DIY Oculus Rift - schematics, instructions and build pics!

[Endothermic]

And with Apple you'd know they'd ensure it was estrically pleasing in design.

Don't forget that it would only work with Mac, iPhone and iPad, would have very strict policies in place for what you can do with it and what your actually allowed to make for it and half the things you submit will probably be disallowed anyway and then they'll sue every other HMD on the market for violating their patents and innovation

I think if the consumer RIFT doesn't flunk when it gets released and shows there is a market for such a thing then you can be sure Sony will be brining out a VR HMD (whether PS4 only or universal is the question but I'd bet on universal) rather then just a personal viewer. Chances are they probably have a guys fooling around with VR already but its not a full fledged project yet till they decide there is a big enough market they could make money off and go ahead and put a full department on R&D and get it in production.

[BHawthorne]

And with Apple you'd know they'd ensure it was estrically pleasing in design.

Don't forget that it would only work with Mac, iPhone and iPad, would have very strict policies in place for what you can do with it and what your actually allowed to make for it and half the things you submit will probably be disallowed anyway and then they'll sue every other HMD on the market for violating their patents and innovation

I think if the consumer RIFT doesn't flunk when it gets released and shows there is a market for such a thing then you can be sure Sony will be brining out a VR HMD (whether PS4 only or universal is the question but I'd bet on universal) rather then just a personal viewer. Chances are they probably have a guys fooling around with VR already but its not a full fledged project yet till they decide there is a big enough market they could make money off and go ahead and put a full department on R&D and get it in production.

People love to bash Apple and seem to gloss over what they're good at. Apple has the fabs and manufacturing base to do some pretty insane things with HMD if they ever wanted to. They also have the retail chain and marketing to back it up. I use Apple purely as an example though. Any of the big names have advantages over garage projects but the big names tend to shun any idea that didn't spawn out of their own R&D departments as irrelevant.

Sony has been making HMDs for awhile now...it's just the FOV sucks.

[rfurlan]

So this exists now:
http://www.panelook.com/COM48T4Mxxxxx_O ... 15895.html

Panel Brand : ORTUSTECH
Panel Model : COM48T4Mxxxxx 
Panel Size : 4.8 inch
Resolution : 1920x1080
Display Colors : 16.7M   (8-bit)
Signal Interface : LVDS (2 ch, 8-bit)  
Input Voltage : 5.0V (Typ.) 

Trying to buy a couple, will keep you all posted

[TheRealistWord]

Wow, nice find! Insanely huge PPI of 458!! It's something, to think these panels were announced nearly 2 years ago and only now they're beginning to ship them. But, couple questions though - did you actually find a supplier? Because the link you shared doesn't seem to point to anyone selling the panel. Also, with the 5x lens, wouldn't the screen be a bit too small? I've been using my cell phone's display as the screen for now (tried it both with the 5x and 7x lens), and it's pretty impressive, but of course it's way too small (4.1"). I have no idea if 4.8" would work or not for a single screen, unless you use a separate panel for each eye? I have no clue (I'm bad at measurements usually, lol). Keep us posted!

[Dycus]

Too small without some crazy optics... but that's a pretty sweet screen.

[MrGreen]

What would be the perfect screen size Dycus?

[Chriky]

With the current style of optics, the limit for 100% overlap is about 13cm horizontal (2x the lens spacing). You could add on some more either side but it would only be for the eye on that side (this actually reflects normal human vision). I would say 7 inches would be the maximum.

[zacherynuk]

So this exists now:
http://www.panelook.com/COM48T4Mxxxxx_O ... 15895.html

Panel Brand : ORTUSTECH
Panel Model : COM48T4Mxxxxx 
Panel Size : 4.8 inch
Resolution : 1920x1080
Display Colors : 16.7M   (8-bit)
Signal Interface : LVDS (2 ch, 8-bit)  
Input Voltage : 5.0V (Typ.) 

Trying to buy a couple, will keep you all posted

I asked ortus directly some time ago when they started shipping samples, but nothing. Where have you had joy?

[Dycus]

The screen size is limited by the magnification. We could use a 4.8" screen if we had higher magnification lenses. Problem with that is then the exit pupil gets smaller and you get more distortion. Plus, there aren't even lenses like that yet.

The perfect screen size would be of an aspect ratio they don't mass produce yet. At 16:9, though, 7" is great.

[zacherynuk]

The screen size is limited by the magnification. We could use a 4.8" screen if we had higher magnification lenses. Problem with that is then the exit pupil gets smaller and you get more distortion. Plus, there aren't even lenses like that yet.

The perfect screen size would be of an aspect ratio they don't mass produce yet. At 16:9, though, 7" is great.

Getting slightly OT, but:
Using single lens / current Rift design; perhaps. However, my view, building LEEP style plastics optics for anything UNDER 5.8" is easy enough - down to about 1.5" - aiming for an exit pupil of 30mm+
The reason the 5.8's are so good is that the IPD is already near perfect average.
A single 4.8" screen would have an almost unusable IPD of about 5.3cm (vs current 6.4cm) 2 screens of 4.8", in portrait, bring a side by side IPD down to a workable minimum 5.9cm.

(As I already have a head big enough for two side by side current 5.8 screen, for me this is moot - and the aspect ration detracts from the immersion)
However, the reason I wanted these screens is that they are the first small HD screens to do 120hz - this means I could build 3 in portrait running Nvidia surround in 3D with shutter glasses built into the helmet design. I made a 3 portrait mono rig with the current screens but it was A) bulky and B) Not 3D (though 30hz 3D possible it stinks)

Also, I would say a 7" screen is too big for comfortable side by side, unless you have a gargantuan planetoid of a head. Natural IPD of 7" diagonal screen is 7.7cm

Source: http://en.wikipedia.org/wiki/Pupillary_distance & http://www.prinds.com/tools/screenDimensions.htm

[Dycus]

Well yes, if you don't move the image at all. But who says the center of each image has to be centered in its respective half of the screen?

With the 5.6" we've been showing off, you can just see the left and right edges if you look. You can't see them on a 7".

[zacherynuk]

Well yes, if you don't move the image at all. But who says the center of each image has to be centered in its respective half of the screen?

With the 5.6" we've been showing off, you can just see the left and right edges if you look. You can't see them on a 7".

Granted; at 100% overlap you could lose the borders, but then why waste money on the higher resolutions?
However, with wider optics and partial overlap you could do cool stuff with a larger screen - I have yet to find anything which would process partial overlaps though, to my knowledge neither IZ£D nor Tridef does it. It's eay enough in principal and hopefully something that will be part of the SDK, though as Rift is 100% overlap I doubt it...

[brantlew]

@Dycus: Hmmm Oculus guy. It sounds like you have some real hands-on experience with a 7" screen.

[Dycus]

@Dycus: Hmmm Oculus guy. It sounds like you have some real hands-on experience with a 7" screen.

I have one, yes.

[MrGreen]

The screen size is limited by the magnification. We could use a 4.8" screen if we had higher magnification lenses. Problem with that is then the exit pupil gets smaller and you get more distortion. Plus, there aren't even lenses like that yet.

The perfect screen size would be of an aspect ratio they don't mass produce yet. At 16:9, though, 7" is great.

When you guys picture yourselves in the relatively near future, do you see VR taking off enough to have custom panels manufactured for it?

What kind of FOV do you get from a 7" screen?

[Malfate]

@Dycus: Hmmm Oculus guy. It sounds like you have some real hands-on experience with a 7" screen.

I have one, yes.

something tells me this has something to do with Palmer saying that the developer kit may end up having a slightly greater horizontal FoV. Instead of the current, 90, he was saying it could get closer to 100-110 for the developer version.

[Dycus]

When you guys picture yourselves in the relatively near future, do you see VR taking off enough to have custom panels manufactured for it?

What kind of FOV do you get from a 7" screen?

I actually have no idea there, and if I did I probably wouldn't be allowed to say. But I really hope we get big enough for custom panels. That would be awesome!

FOV is a little bigger horizontally, not sure how much, and you can't see the edges any more.

[crespo80]

Dycus, do you guys think to ship a standard complete consumer system (ie the rift + a hydra-like controller) to make it easier for developers to design highly optimized rift-ready games for a shared ecosysem, or you'll sell the rift alone and we'll see different games that will need different input control systems?

[IGameArt]

That's great news, i honestly don't care too much about getting the fov much wider because everyone says the immersion is spectacular regardless, but if you can make it so the edges are undetectable for just the devkit, then i'm gonna be in heaven come December.

[Dycus]

Crespo - I doubt it. We're letting developers figure out what kind of control schemes work best for their games and whatnot.

[IGameArt]

the control scheme is really something that should be dependent upon the game, you wouldnt want to play a racing or flight sim using a razer hydra, but playing something like skyrim would be great with that combination, where each wand controls each hand.

[mahler]

Dycus, do you guys think to ship a standard complete consumer system (ie the rift + a hydra-like controller) to make it easier for developers to design highly optimized rift-ready games for a shared ecosysem, or you'll sell the rift alone and we'll see different games that will need different input control systems?

I'd like to wait and see what the most popular gametypes will be after December, before considering this.
It could be that somebody makes a proof of concept of something amazing using something totally different.
But I've already seen 3 awesome hydra-demos sofar (one, two, three), so you may be right about it becoming an important part of VR games.

[3dvison]

Could somone who has used a hydra along with the type of tracker that was used with the RIFT say for sure if they work well together or if there if a problem with magnetic interference ?

[cybereality]

I've used the Hydra with the Vuzix tracker and it seemed to work OK. Didn't do any scientific study but it worked good enough.

[android78]

Could somone who has used a hydra along with the type of tracker that was used with the RIFT say for sure if they work well together or if there if a problem with magnetic interference ?

I thought that the Rift tracker only used accelerometer and gyroscopic data, not magnetic sensor - hence the inability to have absolute direction tracking. So there should be no issue using the hydra with the Rift.

[brantlew]

We only know the tracker that was used for the demo prototype - and that was almost purely out of coincidence because it just happened to be the tracker that Carmack had fiddled around with and was able to quickly integrate into the demo. But just as Oculus is reconsidering other possibilities for screens and lenses, we should expect that they are looking at all available tracker solutions - including a "driftless" 9DOF tracker.

[FingerFlinger]

Cool, I just found out the 7" panels we have at work use a discrete driver, so I should be able to inject my own signal. I'll have to bring one home mess with it.

[MSat]

Could somone who has used a hydra along with the type of tracker that was used with the RIFT say for sure if they work well together or if there if a problem with magnetic interference ?

I'm sort of curious about this too. Having done a quick google search has shown that MEMS gyro/accelerometer structures could be vulnerable to magnetic interference. However, I have my doubts that the coil in the hydra is powerful enough to be a problem.

[3dvison]

I was just using the headtracker that was in the Demo RIFT, as a worse case example. So should it be the one that does get used in the Dev Kit RIFT and I buy a Hydra now, will it work with that tracker ?

[brantlew]

I was just using the headtracker that was in the Demo RIFT, as a worse case example. So should it be the one that does get used in the Dev Kit RIFT and I buy a Hydra now, will it work with that tracker ?

I think the Demo Rift (Hillcrest) tracker would be the "best-case" scenario since it does not include a magnetometer.

But you bring up a really good question though. Can magnetometers and magnetic trackers be used together without interference? In other words, if Oculus includes a 9DOF head tracker how will this affect users that are planning to incorporate Hydra-type devices in their setups? You can calibrate magnetometer-assisted IMU's against a stable magnetic background so in theory you should be able to use them together, but I wonder how confined your movements must be. Could you move a meter to the right or left or would this throw the IMU calibration off? How about a head mounted Hydra? If the magnetic field generator and IMU were locked in the same reference frame, would this provide a stable background field? I can only speculate. I still have a Hydra sitting in a box that I desperately need to take out and do some testing with.

[3dvison]

Ok thanks brantlew & android78 , I did not know that.
So the Vuzix tracker does have a magnetometer ?
And if so, I guess Cybers answer is a bit more inline with what I and brantlew were thinking about, as far as crossed signals go.
Hopfully this won't even come into play with whatever tracker they pick to use with the RIFT.

[brantlew]

yes, the Vuzix tracker does include a magnetometer so cyber's experience is the only relevant data point that we have

[crespo80]

Crespo - I doubt it. We're letting developers figure out what kind of control schemes work best for their games and whatnot.

I agree with you for the dev kit to be released in December: the best thing is to let the devs free to experiment any sort of control scheme and figure it out which works best for their games.
But don't you think that, for the consumer version, would it be best to make a selection among the different approaches and choose the best one to make it the "standard Rift control set"?
Because if the consumer rift will be sold alone, the most plausible scenario is that every "rift" game is just a normal mouse/keyboard game that has a "rift mode" that only lets you turn your head around like in the doom3 demo, and this would be a pity in my opinion.

If you made a consumer bundle even with a low-cost hydra-like controller that doesn't raise the price up too much, all the rift games could be highly optimed for that and bring a far superior VR experience just out-of-the-box.

Just my 2 cents

[Dycus]

I dunno, I guess we'll see!

Also, check out the controller this guy made!

http://imgur.com/a/7hlI2#0
http://www.reddit.com/r/electronics/com ... _designed/

I've been meaning to get into microcontroller programming for a while now... except I would want to build mine in a Zapper.

[crespo80]

I've been meaning to get into microcontroller programming for a while now... except I would want to build mine in a Zapper.

I want that, and please make a duck hunt mod in which we can finally shot that damned dog

OT
On a side note, I just noted that the oculus image of the stylized eye is identical to the "watch item" icon of the ebay android app, I hope that won't cause any legal issue

[MrGreen]

Crespo - I doubt it. We're letting developers figure out what kind of control schemes work best for their games and whatnot.

I agree with you for the dev kit to be released in December: the best thing is to let the devs free to experiment any sort of control scheme and figure it out which works best for their games.
But don't you think that, for the consumer version, would it be best to make a selection among the different approaches and choose the best one to make it the "standard Rift control set"?
Because if the consumer rift will be sold alone, the most plausible scenario is that every "rift" game is just a normal mouse/keyboard game that has a "rift mode" that only lets you turn your head around like in the doom3 demo, and this would be a pity in my opinion.

If you made a consumer bundle even with a low-cost hydra-like controller that doesn't raise the price up too much, all the rift games could be highly optimed for that and bring a far superior VR experience just out-of-the-box.

Just my 2 cents

Yeah if there's no driving force behind it I'm afraid most devs will continue to support gamepad + m/kb and call it a day. On a side note, mouse and keyboard, which is considered the preferred control method of gods, is an absolute no-no with the Rift. How ironic... Problem is, the simple thought of having to use anything else could be a deal breaker for a lot of PC gamers. At this point the Razer Hydra seems to be the best choice. The most realistic one anyway. But how many devs will properly support, let alone design for it, two expensive optional peripherals?

Anyway, I don't wanna sound more pessimistic than I am. Hell, I'm not pessimistic at all. I'm merely thinking out loud here on what it would take to bring the Rift to mass market. Mass market in gaming terms anyway...

I'd bug the hell out of the big 3 and this point. Microsoft in particular. Rift + Kinect 2 = profit.

[mahler]

I've been meaning to get into microcontroller programming for a while now... except I would want to build mine in a Zapper.

I want that, and please make a duck hunt mod in which we can finally shot that damned dog

OT
On a side note, I just noted that the oculus image of the stylized eye is identical to the "watch item" icon of the ebay android app, I hope that won't cause any legal issue

For illustration



and



But I don't think it's that big of a deal.
Besides... I think they eBay icon looks an awful lot like a free stock-icon

[rfurlan]

Latest build pics:





By the way, these are the goggles I am using, you can order it from Amazon for just $11:
http://www.amazon.com/gp/product/B0041PXQ38

[rfurlan]

After playing with my latest build for a bit, here are my comments:

(1) With head-tracking the feeling of immersion is absolutely incredible, it is almost as good as the nVisor SX111 for a fraction of the price.
(2) Make sure to design your case so you can open it to reach the screen without having to remove too much tape
(3) One problem with using ski goggles is that it makes it harder to implement replaceable face plates. Not a big problem if you are going to be the only person using your HMD, but I have a very wide IPD which makes it impossible for me to share the same face plate with my wife.

Conclusion: If you haven't built your HMD yet, I recommend that you use TheLostBrain's CNC design (http://bit.ly/S4nEYV) - instead of ski goggles - because: (1) it allows for easily replaceable face plates; (2) easy access to all internal components for future upgrades; (3) doesn't require duct tape!

[MSat]

Latest build pics:

Are those the 2" 5x aspheric lenses?