalent of photo sets but with laser-scanned or SFM-driven data.
"OK, baby, that's hot; just hold that position while I set up the LIDAR..."
You'll probably be fine. I have absolutely horrid nearsightedness without my glasses - eg, from where I'm sitting, without my glasses I can barely read the "The Register" banner, and 'Biting the hand that feeds IT' is just a blob - and I can pretty much use my Rift fine without my glasses on *and* with the standard lenses (it comes with different ones for people with different vision characteristics). Granted, part of that is because the Rift's resolution is low enough that there'd be no point in tack-sharp vision, but even I can focus pretty well on the screens' pixel structure if I try, which suggests that there's a pretty wide range where you can be comfortable.
It's also possible to use the Rift while wearing glasses; it's designed cleverly enough that it's completely comfortable once they're on. The biggest problem is that it's a bit hard to put them on without smashing your glasses up against your eyeball and smearing them, and then you have to take off the Rift...
Basically, I wouldn't worry. You'll probably be fine with the contacts, and could probably just take them out and use the alternative lenses at worst.
And actually make use of the hardware that is superior and available now.
Optical head tracking (eg TrackIR) is vastly more superior to OR's inertial tracking, plus you can build one yourself, there are plenty of instructions on the internet and it will work with any display - even a wall of 4K monitors.
3D is subjective. For competitive gaming, it gets in the way. You certainly would not want to play a 6 hour marathon with a VR headset. Players of games with lots of controls (flight sims, MMOs etc) will not likely buy them, because you cannot see the keyboard while playing. I can also see parents not buying them for kids because they perceive that it will cut them off from the real world more. I can also see safety issues - people will try it standing up and fall over and into the telly etc.
An expensive gimmick.
"Optical head tracking (eg TrackIR) is vastly more superior to OR's inertial tracking"
I've used both. In my experience, the OR's is at least equal to the TrackIR. I'm actually not sure it could *get* much better; it's blazing fast, high frequency, smooth, not jittery - it disappears entirely when you're using the RIft.
Yeah, a TrackIR will work with a wall of 4K monitors, but you must have a hell of a neck to take advantage of that in this context... ;)
You're right that if you want to go online and pwn some n00bs in TF2, the Rift will, by its nature, be a handicap - I think it's hard to beat mouselook for efficiency in aiming. But just because something is the best technical way to achieve a result doesn't mean it's *the best way to do games*.
There are solutions to the other issues, too; it's just that they haven't been implemented because nobody has had a reason to implement them yet. You mention flight sims; there are already ways out there to track your hands and fingers, so it's *very* easy to envision combining that with an in-game render to replicate your body so you can flick switches and press buttons with your real hands (but without anything actually there). It's certainly plausible; the Rift does such a good job of planting you in the world that when I was using it I was pointing at on-screen objects while talking with my friend, who was highly amused to say the least.
Yeah, there are some genres of game that won't work well with the Rift. So what? There are also a lot of genres of game that don't work well *without* the Rift, and they haven't even been made yet. And my experience suggests that there's going to be a lot of them.
Even if you have to use a keyboard while playing, the result isn't as awkward as you might think. Chances are you're not staring at it while playing anyway. If that were the stumbling block, the benefits of the Rift easily overwhelm it.
I'm sure that your point about parents not wanting to turn their kids into some kind of cyberpunk zombies has merit - for a small and paranoid segment of the population. But there are people who think Harry Potter books are the world of the devil, and they've sold reasonably well nonetheless.
As far as a 6-hour marathon - it might get a little hot, yeah. But even the prototype is *VERY* comfortable to wear. If you're basing your judgment on things that have tried to look sexy and that hook over your ears, you're saying that a fine leather chair must be uncomfortable because you sat on a cow skeleton once and it wasn't all that. They really did a stellar job; the details of implementation are numerous and have to be experienced to be understood.
I've said from the start of this that one of Oculus' biggest difficulties after actually making the Rift good would be to convince people who can't imagine that a company could do something without making the mistakes everyone else has been making for years. Your post suggests that barrier is a real one. But trust me on this - it really is that different. Sometimes people just build stuff Right. It's rare, but this is one of those times. Don't count them out just because everyone else has screwed it up so badly!
" because you cannot see the keyboard while playing"
They could always develop a keyboard where there are sensors around the edges so that a tracking unit on the OR can place it, then all they need to do is render it on-screen for you :) Job done.
Quite how CoD would look with a keyboard off to one side though is another matter :)
which competitive gamer cannot touch type?
HOTAS setups for flight sims are fine, i never look at my buttons in combat, i only use the keyboard to messages at the people dumb enough to fire through friendlies to score a kill.
how many buttons do you need for a racing sim?
A spaceship sim could be expected to have a few more button requirements though - although if the OR is all it's cracked up to be when it comes out I will be building a rig especially for Star Citizen including a dedicated button/switch console that maps to keys for the in-game controls.
Not sure I'll go as far as a full-on cockpit, but I can see a comfy chair and well-placed controls in my future :)
Oh my, a lot of fanbois voting me down it seems without even considering how these things work.
"You're right that if you want to go online and pwn some n00bs in TF2, the Rift will, by its nature, be a handicap - I think it's hard to beat mouselook for efficiency in aiming."
Dumb comment. No-one is suggesting that OR be used to control aiming, that would clearly be stupid.
In response to saying you cannot see the keyboard - "which competitive gamer cannot touch type?"
Are you SERIOUS? Go play a flight sim, put a cloth over the keyboard and try to hit the myriad of controls without hitting the wrong key by mistake. Go on, try it now. Rest your hand on the WSAD keys, close your eyes and see how long it takes you to find numpad 5.
And forget about if you do not use a computer for a living - I have mates who play FPS and still need to look at the keyboard to the correct number key.
"especially for Star Citizen including a dedicated button/switch console that maps to keys for the in-game controls."
Yeah good luck with that, it's a very tall order. You will be limited to one kind of ship only - if the ingame console moves, you are screwed. Furthermore, most likely you will want some kind of amplification of your head movements, else it will get very uncomfortable to try and look over your shoulder whilst operating the stick and throttle. If you have any kind of amplification, the virtual buttons will not correspond in 3D space with your physical console.
I still say that will not work as well as you think.
I reckon you would be better off with a HOTAS system in your hands, and a joypad on your lap for extra buttons. At least then when you grab the thing, you can instantly tell where the rest of the buttons are by feel.
A flat panel with arrays of switches and stuff would be too hard to operate blind - you are going to need some kind of haptic feedback that tells you with one touch where your hand is in relation to all the buttons on the panel - otherwise you will be devoting too much concentration to find the control you want.
""You're right that if you want to go online and pwn some n00bs in TF2, the Rift will, by its nature, be a handicap - I think it's hard to beat mouselook for efficiency in aiming."
Dumb comment. No-one is suggesting that OR be used to control aiming, that would clearly be stupid."
On the off chance anyone will see this: The OR *can* be used to control aiming; there are various implementations, in particular with TF2. But no matter how the *aiming itself* is achieved, you're still *looking around* with the OR, which means that if you're not also aiming with it, aiming and looking are decoupled (as in real life). But with mouselook, aiming and looking are the same, so, for practical purposes, mouselook means you've got the movement of the whole screenful of graphics to correlate your aim with. It's much more difficult to aim without the static reference point of the monitor frame, and the static internal reference of your stationary head.
The only way to recouple those things is actually *to* aim with the OR. But that's impractical, because the mouse is necessary to do rapid turns; you can't be spinning around in your chair for obvious reasons.
This shouldn't be construed as a knock on the OR - it's just that it won't be an idea means with which to compete in FPSes as they are currently implemented. Ideally, though, the OR is the kind of device that will, quite literally, be a gamechanger, in that the games themselves - whole genres - are going to be reshaped to match its capabilities and characteristics. It's a completely different world.
Biting the hand that feeds IT © 1998–2022