Posted a thread for the PR3:
http://www.mtbs3d.com/phpBB/viewtopic.p ... 79&start=0" onclick="window.open(this.href);return false;
@Syntax: That could work with an MRG2.2, but for something like this, you need a display that puts the center of the view in front of each eye. Yes, you can get away with converging (Like the My3D for the iPhone), but it ruins a lot of the effect.
@Pierreye: Very interesting gadget. Of my own budget, but I will see if my work will buy one.
@Foisi:
1) Yes, it does make a difference. Works fine without it, but you are able to see the individual pixels if you look closely.
1.1) You are going to laugh... I got it out of an old Powerbook G4 LCD.
The backlight has a number of diffusion elements, and one of them is very fine grain. I have my hands on some diffusion material that is used in the Wide5, going to be trying that next.
1.2) Blur is a bad way to describe it. It just diffuses the light a tiny bit, your eyes do not strain. Besides, suppose you are looking at a blurry image on a computer screen: Do your eyes strain to focus? Nope!
2) These lenses are acrylic, but we have a diamond saw at my work for when we want to cut lenses.
3) I made the lenses using a vacu-form mold, same way the original LEEP shell was made.
4) I will take some pictures when I get the chance.
5) I calculated the diagonal FOV with the help of the tracking system we have at work. We have a virtual environment with a grid, and you can move the HMD around, noting at exactly what point the grid lines are not visible anymore.
6) Fun trick: You use a DSLR camera, and point it into the lens of the HMD. You see, most DSLRs are cable to tell you what distance the focus is set at! Just adjust the focus until you can see the pixel grain, and you have the distance of the display.
6.1) The PR3 is not adjustable on its own. It does, however, have a swappable lens faceplate! If you made a faceplate with adjustable optics, then yes, you could have it. I do not think it is really needed, though, LEEP lenses are very tolerant of eye spacing and relief.
7) Same as 5. It is not a good "budget" way to calculate it, I know, but it is extremely accurate.
@Bobv5:
1) The parts that are not off the shelf were made using a vacu-form mold. You cannot go out and buy them, you have to make them yourself (Mainly the lens holding assembly).
2) It is a pretty low resolution, but you can get around that somewhat if you are willing to stretch the image a little bit vertically. As for making a 2D version, it would be a piece of cake! The PR3 has swappable lens faceplates, if you want, I can make a few 2D faceplates using actual MRG optics, and a few alternative lenses. This would make it very similar to the PR1. One foot apparent distance is definitely possible, but I would consider 3 feet or so to be the bare minimum for usability, something I believe I can accomplish.
2b) Foisi gave you a great technical answer. The pixels are actually very, very small. Without a diffusion filter, you can not see individual pixels so much as the image having a slightly grainy overall look. With the diffusion filter, you can not even pick out individual pixels, they are invisible to the naked eye. I know it sounds low, but this panel is much higher quality than most microdisplays, and the resolution actually looks very nice. Recall that the MRG2.2 was about a fifth of the resolution, and people still like it!
3) I am still doing tests with this. Right now, it is very, very promising, and I would say that you have 70-80% of the immersion when using the PR2. The most important thing is that you have pitch and rotation tracking, and that the virtual FOV perfectly matches the FOV of the HMD, so that objects look life sized. Not to brag, but the PR2 actually has better color reproduction and contrast compared to the Wide2, and it certainly shows. FOV is much more important than those things, though (I mean, people still feel "immersed" in the real world when they wear tinted sunglasses, which ruin your color and contrast). The other exciting thing is that I might be able to push the FOV of my system up near 140 degree FOV, which would put it very, very close to the Wide5. More as it develops.
4) Actually, the edges of the screen are touching the edges of the lense, so even in 640x480, you still have almost the same FOV, somewhere between 115 and 120 degrees. You only lose a little bit of vertical FOV! Only about 640x720 of the display is visible if you strain your eyes to see, more likely is 640x640.
@InvaderZIM: Consider yourself on the list to try these things out, once I make some more improvements. If you like them, a good review would really help, and if a lot of people are interested, maybe I could see about making more to sell (Maybe a kit?). My goal is not really to make money, obviously, or I would not be telling everyone exactly how it works!
That said, it would be nice to sell some HMDs to VR enthusiasts. Made a thread for the PR3, go see it!
Phew, all replied! Sorry for taking so long to do so! Let me know what you think.
can you explain this please 
The problem with the low res in 3d is that the instruments in any simulated vehicles is impossible to read. When I said 1 foot, I only meant that I could cope with it, not that I wanted it. Not sure if I made that clear. I do wonder though how it would even look in 2d, with the interior of vehicle suddenly being spherical! It looked a bit odd on the low fov hmd's, but everything looks odd at 30fov.....
Luckily, the panels are available elsewhere. The control boards are actually made by a company called Roward, all Vitrolight does is program them. Worst case scenario, we might have to get a group buy together for panels and control boards.
One filed a claim saying the screen I shipped him was defective, and he sent it back with a torn ribbon cable. Ugh. Anyway, your idea would definitely help with that.