Oculus "Rift" : An open-source HMD for Kickstarter

[MSat]

I figured that if there was IPD adjustment, it would be made from the knobs located on either side of the HMD. That way there would be indicators (numbers/letters) you could use to adjust the software accordingly.

[Owen]

Well in the consumer version you would want to make sure that it has encoders that report the IPD to the driver and automatically adjust the rendering. If users have to calibrate it, they generally won't, and then they will complain about how it looks messed up.

[Coasterpro]

I figured that if there was IPD adjustment, it would be made from the knobs located on either side of the HMD. That way there would be indicators (numbers/letters) you could use to adjust the software accordingly.

There is indeed an indicator above those knobs. But I'm not sure that it makes a lot of sense for them to be for IPD given that there are indentations embossed into the ring, apparently for better grip. It seems more plausible that IPD is meant to be adjusted by turning the rings directly and that the knobs on the side are for focal adjustment or something. ??? Hopefully the next update will shed some light on these questions.

Well in the consumer version you would want to make sure that it has encoders that report the IPD to the driver and automatically adjust the rendering. If users have to calibrate it, they generally won't, and then they will complain about how it looks messed up.

Ah okay I see what you're saying. Yeah that would be a nice feature. Certainly a convenience at the very least, if not, as you suggest, a necessity.

[MSat]

I don't see any grip-like texture on the lens rings - they just look like housings that may be connected to the adjustment mechanisms on the outside. If there are grips around the lenses, I presume it would be to screw them up or down perhaps for focus. At any rate, Oculus has said that the unit in the images (including the adjusters on the side) is not the final design, so that's why their functionality hasn't been specified. We'll know for sure once it's all set in stone (er, plastic).

[Coasterpro]

The grip-like texture is definitely there. This illustration should make it more clear. Since there's no stylistic reason for putting those indentations on the rings they must be for something. And given that the lenses are offset within the rings it seems logical to conclude... if tentatively... that the rings are for adjusting the interpupillary distance somewhat akin to how one would adjust a pair of binoculars. Seems a good way to do it but they may just be experimenting as you suggest.

[MSat]

To my eye, that looks like empty space for the lens + housing to move around in, and the pattern is just a reflection off the LCD. Why the cut out is a circle instead of an oval, I don't know. Maybe to allow for different sized lenses?

[android78]

my guess would be that the adjustment on the sides is for interpupillary distance and you can screw the lenses in and out a fraction to adjust the focus a little... just a thought.
We also can't say for certain that those lenses aren't positioned in the center of the recess, just sticking out in front of it a little. Perspective would make them look off-center.

[EdZ]

It's not just a trick of the light, you can see the edge of the grooves here:

[marbas]

hmmm notice how lens's are off-centre in the black rings holding them . Wonder if that has to do with IPD adjustments just turn the rings until the space between lens's are right.

To me it looks like the lens's are off-centre illusion is actually due to the natural perspective shift from the lens taking the picture.

[ChrisP]

@marbas
well maybe. but that would still be a cool way to do a slight IPD adjustment. thick side out small IPD, turn 180deg thick side in max IPD.

[BOLL]

@marbas
well maybe. but that would still be a cool way to do a slight IPD adjustment. thick side out small IPD, turn 180deg thick side in max IPD.

Seems plausible when looking at the images, good detective work people :P Also, unlimited settings if the adjustments are discreet enough so it's not important to have them centered vertically. In that case you just have to make sure they have mirrored angles, nice!

[crespo80]

WOW
Very nice find EdZ, I put a red circle to make it visible to those who missed it

[ChrisP]

i just type without thinking sometimes.
... wish there was a delete post button haha

[Randomoneh]

To get this out of the way.

It is optimal for eyes to look through centers of the lenses in any case, right? Otherwise, you get possible low inner or outer FOV (or low stereo effect) + unpredictable asymmetric distortions that are different for each eye and therefore require different algorithms for each half of the display. For me, this alone should be enough for implementation of physically adjustable lenses.

Am I right?

[EdZ]

Yes, that is correct. When your pupils are not centred with respect to the lenses, the distortion will not be the same as when they are, and a different shader would need to be used. Additionally, the area where the image is 'most flat' (i.e. has the least distortion and highest pixel density per solid angle) will no longer be straight ahead, and will be in the same location for each eye.

[Randomoneh]

Yes, that is correct. When your pupils are not centred with respect to the lenses, the distortion will not be the same as when they are, and a different shader would need to be used. Additionally, the area where the image is 'most flat' (i.e. has the least distortion and highest pixel density per solid angle) will no longer be straight ahead, and will be in the same location for each eye.

So, would you agree that physically adjustable optics would be the obvious choice and standard feature of future HMD's? Any other way seems pretty backwards way of doing things to me.

[C3DPO]

To get this out of the way.

It is optimal for eyes to look through centers of the lenses in any case, right? Otherwise, you get possible low inner or outer FOV (or low stereo effect) + unpredictable asymmetric distortions that are different for each eye and therefore require different algorithms for each half of the display. For me, this alone should be enough for implementation of physically adjustable lenses.

Am I right?

Yes, if you aren't dead center you create both prismatic and chromatic and other aberrations. Because the lenses in the Oculus are plus powered, if you were to set the lenses too close, or too narrow for your face you will be creating Base In prism. Base Out if the lenses are wider than your interpupillary distance. This prismatic effect worsens as the power of the lenses used goes up. I have no idea what power the oculus rift will use. For collimated light, they would have to use about a 5 diopter lens if the screen is about 20 cm away from your face. Maybe someone has already posted how far the screen sits from the eyes I don't know. If its further than 20 cm then you'd need a less powerful lens, and therefore the negative prismatic effects would be lessened per millimeter offset you have in the alignment. And if it's closer, a more powerful lens is needed and the alignment becomes even more important.

What this means is mostly headaches, and this is also where one of the benefits of the RIFT splitting the screen for right and left eyes will help. If they were to use traditional 3D where the screen is not split, but the entire screen is used for both eyes, there would probably be common headaches because of the large discrepancy between the eye's focusing and the eye's alignment (keeping the image overlapped and not double). If they used collimated light (light behaving as if it were coming from a distance), then your focusing system would be relaxed but your alignment would be pretty active. We don't do that normally, and therefore headache. In that scenario, it would be interesting to see if inducing some base in prism would actually help the alignment system to relax. BUT, the good news is with the RIFT our eyes won't have to converge as much because the screen is split. That means focusing will be relaxed, and alignment will be much more relaxed than with a traditional 3D screen (like 3DS or tv).

But the other issue is you also will get an illusion of image size differences. If the lenses are too narrow, then objects will appear larger/closer than they should. If too wide, then things will look smaller/further. This is probably negligible though. In reality, I will be playing with the IPD a lot to see if inducing just a little Base In relaxes me a little.

[C3DPO]

I work with boy scouts and we took them to an aiport to investigate careers in aviation. They had a flight simulator there that cost 6 figures. It was awesome and running it was very cool. But the whole time I was thinking, this could be matched and out-matched by the RIFT. Dang GINA with some good software flight sims could be crazy and so much less expensive! Rift really is the start of something huge.

[brantlew]

Flight sims is one area where you really can't do better with an HMD. I think having a physical cockpit environment beats the hell out of a virtual environment - not to mention that the detail of physical screens placed a meter in front of you will kill the resolution of any HMD - plus stereoscopy is practically a non-issue since objects in flight are 100's of meters away at best. But maybe you can make an argument for quality/cost ratio. If your total budget is $1000 then sure, an HMD rocks. But for pure immersive quality those enclosed cockpits are freakin' amazing.

[Mel]

Flight sims is one area where you really can't do better with an HMD. I think having a physical cockpit environment beats the hell out of a virtual environment - not to mention that the detail of physical screens placed a meter in front of you will kill the resolution of any HMD - plus stereoscopy is practically a non-issue since objects in flight are 100's of meters away at best. But maybe you can make an argument for quality/cost ratio. If your total budget is $1000 then sure, an HMD rocks. But for pure immersive quality those enclosed cockpits are freakin' amazing.

My experience in MS FS-X with my DYI HMD confirms what you say. I found it the least compelling game of all those that I tested. And all this talk about lens placement, adjustable IPD, image center adjustment, etc gives me hope that I may yet get a more satisfying HMD experience, because as it is right now, my DYI unit is sitting unused due to visual discomfort. Hopefully the real Rift will address all the problems I personally have with HMDs.

[C3DPO]

Flight sims is one area where you really can't do better with an HMD. I think having a physical cockpit environment beats the hell out of a virtual environment - not to mention that the detail of physical screens placed a meter in front of you will kill the resolution of any HMD - plus stereoscopy is practically a non-issue since objects in flight are 100's of meters away at best. But maybe you can make an argument for quality/cost ratio. If your total budget is $1000 then sure, an HMD rocks. But for pure immersive quality those enclosed cockpits are freakin' amazing.

well you have great points. But it's not just the money. If you have the space in your garage or dedicated room, If you have the money, and if you have the permission from the wife, and the time to set it up and get it working then yeah, don't stoop to playing any other way, and you should probably just buy a cessna if you are that into it. Except for the fact that once you do shell out the cash, floorspace, and time, it's the only cockpit you get in that spare room. In a virtual cockpit being able to switch the view from a boeing to an F-22 to a crop duster to a jet pack to a helicopter is a huge plus.

It's true, stereospcopy is trivial, but looking out your side windows isn't. That's a premium I didn't have at the flight school simulator. I'm sure there are nicer ones out there that may wrap around or project to the side. But in an HMD, it's done.

The biggest advantage to a real cockpit is the quality yoke and foot pedals. Those would be a pain in the butt to recreate. But other than that, give me the flexibilty of the headmount.

[C3DPO]

That being said I have only arcade experience with VR, so maybe it's wishful thinking. But I would still like to try a flight sim on the RIFT

[android78]

Did anyone else see the following headline and have a little heart attack:
http://au.gamespot.com/news/layoffs-at-rift-dev-6401430

... it's ok. Rift is a GAME!

[Owen]

I would agree, if your cockpit setup has 360 degree field of view. Being able to look in any direction is huge, for both immersion and gameplay.

But sure, if you are playing the type of sim where you are in a modern fighter and just fly by instruments with self-guided weapons a mile from the target then you only really care about the cockpit. For those kind of sims, the view outside the window is basically just a pretty horizon marker.

[STRZ]

Wouldn't it be possible to overlay your physical cockpit over the virtual environment, AR style? If you use a camera and some software tricks?

[brantlew]

Wouldn't it be possible to overlay your physical cockpit over the virtual environment, AR style? If you use a camera and some software tricks?

Yeah, I guess you could green-screen it in real time just like they do in movies. Sort of the best of both worlds. That would be really cool for a fighter game where the windshield wraps over your head and would be really hard to create with physical screens. I guess a CAVE like projection system could be used instead, but for this usage an HMD and a CAVE system would be fairly equivalent. In a commercial jet sim however, there wouldn't be much benefit to the green screen over just mounted LCD's.

[MSat]

It's not just a trick of the light, you can see the edge of the grooves here:

Ok, good catch guys! Maybe it's for interchangeable optics, so people with vision problems can screw in appropriate ones? Any educated guesses on how many different ones would be needed for something like that?

[2EyeGuy]

Well, I need -4 dioptres. I don't know about other people.

[MSat]

I still think an HMD would be awesome for flight simulators. Maybe not the best for commercial flight trainers (where much of it has to do with instrument flying and management), but for home use, especially if you also play games other than that (or as mentioned a variety of aircraft), I don't think an HMD can be beat. Remember, AR could also be used to virtualize your arms and joystick too - no need to see your actual hands and joystick. As for the low resolution, it's been discussed before and would probably be fine by relegating a joystick button to zoom. Just glance over at the instrument and press the zoom button - quick and easy enough (especially if you normally use a mouse or hat switches to look around). Of course, resolution matters. Hopefully that'll be less of an issue with the consumer Rift.

[BOLL]

As this thread sure is about Kickstarter, here is an article with Palmer's comments on just that: http://www.develop-online.net/news/4277 ... rter-rules

[STRZ]

Indiegogo is a good alternative to Kickstarter http://www.indiegogo.com/

[Dycus]

Indiegogo is where rejected Kickstarter projects go.

[marbas]

Can you promote your project on both Kickastarter and Indiegogo simultaneously? Sounds unlikely, but then again I haven't read their rules.

[Kazioo]

Indiegogo is where rejected Kickstarter projects go.

This isn't entirely true.
Kickstarter is only for Americans and Brits. The rest of the world has to use workarounds (US/UK family, friends etc.) or use Indiegogo.

[STRZ]

Stumbled on this game brwosing reddit.com, looks cool, very sim like http://www.enemystarfighter.com/

On Youtube there's a comment from the developer that he will support the Rift as he gets his dev kit, Linux supported as well

[Coasterpro]

Cool! Perhaps not as cool as Star Citizen but I actually kind of like the non-textured retro art style. It also has the added bonus of ultra fast rendering speed (necessary for the Rift). However I noticed the smoke is made up of sprites and that won't look very good in stereo 3D. But if the developer is planning support for the Rift as you say then he'll no doubt make those changes once he sees it in action. I'll have to keep an eye on this one.

[SiggiG]

Hey guys, check out the lens adjustments and IPD in the old VFX1 helmet:
http://www.youtube.com/watch?v=J0n5B3fl-bU#t=6m05

Maybe the grooves you saw in the Oculus are for something similar?

[drifter]

However I noticed the smoke is made up of sprites and that won't look very good in stereo 3D.

Good point.
I'm working with Unity and was wondering how to replace all these 2d particle effects (smoke, fire, etc.) for a proper stereoscopic experience ?
In many games adapted for 3d vision, they simply just removed these particle effects :/
The solution i guess would be a kind of LOD particle system : some 3d particles (very costly i guess) displayed under like 10m, and 2d particles beyond.

[codes]

I'm working with Unity and was wondering how to replace all these 2d particle effects (smoke, fire, etc.) for a proper stereoscopic experience ?
In many games adapted for 3d vision, they simply just removed these particle effects :/
The solution i guess would be a kind of LOD particle system : some 3d particles (very costly i guess) displayed under like 10m, and 2d particles beyond.

That'd prove more difficult than it sounds though, because you would get really prominent popping on the transitions between 2d and 3d assets. Essentially it'd come down to creating a brand new volumetric/model effect and rendering it out with it's animations into a 2d sprite so the transitions are as smooth as possible, but still you will notice them.
Probably best not to use a sprite system at all and let the engine LOD the effect through the model system.
All of this is a hell of alot ore work than generating sprites though

[Owen]

Well the easiest way is to have a separate sprite for each eye. You can render out some 3d smoke into those sprites from slightly different perspectives. Not perfect but it definitely won't appear flat that way.