Slashdot Mirror
Carmack On VR Latency
An anonymous reader writes "For a while now, John Carmack has been pushing to bring virtual reality technology back to the gaming world. VR was largely abandoned over a decade ago when it became apparent that the hardware just wasn't ready to support it. In 2013, things are different; cheap displays with a high pixel density and powerful processors designed for small systems are making virtual reality a... reality. One of the last obstacles to be conquered is latency — the delay between moving your head and seeing your perspective change in the virtual world. In a lengthy and highly-technical post at #AltDevBlogADay, Carmack has outlined a number of strategies for mitigating and reducing latency. With information and experience like this being shared with the game development community at large, it shouldn't be long until VR makes a permanent place for itself in our gaming lives."
Dactyl Nightmare
--- rapper/producer/bachelorette party stripper
You mean "Tock"?
Also, VR will make a massive comback if, as I suspect, Google Glass takes off and competitors crop up. This isn't a new idea, since Steve Mann has been wired for VR since, what? The 80's? I think its time has come.
If computers were people, I'd be a misanthrope.
.... I can't help thinking this
Surely any dedicated gamer would see the value in simply injecting a thickening agent into the endolymph of the Vestibular system. With careful dose control, that should induce a matching lag in the perception of motion, thus providing a highly realistic experience!
*Ability to walk and/or perform normal ocular saccades not guaranteed, please refrain from the use of industrial silicones in medical applications.
There has been 3D movies for decades, (50s or 60s... to lazy to check) but just recently the technology allow it to became mainstream.
Maybe it's time for a VR grand comeback.
(... I'm still waiting for my holosuite...)
Stanford has an elaborate VR lab. The system is 120FPS, and the lag is low, but I'm not sure how low. There's full motion tracking of the subject in a 20 foot by 20 foot space. They have public tours every Friday. Sign up and try high-end VR.
This isn't a graphics lab. It's a psychology lab. Some of the results are scary. They've had kids go through a VR experience of swimming with sharks. A few weeks later, the kids are asked about it, and a sizable fraction of them believe they really did it, adding details that were not in the sim like what they ate while visiting the sharks.
They're always running psychology experiments, and looking for volunteers. Pays $15/hr.
How can they not talk about Carmack's chosen one? This seems to be the best hope for affordable VR for the masses.
http://www.oculusvr.com/
Oh, yeah. Michael Abrash did this two months ago.
sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
yeah. Fred Brooks handled that last millennium.
Google Glass doesn't have anything to do with VR. You are confusing it with Augmented Reality.
`echo $[0x853204FA81]|tr 0-9 ionbsdeaml`@gmail.com
I remember when everyone suddenly got excited about virtual reality in the early 1990s. Of course, back then it was more the concept and the possibilities that triggered peoples' imaginations- actual VR systems and games did exist at that time, but were never really widespread, probably due to the limitations and cost of the then-current technology and the fact it was essentially a novelty.
:-O
One commentator, however, said something that has stuck with me ever since. I can't remember the exact wording, but it was along the lines of...
"Eventually the current fad for Virtual Reality will pass, and everyone will forget about it. Then one day you'll look around you and realise that it's everywhere."
(*) If you remember it too, then yes- it really *was* that long ago
"Slashdot - News and Chat Sites Deviant". (Click "homepage" link above for details).
incidentally Google Glass has very little to do with AR. Its just a transparent 320x200 (or less) resolution display with camera for your phone :/
Who logs in to gdm? Not I, said the duck.
Numerous studies showed that extended use of VR could cause severe problems, namely permanent lazy eye (loss of depth perception). I believe it was Nintendo that dropped a VR product because of their own studies (I'm too tired to go look for the data at the moment). Government studies also found this to be true, so working in VR in Government jobs is restricted (or was when I was there) to 8 hours per week.
More studies need to be done to determine safe levels, and most importantly people should be made aware of the potential risks to health. Currently there are no warnings that I'm aware of and most people have no knowledge of the studies.
-The wise argue that there are few absolutes, the fool argues that there are no probabilities.
Its like the Wii, its awesome for a little bit, but at some point your just going to want to sit down and play a game like normal. Thats kind of like the way VR is, its awesome ... then it becomes more and more of an inconvenience, then one day you clean off your desk cause the damn gear, is in your way.
Balance the weight of the VR screen with counter-balances around the head, and you produce a hideous unwieldy 'helmet' that will still cause neck-pain when the head is 'snapped'.
throughout time military helmets have had a certain amount of weight to them.
It is no coincidence that military flight simulators do NOT use VR goggles, even though they have the funds to do so.
Such systems are about training a pilot to use a real jet. a real jet has an large amount of very small controls.
here's an enthusiast simulator startup video: http://www.youtube.com/watch?v=qBFb45nPSNs
in a military simulator you can physically interact with all of those switches. so I would guess the military doesn't use VR goggles yet because of the latency and the difficulty of tracking exactly where the pilot's hands are.
You see, modern games now render using a high-latency pipeline, with some work for future frames being calculated before the current frame is even done. It is ESSENTIAL that the input loop is low frequency compared to the render system.
I think that's the other way around. at least with race-car games it's the other way around. in those you do several input and physics frames, and render every fourth frame or so. or they can be asynchronous, with the render thread drawing the latest complete physics frame.
Hacking one of the gyro based optical image stabilisers from digital cameras would do the trick. Stabilise the overall frame and lose the motion sickness, lag in moving elements (and perspective changes) won't be noticed or won't be noticed enough to break immersion. Having the entire scene out of sync is very easy to detect, even if you aren't consciously aware of it.
I briefly worked for a VR company last century, 1st day they pointed out that stereo is an option not a necessity, so is resolution. All that matters is minimising lag (which implies good tracking as a prerequisite). Still just as true today. Turns out you can cope with (slow) drift in the tracking as well, it's annoying rather than immersion breaking.
Funnily enough, that's exactly one of the methods the article discusses, except there are some details you hav glossed over that he fills in.
If you've played Nintendo Land on the Wii U you've seen this problem basically solved: you look around by moving the Wii Pad as though it were a window, and there's no latency problems at all (unlike with the Wii, thankfully). Just shrink down the display and make it head-mountable.
I don't think you read the article. He's talking about ways to design new displays that can reduce the head-movement -> display latency to 3ms or so, without going through the GPU.
I wonder if they can combine it with the unlimited detail rendering technology developed in OZ.. That would be kick-ass
Unlimited Detail Real-Time Rendering Technology Preview 2011 [HD]
http://www.youtube.com/watch?v=00gAbgBu8R4
"Tick", hyphen, "tack"? Your 3D printed clock sounds like junk.
But his breath is minty fresh.
we can crash someones brain with a QR code.
In the free world the media isn't government run; the government is media run.
incidentally Google Glass has very little to do with AR. Its just a transparent 320x200 (or less) resolution display with camera for your phone :/
Having a display at a constant spot in your field of view, and a camera perfectly aligned with your field of view, doesn't give you possibilities to quite immersively augment your perception of reality? Come again?
After playing RAGE, I'm pretty sure whatever Carmack is doing at the moment, it's about 10 years out of date.
incidentally Google Glass has very little to do with AR. Its just a transparent 320x200 (or less) resolution display with camera for your phone :/
Having a display at a constant spot in your field of view, and a camera perfectly aligned with your field of view, doesn't give you possibilities to quite immersively augment your perception of reality? Come again?
not if the display is in the corner of your eye instead of overlapping said reality
but hey, Im sure that wont stop Google from redefining AR to "reality with some subtitles in the corner"
Who logs in to gdm? Not I, said the duck.
You mean "Tock"?
Also, VR will make a massive comback if, as I suspect, Google Glass takes off and competitors crop up. This isn't a new idea, since Steve Mann has been wired for VR since, what? The 80's? I think its time has come.
You are confusing VR with AR. Augmented reality systems like Google Glass simply overlay information about our environment in our visual field -- it doesn't replace reality like a VR system is supposed to, it just augments it. Augmenting reality *is* trivial, and the solutions are easily within the domain of current and near-term forseeable engineering technology. Functional VR, otoh, means directly interfacing with the proprioception/kinesthesia network in human neural anatomy that tells the brain what the body is doing in relation to other objects in the mental model of the environment. Modeling those other objects is trivial, as most VR researchers, including Carmack, will assert, but VR researchers are going to also have to figure out how to intercept, decode, modulate and retransmit the electrical impulses traversing the PK network that represent your body's position relative to those objects, which is decidedly non-trivial, especially while suppressing the original signals telling your brain that you are actually motionless, and even more especially doing it in a reversible way. These solutions, IMHO, require way more knowledge of human neural anatomy than we presently have, and will require the invention of new bio-engineering technologies to exploit it once we have that knowledge. AR has a distinct market advantage right now, so I'm certain Carmack and other VR researchers will turn their ingenuity to AR and away from VR, once they realize this.
One: laser ring gyros instead of mechanical accelerometer or visual head tracking systems two: Render a frame larger than FOV and digitally move that before the next frame is rendered.
Not hard.
The first costs on materials, power and research, and the second requires loads of processing power. Cheap compared to the cost of, say, an F35, but expensive for consumer electronics (which I gather is the point for this article).
For your edification as an armchair specialist on laser ring gyroscopes, framebuffers and all things VR, may I suggest that you at least bother to multiply the number of bits per pixel times the number of pixels times the FPS, to at least get a handle on how much data needs to be processed and understand the problem before you post.
The three laws of thermodynamics:(1) You can't win. (2) You can't break even. (3) You can't even quit.
Seriously, though: as a longtime admirer, I have to say his genius would be better used in gaming if he rid himself of the albatross known as id.
Imagine what he could do in any number of R&D areas if he didn't have to ship games bogged down by boring narratives, bland level design and twenty-year old ideas of corridor-based run-and-gun.
I wish he'd turn his attention to improving AI and developing emergent gaming. The next frontier awaits, but our Einstein is bent on rendering the same old mousetrap in ever higher fidelity.
It's not like Google magically broke you jackass. Here is one page, and here is another. When reading that second one, remember what is discussed in the first. Also follow the links in the first article. The Government studies are harder to find, but do exist.
-The wise argue that there are few absolutes, the fool argues that there are no probabilities.