It's high time we as a community build the finest HMD?

[smoothy]

I think the team behind mtbs need to create an initiative to aid manufacturers to really push to create the world finest affordable HMD?

The features I really want to see are the following

stereoscopic RGB laser display that doesn't need focusing such as the Microvision pico projector, for a first time HMD I don't mind if it's not HD, but I want full human visual field of view. 6DOF head tracker, microphone, eye shield to block out all light, wired or wireless connection. I believe the technology is here now but companies really need that creativity to make it into reality. The Microvision MEMS laser pico projector is being handed out to OEM's. Microvision say that the projector to be embedded into a device only raises the costs of the device by $100. The pico projector is the size of a mint. Now obviously hardcore HMD users like us are quite limited in today's industry. So if there was one an HMD of this type made it may not sell as much which is why the HMD manufacturer may charge a high retail price for such a device. That price could be for instance $2500 But I think us hardcore users would be happy to pay that much for full immersion. I mean look we shell out several thousand dollars for our passive projector setups. Now at the same time it would be us paying hardcore members that would allow these HMD manufacturers to show the world user bases that these new full immersion HMD's are incredible which would then excel manufacturing volumes from the hundreds to thousands of units which would then lower the price to say under $500.

Now here is another scenario. I think quite honestly if HMD's are going to get anywhere beyond the tunnel vision that we have been seeing for years. It's only going to change if we as a community make it happen. So I believe firstly there should be a poll on this forum with two questions and a radio button to click yes or no.

These questions should be, Do you want total immersion in an HMD with the full human visual field of view? and the second question. Would you be happy to contribute $10 or more to a special team at mtbs to help develop a full field of view HMD? Now if we were to do this we can get an idea of how many would be interested in these two questions.

I feel if we all donated $10 or more to a forum user who would like to take the challenge of building an HMD of this scale. Then this could really be the start of something fantastic. Of course people don't just hand their hard earned money to strangers. I feel it would have to be a technical guy that likes working on their DIY projects. Who also has a high forum post count and is backed up by the team at mtbs. Then we donated $10 or more to this person or group of persons who join the HMD initiative to be setup. Now with blogging sites like digg, technorati etc we can spread the word and hopefully many donations into the hundreds over time will happen which should give enough money to build a prototype. Then a testing and evalaution period can commence. Perhaps a years time we could have a HMD in our hands. Now obviously this type of project I would like it to be open source. So if we want we can all build it ourselves while the main initative is building it as well. I thought that when ever the intiative meets a milestone it would release plans on how to build that core componenet and afterwards when all coponents have been done we can put the thing finally together.

I think this mode of thinking will help propel this HMD market. We could develop our own DIY HMD through the initiative. Then once it's done and works we could help to get it mass manafactred and sold for those people that don't want to build it themselves. and for those that want a more streamlined device rather than there prototypes they've made. Finally I believe from this project with a final result if we get there. It will show other manufacturers that it was entirely possible to do which would aid in many companies to start developing such HMD's.

Now I think the people on here are the best people who know about making stereo 3d setups and there are countless other forums which have very technical people just like us which I am sure would be more than happy to help. You see most HMD companies have a lot to lose in the HMD market and they don't listen to their customers, they would never bother with forums like this. Stereo projection with passive or shutter glasses companies would listen to us like iz3d does but that's because there is a bigger market than the limited HMD market. So let's all get into this and make this happen. Let's show the big companies what we really can do.

[OllyLuck]

And i give you vuzix new VR1440.
for PC users, the new VR1440 gives you the resolution you’ve been asking for and the VR
experience you’ve only dreamed about.

[smoothy]

And i give you vuzix new VR1440.
for PC users, the new VR1440 gives you the resolution you’ve been asking for and the VR
experience you’ve only dreamed about.

Please post any info on this new VR1440

The only info I found is that the resolution is still 800x600 but the only thing it can do is take a 1600x1200 resolution game and downscale it to it's native resolution of 800x600.

[groda]

YES!

It is really about time "someone" produces a functional and moderately cheap HMD with full field of view (FOV). I'm not sure if this community is the right forum to do the actual production, but, surely, we must be able to forward to the companies that are in this business this is what we (at least I) want.

The advantage must be obvious compared to the limited view of either watching a slightly different computer screen with each eye (polarized glasses, shutter glasses, mirrors etc) or some type of HMD displaying a computer screen seen at some distance. Instead of watching the game world protruting from, and continuing behind, the screen while you're playing a game (like walking around in real life with your head in a box with a hole in front of you), with full FOV you will able to visually experience the surroundings to a large extent like what you do in real life. Of course, head tracker and microphone should also be included in this kind of product.

As I see it, there can be at least three reasons why an affordable product with these specs isn't already on the market.

1. The companies that produce these gadgets don't know that there is a market for this kind of product. (There just has to be a market for it!!!)
2. It is difficult/expensive to manufacture.
3. There is actually not a market for an affordable full FOV HMD.

Please drop a line if are you interested in getting a full FOV HMD, or if you know why you can't already buy one. I, for one, think that this would be "the ultimate s3d device" (until they plug us in for good...), and I support Smoothy's idea that we can contribute to make it become reality.

[smoothy]

Thanks for your input groda!

I don't think it was a good idea me posting in this part of the forum. It seems not many people come here like they do in general stereoscopic 3d discussion. I think this thread should be moved but I don't know how to, if anyone can tell me how that would be great? In the mean time mods please move this to general S3D discussion. Many thanks in advance

[cybereality]

Everyone would love a full FOV HD HMD but it can't be that easy to make. It will require a lot of money and a lot of talented people. Then you have to wonder what is holding back companies like Vuzix from doing this themselves. Again, it can't be that easy, even harder for DIYers. I commend your effort but I feel like this is a losing proposition. I think we will just have to wait for technology to advance and some company like TDV to deliver the goods.

[chrisjarram]

I for one found this post amusing I have to say

[smoothy]

Everyone would love a full FOV HD HMD but it can't be that easy to make. It will require a lot of money and a lot of talented people. Then you have to wonder what is holding back companies like Vuzix from doing this themselves. Again, it can't be that easy, even harder for DIYers. I commend your effort but I feel like this is a losing proposition. I think we will just have to wait for technology to advance and some company like TDV to deliver the goods.

Those companies you mentioned could build such an HMD that I propose and the technology is already available but the reason they don't is because they know the product will be too expensive and know they could only sell a few hundred compared to thousands if they just make a lower spec system. On the other hand we are hardcore users, so making a hardcore HMD that i propose will take time and a lot of money but it is doable if we all contribute to the project. The HMD won't be as well made as if it was done in a factory but it would be good enough for us general users.

[groda]

"Everyone would love a full FOV HD HMD but it can't be that easy to make. "
Yeah, I guess it can't be all that easy, but let's consider how difficult...

First, to make the actual displays for each eye larger should be quite easy.

Then comes a serious problem, namely that the surface might have to be curved to cover the human FOV (a bit less than 180 degrees) totally. However, I have seen advertisements for HMD's with curved displays (not stereo HMD, though), so it can surely be done. Besides, flat displays that are wide enough will cover most of your FOV - like a pair of ordinary sunglasses.

So, what is the problem then? My guess is lack of software. (Or lack of effort from the hmd companies...) A normal computer game hmd simulates a monitor seen at some distance. All that has to be done is to somehow translate the pixels on a monitor to the small screens in the HMDs. I think it gets much more complicated if the small screens can display things outside the range of a normal screen.

Anyway, I don't think that this has to be only wishful thinking, if, let's say companies like iz3d and emagin (z800) could team up and create a cheap and good full (or almost full) FOV HMD with working software.

At least now things are starting to happen in the s3d world. As someone said, 2009 will be the year s-3D starts to become mainstream. Hopefully it will come true.

[Tril]

I'll start by saying that I'm definitely not an expert on HMD. I've tried two and my knowledge stops there.

I think the main issues are related to the optics. Getting a very wide screen to be kept in focus. The human eye only keep in focus a small area in the center of where it casts its gaze. If you have a screen that goes farther to the left and the right to get a bigger FOV, the left and right parts of the displays will be blurry. You might have to make an HMD in which you can rotate your eyes to the left and the right to see the edges of the screen in focus. HMD usually require very precise positionning of the eyes to see the displays properly. It might be hard to make an HMD that works even if you rotate your eyes a bit. It might need a custom lense and a custom curved display. You also want a screen that's seamless (if it's made from combining more than one displays beside each other).

You're free to spend your free time how you want but I think this project is bound to fail, especially if you're not very knowledgeable about embedded design, FPGA, usb, VGA, DVI, smd electronics, opt6ics, HMD, small display, computer programming, etc.

[Okta]

My guess why no one has built a high FOV HMD yet for consumers is they only want to make a multi function device that you can use as a desktop as well as gaming and movies. This makes the optics very difficult to avoid blurry edges of the screen.

If you made a dedicated gaming hmd and sacrificed some edge blur perhaps a larger FOV could be achieved.

[smoothy]

I think the main issues are related to the optics. Getting a very wide screen to be kept in focus. The human eye only keep in focus a small area in the center of where it casts its gaze. If you have a screen that goes farther to the left and the right to get a bigger FOV, the left and right parts of the displays will be blurry. You might have to make an HMD in which you can rotate your eyes to the left and the right to see the edges of the screen in focus. HMD usually require very precise positionning of the eyes to see the displays properly. It might be hard to make an HMD that works even if you rotate your eyes a bit. It might need a custom lense and a custom curved display. You also want a screen that's seamless (if it's made from combining more than one displays beside each other).

You are right about the edges will get blurry but that's on most conventional displays not with lasers. The new microvision picop MEMS laser can project to curved surface and always stays in focus. It doesn't use mirrors, optics or a focus wheel like all the other pico projectors. This makes it very unique as in you can really project and make a full FOV HMD very easily. Now the idea I thought of using a motorbikes helmet visor is because that is basically a curved screen. You can get all sorts of motorbike helmets with varying visor sizes. Choosing the one with the largest FOV that wraps around you is a good choice. Then you can use imaging warping software like sol7 which would stretch the image of the projector all the way round so the projected image aligns from edge to edge on the motorbike helmets visor. Then you can paint the visor with some projection paint so you got one seamless projection surface. I have seen test of the pico projector from microvision while projecting onto curved surfaces like a curved piece of paper and I can tell you the whole image including the edges do net blurry at all. So it would be easy to focus on in my setup. Finally I could wear some laser safety goggles with fresnel lenses added in. and perhaps nvidia shutter glasses underneath to get a stereoscopic image. There I should have a working HMD.

[ngsmith]

Hi Smoothy,

I have been experimenting with this idea for awhile, but just have not had the time to put my full effort into it. Perhaps, after I finish my dissertation. In the meantime, here are my ideas if anyone wants to work on them.
In order to get a larger FOV, you need a slightly bent screen, which with oled will probably be available in just a couple years.
Another solution that can be created today is to use projected light which can be bent around a curved surface. There are a lot of university researchers that have developed many different techniques using projectors to create dome projections. However, again the significant bottleneck is being able to write the code to properly project a 2D image onto a spherical/cylindrical object and get it to work properly with your software. Check out this website: http://local.wasp.uwa.edu.au/~pbourke/m ... nityiDome/" onclick="window.open(this.href);return false; Paul Bourke has a lot of good references. Sol7 is great software that can do everything mentioned here, but is very expensive. If you email them they may allow you to try a demo of their software. Here is the link: http://immersaview.com/sol7.html" onclick="window.open(this.href);return false; Maybe if a working prototype could be demonstrated they may be willing to release a scaled down software version in a reasonable price that only works with your HMD.
In order to create a projected head mounted display, you would need two half domes (one for each eye) and two projection sources. Theoretically you could take two lightweight small projectors and mount them on the top of one's head and use flat surface mirrors to reflect them onto the domes. The domes would need to be a thin white plastic so they can function in rear projection. The smallest projectors right now are those pico-projectors which you could get for $600 a pop with VGA resolution. If you wanted higher resolution you could take two SVGA/XGA/HD projectors and strip them of everything except their lenses, lcd and i/o pcb. Instead of using the lamps you could replace it with LED's, since you will not need a significant amount of light and only be projecting max 2' in a completely dark environment. If a working prototype could be demonstrated than perhaps a manufacturer may be interested in providing the resources to produce these on a larger scale and build it from components rather than stripping components from a another product.

[ngsmith]

I just saw The thread: [REQ] 3D Helmet using mini projectors. Looks to be similar idea but using polorizers. As I suggested mounting the projectors on the head reduces torque and so theoretically it would have less strain on the neck than current hmd which all mount to the face. You can carry a lot of weight on the head with little strain. I still think it would be best to have one projector per eye so you get full resolution, color and lumens, and no ghosting. However, this would be a similar solution. Either way Sol7 looks to be the software to use to project onto a bent surface.

[smoothy]

Thank you ngsmith for your excellence

You seem to know a lot about this area

I can tell you that the microvision picop was design from the ground up as a pico projector that could project to curved surfaces. The reason for this is that sometime down the line microvision want to bring out mobile eye wear that have curved lenses. It is said to be for VR/AR applications in military and gaming use. I have already found a while back that the company behind sol7 were offering their software on a per computer license for $225. I cannot remember the forum but it was some flight sim forums where someone by the name Andy who worked for immersaview who make sol7 he was offering for that price for people who have matrox dh2go and th2go. Well for anyone actually. I can't find the link to that forum anymore. But you know I am sure they are still offering the same price. And as the microvision pico projector can beam to curved surfaces it wouldn't be difficult to make 2 curved displays and would be very cheap to do. 1 sol7 license 2 $600 pico projectors like you said and we could build this thing for under $1500. Then if we can show it to a manufacturer we could get these mass produced. There really is a market for this.

[nubie]

I think what we need is Research and development money and a lab, also some Electrical Engineers and somebody who is experienced in group buys from China/Asia.

It is possible of course, but I think we should focus on minimum 1024x768 display standard HMDs before we go crazy over the top. (IE: create one or two models of "standard" HMD and then start researching the crazy type.)

I would love to see a reasonably priced HMD (in line with standard LCDs, 1280x1024 for $150, stereo for $300, and perfect drivers, although iZ3D has already written them).

I don't know if the future is projection, or if the future is high resolution flat displays, MIDs are getting higher res all the time, but pico projectors might be getting enough resolution to use soon.

[uberdeity]

Would one of these be a solution http://www.sensics.com/? It uses an array of small OLED screens (the same as on the eMagin z800, I believe) to provide you with a near-as-damnit 180 degree FOV.

Earlier in the thread someone suggested that electroniccy people could be a great help- well, I've got a degree (BSc) in it & a decent amount of experience with the whole Stereo / immersive gaming thing. So I'd be happy to help- especially if there are comp-sci types out there who'd be able to help with the software side of things.

I'd suggest going for a frame-mounted version to start with- it's easier to achieve the head-tracking (as you've got a fixed reference point), weight isn't a problem (the frame can counterbalance the HMD if required), power delivery and video encoding that way. Doesn't look as cool, but makes it so much simpler while you're just working on the optics. Then you can scale everything down to a small single-unit HMD.

For simple home-made headtracking, could I suggest using a high resolution rotary slot encoder (or a greyscale style encoder) as used in the old rollerball mice? It's cheap, simple and accurate. Plus it can be interfaced quite simply to a PS2 or USB interface using common components- and with a few slight changes can even be easily drift-compensated (something that's definately needed). You could even have it control the mouse with the headtracker- giving a decent degree of compatibility with existing games and software.
Given the minor extra expense, you might also want to add on a second set of encoders to handle the slanting of your head. This could be assigned to a second mouse in Linux (with MPX) or Windows 7 (if it supports multiple mice), or the slant could control the rate of tapping on a keyboard key or something.

Would it be possible to have someone more familiar with video processing or optics to comment on the chance of getting a box built that would automatically deform the video image going in (to make it match the curved screen)?

[smoothy]

uberdeity

Many thanks for your input to this project, much appreciated

I believe we can do better than the sensics by using a laser projector. You can check out this video which is a clip from the discovery channel on the microvision laser projector that I have been talking about constantly throughout this thread http://microvision.blogspot.com/2009/01 ... nt-on.html" onclick="window.open(this.href);return false;

[bastian74]

Sensics HMD
I believe each cell is an 800x600 display. What I'm not sure is how obvious the seams are when you're wearing it.

[crim3]

what a beautiful piece of tech

[smoothy]

It's beautiful but at an overly expensive $200,000. Can we please keep this thread open to the possibility of making a mini laser based HMD? As these are getting more affordable and seem the best way to creating a panoramic curved screen around the users field of view. It's been over a week since I started this thread and not much has happened. I had hoped perhaps the moderators would of created a poll with the questions I mentioned in my first post on page 1. Then I hoped someone talented here would of made a website to promote the idea to all the DIY sites like hackaday etc and then I hoped some trusted member on here would of started a donations thread backed by some mods on here. Come on guys this really can be done. If we had all sat around and did nothing back in the 60's we would of never made it to the moon.

EDIT: bastian74 I read a review a while back of the sensics and they said when you turn your head with the tracker you can see faint lines coming down where the seams are stitching the screens together but it's uses some kind of overlap system.

[crim3]

Still we need an expert in optics. No HMD on earth uses fresnel lenses (I think...) so there must be a good reason. Maybe the prototype could be built without them and use those custom glasses with fresnel lens already mentioned, and be usable. Some expert could explain us why those expensive lenses are used in HMD's and go on with the project with more confidence... or not, depending the answer.

Those pico-projectors have really brought us a relative easy solution for the display/s. Now, the second main problem, how to look at a screen so close to eyes must be made clear before going on.

[uberdeity]

Still we need an expert in optics. No HMD on earth uses fresnel lenses (I think...)

I'm nowhere near an expert, but I still know that they screw up the image slightly (according to a 10 second wiki search, they cause "a reduction in image quality") as they're not a "proper" lens.

If I remember correctly, HMDs need lenses to trick the eye into thinking the image is further away than it is (compensating for the lensing effect of the human eye). Which greatly lessens eye strain when using the HMD- your eyes aren't focussing 1/2 an inch away from themselves as they would if there was a plain screen in front of them, they're focussing as though the screen was a comfortable distance from the eye.
A problem with wide-angle HMDs is that- to avoid eye strain and nausea- you'd need to design your lenses so that they have the same lensing effect wherever the eye was looking. Which makes for a very, very complex design of lens.
Alternatively, you could have a weird form of contact lens so the lens is always over the eye- it's not quite as simple as a "strap-in-and-go" HMD, but the people who'll use this sort of HMD will probably put up with up with it for the incredible immersion it'd provide. If someone would be kind enough to find the appropriate design of contact lens (focal length etc) this could probably be produced without a huge cost- probably less than the complex lens you'd otherwise need.

I'm pretty sure, OP, that the reason the mods haven't done the fairly large amount of work you suggested is that there's nothing to the idea yet. It's at the "Let's go to the moon!" stage- or at best the "Right, so.... we'll use rockets?" stage. There isn't even a definition of where the money from the donations would go apart from "to the project".
If you want to make this into a decent project and make a decent attempt at it I'm sure the mods would help, but you'll probably need more of a plan than "it's high time that someone...". See if the mods would sticky (even just for a month or two) an "ideas for solution" thread that details the requirements of the project. Meanwhile, I'm off to check out the lenses on my z800.

Edit: Also, I've been looking at the Microvision laser projectors and while they'd be great for a cellphone they look like they'd probably suck for a HMD- you'd need a decent distance (say 12") between the projector and the screen for a 180-degree FOV. And, I think, 4 projectors to cover all the possible angles (front/back left, front/back right) assuming you're projecting onto a curved surface for greater realism. Or two if you've got some fancy optics in the way.
The further from the user the projectors are the less comfortable the HMD will be- and the less practical it'll be too (as you'll end up with a pair of moose-antlers swinging around you as you're playing a game).
Alternatively, I guess you could use a set of first-surface mirrors to get the neccesary distance from the screen by bouncing the image around the user's head. You'll still have a pretty hefty HMD, though.

The only resolutions listed for the microprojector (that I've found) are pretty poor- using a microprojector would be a definite backwards step for HMDs in terms of quality.

Anyone thought any more about headtracking or the interface with the video source?

[smoothy]

uberdeity

Many thanks for your positive and highly interesting input. Unfortunately I do not share your same opinion on how the microvision pico projector wouldn't work well. I can tell you that microvision have designed the picop projector to be so small at the size of a mint as they want it to be embedded into HMD's. They are working on their own mobile display glasses. You can see it here http://microvision.com/wearable_displays/index.html" onclick="window.open(this.href);return false; You do not need as many pico projectors as you say to get it inside an HMD. You could still get stereo 3d from one projector and even though the image will be wrapped around your field of view it would be so immersive that you not having stereo wouldn't be so bad. It would actually be amazing stereo or not stereo. I have seen demonstrations of the pico projector up close and you can fill a curved piece of a4 paper with the projector being about 3 inches from the screen. Now add optics in there and I am sure microvision will get it projecting in millimeters in their mobile eye wear. Of course if we were to build one we would take the consumer graded model apart and just take the mint size projector and mount that. It's power and video signal could be sent down a fiber optic line into a computer. This is actually quite clever. There was an apple patent several months ago circulating the net. They wanted to do just this, make HMD's less bulky by putting all their power, signal and other electronics that are found in conventional HMD's and placing it into another device that connects to the HMD by a fiber optic cable, This HMD apple is designing is based on a mini RGB laser display and using wedge optics (googling that will show that it can increase the field of view greatly in HMD's). Now we need implement these feature because it would greatly reduce weight from the users head which is quite an excellent idea on apple's part.

[ngsmith]

I looked into the microvision pico-projectors and out of all the other pico-projectors coming out they appear to be the best solution, since they stay focused at no matter what distance you have them even when projecting on to curved or angle surfaces which is pretty cool. The only downside is the resolution is only wvga. If they are not extremely expensive they could rival other low-res hmd, but I would like to work on having at least 720p resolution-especially since you are stretching pixels over a curved surface.

[TheLostBrain]

Sensics HMD
I believe each cell is an 800x600 display. What I'm not sure is how obvious the seams are when you're wearing it.

Bastian74

Unfortunately, in my testing I always found the "seams" caused by the tiled fiber optic tapers very noticeable. Even after spending quite a bit of time fitting and re-fitting their xSight model I was never able to attain an alignment that produced a "seamless" view.

I've worked with many HMDs and of all of the optical assemblies I've sampled this one was unfortunately, to me, one of the most disappointing.



To everyone...

Keep in mind that there are many attributes to good HMD design that go beyond Field of View.
In many ways good HMD design is about finely balancing compromises.

For example:

You may design an HMD with an endless array of adjustments (IPD, Eye relief, Diopter (focus), Optical assembly elevation, etc...) in an attempt to fit every possible user, but for this you generally trade expense, weight and most importantly, ease of use. All but the simplest of adjustments are very often double-edged swords - many potential adopters of HMDs are frustrated to non-acceptance by just the "fitting" process itself.

On the flip side you may instead decide to design an HMD that's very easy to use by forgoing any adjustments at all. However, you've now limited your potential market to only those with physical characteristics that fall within the optical tolerances imposed by your "fixed" design. Also, keep in mind that those who sit on the outer edges of those tolerances generally can't take part in extended use as HMD related fatigue can even be (and often is) caused by optical alignment disparities that are visually unperceivable to the user.

Now, let's take a look at the attribute of HMD design that everyone always seems to crave more of - FOV (Field of View). While I'm not going to attempt to discredit the importance of a large FOV when it comes to maximizing the sense of immersion, it's important to be cognizant of the surprising number of trade offs that are imposed by having one. In addition to the commonly know historical challenges of optical design complexity, cost, and weight there are also many other less commonly mentioned, yet equally important, negative impacts. To name three that immediately come to mind: The decrease of visual acuity, the increased chance of simulator / motion sickness, and in cases where you have large field of view but lack headtracking - an increase in eyestrain caused by excessive eye movement.


I've got quite a bit of experience using, repairing and rebuilding HMDs. I've personally owned multiple highend units including an N-Vision Datavisor Hi-Res, a Kaiser Electro Optics Proview XL-35, and most recently an N-Vision Datavisor 80 (as seen in my Avatar).

I've also thoroughly studied HMD design and have designed and built a working large FOV unit, of which I'm currently in the process of patenting. Once a patent is issued I'd be more than to happy to share the design with the community for the DIY'ers who'd like to take on such a project. Also, anyone interested at that time may also opt to purchase a pre-built unit as well.

As for the idea of the laser based projection HMD, it's a thought I've toyed with myself and in my opinion while it does have some optical design challenges that have to be overcome, it's a great concept that has huge potential. I would highly recommend continuing the collaboration on the idea and as time allows I'd be happy to contribute as I can.


For those interested... my current daily use VR configuration consists of the HMD prototype mentioned above and an Ascension Technology Flock of Birds unit with an ERT (Extended Range Transmitter) for 6DOF head tracking. In non-"VR ready" applications I use the IZ3D drivers and also glovepie for emulating a TrackIR, etc. w/ the Ascension-Tech FOB.

If you'd like to read about how I came across the Datavisor 80 shown in my Avatar you can learn more here:
http://www.thelostbrain.com/post/2008/0 ... play!.aspx" onclick="window.open(this.href);return false;

You can also view a walk through of some simple repairs done to it here:
http://www.thelostbrain.com/post/2008/0 ... -new!.aspx" onclick="window.open(this.href);return false;


-TLB

[crim3]

I'm nowhere near an expert, but I still know that they screw up the image slightly (according to a 10 second wiki search, they cause "a reduction in image quality") as they're not a "proper" lens.

Do you mean that 10 seconds wiki searches is all that is needed to develop HMD optics?

[uberdeity]

I'm nowhere near an expert, but I still know that they screw up the image slightly (according to a 10 second wiki search, they cause "a reduction in image quality") as they're not a "proper" lens.

Do you mean that 10 seconds wiki searches is all that is needed to develop HMD optics?

No, but it's all that's needed to find out why no-one uses Fresnel lenses. Which was the information you asked- why no-one uses them.

Come to think of it, you can see the optical problems it causes without searching the web- just think about, say, the lights on the front of a car. Or the lens in front of the light in a lighthouse. You can actually see the bumps in the lens. Would you like to use a HMD where there were hundreds of small scores (each about the size of a couple of pixels, given the size of screen HMDs use) across it?

Why did you want to use Fresnel lenses?


Also, @Smoothy

Sorry, I'd not seen them used in mobile device eyewear, and there seems to be a lack of any actual technical information on the projectors available. The only information I had on how short a throw they needed was "a 40" screen from a few feet away", which is impressive but still means a couple of inches for an HMD size screen.

How would you get stereo from a single projector (given that splitting the image in half- one half for left eye and one for right eye- would result in a resolution about the same as a Gameboy)? Is the refresh rate high enough for a shutterglasses based solution (120Hz or above)?

Coherent fibre bundles would be pretty expensive to make, and- as Lostbrain said above- if you have even a tiny border around them you'll notice the breaks. So you need a single bundle for each screen.
Can you imagine the size of the bundle required to manage a decent resolution of screen? 720p in both eyes would mean 1,843,200 fibres. Even a 1.5m cord from headset to base unit would mean you'd need 1,717 miles of fibre- and you'd need a machine to sort them into the correct order. All 1.8 million fibres. With a diameter of 1micrometer per fibre they'd still be arrangeable into a ribbon cable 6ft wide- and that's without any sleeving (which would be requried on each fibre).

Plus, with a WVGA resolution from the projector you'd still need 2x per eye to get a half decent resolution (800x960 or 1600x480).

[smoothy]

I looked into the microvision pico-projectors and out of all the other pico-projectors coming out they appear to be the best solution, since they stay focused at no matter what distance you have them even when projecting on to curved or angle surfaces which is pretty cool. The only downside is the resolution is only wvga. If they are not extremely expensive they could rival other low-res hmd, but I would like to work on having at least 720p resolution-especially since you are stretching pixels over a curved surface.

My understanding is when you have a projector that can beam to curved surfaces. While always staying in focus it wouldn't really matter if you had a wvga or 720p projector. As in this case to use it in a HMD because the picture the projector would beam to would be so much smaller. There would not be an issue with stretching the pixels around a curved screen. Now on the price issue, the microvision is said to sell in the second half of this year for $400-$500 But they said that the cost of embedding one of these projectors into a mobile phone would raise the cost by $100. Now if we can buy these from microvision at trade price that would be incredible, but probably to do that you would have to become an OEM. And we could only do that if we had a proof of concept for a laser based HMD. Perhaps better just buying the consumer version later this year and turning it into a HMD.

[crim3]

Why did you want to use Fresnel lenses?

I know that I don't explain myself well, even in my own language. Even worse, reading again my post I see that I drift away from what I wanted to say at first. And it was that HMD's just don't use fresnel lenses, so someone who knows about optics should help in this project. But then I began to think that to test the pico-projector a temporary solution could be fresnel lenses of some kind. When you begin a project from zero you can't do a first test with everything working.

[smoothy]

Coherent fibre bundles would be pretty expensive to make, and- as Lostbrain said above- if you have even a tiny border around them you'll notice the breaks. So you need a single bundle for each screen.
Can you imagine the size of the bundle required to manage a decent resolution of screen? 720p in both eyes would mean 1,843,200 fibres. Even a 1.5m cord from headset to base unit would mean you'd need 1,717 miles of fibre- and you'd need a machine to sort them into the correct order. All 1.8 million fibres. With a diameter of 1micrometer per fibre they'd still be arrangeable into a ribbon cable 6ft wide- and that's without any sleeving (which would be requried on each fibre).

Plus, with a WVGA resolution from the projector you'd still need 2x per eye to get a half decent resolution (800x960 or 1600x480).

Here is a site talking about the apple patent http://www.appleinsider.com/articles/08 ... plays.html" onclick="window.open(this.href);return false;

Here is a cut and paste about the fiber optic plan they have

'As such, the company suggests a HMD apparatus that separates the laser engine from the image generator via a fiber optic cable so that the laser engine can be physically decoupled from a headgear.'

'Additionally, the fiber optic cable coupling the detached laser engine to the headgear could also be configured to include an audio line that provides audio signals in parallel with the video images. As such, the head mounted display apparatus may include integrated earphones capable of transmitting audio signals to the user's ears. The audio line could be an electrical line or an optical line, Apple said. "In some cases, the optical line used to transmit the RGB laser light is also used to send audio signals (e.g., multifunctionality).'

So if apple can make it happen, I am sure us DIY'ers can as well

How would you get stereo from a single projector (given that splitting the image in half- one half for left eye and one for right eye- would result in a resolution about the same as a Gameboy)? Is the refresh rate high enough for a shutterglasses based solution (120Hz or above)?

It's my understanding that when doing stereo 3d it does not half the resolution. Also say the hz is 120hz you don't get half at 60hz per eye, both eyes get 120hz. Someone I think on these boards explained this to me a while back or it may of been some time ago on the stereovision.net boards. The graphics card draws the same scene twice, so say 1600x1200 that's drawn twice but the refresh rate stays the same for each eye.

I saw that http://www.headplay.com" onclick="window.open(this.href);return false; they've created an HMD that is 1024x768 it uses a single LCOS display and a splitter that splits the image in to 2 separate images that are stereoscopic 3d. We just need to use their splitter.

[uberdeity]

So their patent is sticking a piece of fibre between the laser modules and the light-combiner / mirror sections? Sounds like a step that would be obvious to someone who knew about these things (which should have made it unpatentable)- hell, I use hardware like this every day at work (CWDMs- they let you send multiple frequencies of light up a fiber at once, but we normally just use IR frequencies as they're cheaper and just as fast).
So Apple are sending the red,green and blue stuff up the fiber along with some sort of audio signal- probably not "proper" digital audio, more likely an analogue waveform that would be converted to sound really easily and cheaply with a well-chosen phototransistor, a few resistors and a small amplifier.

Yes, that would solve the problem. I thought your idea was to have the laser shine onto an array of fibers that would transfer the light straight up to the headgear.

With 120Hz screens and shutter-glasses, each eye gets 60Hz. Frame 1 goes left, frame 2 goes right, frame 3 goes left, frame 4 goes right etc etc etc.

If you took the output of a single projector and wanted to get stereo from it as an image per eye without the headache-inducing shutter glass system, you'd have to send half the pixels to one eye and half to the other- meaning each eye would get 400x480 pixels from the projector.

[ngsmith]

I was looking at the Dell M109S and I think two of these would be good candidates for projected HMD-they are fairly light- wearing two (0.8lbsX2) would not be heavier than wearing a helmet on your head, and they would also allow a svga resolution which is a step up from microvision-they are also available now for $450. Here is Dell's info link: http://accessories.us.dell.com/sna/prod ... u=224-0939" onclick="window.open(this.href);return false;

I also found on Ebay plenty of cheap sub $300 XGA projectors that could be dismantled for DIY projected HMD.

[SlyWax]

If you want to solve the problem with the lens. Just use a semi opaque contact lens, and project the image on it.
You will be totally immersed.
The problem would be then to track the position of the eye to project on it, and to change the image according to where you are looking. And it need to be done fast.
The resolution won't matter much because you will get it from the eye tracking ( human vision is really clear in the center, the surrounding mostly only detect movement ).

But obviously this couldn't work for the mass market. Cause they wouldn't bother to wear contact lens for that. Except maybe people who already wear them every day...
And for a DIY it may be too dangerous to play with laser on the eye !!
But maybe you idea is too dangerous to. Imagine there is a scratch on your rear projection little screen, you just blinded yourself !

[smoothy]

I was looking at the Dell M109S and I think two of these would be good candidates for projected HMD-they are fairly light- wearing two (0.8lbsX2) would not be heavier than wearing a helmet on your head, and they would also allow a svga resolution which is a step up from microvision-they are also available now for $450. Here is Dell's info link: http://accessories.us.dell.com/sna/prod ... u=224-0939" onclick="window.open(this.href);return false;

I also found on Ebay plenty of cheap sub $300 XGA projectors that could be dismantled for DIY projected HMD.

microvisions pico projector is 848x480p, I know that's not SVGA but from the specs I just read on that Dell projector you posted, it's a lot bigger than the microvisions one and it's a 480p resolution. Just like microvisions. What you guys are failing to recognize is the fact that microvisions picop is light, high contrast, no need for a large throw distance, doesn't need mirrors, optics or fans which means it draws less power, silent to run, doesn't heat up, and is over all very small while at the same time can project to curved surfaces. So if we were to go down the route of projecting to a motorbike helmets visor then you would not need any mirrors or optics. Which means it ends up being cheaper to be build and less bulky and more streamlined. The fact that it doesn't need focusing or any object it beams to means that it's the most user friendly and best way to create a wrap around screen for an HMD over all other display systems including other pico projectors.

[ngsmith]

I have no interest in defending Dell's projector, but according to how I read the spec sheet it has native SVGA resolution. I am not seeing on the Dell spec sheet where it says the M109S is limited to 480p (720x480 at 16:9 or 345,600px). Dell is claiming native 480,000px which is 800x600 resolution although theoretically if the actual resolution is 858x600 that is 514,800px. That is significantly more than microvision claims 848x480 = 407,040px. The advantage of having a native resolution of at least SVGA is that WinXP and most games or software can run it with out any problems and no resizing or upscaling involved.

My main reason for suggesting the use of the M109S is coming from the idea that a DIY HMD would be more applicable to the community if it can accept many different types of pico projectors. A more modular design that DIY'ers can do minor tweaks to fit their own pico or small projectors into I think will be very beneficial.

[mayaman]

what ever happened to TDVision? I email them about 5 times over the years to buy some HMD and they never return my email. I want the 720p goggles but looks like they never sell anything. how do they stay in business? Now I went and bought an new hdtv with the nvidia goggles. ugh

[Okta]

How do you go about projecting to a helmet visor? Would you use a special coating?

[smoothy]

How do you go about projecting to a helmet visor? Would you use a special coating?

I would get that special white color projection screen paint and just paint on the exterior of the visor for rear projection or onto the interior for front projection.

[Okta]

How do you go about projecting to a helmet visor? Would you use a special coating?

I would get that special white color projection screen paint and just paint on the exterior of the visor for rear projection or onto the interior for front projection.

It would be interesting to see but would the eyes be able to focus on such a close image? I think this is why all hmd optics make the object look more distant. Maybe using small credit card size fresnel lenses would be usefull but also ghetto becuase of the lines.