Jump to content
Jason_Williams

New Oculus Rift unit at CES

Recommended Posts

 

Ok, so Sir Stephen Hawking might not be the ideal candidate for this, but it's hardly a great hardship for the rest of the community...

 

 

You're right he would definitely be cheating, what with his motorized chair and all.  A few tweaks here and there and I reckon he could do a full 360 in the blink of an eye.  He would just have to ensure he had the wireless version  or there might be a tragic accident if he kept turning in the same direction once to often. 

Share this post


Link to post
Share on other sites

The pattern is identical left to right. So it only knows left to right. Those patterns also go bye bye after the head is rotated beyond a few degrees. IR emitters/reflectors work in one direction, and not even on much of a degree scale unless they are spherical instead of flat (which flat they are for OR).

 

So yes, I am correct. You are still limited to left/right movement, and not full on turn around unless the gyro can accomodate that.

 

No you are still incorrect because the unit will obviously be able to tell you have turned your head one way or the other and therefore will know which side of the unit it is looking at, in case you hadn't noticed the IR emitters cover the entire surface of the crystal cove housing  leaving only the very rear part of the unit devoid of any emitters, as I mentioned before even if you turn your head to full stretch and dare I say really pushed the boat out and slightly shifted in your seat the side of the unit will still be very visible to the sensor and the side facing emitters will be effective, after all the emitters on the oculus won't be making a focussed beam and with the emitters placed on the upper/lower and side of the oculus the emission will be wider than your field of view.

 

I meant to mention that the basic tracking the DK oculus already has is not limited in any way and will sense rotation of the device in all 3 axes indefinitely.

Edited by DD_bongodriver

Share this post


Link to post
Share on other sites

OR doesn't lose positional tracking even if you turn your head 180* because it incorporates internal tracking aswell as the external camera tracking.   The OR is currently partly using the external camera as a fixed position reference point as internal positional tracking has a tendency to drift after time.   The external camera is the best possible solution to the problem they have found so far, but they are still working to find a better solution.

 

One of the best features, and the one that most captivates users is that the OR shuts out most other sensory inputs, and draws a person completely into the game world.  The immersion levels are off the charts.  Personally I don't what to physically, or need to see my controls, as it would ruin the immersion.   A Hotas system will accommodate most required inputs, and voice recognition software  would take care of the rest, while making the experience even more immersive when used for tower, and wingman communication.  Its quite easy to learn all your Hotas buttons, and rotaries without looking, and real WW2 pilots where trained to find everything blindfolded.

 

I doubt if the OR developers will allow anything other than 1:1 positional tracking, to avoid as much motion sickness as possible.  Lag caused most of the motion sickness in the first prototype.   Some of the users of the first prototype did some software tweaks to over speed the movement, and it caused far less motion sickness, which "might" suggest that less head movement may be possible.  Personally I think 1:1 is just fine, but noticed even when driving a car I can't turn my head to backup as far as I could when I was younger.   I think the twenty somethings with excellent flexibility and eyesight should be declared cheaters in online game play, if I'm not allowed to game the game to regain the abilities I once had forty years ago.  ;)

Share this post


Link to post
Share on other sites

You're right he would definitely be cheating, what with his motorized chair and all.  A few tweaks here and there and I reckon he could do a full 360 in the blink of an eye.  He would just have to ensure he had the wireless version  or there might be a tragic accident if he kept turning in the same direction once to often. 

 

Having read this thread and the reluctance of some gamers to move their heads Stephen Hawking must be regarded as some kind of athlete with the ultimate gaming hardware.

  • Upvote 2

Share this post


Link to post
Share on other sites

OR doesn't lose positional tracking even if you turn your head 180* because it incorporates internal tracking aswell as the external camera tracking.   The OR is currently partly using the external camera as a fixed position reference point as internal positional tracking has a tendency to drift after time.   The external camera is the best possible solution to the problem they have found so far, but they are still working to find a better solution.

 

One of the best features, and the one that most captivates users is that the OR shuts out most other sensory inputs, and draws a person completely into the game world.  The immersion levels are off the charts.  Personally I don't what to physically, or need to see my controls, as it would ruin the immersion.   A Hotas system will accommodate most required inputs, and voice recognition software  would take care of the rest, while making the experience even more immersive when used for tower, and wingman communication.  Its quite easy to learn all your Hotas buttons, and rotaries without looking, and real WW2 pilots where trained to find everything blindfolded.

 

I doubt if the OR developers will allow anything other than 1:1 positional tracking, to avoid as much motion sickness as possible.  Lag caused most of the motion sickness in the first prototype.   Some of the users of the first prototype did some software tweaks to over speed the movement, and it caused far less motion sickness, which "might" suggest that less head movement may be possible.  Personally I think 1:1 is just fine, but noticed even when driving a car I can't turn my head to backup as far as I could when I was younger.   I think the twenty somethings with excellent flexibility and eyesight should be declared cheaters in online game play, if I'm not allowed to game the game to regain the abilities I once had forty years ago.  ;)

 

 

And even then, they could use a stroboscopic blink of the led, like they use in cinema prodution. It allows to identify each led, so you just need a couple of contact to know the head orientation. Add to this internal sensors data to double proof and you just got rid of the problem. It is not the case at the moment but the question has been raised and might be adressed

Share this post


Link to post
Share on other sites

Just a stupid thought I had ...

Could they perhaps put IR emitters on the strap so the camera can see when your head is turned 180 degrees.

I think that is already the case.

Share this post


Link to post
Share on other sites

Does the occulus cover your eyes?  How are you able to see your keyboard or a printed out map if your eyes are inside these glasses?

 

You know I think you are right, the oculus really would be better if they just took the screen out the front, then we can just put the TIR track clip on the empty frame.

Share this post


Link to post
Share on other sites

Just a stupid thought I had ...

 

Could they perhaps put IR emitters on the strap so the camera can see when your head is turned 180 degrees.

As I said in a just a couple of posts up the OR already does recognize a 180 degree head turn, using its internal positional tracking.   So IR emitters aren't required on the head band.   Among other things the external tracking camera provides a fixed reference point for the internal tracking.

Share this post


Link to post
Share on other sites

No you are still incorrect because the unit will obviously be able to tell you have turned your head one way or the other and therefore will know which side of the unit it is looking at, in case you hadn't noticed the IR emitters cover the entire surface of the crystal cove housing  leaving only the very rear part of the unit devoid of any emitters, as I mentioned before even if you turn your head to full stretch and dare I say really pushed the boat out and slightly shifted in your seat the side of the unit will still be very visible to the sensor and the side facing emitters will be effective, after all the emitters on the oculus won't be making a focussed beam and with the emitters placed on the upper/lower and side of the oculus the emission will be wider than your field of view.

 

I meant to mention that the basic tracking the DK oculus already has is not limited in any way and will sense rotation of the device in all 3 axes indefinitely.

 

No, I'm not incorrect. The device still detects left/right movement, which I said. The 270, at least, view is done by the gyros - not the IR. So if you do 1:1, and you look behind you - no IR means no zoom/strafe. That was what I said the entire time.

Share this post


Link to post
Share on other sites

No, I'm not incorrect. The device still detects left/right movement, which I said. The 270, at least, view is done by the gyros - not the IR. So if you do 1:1, and you look behind you - no IR means no zoom/strafe. That was what I said the entire time.

270? Do you mean 270 and 90 degrees as in 3 o'clock/9 o'clock?

 

i said that even turning your head for a view behind in the same way you would in a real aircraft still presents the side view of the headset, which means all those side mounted emitters are still visible to the sensor, therefore you still have iR and can still do all that zooming and strafing which is so important to do with your head craned round at full stretch.

 

Yes, if you were in a swivel chair and rotated yourself so you were facing completely away from the sensor then you would lose IR, but what would be the point of being able to rotate yourself 180 anyway? Particularly in a simulated cockpit.

Edited by DD_bongodriver

Share this post


Link to post
Share on other sites

No, I meant 270 degrees as in the degrees of rotation - not as in a compass. I'm sure the gyros can do 360, but the IR won't be able to, it'll be at least limited to 270 but I'm sure significantly less.

 

Zooming and strafing is very important, and it's not very accurate at all with IR at oblique angles. It doesn't matter if it's TrackIr (it has curved IR reflectors so around 80 and 280 should work flawlessly if the OR tech of flat IR emitters should but it doesn't ), OR w/IR, FaceNoIr or the other IR webcam device. I use zoom and strafe frequently in side views and rear views, and even at my current settings which requires I move my head about 35 degrees left or right to attain full rear view - it's still very inaccurate and half the time it doesn't work properly. It'll be the same with OR since it's using the exact same technology for those particular features.

 

So if OR is 1:1, and certain parts become non-functional in circumstances such as looking to your rear left/rear right then it's a broken piece of technology. I would hope it's scaling is adjustable, or they come up with some way to work in 360 degree IR.

Edited by FuriousMeow

Share this post


Link to post
Share on other sites

 

So if OR is 1:1, and certain parts become non-functional in circumstances such as looking to your rear left/rear right then it's a broken piece of technology. I would hope it's scaling is adjustable, or they come up with some way to work in 360 degree IR.

 

I think you're incorrect, or you've just mis-interpreted how the OR works.

 

The camera, pure and simple, checks for ACCUMULATED drift, that is all.  It's the equivalent of the recentre key when using TrackIR except it's automatic and done every time the IR dots are in view.  If you swivelled around in your chair 180 degrees and mucked around for a couple of minutes then your view might be out of sync with your head position because the camera can't see the IR dots, but as soon as the camera caught sight of the dots, front or side, it would check for accumulated drift errors and reposition your view accordingly, that is all.  Given that it is unlikely that, when sitting down, you will not look towards the camera at least once every few second then there will be no error at all.  The problem, using the camera, only arises if the player is standing up and is more dynamic in his or her movements and is not necessarily always facing the same direction, in an Omniswivel thingy for example.  The camera is an imperfect solution for that reason alone and why it is not the preferred option from the developers standpoint, but for any game that has the player in a defined position such as a cockpit it is a non issue.

 

I would strongly suspect that anything other than 1:1  would quickly lead to upchucking on the part of the user when playing a flight sim, sure TrackIR can scale movement but you are also grounded in the real world, not so with the all encompassing Rift.

Edited by HagarTheHorrible

Share this post


Link to post
Share on other sites

You know I think you are right, the oculus really would be better if they just took the screen out the front, then we can just put the TIR track clip on the empty frame.

I think we could take that one step further and just do away with the OR frame and the TIR altogether. That would give you an unimpeded view of your surroundings and - let's face it - you don't get much better 3D than actual reality.

Share this post


Link to post
Share on other sites

I think we could take that one step further and just do away with the OR frame and the TIR altogether. That would give you an unimpeded view of your surroundings and - let's face it - you don't get much better 3D than actual reality.

 

Well raaaid might have a different opinion.

Share this post


Link to post
Share on other sites

I think you're incorrect, or you've just mis-interpreted how the OR works.

 

The camera, pure and simple, checks for ACCUMULATED drift, that is all.  It's the equivalent of the recentre key when using TrackIR except it's automatic and done every time the IR dots are in view.  If you swivelled around in your chair 180 degrees and mucked around for a couple of minutes then your view might be out of sync with your head position because the camera can't see the IR dots, but as soon as the camera caught sight of the dots, front or side, it would check for accumulated drift errors and reposition your view accordingly, that is all.  Given that it is unlikely that, when sitting down, you will not look towards the camera at least once every few second then there will be no error at all.  The problem, using the camera, only arises if the player is standing up and is more dynamic in his or her movements and is not necessarily always facing the same direction, in an Omniswivel thingy for example.  The camera is an imperfect solution for that reason alone and why it is not the preferred option from the developers standpoint, but for any game that has the player in a defined position such as a cockpit it is a non issue.

 

I would strongly suspect that anything other than 1:1  would quickly lead to upchucking on the part of the user when playing a flight sim, sure TrackIR can scale movement but you are also grounded in the real world, not so with the all encompassing Rift.

 

Folks, folks. how often does a fighter pilot look directly behind him, I mean 180 degrees? I mean with his head turned exactly BACKWARD on his shoulders? Never. He turns maybe 120 degrees max, then rudders the tail back and forth to check six. Vast majority of fighter planes have armor behind the head so it limits the view to way below-180 degrees anyway.   I'm guessing the LEDs on the sides of the unit will be more than ample for all six-checking in flight sims.

Share this post


Link to post
Share on other sites

Presumably you didn't read my post, not that I'm complaining or bothered, it's not important. If however you had, you would have realised that the point I was making was that the Rift doesn't work like TrackIR, the camera is only an external datum point. If you like it is a foot on the ground when lying on a bed having drunk more than is good for you, it is your reference point in an otherwise spinning world.

  • Upvote 1

Share this post


Link to post
Share on other sites

The camera is the positional tracker, positional tracking cannot be done with the magnetic sensors the rift uses, you can move 20 feet to the side and a magnetic sensor won't know, they can only sense rotational movement, the camera almost certainly will be used to provide a fixed reference to eliminate the drift issues as well.

Share this post


Link to post
Share on other sites

I mean 180 degrees? I mean with his head turned exactly BACKWARD on his shoulders?

Just to say.... but IRL no need to have the head turned exactly backward to look at 180° from frontview. Side View is a standard eye cepability.

 

That's maybe why rotation scaling isn't maybe not totally stupid with OR, since side view is disabled (it only allowe you a frontal view).

Share this post


Link to post
Share on other sites

the field of view of the DK oculus is good, the only part of peripheral vision you don't have is the part in real life that you would lose if wearing a pair of pilots goggles, the consumer release version is likely to expand on that field of view, the simulations I have used the oculus in myself have allowed me to get a full view of the 6 oclock.

Share this post


Link to post
Share on other sites

 I have used the oculus in myself have allowed me to get a full view of the 6 oclock.

Ok, I was wrong then

Share this post


Link to post
Share on other sites

There isn't only magnetometers , but gyro's and other internal devices used for positional tracking in the OR.   I don't know how well they do 6DOF, but  I do know that there is a slow drift problem that the camera corrects.  I also know they combine all the data from the external IR camera with the internal gyro's etc for the best possible positional tracking "so far". 

 

 They've also said the camera will not necessarily be in the consumer version as they're still doing plenty of R&D to keep the best possible positional tracking internal.  The camera and internal tracking was just the best solution ready to show 6DOF at CES.   It will be interesting to see how the next prototype looks at  E3 etc.   I would imagine there are number of changes and additions to the prototypes in the next six months up to E3.  Hopefully they get lucky and will have the hardware/form factor nailed  down enough to start production of the first consumer version before the end of the year, with many upgraded consumer versions to follow.

 

I would imaging the BOS development team will be able to get their hands on the last prototype before release, as the OR team say they will provide these units to large and small development teams that are adding OR support to their games.   The OR needs as many games as possible ready for OR use when the consumer version is released for best possible results.

Edited by JG27_Chivas

Share this post


Link to post
Share on other sites

Presumably you didn't read my post, not that I'm complaining or bothered, it's not important. If however you had, you would have realised that the point I was making was that the Rift doesn't work like TrackIR, the camera is only an external datum point. If you like it is a foot on the ground when lying on a bed having drunk more than is good for you, it is your reference point in an otherwise spinning world.

I didn't mean to  target only your post but the general discussion of fear that the headtracking in the OR will not be sufficient to check six.  I just happened to quote your post.

 

I'm pretty sure that the OR as demoed at CES works on the same principles as trackIR, with a larger angle of rotation. Infrared LEDs on your head provide data to the camera -> computer which computes where your head is at. Trackir 5 has a max rotational angle of 51.7 degrees.  Now if the rift has led's on it's sides, then those kick in when your head is turned sideways and back, thus giving much greater rotational range than trackIR.   Oh, and the rift will be 1:1 translation: TrackIR: no.

Share this post


Link to post
Share on other sites

I didn't mean to  target only your post but the general discussion of fear that the headtracking in the OR will not be sufficient to check six.  I just happened to quote your post.

 

I'm pretty sure that the OR as demoed at CES works on the same principles as trackIR, with a larger angle of rotation. Infrared LEDs on your head provide data to the camera -> computer which computes where your head is at. Trackir 5 has a max rotational angle of 51.7 degrees.  Now if the rift has led's on it's sides, then those kick in when your head is turned sideways and back, thus giving much greater rotational range than trackIR.   Oh, and the rift will be 1:1 translation: TrackIR: no.

 

No offence, but I give up. The Rift camera does NOT work the same as TrackIR.

Share this post


Link to post
Share on other sites

I think it will be a nice gimmick for when you play short amounts at a time (30min or so), anything more than that is just going to be too tiresome.

 

I have been using TIR for a long time now and still on a regular basis have a stiff neck when using it for an hour or so. It was the worst when I started using it because I was not used to keeping my head still when aiming which resulted in me kind of cramping my neck muscles, now it's much better but still I cannot fly with it more than say 1 - 1.5h without my neck hurting.

 

I imagine this is going to happen even more so with the Occulus because it also adds more weight, I'd be wearing that thing plus my headphones. And if there is no way to adjust response as with TIR and you have to move your head in a 1:1 ratio thats going to be quite tiring...

  • Upvote 1

Share this post


Link to post
Share on other sites

The video I saw had Palmer explaining that the camera element provides the 'Tracking' needed to give the addition 3 degrees of rotation, so with the existing tech and the camera there is now 6dof.

So whilst the camera wouldn't be responsible for providing tracking info for the check 6 axis, it is responsible for providing the tracking info for lean, zoom and raise/lower axes (a bit like TrackIR)

Yes, but with  the additional point that if you've got your head turned to check six and want to shift your upper body a bit to see around a window strut or something then the infrared LED positional system kicks in, getting data from the LEDs on the side of the rift..

 

cheers

Share this post


Link to post
Share on other sites

Been there done it. I can also recommend not laying in a tiled bathroom as the geometric pattern of the tiles does weird things to your perception, i.e. appear to move in and out. Not good when the room is already spinning... :biggrin:

 

Presumably you didn't read my post, not that I'm complaining or bothered, it's not important. If however you had, you would have realised that the point I was making was that the Rift doesn't work like TrackIR, the camera is only an external datum point. If you like it is a foot on the ground when lying on a bed having drunk more than is good for you, it is your reference point in an otherwise spinning world.


Regarding muscle strains using the Rift unit, no one will know for sure until we get the consumer version to play with, but I do predict that there will be cases of 'oculus neck'  :biggrin:

Share this post


Link to post
Share on other sites

Will a virtual body be more important in VR ?

 

At the moment we can happily play without a virtual self in the cockpit, however reading a couple of articles on developing games for VR, the authors talk about having to change the way things are done, particularly the details. It will be interesting to see if a virtual body is an essential requirement to maintain immersion in a virtual world.

Share this post


Link to post
Share on other sites

Will a virtual body be more important in VR ?

 

At the moment we can happily play without a virtual self in the cockpit, however reading a couple of articles on developing games for VR, the authors talk about having to change the way things are done, particularly the details. It will be interesting to see if a virtual body is an essential requirement to maintain immersion in a virtual world.

 

I read somewhere (maybe one of the guys who tried it out with Jason in California) that with the Oculus Rift the "eye" position we're all used to, where you can see the instruments easily, is actually very far back in the cockpit. So this would have to be addressed, I suppose. Perhaps a virtual body would help fill up the empty space. I wonder how close a pilot's face was to the recticle? Would it feel right "in your face"?

 

This VR thing is more compex than you'd think, isn't it! (But even with all the possible pitfalls, I think the sense of 360 degree wraparound would make it a no brainer for pretty much any simmer.)

Share this post


Link to post
Share on other sites

Yes, the human body with it's articulated spine, muscles and connective tissues simply was not designed to move about.

Grins, this is better than the removing the glass solution.

  • Upvote 1

Share this post


Link to post
Share on other sites

Definitely the future for sure..I also trying the prototypes. 

Share this post


Link to post
Share on other sites

Ho! and yes VR gloves are the solution.

Share this post


Link to post
Share on other sites

Its very interesting that a review of the Infiniteye said its using using (off the shelf?) hardware to produce internal 6DOF (no external camera) with no noticeable latencies.  

 

"The team are using YEI’s 3-Space sensor to provide the 6DOF data required for head tracking, which we’ll be looking at in more detail in the next article—for now let me tell you that the responsiveness and lack of latency provides a refreshingly transparent experience. Is it as good as the proprietary sensor found in the Rift? Without more time with the unit it’s impossible for me to call, but during my time with the unit I didn’t give head-tracking a second thought, which speaks volumes."

 

 

http://www.roadtovr.com/infiniteeye-worlds-first-ever-hands-210-fov-hmd-video/

Share this post


Link to post
Share on other sites

Well, well, well, there may be some development money trickling in then. I mean the US navy is actually funded for real, not like computer geeks.

 

Oculus Rift isn't hurting for investment dollars, and have far more than they need to R&D, and produce the initial CV1.   About a month or so ago received another 75 million from a private investment group.

Share this post


Link to post
Share on other sites

Yep but of the Skunk works were buying toilet seats for a million each, then a Rift unit probably goes for 10 million. No wonder the federal bank is broke.

 

Oculus Rift isn't hurting for investment dollars, and have far more than they need to R&D, and produce the initial CV1.   About a month or so ago received another 75 million from a private investment group.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...