Meant to be Seen

foisi wrote:New version :

IMG_1182.jpg

IMG_1183.jpg

[youtube]http://www.youtube.com/watch?v=oICGUy2xzNw[/youtube]

note that there are black pixels at the bottom and at the top of the panoramic image I used to make something displayable with the HMD so vertical FoV is not as it should be in the video.

I didn't paint the inside in black (or tape black paper) for the moment, I just rebuilt the box in a few hours today at lunch time and this evening just to try to prove people that the design is possible.
I took about 20 pictures when I was building the new box so I will be able to make some sort of tutorial if anyone is interested in making a similar HMD (I think it deserves a name, any suggestion ?) and I will release the printables very soon.
My mom who happened to be visiting me today tried the HMD when I just finished the new build and she said something that could be translated by "It's f***ing awesome !" ^^

The new weight is 512g (~50g less than the previous version, the PMMA lenses are heavier than the expanded PVC)

Wow! Really impressive stuff!

I seriously NEVER would have thought Fresnels would have provided an acceptable level of quality...but from the video I feel I may stand corrected.

Can you share (maybe I missed it?) exactly which fresnels you're using now (where to get em, etc.)? I may actually come out of diy-retirement to build one of these to compliment the Rift when it arrives.

Thanks and awesome work!

foisi wrote:@Zoide: Maybe Dycus could make a prototype when I release the build instructions to have Palmer admit that it works ^^

I am going to build it myself, and I really do hope it works.

Hi,

Just Google " aspheric fresnel lens ", and there is plenty of great design and supplier information.

Too easy.

Thanks.

I admit that I don't know much about optics, but shouldn't it be possible to use appropriately designed fresnel lenses in a Leep-like manner, or perhaps in a fresnel/standard lens hybrid setup? It would be amazing to make a compact and light weight high-FOV HMD!

PalmerTech wrote:

foisi wrote:@Zoide: Maybe Dycus could make a prototype when I release the build instructions to have Palmer admit that it works ^^

I am going to build it myself, and I really do hope it works.

I just played with my Nexus 7 and my 5x acrylic aspheric lenses, and also with my fresnel magnifiers. When you hold the lenses parallel to the display surface, it is all in focus. For angled displays as shown here, you just need to angle the lenses, which makes them cover even more of your FoV. Even though I only tested HALF of this display concept, it worked well. If I had another 7-inch tablet it would be a better proof of concept, but I am convinced that it will work with fine focus.

For comparison, I also played with my fresnel stacks again, with a single 7-inch display (my Nexus 7). It really needs to use one-third of the width (about two inches) of each side of the lens as I described. Because the two lens stacks are identical for both eyes (other than trimming to fit nose and eyebrow ridge), there should be no problem with assymetrical focus or stereo convergence as suggested elsewhere. So I believe that my experimental method can also produce a similar wider FoV as the one in this thread (but only estimated 150-degrees or so horizontally). My approach will not give the quality of the one in this thread, but it still give huge FoV and will costs half as much (or almost nothing if you already have a Nexus 7 or equivalent). Of course, there will be latency issues though, so it remains mostly for experimental use. And the lens stacks need the ridges toward the eye, unlike the version in this thread.

Based on past research (in the days when DIY HMDs were very low resolution) it was common to put a diffuser sheet (such as waxed paper) over the screen to minimize screendoor effect, so finding a compromise focus should be fine even IF not all pixels were in focus. Small blurring objects are easier for the brain to recognize than sharp blocky objects, which I verified while adjusting the focus and using the "wide-eye" quake image pair from the Rift Quake thread.

CONCLUSION: I am confident that the HMD in this thread will work well for superior image quality, and my method will work too for a quick 7-inch tablet FoV2Go clone using fresnel magnifiers.

MSat wrote:I admit that I don't know much about optics, but shouldn't it be possible to use appropriately designed fresnel lenses in a Leep-like manner, or perhaps in a fresnel/standard lens hybrid setup? It would be amazing to make a compact and light weight high-FOV HMD!

It should be possible to simulate the effect of wearing eyeglasses in Real Life, where the solid acrylic aspheric lenses provide central view, and the fresnel outer area provides peripheral view. If you see the border between the lenses, it should look no worse than eyeglass frames in RL. And even a shift in focus or distortion should at that border should be simlar to wearing Real Life eyeglasses.

foisi wrote:About foamcore (ca veut dire carton-plume c'est bien ça?) can you tell me the density g/cm^3 so I can compare to expanded PVC ? Don't you think foamcore is a little weak to resist from tearing at the mounting points ?

Yes, foamcore is carton-plume, I don't know the density but it's very lightweight. I was also worrying about the mounting points with it, but I think it can be hardened with some other materials. After all there shouldn't be too much efforts at the attachment points once it's in place.

foisi wrote:Oh and Foisi is a terrible nickname (I will not explain why I chose it), it sounds a lot like "moisi"

I didn't think about that, may not be the best term then...

geekmaster wrote:But then "EyePhone/iPhone" was used in the film "Johnny Mnemonic", so who knows where Apple got their inspiration for that product name

EyePhone was foremost the name of the HMD released by VPL research (of Jaron Lanier fame) in 1989, so it may really not be a good idea to use this name.

geekmaster wrote:

MSat wrote:I admit that I don't know much about optics, but shouldn't it be possible to use appropriately designed fresnel lenses in a Leep-like manner, or perhaps in a fresnel/standard lens hybrid setup? It would be amazing to make a compact and light weight high-FOV HMD!

It should be possible to simulate the effect of wearing eyeglasses in Real Life, where the solid acrylic aspheric lenses provide central view, and the fresnel outer area provides peripheral view. If you see the border between the lenses, it should look no worse than eyeglass frames in RL. And even a shift in focus or distortion should at that border should be simlar to wearing Real Life eyeglasses.

What I had in mind was something more along the lines of http://www.vrtifacts.com/hmds/leep-on-the-cheap/ where a series of lenses are used to create a large FOV from relatively small displays, but possibly replacing one or more of the lens layers with light-weight fresnels.

Hi Foisi,

Congratulation for your work! It looks very promising !!
2 months ago I tried something similar (as well as TheLostBrain design with beamsplitter miror) but stopped because of lack of spare time and software support at this time. And thanks for the 'castorama' tips for pvc sheet! I used pvc pipes flatten with a heat gun to build my case, not very practical nor pretty
Maybe I will resume the work soon now I see it looks like a real alternative.

Bravo!

In terms of chromatic aberration, is there a coating for the lenses one could get like we currently have for eyeglasses to get rid of reflections and glare that might help in addition to even better polishing and smoothing of the lenses from the manufacturer? Would that make it feel as if the lenses are barely there?

@TheLostBrain: I'm glad a guy with such talent at making DIY stuff is interested to come out of diy-retirement after seeing something I designed
the fresnels are the same as in the first post : http://optolife.com/diy_projector_lens/FRL021.html

@everyone:
I'm currently finalizing the plans for the release (and it takes a longer time than I thought), what do you think about the Creative Commons licence CC BY-SA for it ?

The CC BY-SA looks more than fair enough

foisi wrote:I'm not sure to understand why people tend to believe that focusing would change depending on the position of the eye... for what I know the focus is set by the distance between the lens and the screen. This distance is the same for the whole FOV because the lens is parallel to the screen. There is chromatic aberration (that can be corrected in software) and spherical abberation (because the fresnels are not aspheric lenses) but it doesn't have a significant impact because the eye is near the optical center

Yes. There is a possibility to predistort two images that are enclined too (in the two stereo pictures or 3d game images from enclined virtual camera's) as described in previous comments. This way, every optical distortion would be resolved completely.
I wish you a lot of success with your awsome project.

Software. This is the revolution we are seeing.

Before Palmer went public with the Rift many of us including Palmer discounted this and the Rifts designs because of the lack of non 100% overlap and warping support. When Palmer presented the Rift to Carmack I think a tipping point was reached when Carmack coded the warping in to Doom 3.
Then with the blitz of the Rift Kickstarter Cybereality and MaterialDefender and many other game devs began their "Rift" wrappers and games. The next hurdle was reached when Oculus announced the the 7 inch screen. This broke my heart because of the move away from 100% overlap but I slowly realised that none of the software guys had borked at it. It really wasn't a big deal to code in the overlap shift after all

Now the sky is the limit for new designs like this if in theory wrappers like Vorpx could add support without to much difficulty. The next problem is market fragmentation. All the new games coming designed with Rift support may never be revisited to include support for a new design like this if it takes off.

Very nice foisi. I feel I don't even need to get back to the shed with so much ground being covered now

Okta: I see you have a home brew high FoV 2D HMD in your signature. Any plans to try building foisi's design? I'd love to hear other people's impressions of it

Zoide wrote:Okta: I see you have a home brew high FoV 2D HMD in your signature. Any plans to try building foisi's design? I'd love to hear other people's impressions of it

I would love to build it and think its a great design but time is against me. I originally set out to build a HMD with 2x7 inch displays side by side or 2x pico projectors relying on optics to keep 100% overlap but decided there were too many caveats. The Lost Brains Brains similar design is an amazing achievement but it took a lot a work and the end result is a little bulky IIRC. The big/little thing the Rift has going for it is low cost, low weight, high/ish FOV and a little Luck (pun intended). Those are the reasons it is/will be so successful.
It is possible to build another design like foisi's here but...being bigger and more expensive will be a road block in uptake. The the dreaded software support is required, and without a large take up that is a problem.

Having a Rift coming has made me a bit lazy too, been thinking more about VR controllers than displays as the Rift has become a kind of default for most of us and now we have seen a bigger focus on the other aspects of VR and some great results.

I have mentioned a few more ideas I would like to build in older threads...maybe one day.

Guys, I have edited the first post with the current release

@ Okta: you are absolutely right. I've been thinking the same thing for quite a while, PC's have become fast enough to do the predistortion in software, instead of requiring lenses to undistort the image. This is a very large part of the revolution.

I know what you mean about not wanting to work on HMD's with the Rifts coming - there are so many areas of VR that require work that its probably best to focus attention on the parts that are truly lacking currently. In many ways, this is why we are very lucky to all be VR hobbyists/enthusiasts - its one of the few science areas where people without any formal training, but with great ideas, can actually make a real difference and achieve something new and special (like the Rift! - of course, to make this difference involves learning a LOT about the areas you are working on).

foisi wrote:Guys, I have edited the first post with the current release

Perfecto! Thanks for the detailed instructions and pictures foisi!

Unfortunately i don't have the time to test it myself (busy enough with my own DIY project), but i can't wait to hear the opinion of others who try it, especially TheLostBrain and Palmer!

Also i forgot to ask before.. what do you think is the lowest screen size that could work with this type of design, while still maintaining a wide FOV? Could it be adapted to use a couple of 5 inchers for instance? I guess by bringing the screens closer to the face somehow?

I have two 5.6" LCD at home that I previously bought for this project, but since I read here that they are not produced anymore, I updated the project with two 7" instead.
I also have 4 spare fresnel lenses of the same type (FRL021) so I can try to see how is the FoV with this configuration this evening when I get home if you want.

foisi wrote:I have two 5.6" LCD at home that I previously bought for this project, but since I read here that they are not produced anymore, I updated the project with two 7" instead.
I also have 4 spare fresnel lenses of the same type (FRL021) so I can try to see how is the FoV with this configuration this evening when I get home if you want.

Well i think that would be a very interesting experiment indeed!

I believe there are lots of screens that could be used potentially (especially with a custom board) and it would be awesome if we could get similar FOV results but with a lighter and more compact design.

http://vr.wikinet.org/wiki/Displays

Especially now that the market is getting flooded with ~5" 1080p displays.

Great plans, foisi! What application did you use to draw them? I feel honored that you chose to go with the name InfinitEye


I'm really curious how small the display can get, but I don't want to drive this thread too off topic so I'll start a new one in the R&D forum - I hope you'll join in with your thoughts and insight.

@MSat :
I'm using Inkscape, a free and open source vector graphics editor : http://inkscape.org/index.php?lang=en
Thanks again for the name, it sounds really good

How much stereoscopic overlap is there in real-life human vision? It looks like the InfinitEye's overlap is much smaller than that of the Rift. Could changing the angle of the InfinitEye's screens improve overlap at the expense of a bit of FoV?

@zoide:
IIRC it is between 90° and 120°, it depends on the IPD and the size of the nose (mine is ~90°)
You are right, it can be modified at the expense of monoscopic peripheral vision. To get an idea of what is ~60° overlap (if you have ~90° stereoscopic vision like me, you can place a finger along your nose bridge and look around, to me it's not that bad )

foisi wrote:@zoide:
IIRC it is between 90° and 120°, it depends on the IPD and the size of the nose (mine is ~90°)
You are right, it can be modified at the expense of monoscopic peripheral vision. To get an idea of what is ~60° overlap (if you have ~90° stereoscopic vision like me, you can place a finger along your nose bridge and look around, to me it's not that bad )

That is more than the 60° overlap from the catadioptrical hmd of H. Nagahara.

Curious, are the PMMA fresnels that much better (optically) than the standard PVC ones?
Ex: http://www.3dlens.com/shop/fresnellens.php

You and geekmaster have both been doing some pretty cool work in this area. Nice job!

A quick question from the hardware under-informed: are OLED screens significantly lighter than fluorescent-lit (or even LED-lit) LCD panels? Anyone have round ballparks by how much? Obviously they're not yet common or cheap for hobby projects, but I was curious (especially with weight being one of the primary downsides for a dual-display system).

TheLostBrain wrote:Curious, are the PMMA fresnels that much better (optically) than the standard PVC ones?
Ex: http://www.3dlens.com/shop/fresnellens.php

PMMA is a bit better. It becomes a notable issue when you stack three of them. Go for coated, obviously. 6 sides is 0.5% per side as a cumulative(? IIRC), so it's not pretty, in the end. It's resolution and contrast loss, then add in a cheaper plastic. Ugh. The kind of smear to be encountered with PVC, might make me want to clear the gunk out of my eyes all the time.

The problem is the loss of contrast in low contrast scenes. I'm sure many of you folks know this already, though. IIC, the eye is stated to run at about 350:1 CR, but in practice it is higher than that. digital devices have problems reproducing contrast, not so much of a problem in the CRT end of projection. digital is only now getting close to what CRT could do, so it's not such a good idea to loose any on the lensing or in any other aspect. Every bit counts.

If you plan to slap them into a shaping mold and curve them, then the PMMA is probably the way to go. It's good enough for eye use, so it should be good enough here.

I've done large RP and front projection experiments (dual sided as well) with both types of plastic and find the PMMA to be the better choice there.

Regarding patents (reading your sig), I filed one in 2008, it only made it through recently. Things are a bit backlogged in the system, to say the least.

I found this http://www.greatwalloptical.com.hk/opti ... llens.html :

Optical plastics are widely used to make Fresnel lens. PMMA is the best material for fresnel lens as magnifier.

I asked the ebay store "chinatobby" if they can make extended LVDS cables (length 1m~1.20m) because I'm really not good at soldering.. The guy answered that he will check first if it works for these lengths and he asked how many pieces I would need because the cost may be higher for samples.
If anyone else is interested in these extended cables please tell me, then I can tell him and maybe the price can be lowered a bit. (it's for the 7" LCD that other people use for their DIY Rift too, I'll make a link to this post in the DIY Rift thread if the results of his testing are OK)

foisi wrote:I asked the ebay store "chinatobby" if they can make extended LVDS cables (length 1m~1.20m) because I'm really not good at soldering.. The guy answered that he will check first if it works for these lengths and he asked how many pieces I would need because the cost may be higher for samples.
If anyone else is interested in these extended cables please tell me, then I can tell him and maybe the price can be lowered a bit. (it's for the 7" LCD that other people use for their DIY Rift too, I'll make a link to this post in the DIY Rift thread if the results of his testing are OK)

Snap! inquired about this today from this store too.
I would be willing to buy at least one but possibly 3.
He mentioned that the default length was 25cm which Is longer than what the picture on ebay suggests.

Great work foisi, sounds like you're on the way to solving some of the difficult problems. Thanks to geekmaster as well for inspiring folks to look at fresnel stacking as a direction to look in.

blazespinnaker wrote:Great work foisi, sounds like you're on the way to solving some of the difficult problems. Thanks to geekmaster as well for inspiring folks to look at fresnel stacking as a direction to look in.

foisi had his lenses just sitting there ready to stack, but my fresnel lens stacking thread was his motivation to get off his duff and actually MAKE his HMD. And ideas aren't worth much until you make something out of them. foisi did a great job here.

My experiments show that we can build an HMD with the same FoV as this dual display HMD, but with only a SINGLE 7-inch display, by using a 5-layer deep stack of curved (dollar store magnifier) lenses that actually wrap around your face from ear-to-ear, touching your skin all the way around. It requires 10 lenses total, with two lens stacks that meet in the middle between your eyebrows. The downside is that stretching the pixels to those extremes (with partial stereo overlap) really demonstrates our need for LOTS MORE pixels, and we need a much more complex pre-warp algorithm. This is just an experiment for now, but it shows some promise. And stacking lenses that deep is a poor tradeoff, sacrificing image quality for virtually total FoV. That makes this design mostly an inexpensive experiment with maximum immersion than a real product. Because the display is most of the cost, a single display can cut the cost in half. This crude prototype design demonstrates the real possibility to build just such a device with a single layer of a custom wraparound lens, but so far that is just a dream.

foisi's dual-screen HMD is simpler in some ways, and provides twice as many pixels to work with, and uses fewer lenses giving better image quality, and it is a REAL thing that you can build NOW (which is all that really matters)...

Hey foisi, were you able to test this design with your 5.6'' screens in the end? Or are you still planning to do it?

No pressure, just curiosity!

I have been toying with the idea of building a HMD and was quite excited to see this version. It has quite a high resolution using the 2 screens and has an incredible FOV but I really cannot get my head around how it would work in practice: going on the image above - to focus on something in the middle of the left screen (roughly where the two pink colours join) then it appears obvious the left eye follows that path. To focus on the same point on the right screen (again roughly where the blue colours merge) then the right eye appears to need to diverge drastically from the left. I would imagine this would not only be quite difficult to do but also extremely uncomfortable. Or do I have it wrong?

I admit I am quite a noob, having tried to read everything on this site for just the past few weeks, so apologies in advance if this is a silly question

Note that in the drawing, I used the 7" display to scale the Rift but with the 90° HFOV 100% overlap which is less likely (I think it will be something like 115-120° HFOV with 80% overlap but we'll see in march)

Isn't the Oculus Rift still unfairly represented in that image?
The horizontal FOV (per eye) seems less than 90 degrees in that picture, but I didn't measure it exactly.

@mahler : I checked the svg file and you are right, the represented FoV is 85.5° instead of 90° here is a corrected version : @PatimPatam : I'm going to check the FoV with the 5.6" right now

@mrgrey : the light pink and light blue colored areas are only monoscopic, don't worry the eyes don't have to diverge
maybe you can get a better idea on how it works with a 180° panoramic picture (from google images) that I modified to be viewable on the HMD (and I added the blue and pink colors to match the drawing)

foisi: Looking good! Any news about Palmer? Some time ago he posted that he would experiment with your ideas.