I'm continuing the discussion from the ceramic lamp thread about the experimental use of Super High Contrast polarizer film from polarization.com. Elken shows good results can be acheived with replaced polarizers (both sides) without having to mount them directly to the panel. The unique finding is that he mounted there polar axis at 90/180 degrees versus the factory 45 degrees angle and produced very good if not better contrast and color output. Elken, please correct me or continue your findings here if you can.

A pair of replacement Super High Contrast polarizers also show initial tests of 10-24% greater panel transmission. It seems Elken is still looking for negative side effects of this configuration. Here's a couple of thread pages reguarding his findings:

At the bottom of this page.
http://www.lumenlab.com/forums/index.php?s...358&st=1800

http://www.lumenlab.com/forums/index.php?s...358&st=1840

Simul8r, you said your also interested in Elkens SHC Polar findings, what's your opinion about Elken's findings so far?

Is there any disadvantage to mounting these directly to the panel? Is mouting the polarizers to the panel a tedious/hard procedure? It seems to me that once an optimal polarizer setup is found it would be less work to mount them to the panel rather than finding a way to secure them in the projector by themselves.

Also, elken seems to be the pioneer of this but uses a ceramic light, so will benifiets still be seen with the standard 400w T15 lamp?

In my experience with replacement polars, I tried to mount them as close or on the panel. It seems from Elkens experiments that there may not be a real visual benifit to direct/close polar mounting. He's seemed to get good results putting them at the rear fresnel and triplet. I never tried moving them away from the panel like he did and thought they needed to be right at the panel for clean results. If you want to adjust there rotation separately from the panel then it's less work to have their own mounts separate from the panel. I also found mounting the loose polars in the panel frame created uneven wrinkles at the corners, they were minor but may have caused "glowies".

Yes, he does seem to be using a ceramic lamp, it's main benefits are smaller arch and better color/sharpness rendering. I don't know if the polarizers react different to ceramic archs.

mounting it before the triplet might cut down on unwanted light making it into the projection..

Here's a few base lines of discussion:

1) Mounting replaced polars at 90/180 degrees versus 45 degrees.

2) Super High Contrast Polars and their transmission/contrast benefits.

3) Replacement polars locations and mounting techniques.

Here's a good post by arizonavideo:

Elkin: In post http://www.lumenlab.com/forums/index.php?s...st&p=189063

You say that this is at 45deg; does this mean that both polarizers were on the panel in the stock 45deg position? If so why is the picture so bad? Or were one polarizer at 90deg and one at 45deg?

I have a cracked LCD and some polarizer so I will do some testing on the polarization testing but the LCD is cracked and does not work so some test I can't do unless I skin the back up LCD (also bad but still works).

How transmissive is a LCD with no polarizer?
How transmissive is a LCD with both polarizers at 90deg?
How transmissive is the LCD with both polarizers at 45deg?
How transmissive is just two polarizer with no LCD?

And then there is the contrast and color question which will need a working LCD.

We will never get more light through a LCD than the bare LCD - the two polarizer losses.

OK a quick LUX test using just a bare panel with the LUX meter right behind it with a 500 watt lamp shining on it from about 5 ‘ away. The panel has no polarizer on it.


Raw LUX no anything 1925
LUX with no polarizer just the bare LCD 406
LUX with two polarizer at 45deg ( Stock). 146 to 150.
LUX with two polarizer’s at 90deg 151.

I saw no LUX gain no mater what rotational position I used. I ran it through 180deg.

So it looks like the 90 vs 45 degree mounting doesn't change panel transmission. If 90 degree mounting works, it will be easier to cut replacement polars with 90 degree cuts - that's a benefit of 90 deg polar mounting.

Is anyone using a pair of replacement Super High Contrast Polarizers and acheiving better brightness and contrast. I know Makau did some tests, but I'm not sure if he replaced both sides with the SHC Polar. If you do use them please note your mounting technique and estimated gains.

That's very interesting and odd that the 90deg polars are no more transmissive than the 45 deg. Maybe the brighter projection that elken is getting is due to the polarizer being high contrast and has nothing to do with the degree?

Looking at those lux numbers it does seem that getting a more transmissive polarizer would be a GREAT way to increase the brightness of the projection, maybe even doubling the lumens @ the screen!

I didn't think 90deg polars would increase the brightness, though elken did show nice color and contrast with the 90deg config. I think it is because of the SHC Polars, that's what I'd like to confirm. This could be done by comparing a pair of SHC polars with standard replacements or average factory polars.

add me to the list of people anxious to try this (I have to anyway for my 3d stereo project). Also, since orientation doesn't seem to have any detrimental effects, couldn't we orient the polars to adjust for color temp BEFORE backing off the R G or B settings in the LCD (which cuts light)?

hey whoaa

90 degrees hasnt shown to be increased brightness,, too many changes all together have.. mostly usage of SHC polar, and experimental placements and dialing etc.. etc

I didn't think 90 deg polars would increase brightness. It seems the main benefit to 90 degree replacement polars is that they are easier to cut from the sheet since you cut them 90 degrees instead of 45 degrees.

Elken, can you generally confirm these findings?

1) A matching replaced pair of polarizers can be positioned at 90/180 degrees to most LCDs panels' pixel-grid.
You have to replace both polars to make this work. (Most factory polars are mounted at 45 degrees.)
This makes cutting the replacement polar sheets easier and cheeper.

2) Super High Contrast Polarizers have the highest light transmission to contrast ratio, making them most
desirable for polarizer replacement. Some brightness gain may be possible given their specs.

3) Mixing pairs of polarizer brands/manufacturers can lead to more artifacts ie. glowies, barring etc.

4) Replacement polars don't have to be placed directly on the panel.

5) Polars mounted separate from the panel can be rotated or "dialed in" to affect color temperature and black level.

6) The farther the polars are from the panel, the smaller they have to be.
(but not smaller than triplet or the fresnels focal cone.)

7) Proper focal placement and alignment of components (reflector/lamp/pre-con/fresnels/polar/lcd/polar/triplet) is crucial for finding the "sweet spot" with the best projected result.

It sounds like a new prescription drug ad .

Your research is helping greatly, I'm just hoping to summarize what I've observed during my own trials and what you're finding out. I agree that their are so many factors that affect the final picture quality and Super High Contrast Polars are NOT a magic fix-all. I hope you can share more about the location of your replaced polars and the best working combination.

Thanks

Aaron

As far as the brightness increase from the new polarizers, I'm guessing it's just a matter of the light transmission properties of the polarizer. For example, the SHC polarizer from polarization.com says that a single sheet is 43.4% transmissive, two parallel sheets are 37.7% transmissive and two perpendicular sheets are 0.0078% transmissive. I don't know what affect it has but the polarizing coefficient is 99.979%. (As an added bonus, they say it also blocks UV).

I will have to edit this later for comparison of 3dlens.com polarizers because I can't reach the site at the moment, however I believe their polarizers were of lower transmissions (maybe low 30s to high 20s).

I agree, it is a matter of light transmission properties. Not only do they let more light trough, they also block almost all the light when crossed. So their range from open to closed is a bigger window than any other polarizer available by the sheet.

So this leads to the big questions:

1) Is it worth the $75 plus shipping to upgrade average panel polarizers to a pair of SHC, thus making an average panel 7-8% more transmissive? (given the old polars have average transmission specs)

2) Does switching from adhesive factory polars to loose polars cause a drop in transmission because of the spaces between polarizer/panel/polarizer?

3) If yes, is this split polar loss less than 7-8% or the benefit of SHC polars?

4) Who wants to spend $75 and risk the panel strip to find out?

[quote name='Sinner7' date='Nov 30 2006, 08:18 AM' post='189933']
So this leads to the big questions:

1) Is it worth the $75 plus shipping to upgrade average panel polarizers to a pair of SHC, thus making an average panel 7-8% more transmissive? (given the old polars have average transmission specs)
[/quote]
This is THE question that I've been trying to reason with for the past week. If it was a full meter length then maybe.
[/quote]
2) Does switching from adhesive factory polars to loose polars cause a drop in transmission because of the spaces between polarizer/panel/polarizer?
[/quote]
This is why we need some measurements but it seems to be a waiting game.
[/quote]
3) If yes, is this split polar loss less than 7-8% or the benefit of SHC polars?
[/quote]
Very good question.
[/quote]
4) Who wants to spend $75 and risk the panel strip to find out?
[/quote]


Keep the interest going Sin, it's turning out to be good thread.

Couple quick calcs:
Assumptions - LCD trans = 5%
- Orig Polarizer trans (from 2 in parallel) = 28%
- SHC Polarizer trans = 37.7%

So...

5% LCD after removing 2x polarizers = 17.86%
After added new polarizers = 6.73%

Total transmission improvement = 6.73/5 = 134%

So if you had 80 ANSI originally you would now have ~108 ANSI.

You make the call. And it's possible that the original polarizers were better than 28% when using two in parallel.

I for one am more than willing to pay $75 to roughly double the lumen output of the PJ. Especially if doing so also increases the color quality!

[edit]
It's my understanding that the average LCD is ~7% transmissive.. so it would be ~14% transmissive after the SHC polars right?

Here's another way to do the math:

A) 100 units of light = all the light hitting the first polarizer, we want to keep all the units we can.

Average polars absorb 70 units.

C) Average naked panel absorbs 23 units.

D) Average panel with factory polars absorbs 93% of light units or 93 units. The only leaves 7 units for the triplet lens.

E) Super High Contrast Polars only absorb 62 units versus 70 units!

F) Average panel with SHC polars absorbs: 23 + 62 = 85 units

G) Average panel with SHC polars leave 15 units of light versus 7 units. That's more than twice the units of light passing through the panel.

This seems to good to be true. Naked panels may absorb different amounts of light so that's the variable here that can drastically change the numbers. What's your naked panel vs polar mounted transmission?

Does this seem right?

http://www.sid.org/chapters/uki/presentati...splaysearch.pdf

side 14

It states that polarizer(s)? pass through 40% of the light.

the values on the left are showing the light passed through by the certain layer.

The values on the right show the total amount of light passed through.

Now you just have to put the transmittance values of the SHC and standard polarizer to the

“polarizer” and you will get the difference between the,.

For example: (lets forget the “lite pipe” cause we don’t have it)


SHC: (37,7% trans)

100 x 0,377 x 0,95 x 0,30 x 0,90 x 0,90 = 8,703045 %

STANDARD: (28% trans)

100 x 0,28 x 0,95 x 0,30 x 0,90 x 0,90 = 6,4638%


I would get it this way NOT DOUBLE values...

Comments?

Immediately tell me if I misunderstood the chart...

So that comes out to ~35% more light?

I like the results from Sinner7's calculations, but which method of calculating is right? hmm..

Maybe there are even more transmissive polars available?

[edit]
Another question to think about:
Does the rated CR on our monitors have anything to do with the factory polars? (it sounds like the polars are what determine the monitor's CR in the first place) My MVA monitor is rated at 1000:1 CR, so I'm just wondering how much of an advantage I'd get from this.

I wonder if you actually gain more than that though (as Elken has seen). When you remove the polarizer, you are also removing the TAC layers aren't you? If I'm reading the pdf correctly (p4), these TAC layers are shown as CV film (anti-reflective) and WV film (viewing angle enhancement). When replacing the polarizer, you're only adding the PVA layer aren't you? or are you adding the others as well? I don't have a monitor to test on yet...

I don't think you read it wrong, their big factor is the color filter eating 70% of the useable light. My math show the panel (crystals & color filter) only eating 23% of the usable light.

Another question needs to be answered by lumenlab optical engineers

What is standard procedure for testing a panel (with factory polars) for it's transmission value?

How did people arrive at 7% for average panel transmission? I know this is in the forums somewhere.

We can't do any math correct until we have a standard way of getting our transmission measured, with and without polars. I believe people used direct sunlight for a test backlight, was that the common practice? Was the display powered with a black screen? Adding current to the liquid crystals obviously changes the transmitted light.

Lothar, You are right. but I don't still get that double value...

I doubt this is the "standard procedure" but I would measure lux at the screen with the LCD in place (maybe measure LCD on and off) and with no LCD. That should show how much of your light is lost in the LCD (thus transmission value).

Here's the thread on transmissive testing:

http://www.lumenlab.com/forums/index.php?s...c=10164&hl=

I noticed they didn't have a standard procedure such as:
Light source: They used lamps, sun, projector box
LCD Power: Some powered them, some didn't
Lux Meter: inconsistant censor placement

I believe these items may greatly effect the accuracy of the readings. This can of worms is bigger than it looks

Here are some real world figures you can play with.

Two different methods of measuring LCD transmittency on two different LCDs'

1 - Sun test
2 - On screen 9 point method with LCD in and out of projector.

BenQ 567S v2 (1024x768)
(One polar with anti-glare removed and replaced with floating 3DLens, caused unacceptable glowies)

1 - 6.49%
2 - 5.90%

Samsung 740B (1280x1024)
(Both polars and anti-glare intact)

1 - 5.15%
2 - 4.40%

edit: Multiple readings were taken and averaged, there were no large discrepencies in individual readings.

Let's try the Samsung 740B, Both polars and anti-glare intact

I'll use the 100unit logic, meaning the amount of light hitting the first polarizer = 100 units

Assumptions: Average polars pass 30u, each one absorbs 35u or 70u combined

Average panel trans = 4.78 units
100u - 4.78u = 95.22 units of light absorbed by polar/panel sandwich

2 Polars absorb 70 units so 95.22u - 70u = 25.22u is the units absorbed by the panel/crystals/color filter

So the "light unit" consumption of this panel would be calculated:

100 units to Polar A === 100u - 35u = 65u
65 units to Panel, crystals, color filter === 65u - 25.22u = 39.78u
39.78 units to Polar B === 39.78u - 35u = 4.78u

4.78u Output

This panel with replaced SHC polars (not factoring reflective polar gaps)

100 units to SHC Polar A === 100u - 31u = 69u
69 units to Panel, crystals, color filter === 69u - 25.22u = 43.78u
43.78 units to SHC Polar B === 43.78u - 31u = 12.78u

12.78u Output

12.78u - 4.78u = 8u per 100unit increase with new polars

I'm just having fun here, I should be working.

Now that math doesn't mean double the lumens, it means this for a naked panel that absorbs 25.22units:

For every 100 units of light, you'll get an 8 unit increase in brightness from the original polars.

So in theory, if your projection is 200 lumens then switching to SHC polars should increase your lumens by:

200/100 = 2 x 8 = 16 lumens

So you'll get a 16 lumen increase, is that worth it? maybe

If you start with 300 lumens you'll get 3 x 8 = 24 lumen increase

This is a very good point. Maybe the increased brightness elken is seeing is due to not just the higher transmissive polars, but the absence of the CV and WV layers. And even if he hasn't removed those, he could remove them to get a further increase in brightness!

Commercial projectors put out insanely high lumens while DIY projectors don't match up despite our more powerful lamps. The same physics apply to both breeds of projector, so maybe we've finally found what's been holding the DIY community back (from high lumen output).

I don't think that math is quite right. If your projection is 200 lumens, that would be equivalent to the 4.78u output of the LCD with original polarizers. If your new polarizer combo outputs 12.78u then your lumen output would be 12.78/4.78*200 = 534.73 lumen.

However, I don't think that your equation for the difference in transmittance is quite correct... But I don't have better yet.

To add to the stew... The info for the 17" LCD polarizer from 3dlabs.com: Transmittance (single) = 42.5%, transmittance (parallel) = 36%, crossed = 0.05%. Polarizing efficiency = 99.72% (maybe that is part of it...), absorption axis = 45% (another part?).

3dlabs.com 210x300mm polarizer: transmittance (single) = 38%, parallel = 30.1%, crossed = 0.0045%, polarizing efficiency = 99.98% - no data for absorption axis.

And as a reminder, SHC polar from polarization.com: trans single = 43.4%, trans parallel = 37.7%, crossed = 0.0078%, polarizing coefficient = 99.979%.

If he was using the 17" polarizer replacement first, then swapped to the SHC polarizer, I don't see that much difference in transmittance. Main difference is polarizing coefficient: 99.72 vs 99.979... Maybe if it was the other polarizer from 3dlens... There must be something else to it... Replacing the original with either should help though it looks like.

I think the math pans out, it's tricky though I may be wrong. If you divide 12.78 by 4.78 you get 2.67 which looks like 2.67 times the original brightness. But the answer is based on Xu's per 100 units, with a difference of 8u's per 100 units, the answer is much smaller. I wasn't looking for how many times the units multiple, I looked for the increase in units per 100 from point A (4.78) to point B (12.78). Given Elken and Makau's pictures, I see a definite gain but I don't think it's 267%. Maybe there's nothing to this, I'm not trying to start a debate. Just contemplating stripping and using the SHC polars. The 17" polarizer has similar specs but I think it's adhesive. Ahh... feel the obsession!

Just in case anyone is interested in how polarizers are used in LCDs (as I was) here are a couple links that I found very informative:

TN-LCD technology side 2

How do liquid crystal displays (LCDs) work?

These links explain why Elken got a negative image when he had the two polarizers in parallel and a positive image when one is rotated 90 degrees. (If anyone needs further explaination, I'll see what I can do.) I also think that the final contrast is related to the polarizing coefficient. The higher the coefficient, the higher the contrast.

Edit: I should really finish my thoughts before posting...
When Elken was testing the polarizers, he had them at the precon and the triplet. Did he test them directly at the front and back of the LCD? Perhaps the reason that the original polarizers are at +/-45 deg ( / or \ instead of | or -) is because that is the orientation of the LCs. As long as the polarizers are at 90 deg from each other, you should be able to get full contrast I think, but perhaps rotating both polarizers together changes the brightness of the output? When matched with the orientation of the LCs you would get maximum output but as you rotate them it decreases? This is just speculation since I don't have a stripped panel to try on. It would be nice to test this though.

But getting back to Elkens experiments with the polarizers at the precon and triplet, that should work fine (and obviously does) as long as the optics in between (fresnels) don't change the polarization. I don't know if they can or not.

The polarizing layer cannot physically support itself, its too thin, it needs a layer on each side(TAC) to support it. However, if the polarizer is adhered to the panel then the outer TAC layer can be removed exposing the polarizing layer. Others have done this, I think Sim was one but I don't think he noticed much difference.

If I understand correctly, contrast is determined by the polarizer AND the lcd's ability to change the polarization of light. With no change the crossed polars will eat all the light=BLACK, and at maximum change the polars will transmit the most light=WHITE. Greater ability to change the polarization of light=greater contrast.

I recently met a LCD display design engineer when I bought a LCD controller for my Sharp WUXGA panel, who introduced the idea of using additional polarizers to get more light through the panel. I have been given a selection to try out when I start to build my PJ. 3M seem to be one of the brands here in the UK.

One of the things that was explained covered the "percieved increase in brightness" which is caused by the polarizer reflecting the light that does not get through successfully back into the light source which then gets to eventually be reflected back onto the polarizer.

You also need to be careful with your observations when using smaller polarizers that do not cover the whole lcd panel, as this reflected light will also cause an increase to the light intensity hitting the area not covered by the polarizer, making it difficult to see if you are actually getting a more light through because of the polarizer.

Brad

These 'Brightness Enhancement Films' are only applicable to the original 'distributed' backlighting system and will have no use in the 'traditional' lens system we use. If you can create an ultrabright distributed backlight then you might be able to utilise them.

Contrast is the next subject on my mind. And I agree that contrast is a relationship between the polars and the liquid crystals polarizing ability. If the crystals can totally relax then the crossed polars can block the light to their maximun light absorbtion. If the crystals cannot totally relax then a small percentage of light is still projected. Has anyone measure the lumens at there wall- screen while displaying full black projection?

Does this seem right? Maybe too simplistic.

SHC Polars - Crossed transmission is .0078%

So...

If you project a full black screen with a panel that has perfect crystal alignment where 100 units of light are trying pass through the panel, then the projected light units are:

100u x .0078u = .78u,

so you'll still have .78 units of light per/100u which works out to a couple of lumens for the average DIY projector with great black levels. That seems pretty accurate with my projector's black screen, the full black screen still being a few lumens bright.

Another thing I notice about my black screen (my panel has the factory polars with AG removed), is my black screen has a very subtle blue hue to similar to crossing polars in front of a bright light.

The problem isn't so much with your math. The problem is that you figure out increase for 100 units coming into the LCD then you figure the change by assuming the 200 lumens coming out is the same as the 100 units going in. If you want to do it your way you have to figure out how many lumens are hitting the LCD from the light engine and divide that into your 100 unit blocks.

David.

So you're saying this:

13,000 lumens hits LCD

13,000 L / 100 u = 13 u

13 u x 8 u increase = 100 u or 100 lumens

That may be right, this only way to tell is to do some real measure with documented setups. I'm just not sure the 8% increase of SHC polars will lead to increases of 100's of lumens. I might just have to get a lux meter I guess.

Yup, that's the way it would work. Using the output lumens works also. But then you have to figure an 8 unit increase for every 4 units not every 100. It really is over a 200% increase if you follow your math.

And of course you are right. It will take real world measurements to really know what is happening. Maybe Lumenlab should have offered lux meters in the online store.

David.

Edit, except 13,000 / 100 = 130. But I think if you work backward from the 200 lumens output you will find it is probably more like 4,000(edit again) lumens on the LCD anyway.

If you hit 2 polarizers in parallel one after the other with unpolarized light, then the first polarizer will absorb far more light than the second. To be precise, with the Sanritz Super High Contrast Polarizer the first polarizer will absorb 56.6% of the incident light, and the second polarizer will absorb 13.1%. What this means is that the polarizers lose 13.1% due to inefficiency and the rest is just part of re-vectoring the light.

To figure out how much brighter your projector will be with full replacement of your polarizers to the Super High Contrast Polarizers, you simply need to research or calculate what the parallel transmittance of your existing polarizers are.

If you have a very inexpensive panel, you may have the Sanritz Standard polarizers with a parallel transmittance of 30.1%. To replace those polarizers with the Super Hight Contrast Polarizers would give you 25.2% more brightness. Your contrast would also obviously be improved giving higher perceived brightness as well. But this is an extreme example between the worst and the best in polarizers with reasonable contrast.

This does not take into account that you will loose around 5% in brightness now that the polarizers are no longer bonded to the panel.

Does that help? I'm not really sure what the questions are here?

Mark

Mark, thanks for your input.

Isn't that what the value of parallel transmission specs are (37.7% for SHC)?

Doesn't this spec take into account the first polar absorbing most of the incident rays?

Is this the right interpretation, for 100 lumens entering parallel SHC polars, 37.7% is passed and 62.3% is absorbed. Is the parallel transmission spec more complex than that?

That's how I arrived at 70 units / 100 units will be absorbed by standard polars. Or in the case of SHC polars, 62.3 units/ 100 units entering parallel polars are absorbed.

It makes sense what you are saying, I just thought the parallel specs represented a pair of polars working together.

Nope. You have it just fine.The single greatest thing holding a large panel projector back is the ratio in sizes and distances between its components. In the end, the usable area of light at the bulb is so small as to be tricky to create efficient light engines. In a commercial projector, the usable area is bigger, and the arc is smaller. This makes it much easier to build an engine that keeps light within the needed tolerances to reach the lens. This thread attempts to explain those factors:

http://www.lumenlab.com/forums/index.php?showtopic=15812

The other key factor is that we cannot do a 3 chip system easily, which are pretty much 3x brighter than a single chip. Or DLP technology which does away with the polarizers altogether for huge brightness increases.

Mark

So here's the big question:

Has anyone replaced both factory polars with SHC polars (non-adhesive) and documented before and after changes?


I've read most of the threads that mention the SHC polars and it seems people have only tested one side or a small piece on one side. It looks like some people ordered them recently, any results? If I did a total replacement, I would be hoping for deeper black levels using the SHC polars as my current brightness is very sufficient in controled lighting.

My setup is this:

250w 4000k MHL - Icecap E-ballast
8.6" 800x600, 25ms R, 300:1 contrast Sysonic (toshiba) panel
DYPC 18" Triplet
DYPC Pre-Con
Pro Reflector
220 mm Con Fresnel
500 mm Field Fresnel

This is where you will find them when they show up:

http://www.lumenlab.com/forums/index.php?showtopic=15700
Since your panel has just a 300:1 contrast ratio, I think it is safe to say you will see improvements in black levels with the SHC Polarizer. Here is something to bear in mind: You do not need to remove the existing polarizers to find out. You can add the SCH Polarizer in addition to what you have. Of course, you will loose quite a bit of brightness (around a 30% decrease), but it may be very obvious at that point what is to be gained.

But really, I think it is very safe to say you will get improved black levels. And your panel is very small, so you may somehow be able to get some small scraps instead of a whole sheet.
Cool.

Mark

Can anybody clarify for me:

1) Is it safe to say the majority of LCDs have polars oriented at 45 degrees?
2) It is possible to *only* do a front polar replacement at 90 degrees? In other words, front and back polar orientation has to match, correct? Elken, when you talked about 90 degree orientation, you were replacing front and back, correct?

I have a 9 inch Toshiba panel and my AG removal went bad. Back polar is still untouched (though I have to say it is *extremely* reflective and more silver than front polar, making me wonder whether some other layer is on top). I'd love to only have to buy the 10" x 6" piece, but need a 9" x 9" if 45 degrees is required.

Mine isn't.
The replacement polarizer must have it's axis at the same orientation as the original. So you are going to need to figure out at what orientation the polarization axis of the rear polarizer is at. The exception is if you are replacing both, then the game is to keep both polarizers at 90 degrees to each other, but it shouldn't matter how they are oriented with the panel itself.
Could be, but the rear polarizer is typically going to have a nice polished plastic finish (but not metallic looking).

One thing that has been very fortunate so far is that everyone seems to have 45 degree polarizers. There haven't been enough panels done to know which is more likely. What is fortunate is that no matter how you cut the polarizer, it will match the other just by flipping it over. DAZZZLA was asking why some manufacturers choose the 45 degree axis. I think this actually could be a pretty good reason: every 45 degree polarizer cut can act as either a front or rear polarizer. The question then becomes, why would a manufacturer choose 90 or 180 degree polarizers? I would maybe think that this is because stretching sheets of polarizer length wise is easier, even though you have to cut two different sized polarizers in the end. Of course, it could be still be some sort of an optical concern, but just a thought.

edit the only way to know for sure what orientation you need when keeping the existing polarizer is to know the axis of your removed polarizer and match that and it's size exactly. Or to buy a large enough sheet for any orientation and rotate that sheet until the desired image is found.

Mark

Thanks for the complete reply Mark.



I assume there is no way to eyeball this, and that the only real way to know the axis of the removed polarizer is to get some definitive word from the manufacturer? Not particularly likely, I'd guess. In other words, the latter option is the only one available



EDIT: Here's the front of my panel, with AG and polar stripped:


And then the back, which is significantly more reflective and silvery. Is that likely to be AG on the back as well? This is a Hitachi IPS panel, if that makes any difference.

Nope, it's actually quite simple to do. Sort of.

Step 1: Set your panel with the rear polarizer side facing you and such that you can see a light on the other side.

2: Look through the LCD at the light through the polarizer you removed from the front (use the front polarizer as glasses).

3: Orient the polarizer in your hand such that you can see full brightness through the LCD.

4. Now, here is where it gets tricky. You need to flip the polarizer over so you are looking through the other side. So if the glue side was facing you before, you now want the side with the anti-glare facing you. Note the brightness of the lamp. Has it changed? What you need to do is figure out on which axis you can flip the polarizer over in space, and not loose any transmittance. 90 degrees to that axis is your polarization axis for both polarizers.

Make sense? It's a rudimentary thing that is ridiculously difficult to explain.

Note: polarizers are not always going to be at 90 degrees to each other. Super Twisted displays can have seemingly random polarizer orientations. You are always best to know exactly what the axis of the polarizer you removed was and work to replicate that.
I think that must be anti-anti-glare. Actually, that is probably a reflective style (as apposed to absorption) style polarizer. Very cool. Using a reflective polarizer is a form of brightness enhancement. It unfortunately only works if used in conjunction with a diffusing light source (like in the original computer monitor), but I definitely would not be removing that thing, it could be used to enhance your brightness should we find a suitable way to diffuse the returning light. People have been looking all over for those things in the past. It's the first one I have actually seen. Here is a quick experiment to see for sure if it is in fact a wire (reflective) polarizer:

Leave it on the table just like that and find the polarizer you removed from the front. Place that polarizer between you and the panel and see if it can block out the reflection in any orientation. If it can, then 3 things:

1. You have a reflective polarizer.
2. I had no idea they were that specular (non-diffusing and efficient). That is freaking sweet. .
3. Seriously, don't remove that thing -you have been blessed. .

Mark