FOV Modification???

[kylelefevre]

I'm new to High FOV HMD's, but am trying to get a greater grasp on how they work. As far as I've been able to ascertain so far, the biggest limitations at present to highly immersive displays is resolution, FOV, and refresh rate. We can take refresh rate out of the equation for now.

There are obviously HMD's presently that have high resolutions. And as far as I can find so far FOV is primarily controlled by the optics.

Would it be possible to modify other HMD's with high res displays to accomodate a larger FOV? Obviously it wouldn't work in the units the displays are housed presently, but through making your own you could fit any lenses necessary.

Has this just not been tried or is the issue more around distortion...or something else?

Wondering what everyone's thoughts are.....

[Chriky]

I might be wrong here so people can correct me but from my understanding of optics it's basically like this...

To have a wide FOV you need actual light rays coming into your pupil from a wide angle. Almost all HMDs have very small screens that it is practically impossible to blow up into a wide FOV. The RIFT uses a comparatively much larger screen, so you don't need to bend the light from it so much.

Hopefully this pictures helps...

[kylelefevre]

Yea, I've been reading alot about the RIFT and it seems like an awesome project! I'm on board for the kickstarter once it launches!

So the RIFT using one lcd is splitting the screen into two parts, one image for each eye...correct? What is the size of the LCD he is using?

With that question, wouldn't two small lcds positioned for each eye have the same effect? (as the HMZ-T1 has)

[cybereality]

Just get on the Rift train. You are not going to do any better than that for that price.

But to answer your question, the issue with modding existing HMDs is that they use very small micro-displays, and thus are not a good fit for large FOV. The Rift uses like a 5" screen (split in 2) so it can afford to use larger optics which make for the high FOV. You can also look into "Leep on the cheap" but this is if you can build the whole HMD (not just mod a consumer product).

[Okta]

The big current hurdle is suitable displays. If there were high-mid resolution 2.5inch screens we would be happy. Two of those would be near optimal for a DIY 3d hmd. But there isnt. So here we are.

[ERP]

The big current hurdle is suitable displays. If there were high-mid resolution 2.5inch screens we would be happy. Two of those would be near optimal for a DIY 3d hmd. But there isnt. So here we are.

It's not just the displays, iphone screens would be dirt cheap and great, but there is no way to drive them.

[Okta]

The big current hurdle is suitable displays. If there were high-mid resolution 2.5inch screens we would be happy. Two of those would be near optimal for a DIY 3d hmd. But there isnt. So here we are.

It's not just the displays, iphone screens would be dirt cheap and great, but there is no way to drive them.

Yes, i didn't indicate but that was also what i meat by suitable. If iphone/htc screens could be used they would need a special mod also to only use a portion of the display as they are too big to use as a pair.

Back to the OP, i think Crikys picture above really nails the issue with current HMD's.

[kylelefevre]

So why don't we find some individuals who can program control boards for those displays....

I can't believe it would be that difficult. I've seen much more extreme diy projects done in communities. When there is a will there is a way.

Honestly, we could probably utilize social media (linkedin) and find some hardware/software developers that work with these displays and work on convincing them to help us!

Then RIFT could have a high resolution display!

[brantlew]

I'm speculating that it would require some special tools and hardware skills as well. Integrating and testing surface mount components on a board. Something that only electronic hardware hobbyists would feel comfortable doing. Now maybe there is somebody out there that could make a little money selling custom built boards as a Rift upgrade - but this level of assembly is probably out of reach for the average hobbyist.

[AntonieB]

I think we shoulld start something here and combine our efforts to find out how to drive phonescreens. I guess it is pretty hard but who knows.


But are there much beter phonescreen options then the rift is using now? talking about a single screen setup

[Chriky]

No there's nothing better on a single screen, and the screens are slightly smaller than you would want. For an HMD we'd use two separate screens. A tablet screen could be used by itself I guess.

But I totally agree there should be a community effort to drive the Retina display. A control board that took DVI, HDMI or VGA in and plugged into a Retina display on the other side would be gold dust to several hobby-tech communities. It could be a successful Kickstarter. The problem is there is no obvious place to organise around. What we need is a single passionate person with a lot of the skills and knowledge needed.

Maybe we should start pinging off emails to guys who actually work in manufacturing for displays, just to get a definitive list of the problems we would have to solve to contruct the HDMI-2-Retina gizmo. It is my very rough understanding that you could potentially do it with a very expensive FPGA (if you were a sufficient genius), and then you could get the FPGA made into a dedicated board if you had a big enough order size.

I don't know much about it but seeing those incredible screens selling for about $50 is serious motivation.

[Fredz]

I'm not sure using retina displays would be that much of an improvement for the Rift. For now it uses a 1280x800 panel which gives 640x800 pixels per eye, while two iPhone 4/4S would give 640x960 pixels for each eye. So you wouldn't gain anything in horizontal resolution and only 20% more in vertical resolution.

Considering the actual costs, a 5.6" 1280x800 LCD panel with a VGA/DVI/HDMI control board can be found at $125 + $45 = $170 on ebay, while the cheapest used iPhone 4 I could find was at 2x300 = $600. That's $430 more for a 20% overall resolution gain, and still without taking into account all the cost involved in the development of a board to drive these screens.

It'll also be a lot more problematic (if possible at all) to sync two displays than using one single display. And you'll loose the almost perfect 1:1 pixel aspect ratio of the current panel (0.98 for a 90°x110° FOV) when retina displays would give a 0.81 pixel aspect ratio (ie. not squared pixels).

To me that doesn't look even remotely interesting...

I guess it's probably wiser to wait 1 year or 2 for new displays to arrive, such as the Toshiba or the Sharp 6.1" 2560x1600 panels, which give a 300% overall resolution improvement.

[kylelefevre]

Well I think there is some truth to what is being said in both directions. Honestly, we are really close to being able to use these cell phone or tablet LCDS, but maybe not 100% there. You can get iphone 4s displays super cheap though...I'm not sure if they are actually OEM even though they state they are.

I have another question for the group. What is the largest size screen that would be reasonable for a 1 per eye setup, and for a single display setup?

I would imagine optics and weight play into this thought...but lets take weight out of the equation at first.

This is some fun dialogue!

[PalmerTech]

There is technically no limit to what size you could use for a single screen. The ideal size is about 6", but 7", 8", or even 10" would work just fine. The goal is to have something with as high of a PPI as you can while still retaining a manageable size.


Using iPhone displays would not really make sense at this point. You only gain a small amount of vertical res, you would need to make sure that both displays were aligned absolutely identically to each other, and you would have to drive them using two perfectly synchronized inputs, which is tough to do with the way that most GPUs put out an image (Most have a frame of lag between outputs).

Several other people here have pretty much covered it, but to recant: Driving these displays is hard, and often done in a proprietary manner. To drive a Retina Display, you would need to reverse engineer the signal protocol, figure out how to drive it with expensive FPGAs, and then drops tens or hundreds of thousands of dollars to convert that design into an inexpensive mass-produced ASIC. Along the way, you need to hope that nothing you do violates some obscure patent held by other companies, possibly even the one you are sourcing the display from (Apple). If you can get someone to cooperate with you, say, Toshiba, your job becomes easier... But you still need a huge chunk of money to cover the initial costs. Even if the skilled labor needed could be sourced for free, making an ASIC for a cheap consumer HMD is not trivial.

The display in the Rift is the absolute best that can be used in a low volume project like this, unless people don't mind the price ballooning by a few thousand dollars. I can hang around and wait for better displays to come out... Or do this right now, and raise money so that perhaps next year, a run can be made at using those new displays that are coming out. It is on the roadmap, but first steps first.

[Chriky]

@ Fredz

You're right they actually wouldn't be much better quality. I suppose you might use them the other way around which would probably make game compatibility a bit easier, and switch horizontal resolution and FOV for vertical, which some people might prefer.

They are much cheaper than buying a whole phone though, for example this one is £20 - http://www.ebay.co.uk/itm/Genuine-Apple ... 2a1b438ff8

--

From the way different technologies are advancing then it looks like there might be some displays coming on the market in the next year or so, and after that tablets are going to start begin fast enough to render on board.