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

[Namielus]

Yet he is actually playing "my first pony".

[Qoheleth]

What is the resolution of the new 7-inch display?

[mahler]

What is the resolution of the new 7-inch display?

1280x800 in total .. split evenly (640x800) between both eyes.

[Krenzo]

That dude is so stealthy with the camo Rift. It's too bad the enemy will be able see the bright blue LED. :p

[marbas]

That dude is so stealthy with the camo Rift. It's too bad the enemy will be able see the bright blue LED. :p

HMD-SHOT!

[brantlew]

It's been a long time since we've seen a new review. Pulled this off the Kickstarter comments. Too bad they were only showing the old headset, but they've got updated demo software now.

Hey guys. I checked out and tested the Rift today at I/ITSEC playing Unreal Tournament and have to say I'm pretty excited to get the kit in March. The prototype they had there was very comfortable. Also, despite still using the now older screen, the visuals exceeded my expectations and were immersive in a way I haven't experienced in a game yet. It was an interesting adjustment getting used to aiming by looking with the control scheme they were using, but after a few minutes it became intuitive and quicker than aiming with a control stick. I can't wait to start playing with it in Unity in March.

and...

The headset prototype was the same one seen in this video (http://www.youtube.com/watch…). A few people tried it before me and one said he experienced some mild nausea in the demo, apparently one that was created by Carmack specifically for the Rift. The demo was a room with a stage and some camera equipment in it that you could fly around using the control stick for lateral movement and your head to look and fly up or down. The demo was a little better for getting a sense of depth than Unreal Tournament as it was rendered much wider, more or less from the perspective of a fly, than a typical game. Overall, I would say the sense of depth was pretty good. I had one critique, which the attendant wasn't able to answer given his knowledge of the tech, about the depth which was that it seemed to stop on a plane somewhere in front of you. To be more specific, it was almost like I was watching a 3D movie on a giant dome screen, similar to the (old) Back to the Future ride at universal, obviously without seeing any screen edges. I was looking for something coming at me, like they do in cheesy 3D movies, which wasn't apparent in either of the experiences.

As far as the quality of the screen, as I said before it exceeded my expectations based on all of the talk about pixel fill issues. Honestly I didn't really notice any grid or screen door effect, although if you were looking for it you could find it. The bigger issues for me were the brightness and color contrast. The visuals seemed a little washed out compared to typical monitors or tv's, almost like the brightness was turned too far up. It will be interesting to see the improvement in this aspect with the new screen that is specced. Overall though, feeling visually like I was in the world was a unique experience that I think will drive a lot of innovation for gaming.

One other thing that's worth noting was the control scheme for Unreal Tournament. Like I said, you had all the same controls you normally would with an Xbox 360 controller, just with an added control of using your head to aim. The attendant likened aiming to "staring your opponent into the ground", which I thought was a pretty good analogy. After a little getting used to, it became pretty intuitive to look at the enemy and hold down the fire button. Honestly though I don't know if having the crosshair following head movement is something that first person shooters will adopt; it was a little awkward as far as realism. I felt more like I wanted to look around with my head then aim with the control stick, although admittedly aiming with my head was more effective.

I think overall it was pretty exciting to see the Rift working as well as it was for a virtual reality headset, especially when you consider the price point. I think it's something most gamers will appreciate for its immersion and it will be really awesome to see how developers are able to innovate gameplay based on it.

[rmcclelland]

Here is another new preview. I look forward to these flowing again:

http://www.stuff.tv/news/computers-and- ... -on-review

I'm curious about the:
"newer, more advanced game engine that Oculus asks me not to identify"

Cryengine I hope.

[Alkapwn]

Here is another new preview. I look forward to these flowing again:

http://www.stuff.tv/news/computers-and- ... -on-review

I'm curious about the:
"newer, more advanced game engine that Oculus asks me not to identify"

Cryengine I hope.

HOLY cakes of SHIZA! You guys thinking what I'm thinking?????

The experience of floating through the virtual environment is so natural that it’s easy to forget that we're still sitting in an office chair rather than roving through an industrial landscape hunting zombies.

Zombies...
Industrial Landscape...

LEFT 4 DEAD!! Could this be official-unofficial Source Engine support??!! Palmer... Dycus...

Edit: In my Excitement I put Steam support instead of Source Engine

[Bretspot]

I was thinking about this problem.

On games where you pilot some sort of large robot or something, if you are in the robot your IPD will still be human sized, so anything that is further than say 100 feet away or so (Easy in a giant robot) will have very little parallax, especially on a low resolution screen. So it would make sense to artificially increase your IPD in the game. But then you run into the problem where you feel like the environment is not giant 200 foot buildings around you, but instead a little model of a city.

This is also a problem for a space simulator games where nearly everything (Except your cockpit) is likely to be WAYYYY too far away for parallax.

I was thinning of trying to calculate the maximum distance an object can be away based on the screen resolution in order to keep some parallax based on 65mm IPD, but it doesnt make much sense to calculate it if you are using anti-aliasing.

[3dvison]

I think that is how real life vision works (But Not Sure).
The closer somthing is the more 3D it looks but as things get farther away, they look flatter, because our eye spacing stays the same.

If you had a small model of a mountain in your hand it would look more 3D than that same mountain full size, if it was miles away from you. As an object gets farther away from you, the more the (Same) the right eye and left eyes view of that object becomes. All the objects close and mid range will give the area between you and the mountain a feeling of depth and distance but the mountain will be flatter looking than those near objects.
I'm just saying, it is still 3D, just not as much as near objects. Or more to the point, our eye seperation never changes.

[Flassan]

I don't believe humans can see 3D further than about 25 yards, after which they rely on occlusion etc.
I think I recall that when they filmed the northern lights in 3D the cameras were a mile apart.

[3dvison]

I don't believe humans can see 3D further than about 25 yards, after which they rely on occlusion etc.
I think I recall that when they filmed the northern lights in 3D the cameras were a mile apart.

BINGO...
I have used 3D cameras and you are right.
Unlike with our eyes, which we can't pop out and move for more seperation, with 3D cameras they can and do seperate the two view points of the cameras for far off objects.

[Owen]

In my giant robot game I will be setting the IPD to be to scale for the robots. That way the player will feel like they are 30 feet tall, walking between little buildings.

[3dvison]

In my giant robot game I will be setting the IPD to be to scale for the robots. That way the player will feel like they are 30 feet tall, walking between little buildings.

That sounds very cool, and when I read it, my eyes did pop out a little bit...

[Nick3DvB]

I suppose you'd expect to see the "doll's house effect" if you were a giant robot!

That's what stuck me about the reviewer above complaining about the lack of "pop-out",

with infinite focus and no apparent screen plane what are things supposed to pop out of exactly...

[cybereality]

That is one thing the is a little disappointing about HMDs in general (not just the Rift) is that you don't get the same "pop-out" effects you get with 3D monitors.

However real-life is pretty immersive, and you don't get things popping into your eyes either (at least I hope not).

[Nogard]

That is one thing the is a little disappointing about HMDs in general (not just the Rift) is that you don't get the same "pop-out" effects you get with 3D monitors.

However real-life is pretty immersive, and you don't get things popping into your eyes either (at least I hope not).

Agreed, I have a fear of heights, looking down at the streets from the empire state building in spite of the IPD and the 25 foot limit on parallax didn't stop me from seeing how god damned high up i was.
and looking up at tall buildings while I was in New York triggered it as well so its not just standing on them that does it.

Perhaps it is the contrast from 3D to flat that does it.

[Namielus]

As long as you have stereoscopic 3D working on a virtual cinema screen, I think thats a way to experience the popout effect in 3D movies. Then you once again have a visual reference that stays put in the virtual room (the frame of the screen),
and the content of the screen can seem to pop out from that frame.

[BOLL]

As long as you have stereoscopic 3D working on a virtual cinema screen, I think thats a way to experience the popout effect in 3D movies. Then you once again have a visual reference that stays put in the virtual room (the frame of the screen),
and the content of the screen can seem to pop out from that frame.

Oh yeah, it is only the reference of a frame in front of you that causes stuff to pop out, like popping out of a an actual window. In reality things are just moving around in the observable 3D space in front of you, seems so obvious now. With a HMD that frame would not be there as the image fills pretty much your entire view, removing the borders of the screen to BE INSIDE IT! Feels like this idea could lead to a TV-ad, I can visualize it in my head!
Someone is playing/watching content, then the borders of the TV/screen expands outwards, opening the "rift" into virtual world (morphing into an eye? logo?), finally enveloping the player/viewer entirely for total immersion with amazement to follow... Same feeling as the portal to Narnia? Haha!

[tmek]

In my giant robot game .

Oh hey Owen, I just watched your robot game dev diary video using my DIY Rift on youtube.

[youtube]http://www.youtube.com/watch?v=KjtQwk7zg24[/youtube]

[coresnake]

However real-life is pretty immersive, and you don't get things popping into your eyes either

Depends on what kind of 3D 'content' you're watching I guess.....

[greenknight]

Here is another new preview. I look forward to these flowing again:

http://www.stuff.tv/news/computers-and- ... -on-review

I'm curious about the:
"newer, more advanced game engine that Oculus asks me not to identify"

Cryengine I hope.

Hehe, I was thinking it was in fact unreal engine, even though we know about that one already. Leaks happen all the time.

[Libertine]

Perspective i assume plays a large part in distance perception. When i first got my VG278H after having used a 46" 3DTV at 1m, i was stunned at the increase in depth, but couldn't figure out what caused it. I assume it was the smaller imagery. From that i have wondered if they could increase the sense of depth by shrinking the scenery the further into the scene.

[2EyeGuy]

This is just a guess, but I think it's actually that the 3D drivers detect your screen size and maybe try to keep the separation the same in real units. For example, infinite distance should be 6.4 cm of separation, regardless of screen size. But when the width and height are smaller but the depth is the same, it probably feels deeper.

All 3D games support perspective rendering already, so there's no need to specially add that.

[Libertine]

Im not sure either, when i first got my 27", i immediately loaded up screenshots which like movies, obviously don't have their parallax changed when brought to smaller screens (yet that i know of). Tron for example had only about about a .5 inch parallax at the max (on a 55" screen) that i could find. So that may have been what formed my opinion, which could be wrong. Yet you'd think the fact that the actual on-screen separation being smaller causing my eyes to converge closer combined with the screen being in front of my 46" would make it have less depth.

All 3D games support perspective rendering already, so there's no need to specially add that.

Yet rendering engines don't account for changes in in-game field of view, or the display's field of view around the viewer, or the viewing distances effect on size, or the fact that viewing distance effects the sense of depth when viewing stereoscopic 3D. When i back up away from my 46" screen from my normal viewing distance of just under 1m, for example, despite having a parallel (ish?) interaxial that matches my interocular, the depth and volume of the imagery looks more correct.

[2EyeGuy]

I get what you mean now. Yes, you can make that kind of VR display where the screen is exactly like a window and it uses head-tracking to adjust the perspective. That's called fishtank VR. Or a VR CAVE if you make the screens room-sized and have screens on every wall and floor.

[rmcclelland]

Here is a short article on Palmer's talk at the Evolve conference. Not sure if we'll ever see the slides or a video.

http://www.develop-online.net/news/4275 ... challenges

Edit: A better written article:

http://www.edge-online.com/news/evolve- ... ge-gaming/

[ChrisP]

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.

[EdZ]

That would explain their grooved surface (for grip). If the lower edges (obscured by the housing) were geared, they would rotate in sync. The lens centres would move up or down slightly, but if you're already nudging the image horizontally you may as well nudge them vertically too. You could even add an encoder and do this automatically.

[Coasterpro]

A pass through the shadows and highlights filter in Photoshop gives us a better look at these "grips" around the lenses. It appears that you can indeed rotate the lenses to adjust the IPD very slightly; hopefully enough to satisfy most users. I assume there will be software to calibrate the image after making physical adjustments to the lenses.

[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.