RealD LP Clone, Cheap Zscreen, passive polarization rotator

[cybereality]

Mathew Orman very rarely comes here (or if he does he doesn't post). He's much more active on AVS.

[opolog]

Check this out. Cheap polarization alternator.

http://displaydaily.com/2011/04/06/3d-e ... n-rotator/" onclick="window.open(this.href);return false;

http://www.3dexperience.co.uk/RotatorScreens.html" onclick="window.open(this.href);return false;

[AntiCatalyst]

Shutter glasses have a linear polarising filter on one side, liquid crystals on the inside and another linear polarising filter on the outside. If I'm not mistaken, the liquid crystals rotate visible light 90 degrees.

If you can remove the polarising filter on one side, you could use shutter glasses as a polarization rotator. You put the side with the remaining filter toward the projector and the side without filter toward the screen. I see three possible problems.
1 - It might not be possible to remove a filter.
2 - The lense might not be big enough.
3 - The lense might overheat and burn. You need some cooling system.

Since they are shutter glasses, you know that they are fast enough to do the job and there is already a way to sync them.

What'd be great is if you could remove both polarizers and use a regular glass polarizer between the spectacle lense and the projector lens. The polarizer is what absorbs most of the light, so having the LC unit by itself = very low heat buildup.






EDIT: seems like i'm 6 pages late >.< oh well, might as well continue

Oh well, too bad. I thought I was on to something but no. I thought you could wear special 3d glasses that had retarders on them as well as polarizing film and that it might have fixed the color problem and then you could use really cheap LCD glass at the projector. The retarders on the glasses would be unpowered and not shutter. Anyway, I thought the info was interesting so I posted it.

You are basically describing circularly polarized glasses, like the ones RealD and MasterImage use - they have a quarter wave retarder in front of the polarizer. You'd have to use circular polarization(can be done with polarizer+lc retarder) instead of linear, but maybe it'd help your colour issues.

I'm hopeful about this

[tritosine5G]

What'd be great is if you could remove both polarizers and use a regular glass polarizer between the spectacle lense and the projector lens. The polarizer is what absorbs most of the light, so having the LC unit by itself = very low heat buildup.

you have a confirmation bias, lol.

polarization modulator is early 80's tech really, slightly better than LC shutters (and much worse than non LC shutters) but without good digital front end its nearly the same cause you can't use high refresh rate (modulator should work above 400hz, your average LC display panel won't )

[AntiCatalyst]

what do you mean? confirmation bias?




An LC shutter is a polarization modulator, just with polarizers on each side to block the light.

[tritosine5G]

oh uh , and where s the glass polarizer?

[Likay]

Shutterglasses layers: Linear polarizer, electrically controlled retarder (the liquid crystal element which works as a qw or hw polarization modulator) and finally yet another circular or linear polarizer depending on the lcd element.

[tritosine5G]

point is , theres no glass polarizer in ordinary shutterglasses afaik

[Likay]

I believe Anticatalyst mentioned glasses polarizers because they're more resistant to radiation than plastic which would be practical in the application in the discussion.
The active element in shutters is indeed a polarization modulator (if qw then linear pol to circular, if hw then linear to linear rotated 90°). However i'm not certain if the panels of shutterglasses works in a quarterwave or halfwave mode. Both solutions are possible since circular light relatively easy goes through a linear polarizer so it's open to inflick more info.
To make a modulator using shutterglasses you need halfwavemode. It's not worse than needing to stack two panels after another to achieve hw if they're originally qw's..

[tritosine5G]

we need better shutters, I dont like the grey tint of silverscreens, and I dont like the polarizer tint either , so thats about it , and 2D CRT is best, colors could be even better though.

[Likay]

Bouncing back... :

you have a confirmation bias, lol.

Not everything has to do with crt-vs-dlp-vs-lcd-vs-plasma whatever thingy.

[Likay]

Dbl post sry but more info about maybe one approach:
A way to find a bigger polarization rotator is getting a weldinghelmet for modification. There are relatively cheap ones out there or if someone could get a used one where the lcd-element still is working. The risque is pretty high that the drivercircuit needs a total modification if the original shifting speed isn't enough.
Active welding helmets surely sunk in price. A tad expensive for a test (799:-~90Eur) but it's certainly affordable: http://www.biltema.se/sv/Verktyg/Svetsn ... hjalm-WH2/" onclick="window.open(this.href);return false; and definitely cheaper than the depth-q polarization modulator.
Even cheaper (499:-~55Eur)... : http://www.jula.se/svetshjalm-212035" onclick="window.open(this.href);return false;

[AntiCatalyst]

point is , theres no glass polarizer in ordinary shutterglasses afaik

no, they're probably plastic.

[Fredz]

I wonder if anyone has ever tried a rotating polarized wheel synchronized to the vertical retrace of a projector.

I guess it could be much cheaper than electronic LC modulators, but I've no idea if it's easy to synchronize with the vertical retrace. I suppose it should be feasible though considering DLP projectors work this way with their chromatic wheel.

Could be nice for an Infitec/Dobly single projector solution too.

[AntiCatalyst]

a polarizer wheel may sound good theoretically, but in reality it wouldn't work well. Even though it's called circular polarization, there's more to it than that. Since a wave retarder is calibrated for just one specific wavelength, only one color(mostly green) can be made totally circular, the others become more or less elliptical.

In more practical terms, it means you still have to align the filters 90° from the glasses, otherwise you get purple ghosts.

Totally unneccesary explanation:
Since the wave retarders in circular polarizers are calibrated for green, green gets circularly polarized, and is absorbed any way you rotate them. Red and blue get elliptical waves instead, so they have to be "re-retarded" in the opposite direction(ie glasses or filter rotated 90°) to be fully absorbed by the polarizer in your glasses. Red light+blue light=purple ghosts!


an Infitec wheel would probably work great though

[Likay]

The fact that all colors aren't polarized the same way does always happen when using circular polarization. If you ask me it's a minor issue while the "washed out" image and ghosting are bigger ones.
Syncronizing the wheel is certainly not an impossible mission. "In younger years" i made a circuit that controlled the speed of a motor and synced it perfectly to a given signal. Simple said it's a pll-circuit that controls the powercircuit of the motor (the motor functions and replaces the vco of the pll-circuit). In this application the pll should get the signal from the vsync and from the motorwheel. Use a 1-bit counter to divide the signal from the vsync to make the engine half the speed of the vsync. The signal can be delayed from either the vsync or the motorsensor to provide finetuning (or easier mechanically adjust the sensor on the wheel).

I also think that an infitec solution with a single projector with one colorwheel probably would be optimal. It's not suitable for the average diy'er to do it since it also requires different hardware for the projector.

I'm curious of the quality of a polarizationrotator. Since it's basically a number of retarders (3?) i wonder if the polarizationproperties could be any good?

[Fredz]

Thanks for the answers, looks like my question was not totally stupid at least.

I also think that an infitec solution with a single projector with one colorwheel probably would be optimal. It's not suitable for the average diy'er to do it since it also requires different hardware for the projector.

I don't understand what you mean by "requires different hardware for the projector". Basically you would only need a standard 120Hz DLP 3D projector and a rotating wheel placed in front of it and connected to a power source and a VGA port.

Looks more and more interesting since I read the article about Dolby3D talking about lowered prices for their glasses. At $12 the pair this could be a very inexpensive solution.

[AntiCatalyst]

I'm curious of the quality of a polarizationrotator. Since it's basically a number of retarders (3?) i wonder if the polarizationproperties could be any good?

Well, if you've ever seen reald, masterimage or digital imax3d, you've seen a polarization modulator in action. They all use a single projector(i think, not sure about imax).

They do work, i'll tell you that much There is ghosting of course, since it's still polarized, but it's not nearly as much as i get at home on my own rig.



EDIT; fredz, i think he meant that replacing the actual color filters inside the projector(the way infitec do it) is not for the average DIYer. A separate infitec wheel is probably much easier to actually pull off, but still not easy

[Likay]

I don't understand what you mean by "requires different hardware for the projector". Basically you would only need a standard 120Hz DLP 3D projector and a rotating wheel placed in front of it and connected to a power source and a VGA port.

I expressed myself poorly . I meant replacing the existing colorwheel with one that has the 6 basecolors of infitec. Since it's 6 colors instead of 3 i figured that the electronic of the projector needs to be adapted/changed. Ironically when thinking more about it, it should actually work well this way if the beamer is feeded with a stable l/r/l/r...signal and make the original engine with the changed wheel go half the speed. Then the only problem is syncing the wheel to left and right.

While dolby/infitecfilters together with the original filters works, the best is definitely if the original filters is entirely exchanged. I tried dolbyfilters on my own projectors some time ago but my results are definitely not as satisfying as Jackbauers. The red nuance in one eye was almost killed and forced me to damp the other colors quite much to maintain balance which resulted in a quite dark image. I had an idea of trying to modify the projectors but it stopped in my thoughts.
I think these results vary widely depending on what projectors that's used because of different basecolors between projectormodels. I probably just had less luck.

@AntiCatalyst: Thanks for info about real-d cinemas etc. I wasn't sure that a polarization modulator always was used in those cinemas. About ghosting i'm almost certain that they've applied some kind of antighosting algorithm adapted to the cinemarig. It's only an assumtion based on observation though: While there's minimal ghosting when watching the movie, the subtitles when not aligned at screendepth sometimes ghosts just awfully.

[Fredz]

I guess the easiest method would be to have an external color wheel, modifying a projector seems a little bit complicated. And using a 120Hz projector probably won't generate the color problems you experienced I guess. Funny nobody tried to do that yet, maybe it's only a question of time...

[Likay]

Using infitecfilters on an external wheel would bring colorproblems since it cannot be adjusted other than adding colorcorrection in a stereoplayer or stereodriver. If adjusting the driver or colorprofile for the projector, both eyeviews will be affected. Circular polfilters on the rotating wheel would work though.

[Fredz]

Ah yes, I didn't think about that. I guess it can be solved by writing a pixel shader which would use precomputed look-up-tables for color correction. That would need to be done just before the images are displayed though (different LUT for each eye), but I don't know if it's easily done on top of the NVIDIA 3D Vision driver. Food for thought anyway...

[Fredz]

I did some reading about Dolby 3D setups in theaters and in fact they also use a rotating color wheel with a single DLP projector.



I guess it should be quite complicated to find such a color wheel for cheap or create one using Dolby 3D glasses though.

[Likay]

Nice find. The filterrig for the diy'er could maybe be done using lenses to first focus the beam and then use small glassesfilters attached together to form a similar "wheel". Cutting the filters won't be necessary but balancing would. Then attach one extra lense after the focalpoint of the beam to gain screenfocus again. I don't know how much the extra lenses eventually would distort the image though.

[Fredz]

Maybe that could be possible with prisms (DIY with glass and glycerin) to reduce distortions, but finding a color wheel would be less hazardous I think. I didn't find any place to buy one though, maybe I should try to contact Dolby directly. Dolby glasses can be found for less than $20 on ebay now still : http://cgi.ebay.com/DOLBY-3-D-Digital-C ... 0269826543" onclick="window.open(this.href);return false;

[Fredz]

I've contacted Dolby and they sent me their list of resellers for the color wheel. I'll contact them to see if the prices are affordable.

[GPLavoie]

I'm curious of the quality of a polarizationrotator. Since it's basically a number of retarders (3?) i wonder if the polarizationproperties could be any good?

Well, if you've ever seen reald, masterimage or digital imax3d, you've seen a polarization modulator in action. They all use a single projector(i think, not sure about imax).

They do work, i'll tell you that much There is ghosting of course, since it's still polarized, but it's not nearly as much as i get at home on my own rig.

Our Natural Resources and GIS departments each purchased projection systems with polarization modulators. The quality is as good as 3D Vision Pro, expect the screen is huge. Works great in the classroom. Expensive, but a great solution.