Meant to be Seen

jaybug wrote:I think I might have broken my LCD controller

The display will shut off all the time, but it gets back on if I hold it in the right position. But I can't find a consistency in it to allow me to have it working in a reliable fashion.

Sounds like a good chance for a cold solder joint or a broken wire - if you are lucky it's an easy fix I guess.

Any tips for figuring out where the problem is?

One of the solder joints on the back of the lvds connector seems to be broken off, could that be the issue?

Uh, yeah. Probably need that wire.

You sure? seems superflous to me, there's so many other wires

But why does it work fine until I move the panel the slightest inch...

So what does this cable do, and what are my options:

Edit: I think it's a ground, is that correct? Is it then possible for me to move it, if I dont have soldering equipment handy? (Or am not handy with soldering equipment )

This is the project I have been working on for about 4 months now.
I ran through several prototypes
I am using a Galaxy Nexus Smartphone on Jelly Bean and the open source engine of quake 2 (can also be run with the shareware)

http://www.youtube.com/watch?v=-n1nr9C6JMk

what do you think ?

Hi,

what do you think ?

Great work.

This is definitely the way to go for VR.

The sooner smartphone/tablet makers wake up to the possibilities of onboard or streamed games in S3D, the better the VR technology is going to get.

The incentive for software providers is the potentially vast software titles that can be put on the appstores for VR content.

Thanks.

Nice! I was thinking this should be pretty doable.. Awesome to see someone go and do it.

Believe it or not, Apple has a patent on this. Yeesh.

Fantastic stuff! Well made video too.

tek2222 wrote:This is the project I have been working on for about 4 months now.
I ran through several prototypes
I am using a Galaxy Nexus Smartphone on Jelly Bean and the open source engine of quake 2 (can also be run with the shareware)

http://www.youtube.com/watch?v=-n1nr9C6JMk

what do you think ?

Very nicely done! Do you have a download link available?

tek2222 wrote:This is the project I have been working on for about 4 months now.
I ran through several prototypes
I am using a Galaxy Nexus Smartphone on Jelly Bean and the open source engine of quake 2 (can also be run with the shareware)

http://www.youtube.com/watch?v=-n1nr9C6JMk

what do you think ?

Thats very elegant

Good job!

I will be releasing this open source , but I dont yet have a download link, i will fix up some stuff first.

tek2222 wrote:I will be releasing this open source , but I dont yet have a download link, i will fix up some stuff first.

Bummer... Until then it is just a teaser. Not even a half-baked early demo? Vaporware? Place the pie before my eyes and tell me I have to wait for some unspecified period of time before I can even taste it myself? Its great attraction is that we can play with it now while waiting for our Rifts to arrive, but that time will arrive very soon for some of use, and then our attentions will be drawn away from your project.

Others can help you polish this off NOW, you know. No need to be embarrassed by warts an blemishes. Many projects fall into neglect waiting for finishing touches when something more interesting comes along. I would hate to see your hard work go to waste, while and others have to independently duplicate your efforts for their projects while waiting for your release.

Your video shows that the software IS good enough for us to begin to play with and adapt. Better to get it out there now, IMHO, so that others may borrow from and contribute to it... But that is not my choice to make, so all I can do is encourage you to get on with the show. Please do!

I guess i could make the software available. I just dont want the prototype on thingiverse yet.

tek2222 wrote:I guess i could make the software available. I just dont want the prototype on thingiverse yet.

That is all I want! Thanks! I can understand holding back on Thingiverse though. Rough ends on assembly cannot be fixed as cheaply or as easily as in the software side of a project.

Regarding the software, yes, please...

jaybug wrote:You sure? seems superflous to me, there's so many other wires

But why does it work fine until I move the panel the slightest inch...

So what does this cable do, and what are my options:

Edit: I think it's a ground, is that correct? Is it then possible for me to move it, if I dont have soldering equipment handy? (Or am not handy with soldering equipment )

Yeah, that one's ground. From right to left, going in pairs - 3.3v, [space], GND, twisted pair 1, twisted pair 2, twisted pair 3, [space], twisted pair 4. Any other wires are just extra, if you're using the Vitrolight panel.

Technically, it should work with only one of those ground connected, but in reality, the wires aren't quite thick enough to handle the current. You could connect that ground wire to any ground on the board. If you need help with that, I can help you. If that doesn't fix it, though, you probably have some other problem - maybe a different wire loose? Maybe an internal connection broken? I have no idea.

Dycus wrote:Yeah, that one's ground. From right to left, going in pairs - 3.3v, [space], GND, twisted pair 1, twisted pair 2, twisted pair 3, [space], twisted pair 4. Any other wires are just extra, if you're using the Vitrolight panel.

Technically, it should work with only one of those ground connected, but in reality, the wires aren't quite thick enough to handle the current. You could connect that ground wire to any ground on the board. If you need help with that, I can help you. If that doesn't fix it, though, you probably have some other problem - maybe a different wire loose? Maybe an internal connection broken? I have no idea.

Thanks for the helping tips, Dycus, Turns out it wasn't the missing solder point after all. We actually tried reconnecting the ground wire to no avail.
After alot of fruitless voltage testing and general fickeling with the circuit board me and my friend found the culprit.

It was simply a loose jumper. Found by carefully tapping on everything once the display was working, and fixed by bending the pins slightly to make it fit. A little bit of a face palm, but a whole lot of relief.

My next step should be to extend the LVDS cable. Having to keep the controller on the headset is really the worst part of my HMD. So heavy and cumbersome with the powercable. Should save that for when I could do without it, though, as I will almost certainly ruin it. -.-

Hah! It's always the silliest thing. Glad you got it working.

Hi All,

This is my first post so go easy on me. I was inspired by this thread to make my own vr and I wanted to share my results as of now:

1. I'm using the lens recommended by Rod via Palmer, and they work great. http://bit.ly/W9cNx8

2. I haven't purchased a screen yet so I'm testing with a third gen iPad and a nexus 7.

3. I put together a quick SBS 3D rig with Unity 3D. This has worked great because I gave myself sliders on the screen to adjust, IPD, DPI, and FOV to dial in the best settings. Also, I added a cross hair because it really helps to get the image in focus. I'd love to share but the I am using a Gyroscope plugin that I purchased (only $10) to do head tracking so I'd have to remove that. Also, to create the fisheye distortion I am using Image effects that are only available in Unity Pro. I could strip out the fisheye and head tracking if people are interested in the basic unity package.

4. Head tracking using the gyroscope in the iPad works surprisingly well. Not much lag, and more accurate than I would have thought.

5. I integrated the lens into some old ski goggles setting the lens center point at my IPD of 60mm (I know, narrow) and put the center of each eye's image at 60mm as well. the effect was impressive but I was getting a lot of overlap of the left and right eye image i.e. I could see the left eye image from the right eye. I experimented with widening the spacing between the lens and found that having the lens wider then my natural IDP fixed the overlap, increased the FOV and did not significantly distort the image. The lesson was I first thought I would have to adjust the distance between the center of each eye image to match the distance between lens, however, this is not the case. Best focus was achieved with the images spaced to match my eyes and with the lens spaced slightly larger than my eyes.

6. I'm interested in doing an augmented reality version so I was thinking about a way to hinge the lens and the idea struck me to get some welding goggles. Many welding goggles use 2" diameter glass to block the light so the thought was that maybe I could swap in the lens and could use that instead of ski goggles. I just got these yesterday
http://www.amazon.com/Hobart-770129-Oxy ... B001FOPXGG and they are awesome. I can mount the lens in one of 4 different ways to allow me to adjust the distance from the eyes to exactly what I want. (I've got them mounted on the inside frame with the frame turned backwards). My concern was that the ultra width distance between the lens would be a problem. the goggles fix the lens at 77mm apart. To my surprise, even at 77 mm apart everything still looks great and the fov is even better.

7. Here's an image with the lens in:

Thanks for the inspration and I hope my input is helpful.

-Mike

miketalbott wrote: Thanks for the inspration and I hope my input is helpful.

Thanks for your input and yes it is helpful! I'm going to try this too now:

miketalbott wrote: The lesson was I first thought I would have to adjust the distance between the center of each eye image to match the distance between lens, however, this is not the case. Best focus was achieved with the images spaced to match my eyes and with the lens spaced slightly larger than my eyes.

Having the lenses farther apart than your IPD causes the virtual images to move outward, preventing overlap, but it also causes asymmetrical distortion, so ordinary pincushion/barrel correction formulas are not quite adequate. The real problem would be with stereoscopic convergence at the outer edges, but that is much less of a problem than it could be only because of partial overlap. With much more full overlap (like when using stacked fresnel lenses), asymmetrical warp can be much more of a problem because of stereoscopic vertical misalignment.

The partial overlap of the Rift and Rift clones makes the pre-warp function a lot easier to handle. And the asymmetrical distortion from looking through the inner sides of the lenses stretches to outer edges of the horizontal FoV even farther (which is a good thing).

Finally done with diy hmd.

But when i see through the lenses , im still seeing two images? Instead of one.

vrvision wrote:Finally done with diy hmd.

But when i see through the lenses , im still seeing two images? Instead of one.

Are you looking at full screen side by side content like a 3D youtube video or John Carmack's Doom 3 image he tweeted? You're not going to see anything but doubled images if you are just looking at your Windows desktop.

vrvision wrote:Finally done with diy hmd.

But when i see through the lenses , im still seeing two images? Instead of one.

If you are viewing a pair of SBS (Side-By-Side) images and you cannot merge them into a 3D stereo pair, you may need to move the image centers to match your IPD (inter-pupillary distance).

If you are viewing on a 7-inch screen, you may need to bring the image centers inward to match your IPD. There are multiple ways to do this:

1) You can shrink the images so their centers are closer together, matching your IPD. You can do this when viewing in a web browser with Crtl-MouseWheel. Or use a paint program to shrink.

2) You can crop the image inside edges so their centers are closer together, matching your IPD. Use a paint program.

3) You can shift your lenses outward so you are looking through the inner half of them. By offsetting your lens centers to NOT match your IPD, you are performing a "lens shift" operation. Lens shift is commonly done in high-end video projectors so they do not need to be perpendicular to the center of the screen, allowing it to be mounted at other convenient locations such as the ceiling or a side wall. Lens shift allows moving the projected image without requiring keystone (trapezoidal) correction. Lens shift is also commonly done using expensive tilt/shift camera lenses, popular in real estate photography so that tall buildings photographed from nearby do not like they are leaning backwards. On a Rift clone, you can shift your lenses a little to pull the wide-IPD images on a 7-inch display inward to match your eyes. This has a benefit of pulling the virtual image inward over the other eye, allowing even more stereoscopic overlap without seeing the wrong image. The downside of lens shift is some asymmetric pincushion/barrel distortion depending on the lens and the offset. That distortion can be corrected in pre-warp software, but it is not as simple as a regular centered barrel distortion correction. Incorrect pre-warp distorting can cause vertical misalignment at the outer edges of fully-overlapped images, making stereoscopic convergence difficult in those areas.

The Oculus Rift pre-warp software uses the "inside cropping" method, resulting in non-overlapped images, which are not as sensitive to lenses not centered over viewer IPD. You could replace Rift Dev Kit lenses with offset lenses cut from larger lenses, to compensate for mismatched IPD, or to intentionally mismatch IPD to give us more central stereoscopic overlap with less image cropping. They are "Dev Kits" just so that we CAN do such experimental modifications to them, after all...

Im using ipad screen (9,7) and im testing it with cybereality's video (pre wraped)

vrvision wrote:Im using ipad screen (9,7) and im testing it with cybereality's video (pre wraped)

The images will have their centers too far apart on a 9.7-inch screen. Shrink the size down and center to your lenses. If viewing in a web browser, you should be able to resize and position them so they look okay.

If you can adjust your lens IPD, you could try shifting the lenses outward from your eyes. That will pull the image centers inward to match your IPD, but will also cause some asymmetrical distortion (not corrected correctly by the symmetrical pre-warp algorithm). The center view should converge well though. But for a 9.7-inch display, you may not be able to offset the images enough without larger lenses, in which case just resize the video like I suggested above...

You can compensate offset lens distortion somewhat by tilting the lens, similar to how tilt/shift camera lenses work.

geekmaster wrote:

vrvision wrote:Im using ipad screen (9,7) and im testing it with cybereality's video (pre wraped)

The images will have their centers too far apart on a 9.7-inch screen. Shrink the size down and center to your lenses. If viewing in a web browser, you should be able to resize and position them so they look okay.

If you can adjust your lens IPD, you could try shifting the lenses outward from your eyes. That will pull the image centers inward to match your IPD, but will also cause some asymmetrical distortion (not corrected correctly by the symmetrical pre-warp algorithm). The center view should converge well though. But for a 9.7-inch display, you may not be able to offset the images enough without larger lenses, in which case just resize the video like I suggested above...

You can compensate offset lens distortion somewhat by tilting the lens, similar to how tilt/shift camera lenses work.

Thanks it works now by resizing, but what about games like doom 3, do i need to resize for oculus ready games too?

Note to anyone in danger of mashing "View active topics" for hours today in search of shipping news - do yourself a favor and read a few pages of this thread at random instead.

You want a Rift so bad? Why don't you build one??

Dies anybody have pdf or Autocad file of 7 inch case?, because the one in the original post is for the 5.6 inch screen

Anyone have schematics for a nexus 7 enclosure?

I already looked at the 3d printed kit... and nope i rather save my cash for a real Rift

Right now though i just want to tinker on the cheap..

Hey guys, I posted this in another thread, but I will post it here as well:

I put together a quick and dirty package to get your Unity games working in your DIYRift kits. There are limitations if you don't have Unity Pro, like having to run cybereality's Vireio driver instead of the shader. The stereo camera script will actually work with the DIY Rift kits and Unity Free, but there is noticeable warping on the edges (hence the need for an external driver like Vireio).

Disclaimer: I cannot take full credit for all of this work, as it was a lot of cannibalizing from different projects and shaders other people in the community wrote, but I did put all the pieces together in a slim, easy-to-use package that works and is free to download and use.

VRDK: DIY Rift Camera Kit for Unity3d v1.0

Please help me refine this and give me feedback on problems and issues you find.

This is great info and perhaps there will be other companies in the future that come out with such things too . I dont see that as a threat. I think honestly that would be great. The problem before is that the focus on VR was on higher end designs /prices. I guess up till now no one really thought about "gee lets figure out a way to make VR avalible to most gamers?"

Even if some reason Oculus doesnt quite acheive what they wanted perhaps they can be a match that starts the fire of bring back the idea of VR into the average consumers mind.

Im not good with things like this i have fine moter skill issues and to be honest id be afraid id berak something trying to put it together.. (more akin to Homer Simpson) steady......steady...bam.

tek2222 wrote:This is the project I have been working on for about 4 months now.
I ran through several prototypes
I am using a Galaxy Nexus Smartphone on Jelly Bean and the open source engine of quake 2 (can also be run with the shareware)

http://www.youtube.com/watch?v=-n1nr9C6JMk

what do you think ?

So How would the actual Rift be better than your purely mobile solution? It seems like having all of the processing and tracking self contained on the device pretty much wipes out the issue of latency.

antimatter wrote:

tek2222 wrote:This is the project I have been working on for about 4 months now.
I ran through several prototypes
I am using a Galaxy Nexus Smartphone on Jelly Bean and the open source engine of quake 2 (can also be run with the shareware)

http://www.youtube.com/watch?v=-n1nr9C6JMk

what do you think ?

So How would the actual Rift be better than your purely mobile solution? It seems like having all of the processing and tracking self contained on the device pretty much wipes out the issue of latency.

Proccessing power and price are the first two things that come to mind. It's good enough for simple graphics but even a gaming machine would strugle pushing the possible 1080p at 60fps in anything demanding without a big sacrifice.
Still, noone said it was out of the question, I think I recall Palmer saying somewhere that ideally, eventually the rift could be a self contained machine.

rfurlan wrote:I already made a pledge worth a few Rift units @ Kickstarter however I don't think I can wait until December. I just ordered the parts to build a HMD based on what is known about the current Oculus design. Is anyone interested in doing this with me? The cost will probably be equivalent or a bit higher, and it will take some work. However I am sure it will be a great learning experience

You can find the updated building instructions here:
http://bitcortex.com/oculus-libre-open- ... culus-rift

Be sure to review this whole thread as it details a handful of different designs you might like too!

[Advanced]
Also check this thread for a guide explaining how to use an HDMI cable for your LVDS connection:
http://www.mtbs3d.com/phpBB/viewtopic.php?f=140&t=15554

So my Rift should be here in the next week or so but I've been thinking and I'm hoping that you gurus could answer a few questions for me. I know the reasons that the specific 1280x800 screen was chosen for the dev kit but let's say I had some time to play, money to burn, and a ~7" 1920x1080 screen/driver-board to play with. What are the factors involved in swapping that with the one already in the rift (assuming fit isn't a problem)? Voltages, signal types, physical connectors, etc..

I do realize that question may be way too large to answer, especially without having a rift on hand to tear down but then again, I don't know what I don't know and I DO know enough to be dangerous so I figured I'd throw it out there anyway in case someone has the knowledge to provide an answer. Feel free to ignore this if the scope is too large to be a reasonable question or something like that

Take care.

I just got back from my long trip around Asia - it is great to see this thread is still going strong

I am still waiting for my Rift, I hope it will be easy to disassemble and tinker with. Have anyone posted tear-down pictures yet? If not, I will as soon as I get mine!

Welcome back! Hope you had a good trip.

Haven't seen much in the way of a Rift teardown yet.
These blog posts give great insights though...
http://www.oculusvr.com/blog/behind-the ... pilot-run/
http://www.oculusvr.com/blog/building-a ... atency-vr/

Looks like the tracker uses an STM32F103, so it's a nice neat design.

OzOnE2k10 wrote:Haven't seen much in the way of a Rift teardown yet.

Ehem... http://www.mtbs3d.com/phpbb/viewtopic.php?f=140&t=17078

Welcome back, rfurlan!

Finished my hackey rift with the help of your excellent guide while you were away, so thanks a lot

tek2222 wrote:This is the project I have been working on for about 4 months now.
I ran through several prototypes
I am using a Galaxy Nexus Smartphone on Jelly Bean and the open source engine of quake 2 (can also be run with the shareware)

http://www.youtube.com/watch?v=-n1nr9C6JMk

what do you think ?

Very impressive.

I have a DIY one based on a Nexus but no native software. Its quite involved to use Vereio (games driver), Freepie (tracking information), Splashtop or Kany (to send the screen info to the DIYift) to get things to work.

Image quality is impacted due to scaling and compression that the screen senders enploy so native software on the nexus would be a great way to fill the time before my rift arrives.

I tried several of the android quake ports but the stereo code does not seem to have been transfered.

In any even loving your work.

UB

tek2222 wrote:I guess i could make the software available. I just dont want the prototype on thingiverse yet.

Sharing is caring But it would be awesome if you could at least make the software available. I already have my own mount for my phone and have been playing games streamed from my computer with splashtop. It would be nice to have something different (and fun) to play with that's native to the phone!