Hey everyone, I'm starting this thread to discuss fresnel configuration and it's relationship to panel polarization and the glowies. If my research is correct, then another benefit of "split" fresnels is less "glowie" issues, more even contrast and black levels. I'm struggling with the glowies in my "unsplit setup", after reading for hours on this forum it seems people switched to split and solved their problem. I searched for the benefits of split fresnels and the main reason for using it is for keystoning. I have a straight box design with 220mm/500mm fresnels and the old DIYPC 18" triplet. I'm considering switching to split because of my glowie problems.

I made a few diagrams to explain my theory and the basics of fresnel focal properties. If the advance members have any major edit suggestions for these diagrams I will gladly change them. enjoy.

Click to view attachment
The first diagrams show simple fresnel properties and their relationship to the lamp arch placement. They don't include the LCD panel for simplicity. It's very important to place the lamp arch at the rear fresnels focal length.

Click to view attachment
If the lamp is too close the fresnel output expands the light rays beyond the field frenels edges. This is wasted light in the system, dimming the projector output and possibly affecting edge focus or creating "glowies".

Click to view attachment
If the lamp is too far, the fresnel output contracts the light rays inside the field frenels edges. This results in hot spotting and dim corners.

You can test your lamp arch distance by putting a sheet of paper in front of your rear fresnel and seeing if the light output image is the same size as the fresnel. If it matches you're at the right distance, if not you need to move your lamp or fresnel until it does.

Now lets add the panel configurations....

Click to view attachment

Here's an unsplit setup, the panel sits in front of both fresnels. The lamp rays pass through the panel while converging towards the triplet. I believe this may cause polarization refractions which lead to the "glowies". If you don't know what they are look here:

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


Click to view attachment

Here's a split setup, the panel is sandwiched between the fresnels. If the lamp placement is correct, then only parallel rays pass through the panel so polarization takes place as it was designed to. Light rays hit the liquid crystals and right angles instead of converging angles. In theory, inncorrect lamp placement may cause glowies in the split setup due to non-parallel ray effects between the fresnels.


Now let's look and this up close....

Click to view attachment

Here's a zoom of what's happening when non-parallel rays pass through the lcd panel. It seems some panels are more sesitive to this and produce glowie effects. My theory is, parallel rays entering perpedicular to the panel are polarized in a more uniform manner resulting in less glowies. This may be another plus to going split for "glowie victims".
Click to view attachment

Here's a zoom of what's happening when parallel/perpendicular rays pass through the lcd panel. This occurance seems most compatible with the nature of LCD panels.


So what to you think? I will be changing my projector from un-split to split following this logic. Please share your fresnel setup experience and it's relationship to the "glowie" problem.

Thanks

Aaron

I got a vertical set-up with split fresnels but still have a glowie issue. Me? I'm thinking of changing to unsplit fresnels.

You should check out Elken2004's plog here. Go to page 89 or 90 (yeah...it's a long plog ) and you'll see his research on the effect of polarizers. Rather than changing from unsplit to split, his solution seems to be removing the "polar bears" (as he calls them) from the lcd and placing replacement polarizers at different locations throughout the optics chain. Then, with minor rotational adjustments, he works to improve color reproduction, contrast, and picture clarity. It's a fun plog to read and packed with information.

I've read much of that plog, nothing personal but he's hard to understand, kind of sounds like a mad scientist. Anyways, I've done plenty of experimenting with replacement polarizers, changing orientation, rotating, placement in the system and I didn't improve glowie artifacts much. I found they mainly shifted and morphed shapes. Or I would get 70% of the screen to look it's best ever and still have a 30% area of weak, shifting black level.

My theory is that the liquid crystal cell grids utillize perpendicular light rays best. An analogy would be a window screen, if you shine a light perpendicular through the holes you'll more light exiting the other side versus shining angled light through the holes. Shining a light at different angles causes strange shifts in the shadow pattern behind the screen. Now sandwich the screen between polarizers and the phase shifts become even more complex. I'm thinking of the LCD panel as a stacked grid of tiny compartments, each compartment has a front door and a back door, angled light that enters the front door has to bounce off the ceiling, walls and floor. Are liquid crystals designed to handle this effect?

Maybe I'm wrong about this, I've just heard several people who have less glowies with a split setup and proper lens placement.

Just a note, last night I converted to a split setup and eliminated about 70% - 75% of my glowie problem. I'm still tweeking and fixing some minor light leaks around my frames. I will take pictures over the weekend to post here. I really think this is proof of parallel rays being better for even polarization and panel transmission. The only glowies left are minor along the edges, not the corners or middle of the screen. Otherwise 75% - 85% of my black screen has much more even contrast.

If you are getting glowies with a split setup then changing to unsplit will only amplify your problem based on my experience. I would be sure you have your lamp arch positioned at it's best distance to produce paralell light between your split fresnels, this may improve your issues (see lamp position diagrams above.) Also shift the position of your panel between the split fresnels. With the proper arch placement, the space between split fresnels can vary.

he IS a mad scientist!

Sinner - is your lamp mounted vertically or horizontally?

Horizontally, It's an EIKO from DYPC's old catalog and I think its guideline suggested a horizontal mount.

Yeah thanks for suggestions, I'm also thinking that tilting the front fresnel for keystone is also one source of glowie, not sure really.

I guess maybe I don't know what a "glowie" really is, but I have found that edge light was a problem for me to begin with.

Using flat black heat resistant paint and making sure to block EVERY last little spot where light can penetrate through other than the panel seemed to solve most of mine.

I am going to try the electrical tape at the edges of my triplet-side fresnel next to see if it dims the last ever so faint edge light.

I think maybe also my arc could be better centered as well, and that may help it, though with a "hanging" lamp design, that is going to be more of a challenge.

Well after several hours of tweeking this weekend I've learned a lot about the "glowies/backlight edge bleed".

First, the basics of my setup.

A) Panel - 8.4" 800x600 Sysonic, 300:1 contrast, 25ms response, 265K colors
- AG stripped, factory polarizers intact.
Here's my stripping post: http://www.lumenlab.com/forums/index.php?s...&hl=sysonic

Fresnels - LL 220mm rear, Edmund Optical 495mm front

C) Triplet - DYPC orginal 18" EFL 5" diameter

D) 250watt Metal Halide lamp, horizontally mounted with LL Pro Reflector & DYPC Condensor Lens

E) Straight line design, lamp arched placed at rear fresnels focal length.
Here's my design post: http://www.lumenlab.com/forums/index.php?s...&hl=sysonic

Second, my conclusions on "glowies" and "backlight bleeding".

Click to view attachment

A) I came from using a HAMI with a replaced rear polarizer that rotated, the glowies remained so I bought the Sysonic. Installing the new Sysonic, I started out frustrated with poor, uneven contrast in the black levels :angry: . The brightest and focus were outstanding, but those pesky "glowies" made dark scenes very distracting. I tried to vary the unsplit fresnel spacing with little sucess. The pic above illustrates my first configuration black level results. I replaced my old 330mm fresnels with a LL 220mm rear and a $100 Edmund Optical 495mm front. Even with the new fresnels, the problem remained. Note: I wanted a field fresnel a little longer FL than my 450mm triplet, "empty my walet" optical had the best selection of focal lengths.

Click to view attachment

After reading about every post here reguarding the glowie issue, I gathered this wisdom, fix light leaks, vary fresnel spacing, split optics = reduced glowies, reduce pressure on expanding panel and that replacement polarizers can help. After using all of these improvements, I still wasn't content. I also learned some panels suffer more than others with glowie issues, maybe I bought one . My theory was that split fresnels reduced glowies because parallel rays are better polarized by the LCD panel. Conclusion: Yes, parallel rays are better polarized by the panel, but the reduction of glowies by splitting my fresnels wasn't enough for me . I also had a side effect that I call "zoom ghosting", illustrated above, a refraction artifact between my field fresnel and panel. Varying the panel/field fresnel gap didn't help. So, after splitting my fresnels, yes my glowies were reduced with side effects like "zoom ghosting", brightness descrease and slight focus reduction.

Click to view attachment

C) In my second configuration, I did improve my optimal fresnel gaps and placement distances for reduced glowies and even lamp light. I decided to go back to unspit fresnels while keeping my improved fresnel gaps. :angry: The unsplit setup only amplified my glowies like before. So I started searching for LCD backlight bleeding as referenced in the forums. I learnded how the edge of the panel can absorb light and create glowie edges and patches. Since the problem occurs with assembled monitors, it must be part of the original assembly. I found this thread for treating a monitor with edge bleeding: http://www.hardforum.com/showthread.php?t=1032268 ,also referenced on the LL forum . I used my monitors original mounting frame and casing to hang the panel in my projector. These frame pieces are alluminum (reflective) and white plastic (reflective). This means my panel edges are surrounded by reflective material hmmm. I noticed there were small gaps in these frames, allowing edge light to escape the panel and bounce back into the edge of the panel. So I CAREFULLY split my panel from it's original frame pieces to tape it's edges with electrial tape. My tape only needed to be about a 3/16" wide to wrap the edge of the panel. I sandwiched the FFC edge with a piece of tape of each side of the panel. Then I painted the framed pieces flat back to reduce their reflective properties. After putting everthing back together, I was very impressed . The illustration above shows the improvement. There's still very, very faint glowies but it's acceptable now. In theory going split after this step may show even better eveness. I want to keep the focus and brightness benefits of unsplit.

So here's my list of "glowie/backlight bleed" improvements:

1) Tape edge of panel glass with light-blocking tape without covering visible edge pixels. (I used electrical tape).
2) Paint all mounting frames, especially frames around panel edge flat back. Paint any other colored surfaces flat black.
3) Be certain your lamp arch is positioned at your rear fresnels focal point. (creates even, parralell fresnel transmission)
4) Fix light leaks while maintaining sufficient air flow.
5) Vary your fresnel spacing.
6) Rotate polarizer - if replaced.

My experience shows these steps will benefit a split or unsplit setup. I now believe any combination of these issues will make your "glowies" unacceptable.

I promise I will post pics of my projector soon. I need to monkey with my camera settings first to get accurate pictures. Until then, I hope my obsessive behavior will save other glowie club members some time and frustration.

I watched about 10 minutes of Cars last night- a MUST HAVE projector experience .

Aaron

One thing, though: the benefit of unsplit Fresnel's is the clean path the light takes once imaged by the LCD. With unsplit, the image only passes through one lens, and that lens is a high quality Aspherical.

This is unlike split designs where you are passing the image through 2 lenses and one of those is a fresnel (not excellent for imaging).

There are quite a few facts that support everything you have said : Polarizers loose efficiency with increased incidence angles, liquid crystals have varying birefringence with varying incidence angles, and the clear light path itself through the panels is obviously optimal at head on incidence. But it is still a debate as to wether you want a clear image path VS absolutely optimal imaging.

One other thing you didn't get to is the effect of a stressed panel. If you are pinching the panel in any way, you introduce stress birefringence to the glass of the panel which will also cause the glowy effect.

And if the liquid crystals are uneven in distribution accross the panel, or is of an uneven thickness you will also get glowy effects.

Another factor is that optics will experience total internal reflection as the angle of incidence increases. Reflection increases up until the critical angle when zero light will get through. This should cause the corners to be slightly dimmer in an unsplit setup.

Mark

I have a little bit of the glowies too, it looks like your second picture. I want to stay with a un-split setup so the black tape treatment works fine? I was thinking of just spraying the edge of the panel with flat black but tape will work too.

Elkin also said something about the reflector making the glowies worse but I also had a hard time understanding him.

Thanks for the testing.

Sinner, great work!

Mark and the funky bunch,


If one does choose this config, i guess you obviously want the highest possible quality front fresnel.
would the so-called "High Resolution" fresnels help in regards to the quote? Im guessing a finer groove pitch makes them "high res"?
It would seem as if most would opt for on-axis light through the polar given the new test results.

Would you guys agree at this point?