I didn't want to hijack the "LL Projector Lumens Theory" thread, so I'm starting this one.

For those unfamiliar with passive stereoscopic projection using polarizers & glasses, please check this link: http://www.accessgrid.org/agdp/howto/geowa....1/html/req.htm

Quoting the site, "The light emitted by DLP projectors is not polarized and so the Geowall was designed with DLP projectors instead of LCD projectors. The latter are typically pre-polarized, so you cannot control the polarization with filters." This was the same thing that has prevented use of DIY digital projection for passive stereoscopic imaging.

However, if removal of the factory polarizers is now "DIY-able", do factory polarizers need to be replaced with "aftermarket" polarizers of identical orientation/angle? Is the actual LCD designed for a certain polarizer orientation? What would be the effect of stripping a factory vertical polarizer off and replacing it with a horizontal one, or one at a 45-degree angle, even? What about circular polarization (okay, I doubt this is possible, from what I read of the explanation of how the LCD layer works in conjunction w/linear polarizers)?

The second question is: would the fresnels (if going split fresnel route) interfere enough with the polarization that the effect would not make it to the viewer wearing polarizing glasses?

One of the big appeals going DIY (aside from the obvious cost advantages) would be the possiblity of building a two-in-one projector using a single light source to ensure uniform color temperature between the two projections.

Just some food for thought...for now...

Ashe

You'll find this interesting.

http://individual.utoronto.ca/iizuka/research/cellophane.htm

Remember, even if we make a dual projector system capable of 3d images, we still need compatible software to actually use it.

While it wouldn't work for movies, Nvidia cards and drivers can be employed in a dual head configuration for left & right views in games and other digital/computer content. You need to tweak drivers and registry entries, but it can be done.

Ashe

And the journey begins....good luck Ashe, I hope you do well in this thread.
sim

Although it has not been confirmed... it is very possible(even likely in my opinion) that the polarizers can be oriented at any angle as long as the first polarizer and the analyzer polarizer are 90º offset when using a conventional TN TFT LCD.

The field fresnel had a small affect on the quality of the polarized light causing a little bit of a glow. This can be overcome by ditching the keystoning and going with an unsplit fresnel configuration.

Project through one lcd back to front (and the other front to back) then there is no need to change the polarizers!

More software stuff to reverse the image will be needed, but less chance of damage

I agree, when I stripped of the front and back polarizers off my ps1 lcd I was able to rotate them dofferently, didn't seem to make much difference as long as they were the same relative to the other polarizer....

You can pull this off with 1 LCD and a variable wave plate. You could even pull it off without a variable wave plate at all if you had complete sub-pixel control of the LCD, as the LCD IS a variable wave plate.

An LCD with both polarizers removed will act as a variable wave plate. By variable, I don't mean variable phase shift, I mean variable axis.

The easiest design I can think of is to order like this:

1. Normal stock LCD,

2. Variable wave plate LCD (stock LCD with both polarizers removed).

3. Collector Fresnel.

4. Triplet.

5. (True) Silver Screen.

6. Polarized glasses with each eye having a 90 degree polarization axis to the other.

To the wave plate screen you would feed full screen video of a black screen for one frame, and a white screen for the next. Have your stock imaging screen flipping from one eye's image from one frame to another.

You could take your chances and hope the timing signal remains in sync from one monitor to the next, then you wouldn't need any fancy software. Just video with the stereo frames alternated. Or you could use software to keep the wave plate and the video signal in sync indefinitely.

If you wanted you could probably place the wave plate in front of or behind the triplet, and not expect a severe loss of quality. That means your second LCD could be the size of your triplet.

Hope that gets the wheels turning .

Mark.

Stumbled upon this site while searching for polarizers....

http://www.silverfabric.com/html/sf_polarizers.htm

Pretty cool. Dosent look much different from what Elken did with his triplet! I dunno if you need thier screen for it to work. Mark you seems to have basicly nailed it in the above post based on what these guys are doing.


BTW, Simul8r, excellent job with the new summary. I was about to take an iron to my scratched polarizer

Wow! I wonder what the cost is? Those polarizers are wire polarizers: something that I have been looking for. And the quality seems outstanding.

The last product has me intrigued: SPAR Polarizer >75% transmission. >1000:1 contrast ratio. This may be something huge. That would be a polarizer capable of a minimum 69% increase in DIY lumens. I am going to have to think about what exactly it is that they are selling there. It only works with previously polarized light, so there must be a catch. That is the sort of technology that you'd think you would have heard more about, so I am skeptical.

The 1800:1 polarizer at 65% transmittance is probably somewhere around .04% crossed transmittance. So not as good as our film polarizers but really not bad considering. A higher contrast polarizer could be added on top to gain the contrast back anyway.

Mark.

Your welcome, I was banking that it would help others and hopefully releive some of the pm's I've been getting...hehehe. Not that I'm complaining but having to repeat the same info over and over.
sim

Mark, this link....do you think it might be just what we needed for more lumens? If this actually works than all we have to do is remove the front a/g&polarizer and stick one of these up to our triplet. Interesting find yoshuaspawn.
sim

Thought you guys would like that! Whats great too is that they seem to encougage special requests as well. I emailed them with a general query stating the need for a high quality polar film with no layers or components pertaining to the angle of viewing(ag.. ect..). I also mentioned that if this is somthing they can do, that there are potentialy MANY interested parties. You and mark should contact them as well. You will do a much better job of asking for exactly what we need. I will post as soon as I get a re:
Hope for the best!

Well not really. I hope I haven't gotten peoples hopes up too much. I am skeptical. It is just that I have never heard of a polarizer that can break the 50% transmittance barrier without reflection. How we would implement it could actually be pretty tricky now that I think about it. The reality is that our analyzers are pretty efficient. It is the polarizers that are eating up the 50%. And these things only work with polarized light, so replacing the polarizer isn't practical. I'm also thinking that these would not work as a standalone analyzer, as I'm guessing they simply recombining the x and y polarization states.

If I were to wage a theory I would say this polarizer takes advantage of the prism beam splitting polarization tactics of a standard LCD projector. The two polarized plains merge at the lens. So what this thing probably does is spins the one axis to be parallel with the other. It is probably constructed on the principle that a wave plate only acts properly on one plane, so one axis of polarization can be altered while the other is basically untouched.

But that would mean that these polarizers must be mounted at a 45 degree angle to both merged polarization axis, as each axis represents different colors. And I don't see why you would need an additional polarizer in the first place unless you have a 3 CCD projector, or want circular polarization. I guess the problem is that I have not read anything lengthy about exactly how you set these things up.

So I am actually leaning towards these not being of much use. But my initial reaction is that it is pretty cool that there is a polarizer that can break that barrier even given limitations. I'd need to know more about it to see if we can adapt it somehow.

If the standard models are cheap, though, they could represent a better alternative to film polarizers for those who are willing going with a triplet mount.

Mark.

Hey Guys, I got a reply from silverfabric.....
here it is,

We do have high-quality polarizing foils as you describe. These materials
are used for the production of our filters.
Normally we do not sell our raw materials. But please let me know the size
and the quantity of polarizing foil that you need, and we will see if we can
sell this to you.
Let me add that handling of our polarizing foils requires expertise.

Dr. Yitzhak (Itsik) Weissman
SF filters technical support
email: dr.yitzhak.weissman@silverfabric.com
web: www.silverfabric.com

Mark, Simul8r,
It cant hurt for you guys to give the 'good doctor Yitzhak' a shout at this point. Your probably correct Mark, but we cant know for sure until you guys give this product a fair probe right? Even if if the product is not that great, I feel like this guy could be of some help to us in finding somthing that is, or at the very least could offer up an idea or two.
Yosh.

Check this out... 3D, Paramount pictures style!
Interesting item at the shed....
http://www.surplusshed.com/pages/item/m2234.html

Assuming you plan to use standard polarized glasses to view the images, the internal polarizers make LCDs useless for standard stereoscopic projection, no matter if there is one light source or two.

I've heard idle talk of replacing the final polarizer with a 1/2 wave or quarter wave plate (retardation plate) to make each projector compatible with paper 3d glasses, but I've never seen one produced. The damned things are expensive.

I can't think of a workable application of 3d using standard TFT color active matrix LCD panels.

Tom

The text says the Paramount widget uses polarization to differentiate between the two stereoscopic images. That makes it incompatible with the LCD panels being used in LL projectors. LCD panels themselves have internal polarizers, so the light emerging from them is pre-polarized.

There are dozens of ways to make stereoscopic images, but none I can think of are compatible with LCD panels.

Tom

So you're saying that if you were to strip the polarizer off the LCD that you would still be able to project light that was polarized? It was my assumption that was the purpose of the polarizer, otherwise you could just strip down to the glass and project away.

I'm not sure what you mean, there is a typo in your message.

If you meant "strip one of the polarizer off the LCD" then the answer is yes. If you meant "strip the polarizers off the LCD", then the answer is no. Neither one result in a useful method of projecting 3D images.




You can, but there will be no picture. Perhaps you are under the impression there is just one polarizer. There are either two or sometimes three.

The idea of projecting three-dimensional images using LCD panels or screens has been repeatedly suggested on this and other forums. What follows are some ways this could be attempted, and the reasons why the effort would be unsuccessful or impractical.

1. The title of this thread is "swapping polarizers". Completely impossible. You certainly could replace both front and back polarizers, but you can't replace the liquid crystals themselves, and their orientation must match the polarizers. The only reason I can think of to replace a polarizer is to fix a botched diffuser removal.

2. If the polarizers are oriented 0 or 90 degrees, then you would have to rotate one of the LCDs. You would have to have special 3D glasses made to match this non-standard orientation. Not impossible, but they cost more than standard, and you can't just have a few dozen made. You would also end up with a square image, meaning that all of your images would need to be cropped, and the cropping would somewhat mess up the resulting 3D separation.

3. If the polarizers in your LCDs are oriented 45 degrees, you could theoretically turn one of them backwards. This would give you the same orientation as standard 3D glasses. I've never tried projecting backwards though an LCD, so I don't know if the resulting image would be identical as projecting forward. To keep colors from bleeding, etc the geometry of LCDs are pretty critical.

4. A variation of the last idea is to use a 1/4 wave plate to re-orient the polarization of one of the projectors. These items are expensive, and one of the projectors would be significantly brighter, which is a no-no.

5. Build a single projector with a single light source and two LCDs. Can't think of any way to make this work. (If you use a DLP projector and strips of polarizer this works great for video, since the current american video standard works exactly this way. Put left and right pictures as alternating fields (not frames) and you're done. But it will not work with an LCD.

6. In theory, you could cut strips of quarter wave material, and apply each strip to all the odd or even scan lines of an LCD panel. (Very tiny strips!) Then you would need to interleave your two stereoscopic images into a single interleaved one. Additionally, this would lock you into a certain screen size. If you tried to show the resulting images on another projector, the apparent stereo separation would change.

Regardless of whether these idea would work or not, they all would require a "silver" screen to maintain the polarization. You would also need to invent software that could drive two monitors simultaneously with two separate but related images. Easy to do by hand for a few still images. A working solution that will simply play video or display a folder of images is much more difficult.

I've built 3D still cameras, a 3D video camera and a 3D projector, and you cannot believe the physiology, mathematics, physics and fiddling involved. There are also slippery issues relating to the way human brains perceive stereoscopic images. Seemingly reasonable solutions which do not work are commonplace. And no matter which method you gamble on, you still need to build two identical projectors, tear the diffusers off two identical LCDs, work out keystoning on two axes, and write some custom driver-level software.

If you figure out everything else, the only problem remaining is finding a place to put a stereoscopic projector the size of a piano. :^)

Tom

If you were willing to build a very odd looking projector with some custom parts (if you want 3d, this is probably already the case), then you could try this:

Strip BOTH polarisers from the LCD (front and rear). Replace the rear polariser with half the screen horizontally polarised, and half vertically (I'd reccomend using a 16:9 panel, and orienting it vertically. That way, the two displays will now be one above the other, and near-4:3 aspect).
Next, add in the correct front polarisers (this depends on the original configuration. If your stock LCD had the same polarisation plane for front and back polarisers, then each half should have the same front and back orientations, and vice versa for 90degree rotated polarisers).
Next, use a SINGLE fresnel as the COLLIMATOR (fresnel between the bulb and the screen)
Next, use TWO fresnels (SEPARATE fresnels. you cannot split a single fresnel for this, as they must focus onto two different triplet lenses) to focus onto the aforementioned TWO triplet lenses.
Finally, you must keystone the pair of fresnels so the image overlaps EXACTLY. Failure to do this will result in a very fuzzy 3d image.

What you now have is a 3d image, with the left and right images polarised at 90degrees to each other, and at half the resolution of your original panel, in a near-4:3 image. For example, if you started with a 1280x720 panel, you would have a 720x640 (which just by happenstance is close to DVD resolution) image.


This design has the same effect as projecting two projectors onto the same image, but instead of using two full-res LCD panels, you use a single LCD panel. the advantage of this is you do not need a dual-head graphics card, as all the image processing is done in software. Additinally, the projector itself is smaller.

You still need the polarised 3d glasses, but these are ridiculously easy to make (buy any el-cheapo frame, and cut a sheet of polaroid to fit each lense-hole. One at 90degrees to the other, of course).

Of course 3D possible with two proyector DIY flipping one the polarizer details at:

http://www.lumenlab.com/forums/index.php?s...15333&st=40

None active glasses works with any LCD (neither 5ms,it is too slow LCD refresh rate (black, not full color, so ghosthing (tested)) actually.

For 3D needed stereo proyector (two proyector 1 for each eyes)
How to build your own home HDTV IMAX 3DDIY at :
http://www.lumenlab.com/forums/index.php?s...15333&st=40
imax3d
The most difficoult is how to make a good 3D silver screen DIY real flat.