2 LCD projectors -internal polarization 4 maximum efficiency

[martinlandau]

http://www.stereo3d.com/Barco_Stereoscopic_proj.pdf" onclick="window.open(this.href);return false;

Looking at this link, Barco made 2 LCD projectors that are exactly identical expect for the polarization of the LCD panels. This gives the maximum efficiency without the use of expensive filters. The LCD panels are already polarizing the light. From the link you are getting almost 60% efficiency versus 38% for the next technology path.

So we need to build a database of consumer level LCD projectors and thier polarizations. Perhaps we can find 2 that will complement each other and not have to buy expensive SPAR fiters and such for a very cost effective passive polarized setup.

Okta, Tril, anyone else? Get us started, how can we go about collecting the data on all the LCD projectors available out there and crowdsource this info?

[Likay]

I like the idea too!
There's a problem with conventional lcd'-projectors though. The colors aren't aligned at the same angles.
The projectors i have have colors has the green color aligned at 0° and the red+blue at 90°. It's probably only a matter of fabrication though but it puts sticks in the wheels trying to use standard projectors.

And agrees about stereopol filters: the spars gives value for the money but the cost is high for two glasspieces. I don't regret in any way getting them though. It's nice having 100" brighter than a standard 19"er. 8)

Here are are the specs on polarization for Hitachi cp-x260:
Green color at 0° and red+blue at 90° or vice versa. I don't know the exact polarization angle of the testing film.

[martinlandau]

Likay do all LCD projectors polarize the green at different angles than the red/blue or just your brand of LCD projectors? Barco has proven you can polarize however you want. Perhaps if we stay persistant, we can find one LCD projector that is polarized at 90 and another at 0, I just don't know where to look for this info. Projectorcentral doesn't list polarization angles of the LCD panels in their specs. Maybe we should contact several LCD projector makers through email and see if they can tell us thier polarization specs.

[Likay]

Lack of knowledge on this sorry.

I don't know about other lcd'projectors, but since stereopol- as well as stereobright-filters are made especially for lcd's there might be a risk for it.
And no, noone seems to know the polarizationproperties for their projectors. I haven't heard with the manufacturer itself though. I read the specs on my beamers and there is no info on polarization either. Not even a mention that the output light itself is polarized. As is now it seems like polarizationproperties is some sort of unusable information and is left out to keep the manual a little bit smaller.
I don't even know if the separation of green at one angle and red+blue at another is a form of "standard" or only specific for my hitachi's. It would be strange if projectorpolarization differs between fabrics. I have trouble understanding how spars and stereobrightfilters would work with different lcd-polarizations without mods or special manufacturing. Those filters should then work good for ordinary dlp's as well but is clearly stated that they don't...

Lol: Seems like it might be a standard: Found this site: http://www.advisol.co.il/StereoBright%20home.html" onclick="window.open(this.href);return false;

Maybe some others joins in with info.

[DavidGhast]

im sure theres a photometric you can buy to test the polarization of each channel.

[martinlandau]

Likay this came from that link you sent:

StereoBright setup relies on the native polarization properties of the projectors. Although all LCD projectors emit polarized light, the degree of polarization varies between different projector models. But even with projectors with excellent degree of polarization, the ghosting in StereoBright will be higher than in conventional stereoscopic setup.


So there is hope. Perhaps you could take your polarized glasses and go into some retail stores where they have several projectors on display and test if the polarizations are the same.

Likay, for 80 dollars, those stereobrights are cheap, but why is the ghosting worse versus other solutions?

[Tril]

I got an idea for you that would work with the exact same projector. I haven't thought about the details so I don't really know how you would actually do it. Project the picture from one projector directly on the screen. Place the other projector facing the direction opposite the screen. Make it project on a mirror and reflect on the screen. A mirror preserves polarization. You'll need a good mirror if you don't want to deteriorate the picture quality. The picture will be mirrored but this is easy to fix because normally any projector has an option to project a mirrored picture.

If the green are at different angles, you can use a software solution to fix the problem. For games, ask your favorite stereo driver company to make a dual projector mode in which the green are switched. I think that's something easy to do for them. The website Likay linked says that Stereoscopic Player has a mode to output the green switched like that.

[Likay]

@Tril: Iondrive@nvidiaforums and here mentioned this principle for me some time ago. When mirror-shifting polarization the angles needs to be at 45° and/or 135°. The polarisation from my projectors are at 0 and 90° so polarization will unfortunately be the same even after mirroring. It can be "solved" by tilting both projectors 45°, sending the leftimage-red+blue +rightimage-green to the upper projector and vice versa. You also need to turn the image 45° from the computer to match. There will be an on top standing square with a 3d image. Having both projectors at the same alignment 45° works when reflecting one of them but since you need to tilt any way...

@martinlandau: I can't tell for sure but i think stereobright filters consists of a few retarderlayers used to to realign the 0/90° image to 45/135°. The result will be a beam "exactly" like the original one but with polarization turned 45° (both axis). This is why you cannot use a rectifying polarizer since it will of course block either the green or the red/blue.
As an experiment i tried the extinction of the different colors with a polarizer and the red is definitely the one that ghosts less (best polarization) while green and blue ghosts pretty much (more depolarized). Blue is an assumtion since it's hard to see the brightness of it. I could still see the blue on a black background when it was suppoused to be blacked.
Stereopol filters has a rectifying polarizer at the end of the retarderstack and therefore polarization is "sharp". Since polarization isn't perfect from the projector and that usage of retarderlayers in the stereobright filters (might, maybe not cause more depolarization?) then it's given why a stereobright rig ghosts more than a "single-axis-polarized" rig. I base this is on what i see on the site so don't take me too serious on this! I think the theory matches though.

Stereobright preserves the lightoutput through the polarization process and guarantee a very bright image. Colors are aligned the same as the original beam but rotated 45°.
Stereopol does the same but with all colors finally aligned the same.

I think it's possible to build your own stereobright filters using retarders. This should be the way to go if going the definite diy route.
I tried experimenting with retarders once to achieve a stereopol effect. It's not easy to get it right and in the end i didn't have the patience for it. Stereopol is said to have about 15 precise aligned retarderlayers followed by a rectifying polarizer. I had a lot of reflection losses through the layers so besides figuring out the retarderalignments a nondepolarizing glue is a must. I have a slight hint that stereobright filters should be way easier to make yourself though.

[martinlandau]

http://www.mtbs3d.com/phpbb/viewtopic.php?f=26&t=3309" onclick="window.open(this.href);return false;

Likay this guy you replied to earlier seems to have exact opposite polarizations of your projector, so we need to be certain what the RGB polarizations are of yours to make sure these 2 can work together.

I don't like doing stuff in software that uses CPU cycles, can't you just make a special video cable that take the green from the DVI output and goes into the other video cable and then vice versa or am I not understanding the math again (sigh)?

[Likay]

Actually the colors on mine CAN be the same since i only have a square polarizing film to try with. I know it's aligned in a manor 0° but unfortunately not in which direction. I can determine that the green has one polarization and the red+blue shares another and of course the differing angle.
The 0 for green and 90 for red+blue was an example to make it "clearer". Sorry for the confusion.

Wiring:
You got it right but crosswiring the wires can only be done if the dual outputs are in perfect sync. I have no idea if this is the fact though. It's unlikely that the projectors or graphcard will take any damage for as long as hotconnecting/hotswapping is avoided, even if the outputs are not in sync.

[martinlandau]

Actually the colors on mine CAN be the same since i only have a square polarizing film to try with. I know it's aligned in a manor 0° but unfortunately not in which direction. I can determine that the green has one polarization and the red+blue shares another and of course the differing angle.
The 0 for green and 90 for red+blue was an example to make it "clearer". Sorry for the confusion.

Stereobright said different model LCD projectors have different polarizations, so we are back to square one. We need to build a database of which projectors polarize at what angles and find 2 different models that complement each other. I like the mirror idea, but since I plan on using a dome and spherical mirror, I want to ty and keep the mirrors to a miimum.

Ok, so what do we need so that me and anyone else can go test some LCD projectors and get the polarizations? I am trying to follow along but the optics math is fuzzy after my 2cnd bottle of sake. We can't use the iz3d glasses? Or the real d glasses? Or the dolby3d glasses? I have all of those, plus some polar optics polarized sunglasses I bought at wal-mart. If none of those work where could we get something cheap to test the polarizations? No ghosting is the holy grail of course.

[kekewons]

Well, if nothing else the idea of the "dome" certainly caught my eye (because I've been thinking the same thing myself for quite a few years now.

"Personal Planetarium?"


k

[Okta]

I think it will be very difficult to use a mirrored dome with 2 projectors thats why i opted for dlp.

[martinlandau]

octa why do you think 2 projectors will be difficult? The stereo could be horrendous on the periphery, but so what, straight ahead it should look fine no?

1080p seems to be several years off for a single projector solution. If I can find 2 models of projectors that polarize their lcd panels differently it should give much greater efficiency as well.

[Okta]

octa why do you think 2 projectors will be difficult? The stereo could be horrendous on the periphery, but so what, straight ahead it should look fine no?

1080p seems to be several years off for a single projector solution. If I can find 2 models of projectors that polarize their lcd panels differently it should give much greater efficiency as well.

I assumed you mean to use 2 projectors in a passive setup reflected off of a dome mirror to a dome screen? I thought it would be very difficult to align the 2 images into the same postion from slightly different angles to the domed reflector. Also not sure what affect this would have on the polarization.

[martinlandau]

I assumed you mean to use 2 projectors in a passive setup reflected off of a dome mirror to a dome screen? I thought it would be very difficult to align the 2 images.

Yes this was the intention, 2 1080p projectors reflected off a dome mirror to a dome screen or each with thier own fisheye lens onto the dome screen. Perhaps it would be better to do 2 dome mirrors, one on top and one on bottom of the dome.

[martinlandau]

Likay I just got back from wal-mart. They didn't have any LCD projectors, but they did have lots of LCD tv's. Went over to their sunglasses section and got a pair of polar optics sunglasses then walked over to the TV section. Most of the brands were polarized the same way, but the philips and vizio line had different polarizations than the other brands (samsung, sanyo, etc etc)

Interestingly RCA had a 26inch LCD that was polarized differently than their 22" LCD, so even with the same manufacturer, they use different polarizations. I think given this, it should be relatively easy to find 2 lcd projectors that have different polarizations.

[Likay]

One more week of work and overtime... Then VACATION!!!

I'm going to try some myself if storekeepers allows me to. I know quite a few that sells projectors but i haven't seen any projectors showcased though. It's probably because of space but i'll keep my eyes open.
Regarding polarization for projectors i've made an experiment: I've tried rotating the spar-filters 90° on my projectors and that is certainly a nogo. Colors get's weird and i loose a lot of light. It would be very strange if the manufacturers of these filters says they work with every lcd-projector (i didn't have to say which projectors i had when ordering the spars. However they did wanted to make sure that i had lcd's before my purchase). Another thing that indicates that projectors may have the same alignment is that Stereoplayer with stereobright option only have one config, red+blue+other green and vice versa. I checked using dual vga and there was doubling. When putting standard polarizers thus blocking the green from one projector and red+blue from the other and looking without polarized glasses (to check alignment) it's pretty obvious i should align my beamers better. It's not a big issue, it looks like a very slight misaligned colorprinted monopicture. That part work i think.
Stereobright filters would be impossible to make if you want to turn the colors +-45° and at the same time guarantee that the output red+blue and green always have the same alignment (which is needed for the stereobright algorithm to work) regardless of projectorpolarization. If lcd-pj's did have different alignments there would be need to shreddermake filters for different projectors and i'm quite sure it's the same thing for spars. However i agree with you that the principle itself allows to make projectors with different alignments without problems.
Is it maybe possible to alter an lcd-projector by simply rotating an internal polarizer (or two) 90°? (I WONT try that with the ones i have! ). Projectors are very sensitive for dirt and i had those for quite some time. Disassembling them would probably render them useless.

Now to the another appealing task:
As mentioned i think stereobright (quite expensive but far from spars) filters are quite easy to make yourself. I only have qw-retarders (very cheap) to try with but i think that should be enough. I have a slight hint this is the most suitable way to go but you will of course not get as a high light output by just using projectors without filters.

Software support if crosswiring cables won't work is another issue. Peter Wimmers player supports stereobright but stereogaming is worse. However: If there's demand for it i'm pretty certain that the iz3d-team will code in stereobright in their driver.

[martinlandau]

Just got back from best buy. Most of their laptop LCD screens had different polarization angles. Interestingly 2 sony laptop models had 2 different angles, so again even in the same manufacturer, there are different polarizations. Dell had the only laptop that was weird. Using my polar optics polarized driving sunglasses, it was the only one that did not go opaque at some angle. Turning the glasses clockwise gave a blue tint, turning them counterclockwise gave a red tint, but never opaque. No other LCD screen has ever done this in my testing, what does that mean?

Then went over to the LCD TV section. All the very large LCD TV's (over 50 inches) were polarized at the same angle. Most of the smaller ones had different polarizations - under 26 inches. They also had an LED tv that was really thin, it was polarized at the same angle as the large regular LCD tv's, turning my polar optics 90degrees clockwise made the screen go opaque.

Another carrot to chew on, some of the LCD laptop screens were more opaque than others. IE, in some there was just the faintest amount of light coming through on high contrast scenes, but on some of the LCD screens, it was practically pitch black.

[Likay]

Lcd-monitors and -tv's have only one alignment and can not in any way be compared with polarization properties on lcd-pj's. It's well known they have different polarization properties as well for instance when trying to build a regular planar rig. This works best without hazzle with equal monitors if the polarization is 45 or 135°. 0/90 works too but you need to do tricks.
You need to test lcd-projectors before knowing for sure. As said i'll keep my eyes open myself but showcasing projectors doesn't seem to happen too often here.
I still believe it seems to be a "standard" how lcd pj's are polarized since you won't need to match high output filters with lcd-projectors. I can be wrong though and i get back with more info if i get any.

[martinlandau]

http://forums.nvidia.com/lofiversion/in ... 31636.html" onclick="window.open(this.href);return false;

Check this out Likay, 2 years ago on nvidia forums. you, nubie, hornet, and a few others discovered using 2 LCD projectors with internal polarizations - 1 with a mirror - would work - UGH!! Why let me re-invent the wheel and waste so much time when you already had all this knoweldge in your head? Does this only work because he is using a special LCD projector with 45degree polarization?

nubieMar 26 2007, 09:34 PM
I am attempting a 3D setup with 2 http://www.lumenlab.com/eVo" onclick="window.open(this.href);return false; projectors. They use a single LCD polarized at 45�, so I will be mounting them to a board together and one of them will be mirrored to flip the polarization 90�

Again you keep restating the belief that all LCD projectors polarize the same, when the very stereobright people you linked to say on their own webpage that is not the case, so why do you persist in that belief when they don't agree with it? It is like I am talking to a religious fanatic that thinks the world is only 6000 years old!

[martinlandau]

Here Likay, these guys got it to work, I can't find the full text online though, maybe you can.

http://adsabs.harvard.edu/abs/2006OptCo.266...55K" onclick="window.open(this.href);return false;

Title:
A novel configuration of LCD projectors for efficient orthogonal polarization of two projected views
Authors:
Kim, Seung-Cheol; Kim, Eun-Soo
Affiliation:
3D Display Research Center, National Research Laboratory of 3D Media, School of Electronic Engineering, Kwangwoon University, 447-1 Wolge-Dong, Nowon-Gu, Seoul 139-701, Republic of Korea. Corresponding author. Tel.: +82 2 940 5118; fax: +82 2 941 5979.
Publication:
Optics Communications, Volume 266, Issue 1, p. 55-66.
Publication Date:
10/2006
Origin:
ELSEVIER
Abstract Copyright:
Elsevier B.V.
DOI:
10.1016/j.optcom.2006.04.018
Bibliographic Code:
2006OptCo.266...55K

Abstract
In this paper, a novel configuration of LCD projectors for efficiently polarized stereoscopic projection with improved light efficiency is presented. In the proposed scheme, polarization orientations of the analyzers in the LCD engine of one projector are rotated by 90° with respect to those of the other one, so it causes to rotate orientations of polarized outputs by 90° as well as to change operational modes of LCD panels from the normally black mode to the normally white mode or vice versa. Then, by exchanging the green color components between the modified and original LCD projectors, two orthogonally polarized views for stereoscopic projection could be obtained without use of external polarizers. Analysis on the operational mode conversions of the LCD engines with rotation of the analyzers is discussed and its experimental results are provided.

[Likay]

Forgot about that thread lol! Sorry.
And yes: Nubie managed to find projectors (lumenlab) which actually seems to have alignment for the colors and aligned at 45° as well. This means he gets 135° by using a mirror. I don't know if these projectors are standard or specialized though.

Hornet uses shutterglasses and 2 projectors (not polarized!, they're aligned next to each other horisontally to give a bigger picture, like extreme widescreen) and have a quite great experience. Would love to see that one in real too.

@thread before this:
I just took a quick look and this seems to do with either modifying existing projectors or alter the fabrication procedure.
As mentioned before i think there might be possibilities to modify an existing projector but i will not try that myself.

[martinlandau]

@thread before this:
I just took a quick look and this seems to do with either modifying existing projectors or alter the fabrication procedure.
As mentioned before i think there might be possibilities to modify an existing projector but i will not try that myself.

Yes I think they modify one of the 2 projectors with a different polarization panel, if those 2 korean engineers can do it, why do you have so little faith in yourself?!?! I have faith in you Likay, you can do it! Barco did it, Nubie did it with lumenlabs projector, you can do it too!! Imagine the increased efficiency! Life without risk is not worth living no?

[Tril]

In this paper, some experiments have been performed to test operational characteristics of the proposed novel Type-1 LCD-based polarized stereoscopic projection system. That is, operational mode conversions of the modified LCD engine due to 90° rotation of polarization orientations of the analyzers have been experimentally confirmed. Practically it was not easy to modify the internal LCD engine of a commercial projector for our experiments that a LCD engine module was externally implemented by using a Type-1 LCD panel and a pair of polarizer and analyzer.

They did not modify a projector, they used an external setup that's gives equivalent results as modifying a projector.

In the experiment, a LCD engine module for one color component of the image has been implemented and tested to prove a feasibility of the proposed method, but the LCD engines for the other colors can be also tested in the same way. Accordingly by combining all color components from three kinds of LCD engine modules full-color stereoscopic images can be finally viewed with much improved light efficiency.

Looks like it worked in their experiment. That article is three years old. We might eventually see a commercial product using the proposed method.