Widescreen Gaming Forum

On the subject of projectors, I wonder if there are some type of OEM projectors that don't include all the bells and whistles of file playback and such.

Are you trying to use projectors that are already on the market? Or would you be designing projectors for it? If the latter, you could design them like modules and be able to support any number of projectors that could physically fit in the assembly.

_________________
System Core: | Intel Core i5-2500K + ASUS P8Z68-V + 16GB Corsair XMS3 DDR3 1333 MHz | Win7 x64 | MSI R7970 Lightning 3GB [1105/1400] |
Display: | 3 x Dell Ultrasharp 3007WFP-HC @ 7680x1600 | Dell u3011 |

I am a bit out of the loop with the actual system( but I can give opinions on the finished product.

1 Anything in the 120-140 range is fine. 120 is not a floor either simply the beginning of the sweet spot to hit. Anything lower and you lose the full range of view, too high and you have extra screen you have to strain into your field of view.
2 How many games actually use it? A serious question, I really don't know. This wikipedia article (http://en.wikipedia.org/wiki/List_of_st ... ideo_games) seems rather sparse to put extra money on a feature with limited use.
3 Your average desk is ~30" deep and ~55" wide. You still need room for mouse, keyboard, speakers. And some people still like their desks to be desks. My current setup with 3x24" monitors leaves me plenty of room to spare on a desk that size.
4 Output quality is going to trump price, however, you are in a niche market. You do not want to price out too many people. A fairly low-mid end eyefinity setup is gonna run you around 700 dollars.
5 Again, niche market who generally do all their own building as it is. We can handle it
6 Assuming if we are putting it together, we can also take it apart. But portability should not be a consideration if it damages the overall product. It should be solid as a rock once it is set up. If sacrificing portability is a causality of that then so be it.
7 True to life. I assume few people go triple wide solely so they can see whats coming in their peripheral vision.
8 Pretty much the same as #5.
9 I play everything with shooters being my most played genre.

#1, 3, 4 are the big factors in determining your market. Get the quality, footprint and price right and you will have a winner

1) As others have responded, I agree that ~135 deg is what you should be shooting for. You definitely have to take current game support in to the equation and already 135 deg turns off a lot of people (not me!) because of stretching on the side monitors. Going with a larger FOV is only going to make this worse so I agree that 135 is the "sweet spot"

2) Yes 3D is important. Again, game support is also a concern here as many games have broken 3D support. That being said it is definitely an excellent feature and the ability to use 3D or not curtails any disadvantage. The main reason for surround gaming is immersion and 3D only adds to that in my opinion. We're also getting to a point where GFX cards will be able to run it with decent framerates at the resolutions you're spec'ing this display for.

3) Here I'm going to say ~6' left to right. That's my current setup and around 2-3' from screen to viewer position is what I use and don't have any issues straining to see. On unit depth I should think you'd want to aim for "as narrow as possible" and make what ever concessions are needed to display correctly.

4) This is tough to answer. I personally have around $750 in my 3 displays, $150 in the stand, so competitively that would put you at around $900-1000 which in all honesty doesn't sound feasible. Depending on fit-and-finish though you could double that and personally I'd be comfortable with it however 6mm plywood doesn't sound like something that appeals to me. 1/8" aluminum does though, or a laminate-covered 6mm plywood with a few different finishes available. As far as making the "plans available" for do-in-yourselfers, 6mm plywood sounds great however as a finished product I'd be aiming a little higher than that and charging for it. It IS worth something to the customer to have a quality look/quality built product.

5) Self assemble. Packaging and other factors would make ready-to-rock prohibitive and for anyone purchasing such a display, they should be able to follow well thought out directions for assembly.

6) Personally, the less you have to move such a setup the better. If the boxes and plastic foam shipping pads that come with the product are well-made and usable over and over again (read --> NOT STYROFOAM) then you can move the setup in its original packaging without much thought. Personally a folding setup would compromise the rigidity of the setup.

7) True to life. To reiterate, the competitive advantage is there anyway and immersion is what we're seeking (at least me) Image quality and response time are key here as well.

8) This would be a nice feature I suppose however for me, I would have my setup a certain way and once its set up I don't want to mess with it. If I need a single projector for watching movies or something else that would be a separate purchase. I agree with other posters that on either a 3x1 or 3x2 setup it would be nice to be able to turn off certain displays and give the set up some flexibility in that regard.

_________________
i7 sandy 4.4ghz with a snorkel
1080gtx
8gb RAM
Gigabyte P67-UD7 rev. B3
OS -Samsung 840pro 256GB ssd
Games -Samsung 840pro 500GB ssd
Pagefile -crucial 60GB ssd
Storage -2x 4TB WD
PWS -Seasonic Prime 1k Titanium

When you think about it, all flat monitor based systems are inclined to do this, i.e represent a wider angle on screen than in the physical field of vision (feel free to disagree, share your setups dimensions!). We are so used to it that we don't think about it. This provides a gaming advantage, dragging the peripheral vision into the primary field of view. Interestingly, the peripheral vision is wired to be more responsive to movement (I've heard!), to give us time to react when a predator approaches from the side, for example your boss when you're busy reading reviews on WSGF...

To offer what I'd describe as true geometry of vision the screen should show things life-sized, and at the same angle to the eye as they would be in reality. That's always been a goal of Infinitx, though it may be misguided. The theory is/was that the brain doesn't have to work so hard to create the immersion. Call it removing the brain lag

This is the human field of vision if you perceive it as a hemisphere rolled flat out, just like a world map.


If we plot the area of field of vision covered by the 135 degree arc screen, you get something like this.


And the initial goal of infinitx from a mathematicians point of view..


Worth noting that here we're talking the field of vision without moving the head, which adds another 90 degrees even when sitting.
(Here's some better illustrations http://www.dsource.in/course/display-design/module-2/vision/vision.html)

As you all know (and I'm slowly being persuaded) we don't need total coverage of the human field of view to feel immersed. 135 degrees is clearly leading the polls so far.



Here we start getting into the Economics of Scale. If we order 300 projectors (that's 100 infinitx screens), we can generally negotiate a better price. If we order 1000 projectors then we can begin to define their specification, and have something custom built. Some minor alterations (like removing tv tuners, analog inputs, card readers etc.) might not require such volumes. If you've ever bought a projector from ProjectorDesign, you'll know that you can get just about any projector specification you might require, but make sure you're sitting down before you ask the price. We are relying on using hardware designed for the mass market for the foreseeable future, and chose a partner who is willing to hear our suggestions for future versions.


First I should mention, in the current design there is no plywood visible when the screen is fully assembled, only lush black and clear acrylic from the front, and a flame retardant stretch fabric shroud on the back. Why fabric? It facilitates heat dissipation, dampens sound and allows air to pass through the cabinet, while filtering out dust.


hi Haldi Crowdsourcing is what we're doing right now, i.e. developing the device together with the community it's designed to serve.
Crowdfunding is definitely an option later, and would certainly help to reach critical mass.

Please keep your comments and suggestions coming. Already I feel we're focusing in on the ultimate gaming screen

Sounds like a great solution Thanks for elaborating on it.

_________________
i7 sandy 4.4ghz with a snorkel
1080gtx
8gb RAM
Gigabyte P67-UD7 rev. B3
OS -Samsung 840pro 256GB ssd
Games -Samsung 840pro 500GB ssd
Pagefile -crucial 60GB ssd
Storage -2x 4TB WD
PWS -Seasonic Prime 1k Titanium

Whoa, did not know the marginal cost going from WXGA to full HD for LED projectors was so steep. LED projectors are a fairly new technology, right? In that case, it may be advisable to wait a few years for LEDs to become competitive with conventional lamps. The reason I say this is because as LuckyNoS7evin alluded to earlier, 1080p is really the standard du jour; I doubt there are many in this community who would be willing to sacrifice more than half of their current pixel count, even for a bezel-less 3x2L 3D display. This is especially true considering that for about the same price as single full HD LED projector you can build a setup like this. I still think this project holds a great deal of promise, however; it offers the best of most everything one could want in a display. When I build my first rig several years from now, Stromberg Industries will definitely be on my radar.

ahhhhh thx for your detailed explanations.

So this would be the Hardware FoV if there is no picture shown on the Screen! But what about Games ? They need to show the FoV thats supposed to fit! YOu might have heard about the console games on a 50"TV with 60° FoV which seems the same as a 23" PC Monitor with 90° FoV because you're much nearer to the PC Monitor. But some Games are bad ported to the computer and end up with a FoV of 60°! Same goes for Far Cry 3. In Eyefinity we had some Problems with the FoV stuck on 110°.
So if you stick a 110° FoV image on this (maybe) 135° Wall it would look totally off ?

The 2x3 180° Wall sure would be awesome for Flight Simulation, for people who have over 250 buttons to press and joysticks to handle! For highly optimized Systems/Games/Simulators. But for the average gamer i'd say the Occulus Rift provides better immersion (mainly because it's around 300$).

So the FoV also depends on how far away you are from the Monitor! on a 180° Setup i assume you're in the middle and the distances is set by the radius.
But how does that work for a 135°C You can sit nearer or further. To far away and the FoV decreases and the picture gets smaler? To near and the FoV is to big and the Items on the screen become to big ?




About the 720p/1080p discussion i think you're gonna have a hard time explaining why 720p Projectors are better suited than 1080p. Sure the price does a lot on that. But people will say "what ? only 720p? No way!" If you offer a 1080p version they will say "what? so expensive? no way! i'd rather buy 720p projectors"
I don't think that many people will be willing to spend so much cash for 1080p Projectors. but just having it will smooth the community and decrease bad advertisement like "this system would be absolutly perfect if it only had 1080p"
So the question would be. Is it worth developing a 1080p version even tough rarely nobody is willing to buy it due to the huge cost?



And sorry, i mixed up Crowdfunding with Crowdsourcing. It was late night/morning... to much crows!
But Kickstarter is definitly a good way. Not because of the funding, but because of the advertisement you get.
You can reach all the fans of tech gadget who don't visit Gaming Blogs every day but are willing to spend their money.

_________________
We gonna send it to outa space!

Those are very valid points Haldi, and surely sum up well the viewpoint of many gamers.

If the Rift proves one thing though, it's that resolution isn't the be all and end all of immersion. After all, in it's current incarnation (the good lord only knows when a consumer model will emerge) is a mere 640 x 720 pixels per eye. http://www.youtube.com/watch?v=xpV7qq1vyd4.

I have a few other misgivings about displays that render the user blind to the real world. The gamers I've observed (mainly my son Jack) play for hours, if not days on end and have chat windows open, they eat, drink, smoke (you just dare Jack!) and answer the phone while gaming. Not easy with a monitor strapped to your face. (maybe I'm just a little jealous of their development budget )

In reality (and virtual reality) resolution is a number. Numbers are the easiest things to compare, but they don't tell the whole story. They don't describe the experience. Otherwise the Rift, and the Nintendo Wii for that matter, would have been dead in the water.

A 3x2 Infinitx would provide you with a resolution equivalent to a 100" 4K TV, both curved, and at a fraction of the cost.
Take for example the Samsung's 85". A snip at $39,999! http://mashable.com/2013/03/22/samsung-4k-tv-s9-price/

To infinitx, and beyond!

Hey there. I'm a 5x1P user with screens that wrap around to an approximate 100 degree arc. in terms of coverage, that's about proper. If you're going for immersion, I suggest 170ish. Maybe 165. But, I also warn of this sort of setup. Games are NOT designed to fit onto curved displays and the camera in the software would have to be warped specifically to fit it. The 3D element of games have always, as far as I've seen, assumed a flat display that can grow off until infinity. This limits you to 180 degrees inherently, but that number is not achievable and you'll have to follow a diminishing return law when considering the vertical as it moves towards the edges. This comes out looking roughly eye shaped with sharp corners near the left and right edges, in which your view is largest in the central portion. This is how regular screens work.

Set up 5 TV's in a row, match the FOV of the game to your FOV of the screens (usually by sitting the appropriate distance), and you will discover that there is nothing being warped from that perspective. This, again, is how the game expects things to be. With curvature comes warping, which reducing the immersive aspect of this system. And I get the feeling that immersion is the whole point of it. What they call "fisheye" only happens when the screens are turned inwards to wrap around the player, and it's a well known phenomenon. A curved screen would only exaggerate this.

So, onto the questions:
1.What is the optimal arc for a wide screen gaming display? (Should there be different options?
180 degrees as a maximum, 90 as a minimum.

2. Is stereoscopic 3D important to gamers? (Does anyone actually use it?)
Yes, but only gamers interested in immersion. FPS gamers do not tend to like stereoscopy. Racing enthusiasts are hit or miss with the technology, as it can give you a better feel of the speed of the car. But stereoscopy doesn't work properly with curved screens, it's designed for flat screens. Check it out yourself if you don't believe me, but it's going to mess itself up. Millions of dollars have been spent in getting the math right for all of the flat displays out there, not the tiny tiny handful of curved displays.

3. What’s the maximum acceptable physical size for such a display?
In this crowd, surely this number doesn't exist. Let's just say that it has to be small enough to fit into a small country.

4. What should it cost? There are several places where it’s possible to compromise on materials and components.
Projector costs included? 1500-2000. The screen setup by itself? 500.

5. Which weighs heavier, the price or the output image quality?
Image quality. Seamless transition between projectors is important as is having everything in focus. If projectors hit a screen at an angle (and with your curved screens, most lenses are designed to focus from a flat plane, not a curve, so you might want to get new optics for this specifically) then parts of the image will be more in focus than others, and this could be a bit jarring up close.

6. Should the screen be self assembly (think IKEA) of ready to rock?
Both options would be good. You could design it for self assembly but sell a fully assembled version that would make up for the cost of shipping and labor, plus a little extra profit. Having a self-assembly option from the get-go gives you both options for selling. Some people would rather save money, some people would rather not have a hassle. I'd rather save money. My whole computer, with 5 screens, cost less than 2 thousand. It took a lot of work getting it together, but that money I saved on service could be put into getting better stuff. If you have a larger version available that's the same cost as the pre-assembled lower model, I'd definitely get the larger one.

7. Is portability a factor that should be considered? Should you be able to pack the screen down either for travel or storage?
I imagine it would be very difficult to design this into what your'e looking for. I suggest holding off on this idea and designing the first version without this in mind to save on costs for both yourself and the consumer. The majority of users won't be interested in this, and the handful of users that would be are also probably used to and willing to deal with complex setups/teardowns. Again, I have 5 monitors. I know what I'm talking about when it comes to frustrating teardowns.

8. What’s most important, a display which is “true to life” (geometrically) or one that gives you the biggest advantage over your competitors?
True to life, easily. What's the point in covering the majority of your vision if you aren't trying to trick your brain that it's actually in the game? Keep the geometry right and leave the advantages with the 240Hz LCD displays. I don't need that stuff to play Skyrim or race between trees in Dirt. Keep that stuff pretty.

9. Should the projectors be removable for use interdependently or in other setups?
An easy release system might have other benefits as well. Not to mention that it could make the setup more modular. If you have it designed in such a way that you could choose between 1 and 6 projectors in whatever configuration you'd like, that would suit the largest number of people. If you design an interlocking system for them, I suggest designing a method to stack them vertically for those people that would like to run 6 of them in a 3x2L setup.

_________________
(imagination not allowed in signature)

Thanks Zencyde for a most informative and in-depth analysis.

I'm aware the one-point perspective devised in the Renaissance and still used almost ubiquitously in 3D engines is inadequate for wide flat screens, when considering the FOV's and viewing distances desirable for gaming. I know you have a very firm grasp of this scenario.
So you have elected to arrange your monitors in a 100 degree arc, in the knowledge that a straight line is the only appropriate position for standard perspective 3D rendering, ignoring stretching in favor of a more encompassing display. Would that be fair? So overall feel triumphs over geometric accuracy, and conversely geometric accuracy doesn’t necessarily give the best feel.

I think we need to rename what’s being called the “fisheye effect” to describe objects stretching towards the edges of the screen. It’s actually the lack of fisheye effect that is being referred to. In a fisheye, objects become smaller towards the edges.

Here’s an example using a 3D fisheye simulation in live 3D. The cars are in a straight line in front of the camera.


Having this type of lens within a game engine opens up for a type of immersion rarely seen outside of expensive custom built/programmed simulators.

The brain is apparently willing to chip in and interpret our primitive attempts at visualization into a virtual world that makes sense, not because it's accurate, but because it's consistent. The geometry/screen/eye/mind pipeline is indeed an interesting one, and probably more straightforward than it sounds. The screen introduces the errors and the mind accepts or rejects the tolerances. We can tune many titles for FOV, but as you rightly point out, apart from a handful of titles (iRacing, any others?), a flat viewing surface is universally assumed, and a barrier for experiencing 3D that wasn’t specifically designed for a wrap around screen.


Side screen as rendered on the infiinitx D 135 degree

I don’t foresee the issue being exaggerated by a single curved display contra an array of monitors, unless of course the physical arc is far greater.

So there’s a tug-of-war going on between immersion and existing gaming titles. That challenges one of the main aims of the project, that is...



Any ideas? I’ve seen hacks done like this one...
http://strlen.com/gfxengine/fisheyequake/

If true immersion, i.e. seeing things beside you that ere beside you in game environment requires mods, it's maybe a more acceptable compromise than making the curved screen flat!

Hello,
I find your proposed project quite interesting. I would say that I would be a potential customer for such a product and as such would like to submit my views on your questions. I have a custom sim-racing rig in which I recently upgraded the display from a triple-screen to a more capable triple-screen. The previous display used three 19” Viewsonics for a 3840 x 1024 display driven using a Matrox Triple Head2Go. The new display uses three BenQ 24” (XL2420T), generating a 5760 x 1080 display, driven by two AMD 7970 cards in Crossfire. Before going for the upgrade I spent quite a bit of time looking at the options available for a projection system and the available components, such as the projectors, screens, blending software, etc. Clearly, I concluded that staying with triple monitors was the best choice for me at this time. A dedicated effort, such as yours has the potential to be, might change the results of my calculation. You can bring resources to bear, and take on such challenges as folded optics that I was unwilling or unable to address. Below please see my answers to your questions and a few thoughts on some additional issues.

What is the optimal arc for a wide screen gaming display? (Should there be different options?)

180 degrees is probably overkill. There is likely (to my mind, anyway) a sweet spot somewhere in the 135-150 degree range. I find that the side monitors don't just provide immersion and a sense of speed. Overtaking situations show their real value as I can see the car beside me well enough to better judge when (and whether) to turn in. Much less accidental contact is the result and better racing. The aspect ratio of the three monitors (48:9), coupled with my FOV (26 degrees vertical in rFactor2), means that many things I might like to see are out of the field of view. I use TrackIR to overcome this, and it also helps in looking to the sides.


Is stereoscopic 3D important to gamers? (Does anyone actually use it?)

I haven't tried it yet. My monitors are 120 Hz which I got not only for their 3D capability but also for the faster (hopefully smoother) frame rate for 2D. Some sim-racing 3D users have reported that 3D is not just eye-candy but helpful in judging distances to opponents, curbing, etc. For now, I am struggling to get frame rates up to 120 fps so 3D may have to wait for the next generation of video cards. I remain hopeful and think this would be good to have but maybe not absolutely essential.

What’s the maximum acceptable physical size for such a display?

Because I have a dedicated rig I naturally think of what fits well with that. At a rough estimate, maybe 5' wide by 2' high by 3' deep. These would be the unit bounds, not the display size.

What should it cost? There are several places where it’s possible to compromise on materials and components. Which weighs heavier, the price or the output image quality?

Image quality. For sure, I understand the tradeoff. Now, I realize that image quality is a function of more than just pixel count but it surely starts with that. I am unwilling to accept an array of 3 x 1280 x 800 projectors that result in a 3840 x 800 display. It is surely possible to use five projectors (in portrait) to achieve 4000 x 1280 (which starts to have sufficient resolution AND helps on the aspect ratio issue), but this may blow the budget. When I was contemplating my choices, I tended to focus on the available cost-effective LED-driven units (e.g., the QUMI). 1920 x 1080 projectors were either projector-lamp driven (which brings in heat, power, and size issues) or were way, way out of my budget possibilities. Three powerful monitors and a decent monitor stand will quickly get you into the $1500 range (if not more). I would hope that you could get the price in a range of $2K - $3K.


Should the screen be self assembly (think IKEA) of ready to rock?

Turning wrenches is fine with me, especially if it saves some dough.


Is portability a factor that should be considered? Should you be able to pack the screen down either for travel or storage?

Not a big driver for me. Could be a part of the DIY aspect of the previous question but I wouldn't like to see a lot of cost go into supporting this feature. See me discussion of reliability below.


What’s most important, a display which is “true to life” (geometrically) or one that gives you the biggest advantage over your competitors?

I don't understand this question.


Should the projectors be removable for use interdependently or in other setups?


Not important to me.




Additonal thoughts.

I am already thinking of my next sim-rig project. I would like to add motion. To do this properly one should make the display part of the platform that is moveable. This leads me to think of the following:

Weight. Keep it as light as possible. Less mass for the actuators to drive is all to the good.

Sturdiness and resistance to shock, vibe, etc.

Mounting. The unit should be easily mounted to standard or custom monitor mounts. Maybe VESA, I have not given this a lot of thought, but maybe you have.

Reliability. The unit should of course be highly reliable for its intended environment.

Power. It would be nice to keep the power budget down for all of the obvious reasons.


I suspect that my needs are somewhat outside the target clientele you are considering. Still, I think there are a not inconsiderable number of sim fanatics like myself. Thank you for listening to my long-winded thoughts.

Dave Bradley

Hi David, thanks very much for your thoughts.
As a simulator builder you are definitely within the target audience, and a very welcome addition to the discussion.

This particular size of screen is thought of to fit on a desktop, or as a close proximity display on an car, or cockpit simulator.
I should probably add some illustrations to emphasize this.
The LED projectors appear to be very robust. At least, we are constantly moving them without switching them off, with no issues. and the whole set-up is comparatively light.
Whether I would recommend strapping one to a fast moving force feedback chair, I'm not sure. I would probably use a larger static screen, with a moving cockpit within.

The "true to life" question is about evaluating the users requirement for a display that attempts to mimic reality, in relation to many other factors.
If "true to life" were a top priority, it dictates many other parameters. What we're trying to achieve is the best balance between all the factors that affect the final design, and ultimately the price point.

Cheers!

Thank you for the reply! I'm glad to see that you're well informed on the subject, and I'm also glad to hear that you've had experience with other curved setups. I've actually argued this issue on the Internet (arguing on the Internet? never) in which the other party, also focusing on curved displays, decided to call me out for pointing out the geometric issues with it. So again, thank you for being completely aware of the limits. I'm also glad to hear that you've done research into modifying game engines to fit better. This sounds like an awesome potential piece of tech and I'm excited to see what it might lead.

Since we're already going to be dealing with wild FOVs and warping with the support coming from the Rift, which by design tears the interface away from the world so that the two are now separate entities, do you supposed some of your curved-screen support might make its way into this sort of system automatically with Oculus Rift development?

I also foresee the Razer Hydra benefiting from this, I'm mostly hoping for a sweeping change in the way that graphics engines are handled, so maybe we can start seeing support for more accessories or strange setups in general.

Again, this looks super awesome. If you want a beta tester, let me know. I have a Steam library with over 500 titles ready to be tested.

_________________
(imagination not allowed in signature)

While we are obviously tracking the progress of Rift, I don't imagine any overlap at present. The Rift is very low res, with the 720p screen being divided in two, half for each eye.
On the other hand, games made for Rift may indeed include some new and useful parameters, both HUD and FOV wise.


The Hydra does look interesting, especially if you were standing up in a FPS, and certainly cheaper than an omni-directional treadmill! We've got one on order...


I'll keep you in mind

Cheers!

I would HIGHLY Be Interested in th 180degree with 3D setup..

IF using a curved screen as someone may have already pointed you you need warping software and I have the perfect solution it is called nThusim..you can check out that sight and the multiple curved screen setup thay have there along with mine I use the same name here as there..

I had looked into the trident as nice as it is it is really expensive.. I am hoping this solution will be in the relms of cost effective..

I just dont see it happening.. the material you are talking about for a Rear preojection is costly then you will need first reflection mirrors.. I dont see how this is going to be an cost effective sloution..

I really hope I am wrong but to get a good rear projection you need the proper material.. I am unsre what type of material you are talking about I hope it is not spandex..

I will kkep this blog linked to keep up with the progress and hopefully see it come to pass..

Hope you can also address some of my concerns..

I've been working on exactly what you are trying to do for a couple of years.
My youtube page has some shots of what I've done: http://www.youtube.com/user/VanFS3K

Short throw is also my goal, but its harder than you think.

I haven't purchased my final projectors yet, just test units. Some 1080p LED projectors come out this fall that I'll have my eye on. The plan is to have 5x of these in portrait. I've learned the best setups of projectors for this kind of thing is either 2 in landscape or 5 in portrait (16:9 or 16:10).
3 projectors in landscape is just too wide, so you'll either lose horizontal real-estate because the screen sizes are too small or most of the vertical screen will be out of your vision because the screen sizes are too big.

Let me know if you have any questions.

PROJECT UPDATE
Finally we have our large bed laser cutter installed


..and tested...


This means we can build begin building physical prototypes.

I can reveal we've opted for the LUXCINE C7 for our initial test set-ups. As well as being affordable, LED (20000+ lamp life), it features diamond shaped pixels and a virtually invisible grid. The result being you can hardly detect any pixels at all. Another pleasant surprise has been that despite it's 500 ANSI Lumen rating, it's plenty bright enough at the 100" screen-size we envisage for the Infintx D, but also under controlled lighting can drive a MUCH bigger screen if you have the room to spare. For example a 3m diameter, 2.4m high screen- That gives you space for 6 people, or a full-sized car or cockpit.

BRAIN STORMING...

(NOTE: The following is just an idea we would like to discuss, and not necessarily a design proposition.)

One or the most obvious challenges with Widescreen set-ups, that everyone runs into, is the "side stretching" of graphics.
It's a widely misunderstood phenomenon and therefor fills considerable pages on this very forum.
http://www.wsgf.org/forums/viewtopic.php?f=68&t=14082
If you want to see what I'm talking about here's a simple demo.
http://www.strombergindustries.com/newPerspective/

...and if you don't have Flash installed here's screenshots.


When you set your FOV for a game to an extreme value (100 degrees plus), you will generally see objects distorted towards the edges. They appear elongated.
This is not an error, but actually because of the mathematical precision of the 3D engine.
If you try the above demo, with your nose up against your monitor (actual distance depends on monitor size), you should notice that the ball looks round from center to extreme left/right, as you are looking at it from the angle it was calculated from. The camera in the 3D engine, the place where the view is calculated, is at that point in 3D space, not way back where you typically have your favorite gaming chair. So why don't they move the virtual camera out to a more natural viewing position? Because that automatically reduces the field of view, which is the reason you lashed out on a triple head in the first place. Here's the simple, illustrated explanation.



So why rake over that age old problem?
We're looking for the optimum shape for the infinitx screen. There's nothing to say it has to be a regular arc. There was an idea to make a flat middle monitor with curved "wings". We're already bending the output image so as to appear correct on the cylindrical screen, why not go a stage further and distort the image to look natural on the screen from a human perspective, rather than mathematically correct, from a flat computer monitor perspective.

It would be possible to build a reflector system that shows a 16:9 area normally in the center of the screen, but then gradually bunches the image up towards the edges, while bending them inwards, compensating for the exaggerated field of view.
In theory at least, that would give you a set-up that shows non-widescreen titles and movies correctly, and super wide fields of view in a more natural looking way.

The screen would look something like this. The checker pattern shows how the wings gradually compress the edges to keep round objects round, when set to the appropriate FOV.





If that makes sense to anyone, I'd be glad to hear about it. Also if it's completely off the mark.

Cheers!

As I mentioned above, 3 in landscape is too wide. Even though it mostly depends on the games scaling capability, you almost always get the "fish eye" effect with 3 projectors. Unless you plan on creating a room with 180 degree projection, avoid using 3.

Wow, 2 posts while I was writing that one

Nephalim > There is no software warping involved here, though you would have the option of edge blending if you need it.
We have invented and patented a mirror array that takes care of that, and costs less than most warping software. The benefits are that it is compatible with all versions of DirectX, OpenGL, and doesn't add lag to the rendering pipeline.

Screen tech is low reflection acrylic, coated with high contrast, light dispersion polymer to remove hotspots.

We are making larger screens in 180 degree format. This particular model is designed for desktop gaming. So far 110 to 135 degrees seems the most popular choice for that application.

Snipe 3000 > Cool stuff
I think maybe we can help on the short throw issue. The screen described here uses a projector with 1.22 throw ratio (not very short throw) and still manages to pack it into 36cm depth. Our larger displays will use genuine short throw projectors (0.5 - 0.7 throw ratio), and manage a depth of 60cm, for a 4.5m wide screen.
The light folding reflectors take care of warping. Aspect ratio is affected, but can be easily rectified.

There's a bit more info about where we're coming from on this thread. Unfortunately half the thread got wiped in the last db crash, but you should get the idea...
http://www.wsgf.org/forums/viewtopic.php?f=72&t=25274

Cheers

I'm very curious to see this work with more than one projector. It would definitely be nice to not need the software. The software's biggest problem isn't price, its compatibility. I have two of them just in case one has an issue with a certain game.

It seems you are using mirrors? Most high quality rear projection screens are glossy on the back side, not completely reflective, but there is still some reflection. How are you guys compensating for this?

I think the following would work better for your "build up as you go" players:

Make all parts modular and then offer them both individually and as packages (possibly with discounted costs). Since this is for the advanced gamer, but will still be easier than building your own H2O cooling setup, it shouldn't be too hard for the people that will actually buy this. Like most gamers, you don't just go out and buy four video cards at 500 bucks a piece, you buy two and then get a third on another paycheck, and a fourth on another.

For instance:

Part Codes break out by revision series and type
P = Projectors
R = Reflectors
S = Screens
1 = Revision 1 (in a few years, maybe do a revision 2 with new technologies)
17 = 17" Size (where applicable, otherwise just extending the next value)
0: last digit (or last 3 digits if size isn't applicable) be an individual part number

Part P1000: Projector LowRes, LowQual $50
Part P1001: Projector LowRes, MedQual $75
Part P1002: Projector MedRes, HiQual $250
Part P1003: Projector HiRes, HiQual $500
Part R1000: Reflector $50
Part S1170: Screen, 17" 4:3 $25
Part S1171: Screen, 17" 16:9 $30
Part S1190: Screen, 19" 4:3 $30
Part S1191: Screen, 19" 16:9 $35
Part S1230: Screen, 23" 4:3 $50
Part S1231: Screen, 23" 16:9 $60
Part S1270: Screen, 27" 4:3 $80
Part S1271: Screen, 27" 16:9 $90
* Prices are estimated, scale to whatever you need to cover costs.

17" 4:3 setups are cheap
Bundle B1170: Single Monitor 4:3 17" setup, includes P1000 x1 ($50), R1000 x1 ($50), S1170 x1 ($25) = $125 - $5 group discount = $119.99 Retail
Bundle B1171: Triple Monitor 4:3 17" setup, includes P1000 x3 ($50x3), R1000 x3 ($50x3), S1170 x3 ($25x3) = ($125x3=375) - $25 group discount = $349.99 Retail

27" 16:9 setups are expensive
Bundle B1270: Single Monitor 16:9 27" setup, includes P1003 x1 ($500), R1000 x1 ($50), S1271 x1 ($90) = $640 - $15 group discount = $624.99 Retail
Bundle B1271: Triple Monitor 16:9 27" setup, includes P1003 x3 ($500x3), R1000 x3 ($50x3), S1271 x3 ($90x3) = ($640x3=1920) - $50 group discount = $1869.99 Retail

This way you can cookie cutter setups together, people can start with 1 or 2 monitors and work their way up to 5 if they want. Screens should have an individual adjustable arc method, using a knob to apply tension (like a bow string bending a compound bow, but obviously not). This way if they can adjust each side however they want to suit their visual needs. In my case, I would probably want flatter middle and more curved sides.

What projectors have you looked into? You mentioned $3000 projectors, but the Dell M110 is designed for short throwing, can do 1280x800, is LED based and is only $419 retail. I'm sure that if you turned this into a major design you could get them cheaper through a Dell partner program.

Have you considered doing an OLED screen instead of rear projection? Grow it to a size/width that would be three monitors and just have a single flowing screen without messing with rear stuff? I don't know how far the technology has progressed on this...

I would be interested in one of these displays. When will they be available for sale?

So when you have a Lens corrected Screen, and then you get something like this:

(Unigine Engine)

It wouldn't work anymore, because you have Software presets for curved screens to work perfectly, and a curved screen with corrected Lens, what would end up horrible ?

And diamon shaped Pixels ? Aren't pixel supposed to be quadratic ? Guess it's time for me to read some more guides about Projectors! Anyone can offer some good guides about Projector technique ??

Edit: Man those Projectors look cool ^^ Android OS ^^ the 1080p version ? I don't get it, what is a native 1280x800 resolution and then you have a expandend 1080p ? but the Airplay/DLNA pat is cool! 3 Projectors and you can beam photos from your Smartphone onto your Monitor you don't need atm. Miracast would be cooler.

_________________
We gonna send it to outa space!

For Portability, sell a "Portability Pack" at a reasonable price:
A. Clips to make the screens mount to each other under break-away mounting. Like Push In, release lock, slide screen back.
B. Reflectors should collapse directly into the back of the screen
C. Back Clip for the Reflectors to solidly mount a projector
D. Handle on the top so you can carry it like a briefcase.
E. Carrying case of various sizes that can hold 3 monitors worth of a setup as well as power cords, etc.
F. Clips that will allow the structure of the Screen and Reflector to be used as a stand for the Projector to point it at a wall for presentations. I could see this being an advantage for small offices where they would want to take a projector into a conference room and use the wall.
G. With this now being so portable, ensure that the plugs work well with TH2Go setups so a person could take their laptop, TH2Go and carrying case to a LAN party and use it however they like.

Touch Support
There are projectors out now with Interactivity capabilities, like the Dell S500wi. This uses a Pen/Wand as a mouse for presentations by tracking the wand position real-time. I believe next gen will actually remove the wand and start tracking actual hand movements. Since this is going against screens on a desk, you could just use touch sensitive screen panelling (like what you would clip over a normal screen to make it a touch screen) as an optional add-on. MagicTouch (Keytech, Inc.) make solutions like these for sale.
http://www.magictouch.com/addon.html

Vertical vs Horizontal Surround Support
Make sure your software and hardware can handle the displays being 9:16 vs 16:9 setups. I personally didn't like the way things were when I did it myself, but others seem to love it.

Custom Sizing
Offer the Screen makeup in square footage (or some size) with instructions on how to build your own screen size. So if you offer up to 27" and a customer wants 31" (some odd number) they can buy and cut and build the screens to what they want and then tune the projectors sync software to what they want.

Thanks Jothay, lots of good insights and ideas there. I think you've hit the nail on the head with a component based ecosystem.

There would still be place for a "standard" set-up, but a web-shop selling components that let you build your own would be a very good idea, allowing hobbyists to build within their budget, and to their specific requirements, and add to the set-up over time.

We can use the same projector cradle with lots of different sized screens, so that's def an option. That allows you to increase your screen size without having to invest in new projectors, or to have 2 different set-ups, maybe one for sitting, one for standing, and use the same hardware.

There are several ways of adding touch interfaces, the cheapest/easiest being infra-red/camera. It's not a priority from our point of view, as touch tends to be less and less practical as screen size increases. Kinect, Razor Hydra, and Leap look like interesting alternatives, while joysticks and steering wheels are surely also here to stay.

Since 1998, when I first started messing with Reality Studio (3D panorama tool), I have been waiting for OLED. Unfortunately the day when we have over-sized OLED screens in diverse radii is still many years away. Too many years to keep waiting. LG are possibly releasing a slightly curved (around 5-10 degrees) OLED screen later this year. For a 3 screen set-up it would cost you £30,000 ($45,000, so start saving), and it's *still* got bezels! I read that LG's entire plant is dedicated to a single sized screen (55") as it's too costly to produce different sizes. Prices will fall. Who knows, maybe in 5 years the entire infintx line will be OLED based, on the other hand 4K laser projectors may cost $300 by then. The best way to predict the future is to build it.

Haldi > Unigine, very interesting engine and a new one on me. Thanks for the link Obviously you wouldn't apply a distortion profile to a screen that handled it for you. I imagine there's still a preset for "flat monitor". Use that and your problem's solved!
Imagine this http://www.youtube.com/watch?feature=player_embedded&v=4yMnsioOpuo#! inside this...



I stress, the "compression wings" concept is just that, a concept for discussion, and may never see the light of day. It is so ultra specific to wide screen, wide FOV 3D engines.

Cheers!

Just don't settle for rectilinear projection on a curved display and everything will be great. Also, if you're doing you own rectilinear to cylindrical blending, use at least 4-5 viewports for 120-140 HFOV, please.

No no no, this is pretty much spot on. I did some pictures with FOV and POV angle matching with Borderlands a couple years back but never actually made a post about it. I see you already have the 3D software ready to do this. Might want to use some of these images for the WSGF so we can explain why people need to stop bitching about fisheye.

_________________
(imagination not allowed in signature)

Randomoneh, point taken. The actual game rendering capabilities are for the most part beyond our control. The reality is unfortunately that for most current titles, rectilinear is all you have to work with. The winged monitor shape, with continuous wide angle warping was a concept for combating that, even though mathematically it's completely wrong, perceptually it may deliver what the brain was expecting to see.



Be my guest Just promise me we'll try and find another term than "fisheye", as it's kind of the opposite of what's happening. The distortion is a product of what Randomoneh refers to, that is a rectilinear projection and NOT a fisheye with is a spherical projection, and would actually work much better for super wide screens.

Cheers!

I've seen the term "fishbowl" used in a similar context.

That is an interesting Video on YouTube. I might be a bit dense with some of the terms, but what is causing the trees in some parts of the video to bend prior to going off the edge and some are just straight? What causes the terrain to bend around/away from the perspective?

Just posting to add support to this project, will keep an eye on it for sure! This is something I've been following for years (obviously previous attempts never really took off) so I'm hopeful this time a real contender will emerge...

Think the new (slightly) curved OLED tvs may help public perception of the idea too. I know it's going to be mostly enthusiasts that will want this at first, but gaining even some small traction amongst the (usually fairly conservative) general population is always good

I hope you will be able to produce a setup that is competitive in price with a monitor setup similar in size. Just the lack of bezels is a huge plus point. Possible better perception with custom screens (wings/flared and such) is another big plus over monitor setups. And finaly if over sea delivery will be possible/affordable you can write me up for 1!

_________________
3xU2311H_portrait+HD6970+2500K@4Ghz+P67Ex4Gen3+8GB+M4_128GB+4x500GB_raid10+Raven_RV01= a little outdated but still nice pc gaming fun

Being brutally realistic, the price point would be around $3000, given current component prices.
That's still less than half the price of say, an Ostendo http://gizmodo.com/5411486/ostendos-crazy-curved-monitor-can-finally-be-yours-for-6500?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed:+gizmodo/full+(Gizmodo)&utm_content=Google+Reader
...and over double the size.

To drive the price further downwards would require a large scale production, and with it some substantial investment.
Kickstarter would be a cool option, if only we were based in the UK or US. They don't currently accept applications from anywhere else.

LED/laser projector prices will fall and resolutions will increase, making it a more attractive proposition as time goes on.

Should anyone require one straight away, we have the facility to build them to order, in regular arc shapes from 60 to 360 degrees.
Just drop me a line

Cheers!

What is the optimal arc for a wide screen gaming display? (Should there be different options?)
Needs to be divisible into 360. There are several reason for that. 360 / 2 = 180, 360 / 3 = 120, 360 / 4 = 90. I suggest 115 degrees per screen for those that might use 3 for a "full" FOV setup with flight sims and want a entry/exit space.

Is stereoscopic 3D important to gamers? (Does anyone actually use it?)
Is it important? No. Do people use it? Yes. Whether it's used or not should be measured on if it's an undue burden to unit cost to add it.

What’s the maximum acceptable physical size for such a display?
I'm bias to wall height panel units at 90" maximum height. Maximum outer box dimension of 10 foot width to fit into most regular rooms.

What should it cost? There are several places where it’s possible to compromise on materials and components. Which weighs heavier, the price or the output image quality?
My personal suggestion would be the baseline for the design be 3x1P. Pixel height is very important. 1280 or 1920 pixel height on the setup would be an excellent target considering PPI projectors have to deal with. 3x1P should be achievable @ 115 degrees with three projectors. It's up to you to work the engineering particulars to assure the projectors work in portrait mode with proper thermals. Since the design is custom the only question is it plausible to source the projectors affordably for that end use?

Should the screen be self assembly (think IKEA) of ready to rock?
Self assembly because the cost to ship will kill it otherwise.

Is portability a factor that should be considered? Should you be able to pack the screen down either for travel or storage?
Yes, and this plays directly into shipping too.

What’s most important, a display which is “true to life” (geometrically) or one that gives you the biggest advantage over your competitors?
Neither, what is most important is to produce something that most people can use within a standard room in a FOV divisible by 360.

Should the projectors be removable for use interdependently or in other setups?
No, that is outside of the scope of the design.

Frankly, I don't see how you're going to pull this off without pre-warping to maintain calibration. If you'd like to work an OEM agreement for a custom version of Nthusim, get ahold of me via PM and we can work details.

_________________
Brad Hawthorne
Product Manager
Nthusim Pty. Ltd. | www.nthusim.com

[quote="BHawthorne"][/quote]

I second that!
Saves a lot of typing

_________________
3xU2311H_portrait+HD6970+2500K@4Ghz+P67Ex4Gen3+8GB+M4_128GB+4x500GB_raid10+Raven_RV01= a little outdated but still nice pc gaming fun

I'm really trying to be constructive here when I say this.. 2880x900 ? I have more pixels on my Nexus 10.

If you were selling this 5 years ago, I'm sure this would be fine, but please get your resolution up before this can be considered as a legitimate enthusiast product.

The only limit on resolution is how much you want to spend, and what projectors are available.
I'm not sure there is an enthusiast gamer market if the projectors alone cost $3300 (for example 3 BenQ W1080ST), or at least not for a desktop sized surround screen. With bigger set-ups, car and plane sims, you are getting a lot more screen for the same money, from the same hardware. This market is restricted by how much space you are willing to dedicate to such a set-up.

Our first model is 200cm x 100cm, and can be delivered in either 2560 x 800 or 3840 x 1080
It's 120 degrees and made to match the output from a GoPro.

So VFOV is ~55, right?

That depends on your source material. If you are running 2:1 ratio, then it's up to 60 degrees.

Wait, this is a cylindrical, not spherical display, right? It's curved horizontally but not vertically.
If that's the case, it can't be 120 HFOV and 60 VFOV. Horizontally, relation between size and FOV is linear but vertically it isn't.

If your source is a fisheye, then FOV's are proportional at the center line, though this would depend on the lens, whether real or virtual.
For example the GoPro black edition has a VFOV of 93 degrees, and a HFOV of 121 degrees, despite having a 16:9 output, which logically would give you a 68 degree VFOV (121/16 x 9 = 68 degrees). When we remap that to cylindrical, the HFOV and VFOV are maintained on the centerline and horizon.

Demo's to come

It's been a while.
During the summer we finished the second prototype. The conclusion was, it's close, but no cigar just yet.
Here's a couple of videos of the construction and the finished screen.
Apologies with the first timelapse, I left the display on on the GoPro and it ran out of battery without me noticing.

Construction:
http://www.youtube.com/watch?v=D-lperdGeXY&feature=c4-overview&list=UUdfmUD5q097W2hrLEzQGvuA

Live Demo:
http://www.youtube.com/watch?v=Bi6mXSgUar0&list=UUdfmUD5q097W2hrLEzQGvuA

The screen material we'd banked on just wasn't up to the job.
First of all it was glossy, as i usually prefer glossy screens. But in our particular design that was a mistake, as the screen ends up reflecting itself. So next time it will be a matt, none reflective finish.

Blending. You will notice on the video that there is a line clearly visible where the 2 projectors meet.
It's always been our goal to blend without software, but here we must admit defeat so far. In the next prototype there will be more overlap and we will also employ software if necessary to completely hide the join.

There were a lot of lessons learned and we look forward to putting them into practice in the next iteration.
Watch this space!

As always thanks for all your input

Cheers!

Cool timelaps Video! To bad it did run out of battery.

And +1 for the matt screen!

_________________
We gonna send it to outa space!