Support NeoGAF

[Onemic]

Well, if serversurfer is right, we'd all just be playing Eve Valkyrie anyway. RIP, roomscale!

I really do wonder how they'll eventually solve locomotion with the given space limitations. Whatever solution they eventually come up with should be pretty interesting to say the least.

 

[plagiarize]

I really do wonder how they'll eventually solve locomotion with the given space limitations. Whatever solution they eventually come up with should be pretty interesting to say the least.

Eventually they will need to. Room Scale doesn't feel like an end point for most things, as I've said before... however now that I've played a few games that do the teleportation thing (Vanishing Realms and Budget Cuts) and are smartly designed around the limits of room scale, I'm very encouraged by how they played in my small space. There is a huge amount of gameplay possibilities within room space that we haven't even begun to explore. So long as teleportation takes at least a degree of concentration and is limited to a relatively small range from your current position, it doesn't feel like a super power at all.

It isn't great in everything. I didn't like how that robot invasion shooting game handled it (you're on buildings shooting stuff). It didn't seem to mesh with the gameplay, and being able to teleport away whenever you get swarmed seemed far too easy.

So it's not a magic bullet, but if it works in something like Vanishing Realms where it isn't implicitly part of the in game fiction... I see it with a bit more optimism than I did before.

I still dislike the chaperone grid (although more space would reduce that annoyance as it wouldn't feel like it was on more often than off). I still wouldn't want to play a game like Dreadhalls or Adr1ft room scale style (both I think are more suited to standing play with controller driven locomotion). But it's definitely a creatively furtile space that's going to offer up compelling and different experiences to the other VR spaces.

Exploration is a big thing for me in games in general. I'm very happy with how much it felt like I was exploring the worlds of Budget Cuts and Vanishing Realms. It hasn't soured me on the other types of VR games at all though.

 

[Zalusithix]

You're dismissing it too soon. There is some promise hidden underneath that orange coat. Probably. Hopefully. Hardware seems to be good (enough), it'll be all about the Software. If this thing supports OSVR/OpenVR this could become a viable Option (amongst other headests that will undoubtly See the light of day in the near future).

To adress some of your questions. This thing has two modes of operation: Untethered/Mobile and tethered to a PC.

The first Mode requires that you buy the package with the "Smart" gamepad which will be connected to the Headset via USB. No wands or positional tracking in this mode. So let's quickly forget about that, there's cardboard for mobile porn.

Second Mode of operation is PC Mode (you can buy this without the snapdragon loaded gamepad for 200$ less!). You connect the HMD via a single USB cable to your PC (same Port on the HMD as for the gamepad).

The TrackingStation with two undetachable cameras can be used wireless (and most probably wired) and connects to your PC via 802.11ac dual band WIFI. So now you got the cameras and HMD connected to your PC where the Sensor Fusion happens. The wands connect to either your PC or probably to the Tracking Station via Bluetooth.

That should be it.

Some other thoughts and info from the other thread.

Some other interesting tidbits:

Thanks for the info since the original links were light on details. That said, a WiFi camera? Ehhh... A camera -> encode -> transmit to AP -> route to computer -> software stack (decode, analyze) chain might introduce a bit too much latency for VR. Hopefully it has a wired option.

Anyhow, as you mentioned, a single camera solution has occlusion issues, so the use cases for this will be similar to PSVR, not the Rift or Vive. Assuming PSVR does support the PC anywhere near release, it'll be the primary competition. In that case the price advantage becomes lower. Granted, it has some hardware advantages against the PSVR, but until people get eyes on both, it's too early to call it one way or the other.

Ultimately the proof will be in the end product. I'm quite skeptical about startups in general being capable of introducing quality hardware, let alone little known Chinese ones where crap knockoffs are unfortunately the rule rather than the exception. If Oculus/Facebook and HTC/Valve are tripping with their products and implementations, I don't have much faith in a small no-name faring well.

Also, supporting ASTW (aka reprojection) is a software thing, so if they're supporting OSVR/OpenVR APIs it comes with the package. It's not really a feature worth mentioning.

 

[Croatoan]

I really do wonder how they'll eventually solve locomotion with the given space limitations. Whatever solution they eventually come up with should be pretty interesting to say the least.

Hover Junkers already shows what we will have in the future. Directional (WASD) movement based on either where your head is facing or where you are "Pointing".

Once we have better screens (Near Lightfield Tech) and better performance (Foveated rendering) more sickness will be alleviated.

Also, the human body can get use to stuff like this. I mean, we got use to moving in cars.

Its possible future headsets might find a way to stimulate the vestibular system as well.

Basically, this tech is going to change A LOT over the next 10 years. I don't think VR will be mainstream for at least another 5 years and it wont be the main method of gaming for 15-20 years.

We are in the NES days of VR really.

 

[SinSilla]

...I'm quite skeptical about startups in general being capable of introducing quality hardware, let alone little known Chinese ones where crap knockoffs are unfortunately the rule rather than the exception. If Oculus/Facebook and HTC/Valve are tripping with their products and implementations, I don't have much faith in a small no-name faring well...

I heard that "Pico is funded by Goertek - the manufacturer for Rift and PSVR."

If that's true then thats quite the advantage over the regular Chinese clones.

...Also, supporting ASTW (aka reprojection) is a software thing, so if they're supporting OSVR/OpenVR APIs it comes with the package. It's not really a feature worth mentioning...

We don't know yet if they support OSVR. ASW support is therefore currently a feature of their own SDK.

 

[Candescence]

Just wondering, for those in the know, how practical would it be to adapt the Vive's lighthouse tech to all sorts of wearable things that can track various body parts? I imagine it'd be feasible for limbs and torso, and maybe even the hands and feet, but not for individual fingers.

 

[Zalusithix]

I heard that "Pico is funded by Goertek - the manufacturer for Rift and PSVR."

If that's true then thats quite the advantage over the regular Chinese clones.

If that's true, then I rather dislike the whole situation. A manufacturer of other companies technology funding competition to that tech? That just rubs me the wrong way.

 

[Cyriades]

I heard that "Pico is funded by Goertek - the manufacturer for Rift and PSVR."

That's interesting to say the least.

 

[viveks86]

It isn't great in everything. I didn't like how that robot invasion shooting game handled it (you're on buildings shooting stuff). It didn't seem to mesh with the gameplay, and being able to teleport away whenever you get swarmed seemed far too easy.

So it's not a magic bullet, but if it works in something like Vanishing Realms where it isn't implicitly part of the in game fiction... I see it with a bit more optimism than I did before.

Yeah, there will always be some hybrid solution that is applied on a case by case basis. Eventually devs will converge on some standardized options, such as:

Roomscale + teleportation
Roomscale + analog stick movement (not turning) + some technique that alleviates motion sickness (like tunnel vision)
Roomscale +walkabout
Roomscale + jogging/walking in place
Roomscale + "ratcheting"

It's just another way to play games and can quite easily be a superset for all games, since it inherently supports standing and seated arrangements anyway. Never really understood the skepticism around it, which mostly seems to stem from "oh I don't have space for it, so it will never catch on" knee jerk reactions. The way I see it, roomscale is an all encompassing tracking technology that EVERYONE should implement. Whether it's standing or seated or moving is a genre and game specific implementation of roomscale. All other forms of tracking are destined for obsolescence.

 

[Onemic]

Yeah, there will always be some hybrid solution that is applied on a case by case basis. Eventually devs will converge on some standardized options, such as:

Roomscale + teleportation
Roomscale + analog stick movement (not turning) + some technique that alleviates motion sickness (like tunnel vision)
Roomscale +walkabout
Roomscale + jogging/walking in place
Roomscale + "ratcheting"

It's just another way to play games and can quite easily be a superset for all games, since it inherently supports standing and seated arrangements anyway. Never really understood the skepticism around it, which mostly seems to stem from "oh I don't have space for it, so it will never catch on" knee jerk reactions. The way I see it, roomscale is an all encompassing tracking technology that EVERYONE should implement. Whether it's standing or seated or moving is a genre and game specific implementation of roomscale. All other forms of tracking are destined for obsolescence.

Are there any games that do roomscale plus analog movement with no turning? That seems like it would be pretty cumbersome, but if anyone could post some impressions of how it feels that would be great.

 

[AbortedWalrusFetus]

Are people getting motion sickness with games with vehicles? Or is it just situations where its you as a character that is moving?

 

[SinSilla]

If that's true, then I rather dislike the whole situation. A manufacturer of other companies technology funding competition to that tech? That just rubs me the wrong way.

That's interesting to say the least.

Thats what the Internet has to say:

GoerTek is well positioned, in our view, as the major EMS for Oculus Rift, Sony PSVR and. LeVR.

Goertek provided its advanced optical solution, system integration and world class supply chain support.

It's indeed a delicate situation as there are quite obvious similarities with possibly even more under the hood. They were clever enough to alter the design enough for not being a direct Plagiat. I'd bet money that the shape of the facemask is exactly the same as the Rifts (with a cheaper cushion though).

 

[viveks86]

Are there any games that do roomscale plus analog movement with no turning? That seems like it would be pretty cumbersome, but if anyone could post some impressions of how it feels that would be great.

Windlands does it and I've had no problems with it, but YMMV. Some people are definitely getting sick. I find it extremely enjoyable, as long as I don't turn using the other trackpad. Since it does roomscale, I can simply turn in real life and walk/balance myself whenever it feels right. It's so liberating!

The hope is that with stuff like tunnel vision (or some other technique people haven't thought about yet), they'd be able to reduce chances of motion sickness further. Personally, I can't wait for a massive game like Cyberpunk 2077 or Star Citizen with roomscale + analog movement. It's the least disruptive method for keeping traditional level design but still getting the benefits of roomscale and motion controls.

 

[plagiarize]

Are there any games that do roomscale plus analog movement with no turning? That seems like it would be pretty cumbersome, but if anyone could post some impressions of how it feels that would be great.

Well. Sort of. Kinda. I mean most VR games let you just noclip wherever the hell you want. When I'm playing Ethan Carter, I can play it standing up, and use the joypad to walk 'forwards', yet still walk around my room to investigate something.

 

[Onemic]

Windlands does it and I've had no problems with it, but YMMV. Some people are definitely getting sick. I find it extremely enjoyable, as long as I don't turn using the other trackpad. It feels so liberating!

The hope is that with stuff like tunnel vision (or some other technique people haven't thought about yet), they'd be able to reduce chances of motion sickness further. Personally, I can't wait for a massive game like Cyberpunk 2077 or Star Citizen with roomscale + analog movement. It's the least disruptive method for keeping traditional level design but still getting the benefits of roomscale and motion controls.

Oh, so Windlands lets you rotate with analog controls, you just intentionally dont use it?

Well. Sort of. Kinda. I mean most VR games let you just noclip wherever the hell you want. When I'm playing Ethan Carter, I can play it standing up, and use the joypad to walk 'forwards', yet still walk around my room to investigate something.

Ah. I wonder which games will actually restrict analog movement to forward/backward motion only with roomscale as standard then. Are devs actively talking about using this control scheme?

 

[Tain]

Windlands does it and I've had no problems with it, but YMMV. Some people are definitely getting sick. I find it extremely enjoyable, as long as I don't turn using the other trackpad. It feels so liberating!

The hope is that with stuff like tunnel vision (or some other technique people haven't thought about yet), they'd be able to reduce chances of motion sickness further. Personally, I can't wait for a massive game like Cyberpunk 2077 or Star Citizen with roomscale + analog movement. It's the least disruptive method for keeping traditional level design but still getting the benefits of roomscale and motion controls.

I haven't tried roomscale yet, but I can handle standing VR with analog movement pretty solidly (though sticking to full movement speeds seems to feel slightly better). I'm hoping for the same thing: some large-budget traditional-style games with VR modes that, if they have to, allow use of the traditional stick for those that can handle it.

 

[viveks86]

Oh, so Windlands lets you rotate with analog controls, you just intentionally dont use it?

Correct. They have an option to snap turn fixed degrees or turn full analog. But I can turn in real life, so they are completely redundant for me. I can do a 180 faster than an analog stick anyway

It's bound to reach a point where that option may get completely disabled (thereby freeing up a trackpad for other actions), since analog turning in VR is pretty much universally criticized. People would still need something for seated arrangements, but it seems unnecessary for standing.

 

[plagiarize]

Correct. They have an option to snap turn fixed degrees or turn full analog. But I can turn in real life, so they are completely redundant for me. I can do a 180 faster than an analog stick anyway

It's bound to reach a point where that option may get completely disabled (thereby freeing up a trackpad for other actions), since analog turning in VR is pretty much universally criticized. People would still need something for seated arrangements, but it seems unnecessary for standing.

Standing and physically turning around is definitely the way to go for most stick driven movement. I have zero issues with it. The track pad will likely be much better than a stick for turning though. I had zero problems turning using a mouse in sessions that lasted hours, but turning with the pad gets to me in about fifteen minutes.

 

[viveks86]

Well. Sort of. Kinda. I mean most VR games let you just noclip wherever the hell you want. When I'm playing Ethan Carter, I can play it standing up, and use the joypad to walk 'forwards', yet still walk around my room to investigate something.

The issue with Ethan Carter is the head tracking isn't tied to your movement at all. So you end up with situation where your head is in one direction and analog forward pulls you in a different direction. So yeah, "sort of kinda" is exactly how I'd describe it. Very disappointed with their lazy ass implementation on the Vive. Add atrocious image quality to that (with or without AA) and it was time for a refund

 

[RealMeat]

The issue with Ethan Carter is the head tracking isn't tied to your movement at all. So you end up with situation where your head is in one direction and analog forward pulls you in a different direction. So yeah, "sort of kinda" is exactly how I'd describe it. Very disappointed with their lazy ass implementation on the Vive. Add atrocious image quality to that (with or without AA) and it was time for a refund

Weird. On the version on the Oculus store, using a rift, up on the stick always went were you were looking, so you could play it by physically turning if you wanted.

 

[plagiarize]

Weird. On the version on the Oculus store, using a rift, up on the stick always went were you were looking, so you could play it by physically turning if you wanted.

Yeah. That is weird. I wonder why they dropped the ball on the Vive version like that.

Still ideally detached movement and aiming should always be an option for the player, imho.

 

[Onemic]

Listening to the Giant Bombcast and they're gushing over this thing called the Magic Leap, saying that it's the real deal over all the other HMD's. What is it and what do you guys think about it vs the Rift and the Vive?

 

[Zalusithix]

Listening to the Giant Bombcast and they're gushing over this thing called the Magic Leap, saying that it's the real deal over all the other HMD's. What is it and what do you guys think about it vs the Rift and the Vive?

Everything I've seen from Magic Leap has been AR, not VR. While there's overlap between the two, they're largely different things.

 

[viveks86]

Weird. On the version on the Oculus store, using a rift, up on the stick always went were you were looking, so you could play it by physically turning if you wanted.

Hmmm. Think i'm remembering it wrongly. May be the issue was object interaction? I remember my head being somewhere but the game thinking my body was somewhere else. I'd try to interact with something, but nothing would happen till I moved my body from where it previously was. It just felt terribly tacked on.

On top of that, it wouldn't even support the steam controller, let alone the vive controllers. Even Mind: path to thalamus supported the vive controllers for movement. And that was FREE.

Listening to the Giant Bombcast and they're gushing over this thing called the Magic Leap, saying that it's the real deal over all the other HMD's. What is it and what do you guys think about it vs the Rift and the Vive?

It's definitely really cool tech and has a place in vr, though it's a much better fit in ar. However, it's use cases will be very different from that of motion controls. Holding a gun, for example, is infinitely better with motion controls and haptics. The real future is when all these technologies converge. Then you end up with hybrid solutions that can cover every use case. I don't think one needs to replace the other

 

[Paganmoon]

Listening to the Giant Bombcast and they're gushing over this thing called the Magic Leap, saying that it's the real deal over all the other HMD's. What is it and what do you guys think about it vs the Rift and the Vive?

Looks like some AR thing, https://www.magicleap.com/#/home
If there wasn't text about it on MIT's Technology Review I'd have said it was some type of hoax. Seems Microsoft is involved also, could be related to HoloLens? Though from their website, no HMD is in view.

https://www.technologyreview.com/s/534971/magic-leap/

 

[Arulan]

Listening to the Giant Bombcast and they're gushing over this thing called the Magic Leap, saying that it's the real deal over all the other HMD's. What is it and what do you guys think about it vs the Rift and the Vive?

Magic Leap is Augmented Reality. It's on the rise alongside VR, but I believe they're still on the "let's prove this works" phase, although already to some success with Magic Leap and HoloLens demos, but will likely need further refinement and time for a consumer release. Additionally, due to AR's much more untethered nature, local computing power will likely be the norm. Then, as for applications, I don't think AR goes hand-in-hand with video games. There are certainly exceptions, but I believe AR will initially be focused on productivity and daily-use applications.

 

[GuitarAtomik]

Looks like some AR thing, https://www.magicleap.com/#/home
If there wasn't text about it on MIT's Technology Review I'd have said it was some type of hoax. Seems Microsoft is involved also, could be related to HoloLens? Though from their website, no HMD is in view.

https://www.technologyreview.com/s/534971/magic-leap/

It's real deal. They have a lot of investors but Google is the main name that gets thrown around because they've gone in heavy. They're being really secretive about the actual headset but they've shown the lenses and everyone who's gone to them and demoed it (with their prototype headset) came out a believer. Here's a thread on a recent Wired article about it: http://www.neogaf.com/forum/showthread.php?t=1209392&highlight=magic+leap

Recent video: https://www.youtube.com/watch?v=GmdXJy_IdNw

 

[plagiarize]

*snip*

Nothing to see here. It is just a mirage.

 

[Turin Turambar]

Was this posted? A prototype? of a tron disc-like game from CCP.

https://www.youtube.com/watch?v=riCYfnHGA5k


I think it's the first time of a Rift game using mixed-reality marketing videos, right?

 

[Onemic]

Magic Leap is Augmented Reality. It's on the rise alongside VR, but I believe they're still on the "let's prove this works" phase, although already to some success with Magic Leap and HoloLens demos, but will likely need further refinement and time for a consumer release. Additionally, due to AR's much more untethered nature, local computing power will likely be the norm. Then, as for applications, I don't think AR goes hand-in-hand with video games. There are certainly exceptions, but I believe AR will initially be focused on productivity and daily-use applications.

What makes this different from hololens?

And ya, an HMD that combines AR and VR would be godly

 

[GuitarAtomik]

What makes this different from hololens?

Same idea really. Though apparently Magic Leap can occlude the virtual objects if I read the Wired article correctly.

 

[Paganmoon]

It's real deal. They have a lot of investors but Google is the main name that gets thrown around because they've gone in heavy. They're being really secretive about the actual headset but they've shown the lenses and everyone who's gone to them and demoed it (with their prototype headset) came out a believer. Here's a thread on a recent Wired article about it: http://www.neogaf.com/forum/showthread.php?t=1209392&highlight=magic+leap

Recent video: https://www.youtube.com/watch?v=GmdXJy_IdNw

Yeah, after seeing the MIT article, I sorta figured it had to have some meat to it. But their own homepage is treadbare with info, and the blog really didn't help. But reading the MIT piece on it, it seems the founder is eccentric as fuck, so it explains the blog.

 

[cakefoo]

It's real deal. They have a lot of investors but Google is the main name that gets thrown around because they've gone in heavy. They're being really secretive about the actual headset but they've shown the lenses and everyone who's gone to them and demoed it (with their prototype headset) came out a believer. Here's a thread on a recent Wired article about it: http://www.neogaf.com/forum/showthread.php?t=1209392&highlight=magic+leap

Ooh, some glaring misinformation in there about VR headsets:

"The phonelike screens used in the majority of head-mounted displays created a nagging problem: They were placed right next to your eyeballs. If the device is generating the illusion of a blue whale 100 feet away, your eyes should be focused 100 feet away. But they’re not; they’re focused on the tiny screen an inch away."

 

[Zalusithix]

Ooh, some glaring misinformation in there about VR headsets:

"The phonelike screens used in the majority of head-mounted displays created a nagging problem: They were placed right next to your eyeballs. If the device is generating the illusion of a blue whale 100 feet away, your eyes should be focused 100 feet away. But they’re not; they’re focused on the tiny screen an inch away."

I'd like to see that person try to focus on something an inch away from their eye. (Without the use of optics.)

Of course focus depth is an issue with current VR, but more to the exact opposite case where everything is equally focused at a distance. Short of light field displays, that's not going to change though.

 

[Phatcorns]

I now have both a Vive and Rift. My thoughts:

The Vive has a much better field of view and the room scale VR is the future of this stuff 100%.

The Rift has a box-like field of view that is very narrow. You still get the nice VR experience, but that is even narrower than the Vive's already narrow field of view. The headset itself though is MUCH more comfortable and the built in headphones are a lifesaver, much more convenient than the Vive.

Anyway, just my two cents for those interested.

 

[Guardian Monkey]

Seems to be par for the course in what everyone is saying. Imagine with a smaller fov the rift probably has smaller screendoor/pixels?

 

[Later Alligator]

with the cv1 and vive out now (though in limited fashion), is it worth it to grab a cheap dk1 or dk2 off of craigslist to mess around with for those who have not had the chance to try VR out yet?

dk1's are going for around $150 while dk2s seem to be in the $350 range. those prices are in CAD

 

[Zalusithix]

with the cv1 and vive out now (though in limited fashion), is it worth it to grab a cheap dk1 or dk2 off of craigslist to mess around with for those who have not had the chance to try VR out yet?

dk1's are going for around $150 while dk2s seem to be in the $350 range. those prices are in CAD

Might as well just grab a new OSVR hacker dev kit for $300 IMO.

Edit: Noticed that I glanced over the whole CAD part. I guess that would make it $450 then. Still, the DK1 is really outdated tech, and shouldn't be a consideration IMO. That leaves it at $100 difference from a used DK2. If it weren't for the (current) built in support of the DK2 in the Oculus store, I'd say it'd be an easy decision.

 

[TheChewyWaffles]

Vive is coming on Monday. Woohoo!

 

[bj00rn_]

Ah well. Hopefully Rift and Vive CV2 ups the resolution of each eye. I wonder how long it will take to release CV2's of these headsets? I'm thinking something like 3-4 years.

That's way unrealistic. Sometime between a year or two in my personal opinion. I suspect HTC and Oculus will have to react to early foveated birds at some point.

 

[Tadie]

That's way unrealistic. Sometime between a year or two in my personal opinion. I suspect HTC and Oculus will have to react to early foveated birds at some point.

I think it depends on the available hardware power.

If nobody has the hardware to power the new headsets it would be ridiculous to launch a new version with higher pixel rate.

 

[Durante]

It's just another way to play games and can quite easily be a superset for all games, since it inherently supports standing and seated arrangements anyway. Never really understood the skepticism around it, which mostly seems to stem from "oh I don't have space for it, so it will never catch on" knee jerk reactions. The way I see it, roomscale is an all encompassing tracking technology that EVERYONE should implement. Whether it's standing or seated or moving is a genre and game specific implementation of roomscale. All other forms of tracking are destined for obsolescence.

Agreed.

Just wondering, for those in the know, how practical would it be to adapt the Vive's lighthouse tech to all sorts of wearable things that can track various body parts? I imagine it'd be feasible for limbs and torso, and maybe even the hands and feet, but not for individual fingers.

It's eminently practical, because lighthouse is inside-out it scales basically infinitely to more tracked points.

I think it depends on the available hardware power.

If nobody has the hardware to power the new headsets it would be ridiculous to launch a new version with higher pixel rate.

A higher resolution doesn't necessarily mean that you would need to render all demanding games at that resolution. It would still be very advantageous for virtual desktop and movie applications, physical pixel density reasons, and scalability to future GPUs.

 

[bj00rn_]

I think it depends on the available hardware power.

If nobody has the hardware to power the new headsets it would be ridiculous to launch a new version with higher pixel rate.

I'm slightly confused. Foveated rendering has the potential to mitigate both resolution and power challenges (to a certain degree), that's the point.

 

[Durante]

It also introduces other challenges though. in particular, even if you solve the tracking issue perfectly in hardware, you'd need to then render 200 FPS to make it work as far as I know. Probably almost every non-trivial VR game would be CPU bound on many systems when trying that.

 

[bj00rn_]

It also introduces other challenges though. in particular, even if you solve the tracking issue perfectly in hardware, you'd need to then render 200 FPS to make it work as far as I know. Probably almost every non-trivial VR game would be CPU bound on many systems when trying that.

Maybe. But in any case I'm pretty sure that in one to two years I'm sure we'll at least have seen some sort of announcement from htc or oculus. I just don't think it'll take three to four years for htc or oculus to come out with new technology if they really want to survive in the market. Of course, at some point I also thought the CVs would release in 2015, so I'm not denying there's wishful thinking involved. .

 

[viveks86]

It also introduces other challenges though. in particular, even if you solve the tracking issue perfectly in hardware, you'd need to then render 200 FPS to make it work as far as I know. Probably almost every non-trivial VR game would be CPU bound on many systems when trying that.

Is there a source for this? From what I can tell, the tracking and rendering don't have to be at the same rate. Our brain already compensates for saccadic movement. The tracking needs to be high frequency so it knows every position to approximate direction of gaze, without which the estimate can go completely off. The rendering should work just fine at 90 fps, where each frame uses an average (or some other form of approximation) of 2 or more samples. If our brains can handle 90 fps rendering without eye tracking, it should be able to do the same with eye tracking.

 

[Wollan]

If our brains can handle 90 fps rendering without eye tracking, it should be able to do the same with eye tracking.

Remember though that the eye always lands on 100% quality image with todays rendering method which could be a factor of why 'only 90fps' is doable in the first place. Source for saccadic eye-tracking 240hz or above: http://uploadvr.com/smi-eye-tracking-foveated-rendering-exclusive/

 

[Crispy75]

Remember though that the eye always lands on 100% quality image with todays rendering method which could be a factor of why 'only 90fps' is doable in the first place. Source for saccadic eye-tracking 240hz or above: http://uploadvr.com/smi-eye-tracking-foveated-rendering-exclusive/

It's the eye tracking itself that has to be done at 240Hz in order to accurately detect movement.

Very usefully, your vision is blank during rapid eye movement (https://en.wikipedia.org/wiki/Saccadic_masking), which means you wouldn't see any intervenening lowfi imagery as you flick your eye over the scene. However, when your eyes came to rest, you'd see the low quality frame; the retina detects the lack of motion blur and within 1ms resumes sending images to the brain. 11ms later, a new frame is ready with sharp imagery in that direction. You'd immediately notice this as a "pop" of lowfi imagery every time you flick your eyes from point to point.

But if you track the eye fast enough, you can detect the deceleration before the eye comes to rest and anticipate the final gaze direction >11ms before it gets there, giving you time to prepare the sharp imagery exactly where you need it before the eye comes to rest and starts seeing again.

EDIT: Here's some detailed explanation of the characteristics of saccadic motion. It's very well modelled so you can make good predictions with good enough sensor data. https://www.liverpool.ac.uk/~pcknox/teaching/Eymovs/params.htm

 

[Wollan]

That's a cool built-in human 'weakness' making eye-tracking VR more attainable.

I wonder if devs will shape the 100% quality area of the image based on velocity/trajection as in: If you're eye is moving at medium speed from top left to top right then the area will have an ellipsis form in that direction and not a simple circle.

Also: Carmack has been talking about tricks like garbage collection during eye-blinks. Maybe this can be done during saccadic masking as well?

 

[Crispy75]

HOWEVER: http://blogs.valvesoftware.com/abrash/down-the-vr-rabbit-hole-fixing-judder/

It’s a widespread belief that the eye is blind while saccading, and while the eye actually does gather a variety of information during saccades, it is true that normally no sharp images can be collected because the image of the real world smears across the retina, and that saccadic masking raises detection thresholds, keeping those smeared images from reaching our conscious awareness. However, low-persistence images can defeat saccadic masking, perhaps because saccadic masking fails when mid-saccadic images are as clear as pre- and post-saccadic images in the absence of retinal smear.

So you'd actually see the lowfi imagery strobing as you flicked your eye over the scene. So maybe what they need to do (if the hardware can do it) is turn off low-persistence during a saccade.