Damm I still can’t see it. The only thing I can think of by your description, other than the angle of incidence at the polar bear, is the LF screen may be causing it. Try holding a sheet of paper at that position to rule out the screen itself.

DJ

Hang on I moved the pic across to my other monitor. Is this it?

DJ

Dazz look past the video box line,,, just past it too the right,, you will see what looks like a whitish bar also it is on left side too,,, but right on border of vert video box





Move more to the right beyond where ya arrow is pointing

its not the screen,,, ruled that out already,,, you may be right about steep angle of polar bear,, relative to theray path,, must be a critical point, where is fails

I would bet if throw was 1.5:1 it would not be there,,, this 1:1 is at limits anyhow,,,

there is not a trace along top or bott of screen

.

3D system: passive 3D systems currently function basically as mentioned (Though active systems are better because of the lack of poluting axis diffusion). No need to reinvent the wheel.

mikyd: I was only trying to say a true silver screen is needed to preserve polarization. Normal screens will not.

Large polarizer on the screen: Not good, if you ask me. Indefinately at least some light will be diffused back away from the polarization axis when hitting the screen surface. Brightness will therefore decrease with no predicted improvements so far as I can deduce. The extreme of this is to imagine placing a large polarizer over a fully diffusing sceen (like paint/blackout cloth):

1. Light passes the polarizer, and only the energy encompasing the polarization axis exits.

2. Light impacts the screen, and returns to the polarizer as unpolarized light.

3. Light passes the polarizer a second time. Only the energy encompasing the polarization axis exits.

I deduce this would result in a considereable loss of contrast/brightness.

Artifacting: Any one of:

a. Polarizer viewing angle inefficiency.

b. Any form of polarization change at the fresnel. Be it stress birefringence, birefringence, or simple diffusion.

c. Or just a simple glare issue.

These issues can simply not exist when isolated before the fresnel.

Cooler Panel: same reasons mentioned Here

Cooler Light Engine: . Unless this is a mis reading, I'm with DAZZZLA on this. You have not only isolated the heat from the panel assembly, but have also isolated it from your cooling system. The front polarizer won't generate anywhere near the same heat as the rear polarizer, though. And only on dark scenes. 20 degrees sounds a bit impossible.

Quarter wave retarder: Would produce circularily polarized light. This cannot be the case. Half wave retarder would twist the polarized axis of vibration 90 degrees. Anything else would give you elliptically polarized light, so far as I can tell.

Yep. Some confirmation here would be nice:

Elken:

If you hold the polarizer on one side of the fresnel, and dial it in to a high contrast image, does that angle have to change if you switch to the other side? If there is a 90 degree difference it has quarter retardation properties (Quarter Wave Plate).

If so, which side seems to be the closest to 45 degrees (or any multiple) to the panel?

And were you able to salvage your BenQ front polarizer?

Mark.

always 45% as per the clean test piece,,, bit of a pain would luv to recut,, but you can see thru pic hangs there nicely ,,,hmmmmm

either side of fresnel is still 45%

yes Benq orig is in pristine condition,,,

already did a halfand half test to compare,, as i said earlier min 20% increase in briteness,, no more than 45%

its the colours and whites that are stunning,,, and browns,, which usaully dont show well in dark scene's are coming fantastic,,

blacks are quite good too,, that was the biggest worry,,,

its a bit like going from video tape to dvd, in effect....

Elken: are you sure the edge brightness goes away when placing the polarizer between the fresnels? If this is an error, then I would think you may have an exagerated light leak at the edges. Now that the edges are more transparent after removing the polarizer, any light leak there that isn't masked could bleed into the image (reflects off the condensor).

I hope you mean you have tried the polarizer seperated and between the fresnels now (as per The Theoretical Ultimate).

Edit: and it's all about The Nemo.

Mark.

yes gone when between lcd panel bare side of course and field fresnel

width of screen compared to hieght seems to show that there is a critical limit on screen side of filed fresnel

so close it is

tomorrow i will test this fully,, been to bust watching various movies and etc. to get a good judgment of any funnies,, so farvery positive,,, other than the glow's left and right sides.. as I have said 100" diag screen,, only 4" glow,,, hehhe

clive


I know its a silly point,,, but I reversed my panel so testing bare side is easier, with smaller clean polar bear

had all sorts of probs getting image to mirror,, grr found it hiding in keystoning,, (nvidia),,,, hated doing it butsuch as life...
clive

Mark no matter what,,

Proof of concept,, is now proven for sure,, no doubt whatso ever

best position so far,,hmmm take front fresnel and attach new full sized polarizer to it,,, keystoning,, hmmm maybe not, nah the amopunt of angle is not that bad


thoughts on circular polarizers, for just behind triplet??
might overcome the glow left and right,,

geez watching ep of 4400,, the guy who ran the 4400 centre is wearing a blue shirt the kind that stands out in real life,, well it stands out with pop, here,, watched this ep several times before,, use it as reference,,, never seen such pop and clarity before tonight,,, this is a big bonus,,, mind you I figured that ages ago too... logical

clive

heheh ok got that nemo,, but will do morrow,,, my network is driven by a 2003 win server... cant be bothered right now to put on my public folder and then display then photo the port cam then resize it then post gawd, just thionking about that makes me tired'er

it is painful to experi, then pic then post,, hehehe takes four times longer to do anything

clive


but this has been an international time disconnected experiment,,


hey what i say,, no more an experi,, it's darn real for me,,

You still need both polar bears right?

This is what I'd like to know. Will the polar bear work when strapped down to the field fresnel and tilted up to 15º? And is the viewing angle of the panel wide enough that it could be tilted as much as 7º?

With the films removed; will all of our panels be pretty much the same conscerning viewing angle?

Ahhhh. That feels better . A lot better.

Just to be sure: you have it on the panel side of the fresnel?

The bright edges should disapear? No?

Yeah, we will need to figure out the keystoning issue. I have no ideas here that don't involve iether just putting the thing up against the panel again (not good, as deduced earlier) or adding a new layer of glass. Although, seeing as how you are having descent polarization with the thing all floppy like that, one would think a frame could be built no problem, without a layer of glass.

QUESTION: are you finding the screen to be any sharper? If so, do you feel you can attribute it to the panel reversal, or just to the modifications?

QUESTION: do you feel This Polar Bear would be big enough? Please keep your response seriouse. This is very important. I spent a lot of time looking for it. I have it on my desk right now. Looks to be 90 degree polarized. But there have been a couple unexpected issues.

Mark.

I am laughing so hard here, i nearly fell off chair,,,


my gawd you found the daddy of all polar bears,,,,,


I LUV IT........... saved for a cold day...

hmmmm we think of warm blondes,, and bears well !!!!!!

what makes it funnier is I just reread your post, and laughed even harder,,,

double catch on me hehehehehhehehehehehe

It looks like you've got this concept working now.

Perhaps you should go back to the other thread and make a basic step by step tutorial for how to do it. Where to order the polar bears, and where you can, and can't place the polar bear. Also some tips about safely removing the antiglare surface.

Dam, that would be a huge 'Haas' (get it?) box.....hehehe

The antiglare filter and polarizer are removed from the front of our lcd's (which is actually the back in our projectors, the side facing the bulb) why can't we simply put the new polarizer back on the lcd right where it was?

It looks like your saying there's some uneven lighting at the edges where light is moving in at the sharpest angle. Wouldn't it be better to place it behind the lcd, after the collimator (Where it was originally) where the light is moving straight through it? (collimated)

didn't your momma tell ya not to eat and think at the same time?

especially when ya got a nanna in your mouth...

Mikau: Post 558 and on has already cleared that up. It has been the ultimate plan to keep the polarizers within the collimated area (but not against the panel).

As for the instructions: Unfortunately there are still some missing peices. Just not quite ready for a full blown summary.

To anyone interested in the 3M Vikuiti prismatic filters. Your light engine would need to be surrounded in diffuse reflective material to be of any use. The current predicted ideal would be 2 crossed prismatic filters, followed by a reflective polarizer, with a diffused lamp reflector, and every other area of the light engine made to be diffuse reflective. Those films all require diffusion. Without diffusion, they would only be a hindrance (But for the fact that those polarizers may have awesome efficiency. We already know they have great transmittance.). It has been predicted that the use of Prismatic filters could eliminate the need for a Collimating Fresnel, and a point source. The worry here is that the filters are cut with the wrong angle prisms (90 degree) to allow for light that is collimated enough. Currently, light will be exiting at up to 35 degrees off collimation.

SIMUL8TOR: once again, lets try to keep the comments on this as serious as possible. It is imperative that we stay focused here. .

Mark.

Mark: do you think ozstang has something here?
(edit)...Once again Mark, you beat me to it base on the previous explaination.




Elkin: does this still hold in your presumption as to the difference of brightness between either methods?

Aren't those prismatic filters designed to be placed on unstripped lcd's to increase brightness and contrast? If so, unstripped lcd's have that clear plastic diffusser panel behind the lcd. The prismatic filters may just undo what that thing does. So if we use that on our lcd's with our diffusers already taken off, it may do nothing or even make things worse.

We discussed all of the functions and requirements for each of those films early in this thread. Again, yes you do need some form of diffusion or redirection with the prismatic filter. I don't feel there would be a significant decrease in image quality if you did not satisfy that requirement, but it certainly wouldn't do anything beneficial.

Progress : I just learnt (thanks for the recycler link DAZZZLA) that when circularily polarized light reflects off a mirror surface, it returns rotating in the opposite direction. This means the discussion of a quarter wave plate light recycler is extremely valid (Sorry Phutton). What we need is this:

Reflector,
Lamp,
Quarter Wave Plate (the Fresnel has been confirmed to not have retardation properties),
Collimator Fresnel,
Vikuiti Reflective Polarizer.

From that point on it is up to the builder wether they make any modifications to the panel (Like removing the now duplicated rear polarizer). This light recycler will still do its job just fine either way.

The path that light will take is as follows:

Light straight from the lamp to the Fresnel will exit the lamp unpolarized, and thus pass through the wave plate unaltered. It will then impact the reflective polarizer. 84% of the energy that is polarizable (50% of the energy is polarizable) will pass through and on for projection. The rest will be absorbed. The light that is not polarizable will reflect back, polarized. This light will enter the Quarter Wave Plate headed towards the Lamp. It will be circularly polarized by the Wave Plate. It will then impact the Lamp Arc Tube. Some of this light will be depolarized by the Arc Tube. This light will reflect of the Reflector and return having the same properties as the original lamp output. The light that is not depolarized by the Arc Tube will impact the Reflector as well, and have its circular polarization rotation direction reversed. Now it will reflect back and impact the Arc Tube where some will be depolarized by the arc tube. This diffused light will continue on, having the same properties as the original lamp output. The light that is not diffused will impact the Wave Plate and be converted from circular polarization, to linear polarization at 90 degrees to the direction that it entered (Because the mirror flipped its rotation). 84% of this light can now pass the polarizer. The rest will be absorbed.

It's hard to imagine something being more optimal than this, unless the Arc Tube could be avoided. Of course, a prism can polarize light with near 100% efficiency (As opposed to 84% Vikuiti), but at our scale that is just not an option. This is much more optimal than a basic diffuse approach as I have been mentioning, as the reflected light is truly managed to the exact plain it should be on. The diffuse approach gives light a random chance again and again, losing intensity with each try. This new design actually has a chance of around a 40% increase in brightness.

I'm not sure where to buy cheap large wave plates, though. May need to test cellophane of different thicknesses.

EDIT: The 3M Vikuiti Product Catalogue. The Vikuiti Reflective polarizer films are available in 17" x 17" sheets. Whichever polarizer is chosen, it is critical that it not be one of the integrated diffuser (including matte surface. I think embossed is okay) varieties.

Mark.

So now to complete the process we need a half wave plate to place on our heat shield or collimating fresnal. If I recall earlier, celophane wrap acted as a half wave plate. Look at post # 219

We need a quarter wave plate. Elliptical polarization aught to work as well. I eluded to using different thicknesses of cellophane to achieve this in my previous post. A half wave plate unfortunately wouldn't work (as explained earlier).

Mark.

Those who are suggesting using a diffused reflector to take advantage of recycled light should consider that by doing this you are effectively creating a non-point source light engine. So you are not controlling the direction of the light, ie it won’t go through the triplet. Diffuse light recycling is fine for direct viewing of a LCD but not appropriate for projection.

With this model of polarised light recycler and an arc length the reflective polariser should be place flat against the fresnel. This will theoretically reflect the light back close to the source. If the polariser were to be place away from the fresnel then the reflected light would be reflected too far from its origan due to the angle of incidence. With the polariser at the fresnel it will still be shifted but only to the other end of the arc.
Just so you are aware, the light being reflected by the spherical reflector from the ends of the arc and the added recycled light won’t all go back to the fresnel and on to the LCD. The edges of the fresnel will receive light from only half of the arc. But this is no different than the way we set up our reflectors normally and is another reason we get dim corners. My only concern is that because we are now recycling a lot more light the dim corners may be exaggerated.

DJ

A possible band-aid to the dim corner problem would be to use a much larger spherical reflector.

DJ

Ok further tests


the glowy effect is not caused by using a diff polar bear from another panel

took quite sometime to find something that would remove glue from original Benq polar

finally found acetone softened it so i used a kitchen spatula to delicaticly scrape bulk of glue,,, talk about gummy fingers, then when most was removed,, used soft cloth wetted with acetone,, then soap hot water and final polish..

whewwwww!!

anyhow removed test polar from front of triplet lens,, put Benq original

absolutely out of focus,, but polarized

ran a piece of tape across width,, ran same test again,, perfect polar perfect focus..

so even so that the polar is against panel in orig config,, it proves how much diffusion was happening,,, also looking at subpixels close up there is a signifigant difference in quality of image that antiglare missing

antiglare is by far our worst enemy when it comes to sharpeness and brightness, without a doubt

I now state that the disabling of antiglare, is now part of our process of building a PJ

consider as brain said yesterday (that in the current configuration the PJ works very well)
well when you see this new technique, there is no going back to the previously good way
This method goes in the ADVANCED version of DIY PJ construction,,, the next level to perfection

also this removal changes the behaviour of all the digital settings we have come to know so well

If I could so state,, it has changed it to like adjusting your normal CRT way

The contrast behaves like a real contrast,,, yeah I know sounds wierd,, not like its a gamma control,, thats the closet I can describe it

also the antiglare must have been playing absolute havoc with the colours, as now the saturation control behaves as I would like it too and the antiglare also must have changed the focal lengths of colours (red being longest focus to blue being shortest focus) that would explain why colours now are in balance and purer, and so damm vibrant,,, IE skins tones are so much better

Now not to detract from all the millions that have been spent making LCD panels
as they are today,, you must remember that it was towards it being used as a backlit LCD Monitor,, NOT FOR as we use it,, the market is not going that direction,, except for the mirco three panel commercila PJ's
This above statement is a statement to clarify that I am not critising any manufacturer or supplier of above mentioned points... Just in case someone takes exception to what we are doing...
Clive ...

Sorry guys if I have been discordinate,,, but hard to do and keep up here

As a final point ,as I stand I certify that these two methods work….

Dazzzla are you dialup?

yep still in the dark ages

Can you elaborate a bit more on this.

DJ

ok cleaned orig BENQ polar bear,,,, PJ running, simply put in front of PJ

totally out of focus,,, then ran stick tape over antiglare side,, put back in front of PJ again,, taped part perfect focus,, untaped part blurry,, not even recognizable image,

then dropped as was between LCD panel and field fresnel,,, complete image again,, but tape showed enhanced image,, as we have done days ago,,

hehehehhe curvy slighly buckled polar bear is just sitting there,, no ill effects

clive

Ah yes. The Clive Garner seal of approval.

Elken: Nemo comparison . I love Nemo. Oh, and Jello.

DAZZLA: That's not the way I see it. Here is my reasoning:

First off, the design I propose calls for a mirror smooth, spherical reflector, as in the current LL design.

It also calls for the arc tube being aligned at the Focal Point of the spherical reflector.

So long as the arc tube is roughly at the focal point of the half spherical reflector, all reflected light originating from the arc tube will pass through that focal point (the arc tube) on return. Therefore any light that comes from the fresnel that passes through the arc tube, will also be reflected back through the arc tube (but for some light that is diffused by the arc tube).

So long as a point light source is at the focal point of the Collimating Fresnel, all that light will be collimated (I write in terms of the ideal). Less light will reach the corners due to the exponential relative angle of attack at the edges. Put simple, the edges sample a much smaller cross section of the 360 degree projection of light from the arc tube.

Therefore collimated light passing back through the fresnel will be focused back to the focal point. As such, and light reflected off the reflective polarizer a long as it is perpendicular to the collimated light, and within the collimated area, will return to the arc tube, and return back on the same path on reflection.

Now, the arc tube is not exactly a point, but the symmetrical design should largely counteract.

I don't see how a larger reflector could provide a different function.

The dark corners will be lightened no less than the center.

Please correct me if I am wrong. Or maybe I just don't see exactly what you are saying.

Mark.

Does the quality of the image change when comparing the original polar plus tape with the new polar with no antiglare? Just curious to see if it is worthwhile to completely replace the polar or just leave it attached to the glass and mask the antiglare.

That’s the problem it won’t all be collimated because of the arc length. I’ll draw some pics to explain, might take some time, I’m a slow drawer

DJ

I need to do real test,, just did a very rough and ready test,,, much of a muchness,,, but need more precise method

Elken: Were you able to clearly identify the Compensation Film, as with the laptop?

I see where you are coming from. If the reflective polarizer is too far back from the Fresnel (like on the panel) Then should the light be offset even slightly, it will end up missing its original point of exit from the Fresnel by a great deal on return (Stupid pythagorean theorem ). You suggest a larger reflector, to collect the truly offset ones.

That said, the plan does call for the polarizer to be up against the Fresnel. Thus any offset will at least not be exaggerated. I haven't ray-traced it, but it seems to me that a good deal of these slight offsets will impact the lamp reflector in such a way as to still wind up impacting the panel. A ray-trace would be very useful here. And, should a true point source be developed, this method should work closer to it's theoretical limits.

Mark.

Would this help, better, worst?: Vikuiti™ Dual Brightness Enhancement Film - Embossed (DBEF-E) 0 degree polarization angle film
3M Id : 98-0440-0359-2
GTIN(UPC/EAN) : 0 00 51138 74990 3

The embossed apparently does not diffuse. It'll either be the same, or worse. Want to stay away from diffusion as much as possible with this design.

Mark.

Here you go. First pic is just an arc and fresnel. Shows what happens to the rays from each end of the arc.

This one is with the reflective polariser away from the fresnel. Notice how the returning rays don’t go back to the arc. If you where to place the reflective polariser close to the fresnel the returning rays would not get a chance to diverge.

And the last one is comparing a large and small spherical reflector. The reflected purple ray will be closer to where we want it to be. Still won’t be perfect but it’s better than the green.

DJ

That last one is a nice insight. Cool.

It's a solid factor.

How much recycled light would you guess would make it to the triplet?

Mark.

Here’s one more just so you can see why the reflector will cause dim edges. No direct light from the arc is being shown. Notice that the purple line doesn’t go back to the screen but the green does. This will cause the centre of the LCD to get more light than the rest. Adding extra light from the reflective polariser will only exaggerate the problem. So by making the reflector bigger more light can be directed at the LCD.

DJ

I don’t mean to put a damper on things just that we might need to do a bit more tweaking. By using a bigger reflector we may be able to compensate a bit. Possibly we could use a parabolic. I haven’t thought about elliptical yet, Hmmm could be interesting, 50watt lamp, elliptical reflector and light recycling. cool

DJ

And anti-glare removal .

Mark.

I'll take it one step further:

Point source Light Engine + Polarization Recycler + Panel Modifications + Panel flipped around = A lot of improvements for one thread that started with a guy who accidentally scratched his LCD = Chaos theory .

Mark.

hey now....

....on that note:





So, we aren't sure if embossed is okay, experiment first then? I originally posted this type of dbef because of it's 0 degree polarization angle feature, or did I miss something where it does not matter now.

(edit) it also provides linear light, no need for 1/4 wave film.

That is a lot of testing. Hey Clive what are you doing for the next month or two.

I'm just worried that even though they say it doesn't diffuse, it is going to have an effect very similar to the anti-glare.

We've confirmed (I hope) that 45 of 0 degree doesn't matter if you are removing both polarizers. You can set the angle of one however you like, and match the other at 90 degrees.

If you are not removing both polarizers, then you must match the replacement polarizer at 90 degrees to the original.

Mark.

You gotta be kidding, I want to get rid of my scratch, NOW!!!