• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

The High-end VR Discussion Thread (HTC Vive, Oculus Rift, Playstation VR)

Mindlog

Member
I forgot what it was called but in an old thread there was a technique discussed where the game imperceptibly changes your direction slightly so you never run into a wall, even if you think you're going more or less straight
I know exactly what you are talking about, but I can't find the specific source that has come up before.

Here's a Wired article on redirected walking.
 

Zalusithix

Member
I forgot what it was called but in an old thread there was a technique discussed where the game imperceptibly changes your direction slightly so you never run into a wall, even if you think you're going more or less straight
With a large enough area and a long enough cord, that could work. Not possible with smaller play areas though where you're only capable of moving a few steps in a given direction.

Do you think a little lower than that would work ok?

The Vive doesn't use cameras in the first place. It blankets the area with sweeping infrared beams. The height recommendation is so the beam sweep can reach from ceiling to floor. The lighthouses emit in a 120 degree arc, so as long as the sweep passes the sensors, you'll be fine.
 
The Vive doesn't use cameras in the first place. It blankets the area with sweeping infrared beams. The height recommendation is so the beam sweep can reach from ceiling to floor. The lighthouses emit in a 120 degree arc, so as long as the sweep passes the sensors, you'll be fine.
Not ceiling to floor, top of headset to floor, which is why it's supposed to be higher than the tallest user (and tilted down). Admittedly, I don't have a Vive to play with...
 
Not ceiling to floor, top of headset to floor, which is why it's supposed to be higher than the tallest user (and tilted down). Admittedly, I don't have a Vive to play with...
Alright cool. Don't think my wife will let me mount them in my living room so above 6 ft may be the highest I can go
 

tr00per

Member
I know exactly what you are talking about, but I can't find the specific source that has come up before.

Here's a Wired article on redirected walking.

Yup that's the one.

With a large enough area and a long enough cord, that could work. Not possible with smaller play areas though where you're only capable of moving a few steps in a given direction.

Of course. I've said it before but part of the reason VR is so exciting is that devs and players have to rethink how to make and play games. It really is a renaissance like the jump to 3D in the 90s
 

Zalusithix

Member
Not ceiling to floor, top of headset to floor, which is why it's supposed to be higher than the tallest user (and tilted down).

Your controllers can go above your head by an arm's length. The placing of a lighthouse above a person's head along with the 120 degree cast ensures that it will cover that amount when the player is any real distance from them. Since we're talking two opposing lighthouses you'll always be a fair distance from at least one of them. It's just an easy rule of thumb for them to give that doesn't rely on inconsistent room and player height.

Edit: The tilt down is so it casts all the way down. Basically at the recommended angle it casts all the way from straight down to upwards from the mount point by about 30 degrees. Likewise it'll cover both adjacent walls when corner mounted at 45 degrees.

Edit 2:
Of course. I've said it before but part of the reason VR is so exciting is that devs and players have to rethink how to make and play games. It really is a renaissance like the jump to 3D in the 90s
Yep, pretty much everything we've learned about game design and control during the past couple decades gets thrown out the window with VR. And that's fine, because VR isn't really going to replace traditional gaming. Instead they'll compliment each other. As time goes on, the experiences VR provides are going to increasingly diverge from what we're used to. What we see now is just the tip of the ice berg and akin to when devs were first working out how to deal with the 3rd dimension.
 

Cartman86

Banned
Hover Junkers devs have talked about much the same thing here along with optimal starting direction in relation to the computer.

You are going to get tied up no matter what if you can turn 360 degrees. It seems to happen to everyone at least a couple times over multiple play sessions with a Vive. It's not a big deal. You shake your leg and it comes off, or in rare situations you will have to take the headset off and actually unravel it, but it is a thing that devs have to accept if they are making a game where the user can theoretically have that freedom. Designing the game to have most of the action facing only three walls may be be a way to limit it, but from my experience it can't fully eliminate it if your goal is gameplay with full room freedom. Which is why I find don't find the front facing camera and resting on IMU data for occluded controllers argument compelling. No you won't spin around in a circle three times at the start of a game. Slowly but surely over the course of an hour you will. Again though like people are saying it's not a big deal. It's really minor compared to what you get.
 
It doesn't matter if Sony knows their wavelength. Filters are cheap physical overlays you put over the sensor, as the Move wand color can vary over most of the visible light spectrum they can only use standard IR blocking filters to eliminate unimportant light. They can't dynamically change their filter to match the current color of the light. Whereas if you're using specific IR LED's you can be certain that they will be localized to a very small band in the EM spectrum and can employ cheap filters to block almost everything except that. The LED's are absolutely not the only source of IR light in a real world environment, but there is absolutely far less IR light to contend with than visible light in most real world scenarios.
They use software to eliminate the unimportant light. They just ignore anything that doesn't fit a specific RGB range. They've been doing that since the PS2. The "new" twist is making the Nerf-ball-on-a-Bic self-illuminate, so it's always the color they want, and largely immune to other lights shining on it, unlike a real Nerf ball.

Also dude it is obviously called Constellation because it's a known pattern of multiple LED's, you know, like a real constellation is a known pattern of stars.
And just as with real constellations, they may be hiding among the stars. Similarly, the Move wand may not be the only glowing, purple sphere the camera can see, but if that turns out to be the case, they can just switch to blue instead. So while Oculus may be less likely to encounter interference in the first place, their fixed-wavelength system doesn't really have any way to compensate when they do.

1. You could absolutely use an IR LED to illuminate a bulb it would just look far less pretty.

2. I can't help but feel you're working backwards from the conclusion that Move tracking is better.
Actually, I was just speculating as to why they used such small markers, given the fact that they're harder to track with a camera. Turns out, while you can make IR bulbs, it's not terribly efficient. These guys use seven IR LEDs to make a marker ~27mm across, and that only doubles the effective tracking range versus a single LED.
http://www.ar-tracking.com/technology/markers/

So I was basically right; it doesn't really work very well. And you're right, it's not particularly attractive either. :p

3. That analogy is hilariously skewed. Here is an actual image of what the DK2 camera sees using crystal cove in a quote "very bright office setting".

Big Dipper on a starry night it is not.
Sure, and as discussed, Move's magnetometer worked similarly well in Sony's simulated real-world environments too. The real world doesn't play as nicely though. You already admitted this, so I'm not sure why you're trying to imply otherwise here.

The real point is though, even at that fairly close range pictured, the individual markers don't cover very many pixels, so they shrink to nothingness comparatively quickly. Bigger targets are better, and making bigger targets is easier — and prettier, yes — when you're using visible light, because you can diffuse it more effectively.
 

Monger

Member
And you'll probably be more tolerant to drift without being able to see where your arms are than you would be normally.

But obviously it's speculation in both directions. In a game like job simulator I know you are often grabbing stuff from behind you, but do you spend extended time looking away from your workstation? Genuine question.

No, because if you have a front facing setup the developers place everything in front of you and not too low where tracking isn't an issue. They also change the space depending on how much room you have. It's pretty cool really. They have a good interview about it.

Otherwise yeah, they have you do stuff behind you.
 
Bad news for some of us.

http://www.computerbase.de/2016-03/...ks/2/#diagramm-minigolf-vr-geforce-gtx-980-ti

If you don't have at least a 970, might not have a lot of fun...and sometimes not even then. I'm assuming the Vive and OR have similar performance burdens (why wouldn't they?), where for games like Elite Dangerous a 970 really is the minimum. In several games the 970 appears to be barely qualifying. Radeon users may want to look away entirely from the E:D results.
wtzgU.gif


Good news: Hover Junkers is good to go all the way down to a 770/280X.
 

Bsigg12

Member
Bad news for some of us.

http://www.computerbase.de/2016-03/...ks/2/#diagramm-minigolf-vr-geforce-gtx-980-ti

If you don't have at least a 970, might not have a lot of fun...and sometimes not even then. I'm assuming the Vive and OR have similar performance burdens (why wouldn't they?), where for games like Elite Dangerous a 970 really is the minimum. In several games the 970 appears to be barely qualifying. Radeon users may want to look away entirely from the E:D results.
wtzgU.gif


Good news: Hover Junkers is good to go all the way down to a 770/280X.

I'm really curious to see if the Rift support for Elite Dangerous means they'll have optimized for a 970 since Oculus is saying anything offered in the Oculus Store will run on the minimum requirement for the Rift.
 
I'm really curious to see if the Rift support for Elite Dangerous means they'll have optimized for a 970 since Oculus is saying anything offered in the Oculus Store will run on the minimum requirement for the Rift.

lol, it won't be a different game nor a different build of the game. It'll just have proper support for Oculus SDK. Perhaps Oculus is making an exception here since it's the only AAA title for VR right now. Whether it can be "played" below a 970 is immaterial; the shit's going to suck if you don't have *at least* a 970.
 

cakefoo

Member
https://en.wikipedia.org/wiki/PlayStation_Move#Technology



As I said, I don't know how good the dead reckoning is. I don't know how quickly they get out of sync, and how soon that becomes noticeable.



I understand all of this. But I stand by the notion that it's speculation right now that PSVR can't do something approaching room scale. Can it do it as well? Will there potentially be issues that don't exist with the other methods?

We can all speculate... but we don't even know what the final thing looks like yet. Sony could debut new motion controllers tomorrow for all we know. It strikes me as a disingenuous reason to try and rule out PSVR as a high end VR setup, and I stand by that.

Maybe you'll have to portal turn in Budget Cuts on PSVR to pick up something that fell behind you. Does that really invalidate PSVR? That instead of solely teleporting to move, that you also potentially do a teleport 180?

TLDR: We know touch will support room scale, even if it doesn't do it as well. We don't know what PSVR will support yet. We presume it has a single camera setup with occlusion issues... but lets wait until we see what's in the box before we throw the baby out with the bathwater.
The PS Move quickly loses positional accuracy as soon as it's covered. Nothing involving fine hand interactions will be possible behind the back.

chw5QT.gif


The only reason The Deep demo was able to track facing away from the camera was because you control aim by rotating and tilting the controller only-- no positional tracking involved.
 
The PS Move quickly loses positional accuracy as soon as it's covered. Nothing involving fine hand interactions will be possible behind the back.

chw5QT.gif


The only reason The Deep demo was able to track facing away from the camera was because you control aim by rotating and tilting the controller only-- no positional tracking involved.

Maybe that glove patent will become a real product and could somehow help with that.
 

Mindlog

Member
So while Oculus may be less likely to encounter interference in the first place, their fixed-wavelength system doesn't really have any way to compensate when they do.
Is that really a problem?
Later, it was discovered by Github user pH5 that each LED broadcasts at specific flashing frequencies with various levels of brightness. This allows the tracking camera to identify the source based on the exact blinking patterns of the lights.
The real point is though, even at that fairly close range pictured, the individual markers don't cover very many pixels, so they shrink to nothingness comparatively quickly.
Which would present an enormous problem when attempting to do room scale tracking.
 

TTP

Have a fun! Enjoy!
Right, but the same characteristics that make the bulb easy to track also make it impossible to glean orientation information from it; it looks the same no matter which way it's pointing. Hence the magnetometer and trying to keep track of "down" as external reference points.

Yes, but I know for a fact that the magnetometer wasn't used in any Move game. You could easily tell if a game was using it by moving a magnet around the wand and see if that affected tracking. The only game that used it at some point was Time Crisis Razing Storm but if was quickly disabled via patch.

I talked about it in my video analysis of Time Crisis: https://youtu.be/HpnYkXrP8So?t=9m37s

Moreover, you can disable the magnetometer in the XMB, right after the calibration process. If it was so important they wouldn't let you disable it no? ;)

If you are interested in more, here is a playsist of videos I did that cover everything related to Move tracking/calibration: https://www.youtube.com/playlist?list=PLFE85FD342D003715 (you'll love "Move vs Light" ;))
 

mrklaw

MrArseFace
You're going to have to pay an additional $100-200 for Touch, likely falling in the latter half of that range. I'm not sure what's considerable about the difference. Do you live somewhere outside the US where the prices are further apart?

The time difference between needing to pay all of it now, vs some now and some in 6 months could be important to many people
 

Durante

Member
The Vive requires you to screw mounts into your walls for the sensors? Wtf?
It doesn't require you to, but it does conveniently ship with wall/ceiling mounts.

Bad news for some of us.

http://www.computerbase.de/2016-03/...ks/2/#diagramm-minigolf-vr-geforce-gtx-980-ti

If you don't have at least a 970, might not have a lot of fun...
After both Oculus and Valve/HTC set their minimum specs at a 970, I do have to wonder how this could be "bad news" to anyone :p
 

mrklaw

MrArseFace
And you'll probably be more tolerant to drift without being able to see where your arms are than you would be normally.

But obviously it's speculation in both directions. In a game like job simulator I know you are often grabbing stuff from behind you, but do you spend extended time looking away from your workstation? Genuine question.

I don't know, but the devs said they designed the workspaces to avoid full 360 degree usage, so you are encouraged to turn around to grab something and then turn back to the main work surface. So you'll have areas to the extreme left and right, but the tasks are designed to avoid you continuing to turn in one direction.
 

alexbull_uk

Member
If I don't have enough space for room-scale VR, does the Vive offer anything else that would make it a better buy over the Rift? I'm torn.
 

mrklaw

MrArseFace
If I don't have enough space for room-scale VR, does the Vive offer anything else that would make it a better buy over the Rift? I'm torn.

All of the motion control games should work fine in smaller spaces. Fantastic contraption devs showed an example of the game when sitting down - looks like you're building a toy vehicle.
 

Durante

Member
If I don't have enough space for room-scale VR, does the Vive offer anything else that would make it a better buy over the Rift? I'm torn.
It ships with tracked controllers right away, which you can also use standing. And I guess if you really want to be nitpicking, it also requires a lot fewer USB ports than a full RIft+touch setup.

It has a front-facing camera, so (I think) you can "press a button" to see where your controller/mouse/keyboard/cup of tea/biscuit/dog are.
Right, I forgot about that.
 

dumbo

Member
If I don't have enough space for room-scale VR, does the Vive offer anything else that would make it a better buy over the Rift? I'm torn.

It has a front-facing camera, so (I think) you can "press a button" to see where your controller/mouse/keyboard/cup of tea/biscuit/dog are. And obviously the Vive price includes motion controllers.
 
After both Oculus and Valve/HTC set their minimum specs at a 970, I do have to wonder how this could be "bad news" to anyone :p

I believe many were hoping against hope, especially as reports of several games requiring quite a bit less than a 970. This adds a bit more clarity and perspective to the conversation.
sad.gif
 

artsi

Member

They seem to play the game sitting down so they need to use the stick turning a lot. I wonder if it's any more comfortable as a standing experience, where you can turn using your body.

Of course there's the issue of forward / backward locomotion left.
 

Man

Member
They're not the only one.
I have seen murmurings on Twitter about one outlet (in the works of) writing a larger editorial about how Oculus has dropped all their principles in regards to Minecraft VR. The game simply does not adhere to good VR design in any fashion as-is and it's set out to do way more damage than good. Oculus however is letting it pass because it's too big a IP get.
Ironically: Oculus (in cooperation with Microsoft) will be 'poisoning the well' in grand fashion here if Minecraft VR impressions are to be believed.

I think they should remove the 'full-immersion' mode completely and only have the 'virtual cinema' mode only.
 
They seem to play the game sitting down so they need to use the stick turning a lot. I wonder if it's any more comfortable as a standing experience, where you can turn using your body.

Of course there's the issue of forward / backward locomotion left.

I think mixing room scale with game where you still are going to have to move around a significant amount is going to be really tricky, and probably more confusing than combing head rotation with stick rotation. Teleportation seems to work, but that's not going to be a good fit for everything.
 

mrklaw

MrArseFace
Minecraft sounds bad, but it presents challenges that would be there for most open world exploration games, so I hope solutions are found.

Is it possible to amplify head turning so you can turn in the game more than you turn in real life? That might allow for 360 turning while seated.

Even with 'comfort' turning I would want to use the head for steering most of the time, and avoid strafing which is unfortunately very common in minecraft. I think it would also be useful to have some in-game mechanism to help you orient yourself facing forward or avoid you getting tangled. Eg it counts your rotations and if you turn more than 360 clockwise, it somehow indicates that you should turn anti-clockwise
 

mrklaw

MrArseFace
I just had an idea for seated experiences to help with being limited in turning due to the cable getting tangled. Could someone build a swivel chair that has the VR connection built into the base, so the signal comes up through the centre? Then you connect your headset to the backrest and can freely spin as much as you want without an tangling.
 

Zalusithix

Member
I just had an idea for seated experiences to help with being limited in turning due to the cable getting tangled. Could someone build a swivel chair that has the VR connection built into the base, so the signal comes up through the centre? Then you connect your headset to the backrest and can freely spin as much as you want without an tangling.

The cable would still twist. It wouldn't tangle, but it would twist until it creates uncomfortable torque on the headset or outright breaks. To avoid that you'd need a custom electrical interface between the top and bottom part of the base that can handle the bandwidth requirements of the video and data. You'd be better off slapping the PC on the "chair" as well. Far easier to handle only the three conductors of the mains power cable than the data lines of HDMI + USB + power. Even then it wouldn't help at all for room scale / standing experiences which is where the real problem is.
 

artsi

Member
Is it possible to amplify head turning so you can turn in the game more than you turn in real life? That might allow for 360 turning while seated.

That's what TrackIR has always done, but I don't how well it would suit VR where the idea is to have 1:1 head tracking.
 
Eventually they are going to need to find a way to make games like Minecraft play in both room scale format and sit down. It's not going to be one or the other. A lot of people who want to play Minecraft VR won't have room scale options so there is going to need to be away to handle it with a controller without getting people sick. For those that can do room scale, that option will eventually need to be there on the Oculus platform or it needs to be playable in Vive.
 
Yeah Minecraft VR really needs a better movement solution. A lot of people are going to want to try it when they hear about it and that would make the worst first impression if it gives them nausea. I get that some games are going to push boundaries which is why they're going to have a comfort rating, but a game with Minecrafts reach is something you want to get right. I'd rather they implement a teleport mechanic by default than what they have now, but leave this as an option with a warning.
 

KingSnake

The Birthday Skeleton
For the seated experience, they could just flood the world of Minecraft and make the movement by swimming. Wouldn't that trick the brain somehow better?
halfjoking
 

dumbo

Member
Yeah Minecraft VR really needs a better movement solution. A lot of people are going to want to try it when they hear about it and that would make the worst first impression if it gives them nausea. I get that some games are going to push boundaries which is why they're going to have a comfort rating, but a game with Minecrafts reach is something you want to get right. I'd rather they implement a teleport mechanic by default than what they have now, but leave this as an option with a warning.

Teleport wouldn't really work in minecraft. In survival mode everything would be trivial/broken (mining a corridor would be 'unfun') and in "build mode" there are better ways to use VR. It's actually a pretty hard game to make work "well" in VR, even if it seems like it should.

Like a lot of FPS stuff, you can't help thinking that a mech with a cockpit is a better idea.
 
Teleport wouldn't really work in minecraft. In survival mode everything would be trivial/broken (mining a corridor would be 'unfun')

I don't see how, just have it so that when you press a bumper you jump to the block you're looking at. So in a corridor you clear away block in front of you, hit bumper, jump forward, clear blocks, hit bumper, jump forward. All you're doing is replacing the time it takes to walk with a quick cut, I don't see how that dramatically breaks anything or makes it significantly less fun.
 
The Job Simulator developers talked about how it's important to interact with things behind you so it feels like a real space, while also using other game elements and design to keep you turning back to the front without doing a 360

Same way as developers design around preventing you from walking outside your space, while at the same time not making you feel boxed in. Clever use of geometry and level design without letting you realize it's happening

So the cord might be a dangle risk, but it's something designers are working around while still letting people have full rotation. And as far as I've heard it's pretty successful. The question gets brought up in just about every Vive Q&A, and there has been pretty universal agreement that the cord rarely causes an issue

I assume the same will happen with sitting VR and keeping occlusion in mind. Something we all worry about now, but game designers will find clever ways to minimize

I'm sure that even with just a few hours of using the Vive that you'll become good at subconsciously avoiding getting tangled up and develop a sort of 'cable sense' too. It might add a degree of artificiality over being able to just endlessly spin, but it's going to be a very very small issue in the real world I feel. Yeah, a few times early on you'll look down thinking you'll be able to see where the cable is when you're getting a bit tangled up, but so long as you start out with your back to the computer and so long as you aren't Zoolander, you're going to do just fine.
 
Top Bottom