Optical qualities of prism sheet and glass substrate (LCD screen)

[anorlunda]

This account reminds me of software debugging. Tracking down the actual cause of these complaints can be very similar to finding a bug in software. Therefore, software debugging techniques may be useful.

Finding a bug in complex software can be a daunting detective task. The most common mistake in debugging methods is to prematurely jump to conclusions as to what causes can be excluded. That is why the ultimate solution is frequently to bring in a "fresh pair of eyes". The fresh person will not start with his/her brain contaminated with certainty about what the bug can not be. When the cause of an elusive bug is ultimately found, it is often the case that it is something that had earlier been excluded by the investigators.

There are parallels between software debugging and cold case murder investigative methods.

The most productive methodical way to track bugs is to begin with the collection of thorough and accurate records of how and when the symptoms appear. (Who remembers Jack Webb. "Just the facts madam. Nothing but the facts.") Any statement like "many people" hints that such record keeping has not yet occurred.

 

[informerkh]

This account reminds me of software debugging. Tracking down the actual cause of these complaints can be very similar to finding a bug in software.

You are totally correct!
Few month ago there appeared another person who was geeky enough to deal with this in a 'detective' way. She is probably less sensitive (her list of bad screens is smaller) but have a very severe reaction. I unfortunately can't share her research (actually I can, it is posted in this thread https://4pda.to/forum/index.php?showtopic=943228&st=0#entry111127837) because everything is written in Russian. GoogleTranslate makes text gibberish but if you want some details I can translate.
It is super detailed and covers probably whatever comes to your mind. Or I can ask her directly for more data. She also posted a few small details on ledstrain.org in English.
She had her first symptoms when purchased a new phone (Samsung Galaxy Note 10 lite, oled) this spring. She had symptoms on the second day. She sold it three weeks later and symptoms disappeared completely. Her safe screen was Samsung Galaxy Note 5 (also oled, very same SuperAmoled technology with PenTile). She used it with no symptoms for many years before and she used it with no symptoms up to this September. Phone died in September and she tried to replace it. She found out that she has same symptoms from every modern phone, both IPS and oled. She tried maybe 50. She have done a lot of tests since September.
I can translate other stories or post something from ledstain.org.

 

[informerkh]

What is interesting is that she also find that looking at those layers is problematic (eventually she disassembled one not so good but also not super bad screen). But it is early to jump to conclusions as she only just started her disassembling journey and she wants more screens to compare.

 

[anorlunda]

Reliable fact gathering can never be personal, and never collected by the persons involved. If there are many people with this complaint, and you are collectively ignored, then the only possibility may be to pool your money to hire a team of experienced investigators to do it the right way. Be prepared for a professional investigation to be very expensive. $1 million would not surprise me.

Perhaps more difficult, those putting up the money must accept the possibility that the evidence may point in a direction they consider offensive and insulting.

 

[informerkh]

Reliable fact gathering can never be personal, and never collected by the persons involved. If there are many people with this complaint, and you are collectively ignored, then the only possibility may be to pool your money to hire a team of experienced investigators to do it the right way. Be prepared for a professional investigation to be very expensive. $1 million would not surprise me.

That seems to be way expensive, I hope we can start from a cheaper way. Also it is hard to be this organized. People instead just purchase and sell 10-20-30 different monitors or phones or laptops.
And also this can't be done by any private investigators, only by a team of experienced engineers or researchers who are well versed in tech. And in optics. And in medicine.
I wonder why simply not discuss the initial question of this thread as it is the biggest cue right now.

I again have very unpleasant feeling that it is expected to proof that people have a real problem that has no severe underlying medical condition and is not a fake or mass placebo effect. I definitely came here not to investigate whether my pain is real or placebo, I sincerely hoped to discuss physics as it is called PhysicsForums
You may believe me or not, that is not the issue. I wonder what these two layers do and how exactly light passes through them, I tried to read the specs but they do not cover enough of details.
We can have a long discussion about medicine but then all the helpful physics discussion would be lost entirely.
Just for an example, when oled screens became common and people started complaining, it took a long time to define that they have eyestrain from a PWM that is used there extensively. Some people thought it is placebo or faking, and some thought that it is because lamps are directed into face or colors are too saturated etc. It was also hard to find cause PWM flicker is invisible to brain (but visible to eyes) and it made no damage that could be found by neuro ophthalmologists. Now this source of eyestrain is well-known, but there is still very limited scientific data about its influence.

 

[hutchphd]

I think it must be flicker related. This turns out to be rather involved for an LCD screen.
https://www.eetimes.com/lcd-screens-dont-flicker-or-do-they/
This can be simply investigated using an (amplified) photodiode and an oscilloscope. The frequencies are sub-kiloHz. You need some actual data.other than "my head hurts"!

 

[informerkh]

I think it must be flicker related. This turns out to be rather involved for an LCD screen.
https://www.eetimes.com/lcd-screens-dont-flicker-or-do-they/
This can be simply investigated using an (amplified) photodiode and an oscilloscope. The frequencies are sub-kiloHz. You need some actual data.other than "my head hurts"!

As I said it was investigated thoroughly.
Here is one example. I soldered a device that allows to smoothly change the PWM of the monitor in the range of 100Hz - 1200Hz. My sensitivity to PWM flicker is:
100-150Hz feels horrible
150-180Hz very bad
180-220Hz bad
220-240Hz OK
240-340Hz good
340-500Hz excellent
500-1000Hz makes no sense to use, no additional improvement
Also those layers from disassembled screen have no ability to flicker. I can take them in hands and... I do not know, wave fast? Then they will flicker

 

[hutchphd]

It is complicated. Please read the article. The amount of flicker depends upon lots of stuff.

 

[berkeman]

Here is one example. I soldered a device that allows to smoothly change the PWM of the monitor in the range of 100Hz - 1200Hz. My sensitivity to PWM flicker is:
100-150Hz feels horrible
150-180Hz very bad
180-220Hz bad
220-240Hz OK
240-340Hz good
340-500Hz excellent
500-1000Hz makes no sense to use, no additional improvement

Wait, that is very significant, no? If you can change the frequency of the flicker and it eliminates the pain, then the problem seems to be directly related to flicker frequency and not any of the LCD layers. Can you buy monitors with 340Hz refresh rates? What are typical LCD backlight flicker frequencies?

And I thought you said that your ophthalmologist(s) did flicker sensitivity tests on you and found nothing?

 

[informerkh]

Wait, that is very significant, no? If you can change the frequency of the flicker and it eliminates the pain, then the problem seems to be directly related to flicker frequency and not any of the LCD layers. Can you buy monitors with 340Hz refresh rates? What are typical LCD backlight flicker frequencies?

And I thought you said that your ophthalmologist(s) did flicker sensitivity tests on you and found nothing?

On the contrary, it shows that I am flicker sensitive only up to these frequencies, and none of my LCD screens have these flicker frequencies. These frequencies are typical not to LCD but to OLED screens (typical PWM frequencies) and many if not the most people are sensitive to these frequencies (I mean solely PWM frequencies).
Refresh rates are not PWM flicker (PWM is stroboscopic full-screen flicker) and they are way lower, but my safe refresh rate is 60 Hz like for the most people.

I said there were ophthalmological tests and flicker tests, not ophthalmological flicker tests

 

[jrmichler]

This can be simply investigated using an (amplified) photodiode and an oscilloscope. The frequencies are sub-kiloHz.

And, based on your flicker tests, a vote for this. It will tell what the screen is really doing, not what you think it is doing. Do this with screens that are both bad and good, or at least not as bad, looking for differences. If you do not find anything, try some color filters over the photodiode.

These type of problems are solvable. Finding the solution requires checking everything, especially those things that you don't think need to be checked.

It helps to create a document that states exactly what is known. The act of organizing that document frequently shows gaps in what is known. Those gaps need to be investigated. That document is useful in communicating what is known to people new to the problem. Such a document has helped me to solve similar difficult problems. It is well worth the time to create it.

 

[hutchphd]

I could not agree more and have had many similar experiences.

 

[informerkh]

And, based on your flicker tests, a vote for this. It will tell what the screen is really doing, not what you think it is doing. Do this with screens that are both bad and good, or at least not as bad, looking for differences. If you do not find anything, try some color filters over the photodiode.

These type of problems are solvable. Finding the solution requires checking everything, especially those things that you don't think need to be checked.

It helps to create a document that states exactly what is known. The act of organizing that document frequently shows gaps in what is known. Those gaps need to be investigated. That document is useful in communicating what is known to people new to the problem. Such a document has helped me to solve similar difficult problems. It is well worth the time to create it.

Sure, and this doc exists in Russian.
This is how we ruled out all the flicker influence (FRC, Vcom, inversion, PWM, refresh rate, software etc. etc.).
Again, I am now more interested in how exactly two non-flickering, very static pieces of material can be so problematic for the eyes.

 

[informerkh]

Ok, let me get this straight. Does anyone here have any experience with these layers or knows how they may work? I do realize that everyone are trying hard to help here and I am grateful, but maybe more experience with this particular tech is needed to help with this optical issue.

 

[berkeman]

Does anyone here have any experience with these layers or knows how they may work? I do realize that everyone are trying hard to help here and I am grateful, but maybe more experience with this particular tech is needed to help with this optical issue.

Sure, I can probably dig up some SID papers that discuss those layers (Society for Infomation Display -- the main technical society that deals with this sort of display technology).

Summary:: Need help with defining optical qualities of two LCD screen layers that can severely hurt eyes when the screen is disassembled (what do they do to light to make it dangerous)

I experimented with monitors a lot, and came to the conclusion that the problem is optical. I completely disassembled the LCD screen, took a table lamp with an incandescent light bulb, applied all the screen layers above the lamp one by one, and directed a lamp light onto a white wall. I noticed that my eyes hurt when looking at two layers - a prismatic sheet and a lower glass substrate. Moreover, my eyes hurt even from the reflected light from the wall! The symptoms are exactly the same as when working with the monitor. Does anyone know what these two layers do to light and why it can hurt eyes? I have found many people who have the same problem, to a greater or lesser extent. I have not found any research other than the effect of PWM on flickering and blue light, so any information will be appreciated.

So those two layers are the ones that you can shine DC illumination through and experience eye pain when looking at the light reflected from a white wall?

 

[informerkh]

Sure, I can probably dig up some SID papers that discuss those layers (Society for Infomation Display -- the main technical society that deals with this sort of display technology).

Wow! Thank you so much! That would be great! I sincerely hope that will help to figure things out! (At least to define set of non-flicker probable causes)

Yes, they are the prism sheet (it might be not a direct cause of pain in the monitor but it definitely unexplainably hurts when it is disassembled) and lower glass substrate (it has a thin film transistor layer on it and they cannot be separated, at least I could not separate them). Glass substrate might be a direct cause and I can't understand why it has effect so similar to a prism lens.

 

[Tom.G]

Here is one example. I soldered a device that allows to smoothly change the PWM of the monitor in the range of 100Hz - 1200Hz. My sensitivity to PWM flicker is:
100-150Hz feels horrible
150-180Hz very bad
180-220Hz bad
220-240Hz OK
240-340Hz good
340-500Hz excellent
500-1000Hz makes no sense to use, no additional improvement

Could you perhaps try lower frequencies for the flicker test?
I ask because common refresh rates are 60Hz and 75Hz, and your data shows greater trouble with lower frequencies.

Typical vision can detect 50Hz with averted vision, but it is not usually noticable when looking directly at the source. 120Hz is not typically visible under any condition. Maybe your eyes just have a faster response time than most! (I wonder if that's a survival advantage, either in the distant past, or in the present with powered vehicles to avoid. )

Also, your comments indicate that there is more than one contributing factor, specifically the refresh rate and the presence of the prismatic layer. You may find that there are other contributions.

Cheers,
Tom

 

[informerkh]

120Hz is not typically visible under any condition

There is a very simple explanation. There are two things participating in vision: eyes and brain. 120Hz is typically invisible to the brain. Person can't see it. But it is detected by eyes. Eyes can detect this flicker nonetheless. A very simple pen (or pencil) test proves this. If you take a pen and wave it in front of something that uses low frequency PWM (probably 120-400Hz, but here I am not sure) you will see not a usual pencil movement, very smooth, but a stroboscopic effect with 3, 5, 10 separate pencils. It is fun to watch. Meanwhile when looking at this source without any pencils you will see a usual normal image. This happens because brain takes an input from your eyes and proceeds it to you as a simple unmoving non-flickering image.

PWM is different than the refresh rate, it makes screen (or a light bulb) to flicker in a stroboscopic manner. So people are sensitive to it in a different way than to a refresh rate. No one knows yet why exactly people do react to PWM flicker. Does it break small involuntary eye movements (saccades) or does it affects the brain somehow or does it affects eyes in other way? Once scientists will find out.
But it is absolutely normal to react to PWM flicker. There is IEEE 1789-2015 - IEEE Recommended Practices for Modulating Current in High-Brightness LEDs for Mitigating Health Risks to Viewers. It says that frequencies up to 400 Hz are somewhat dangerous, they bring temporary health effects and should be avoided.
Here is the video that illustrated PWM flicker (it has a loud music so you might want to turn sound off):


I also tested another very common source of flicker, Vcom, and there is an opportunity to make it so strong that the flicker will be visible. I had no reaction to it. There are also test images for flickering, and my safe screens were flickering more that bad ones.
For example, this is how severe a Vcom flicker can be (upper image has Vcom flicker captured):

I would rather look at this flicker than to my bad non-flickering screen.

Yes, sure, there may be more than one contributing factor, hope to find them all!

 

[informerkh]

I just found this really cool article https://spectrum.ieee.org/visioncorrecting-display-means-no-need-for-reading-glasses. It describes methods of adjusting screen image for people who wear glasses so that they can see this image clearly without prescription glasses. I am now thinking again about that top polarizer with smaller than subpixel light rays. It may be possible to distort an image the other way round.
This is just a thought right now, nothing more.

 

[hutchphd]

I just found this really cool article about https://spectrum.ieee.org/visioncorrecting-display-means-no-need-for-reading-glasses.

I think this is not very newsworthy. In particular this technique has been known for a long time.

https://en.wikipedia.org/wiki/Pinhole_glasses

I don't see why you are so resistent to very good recommendations you have been given. I like to stumble around the internet sometimes but if the problem is related to flicker then why are you not looking at flicker using time resolved detection of some sort? Your methods are clearly inadequate to your purpose. What you say may be salient (or may not) but looking at the layers in a polarizing LCD stack is a very obtuse methodology for looking for frequency dependent effects.
I have no idea what the point of the video is other to illustrate one way to "see" frequencies higher than your flicker fusion rate. Folks with strobe lights know this.. that's why they have strobe lights.

 

[informerkh]

I don't see why you are so resistent to very good recommendations you have been given. I like to stumble around the internet sometimes but if the problem is related to flicker then why are you not looking at flicker using time resolved detection of some sort? Your methods are clearly inadequate to your purpose. What you say may be salient (or may not) but looking at the layers in a polarizing LCD stack is a very obtuse methodology for looking for frequency dependent effects.
I have no idea what the point of the video is other to illustrate one way to "see" frequencies higher than your flicker fusion rate. Folks with strobe lights know this.. that's why they have strobe lights.

Excuse me but I literally just described some tests for flicker sensitivity in detail.
Problem is not related to flicker.
Why do you suppose that it is?
I am not sensitive to any flickers other than the common low PWM rate that can easily be avoided and is not present in most LCD screens. Many LCD screens have no PWM at all!
I wrote here twice that replaced the LCD panel without the circuitry and problem moved with it. LCD panel cannot flicker by itself, it needs software of backlight or circuitry.

 

[berkeman]

Problem is not related to flicker.

Well, part of the problem might be, but you said that you can experience the pain with Static illumination through just the suspect LCD layers in isolation, right?

So those two layers are the ones that you can shine DC illumination through and experience eye pain when looking at the light reflected from a white wall?

 

[informerkh]

Flicker is not part of the problem, I can easily test it and easily avoid it, it is a simple thing to check.
Right, that is exactly the case, I can experience the pain with static illumination. And that table lamp without LCD layers above it is completely fine for me, just like any ordinary lamp.
LCD panel without backlight layers, without backlight, with top polarizer replaced defines whether screen is good or bad for me. That is for sure. There's no other explanation why all the sensations from the screen move together with it. I placed it in another device and sensations moved with it. It is the most important cue.

 

[hutchphd]

I wrote here twice that replaced the LCD panel without the circuitry and problem moved with it. LCD panel cannot flicker by itself, it needs software of backlight or circuitry.

Did you read and understand the article I recommended??

 

[informerkh]

Did you read and understand the article I recommended??

You posted an article about a Vcom voltage. I posted just now! that I tested Vcom by adjusting it to make its flickering totally visible. Monitors (not all) have a resistor that can be twisted and made better / worse up to a pronounced flickering of neighboring subpixels that can be seen easily, just like on the picture I posted.

 

[berkeman]

Right, that is exactly the case, I can experience the pain with static illumination. And that table lamp without LCD layers above it is completely fine for me, just like any ordinary lamp.
LCD panel without backlight layers, without backlight, with top polarizer replaced defines whether screen is good or bad for me. That is for sure. There's no other explanation why all the sensations from the screen move together with it. I placed it in another device and sensations moved with it. It is the most important cue.

I agree that these are important clues.

Since you know others with similar eye pain problems with LCDs, do you know some who would be willing to go through a couple simple tests with you to start gathering more data?

If so, I'd suggest setting up a test where the subject is looking at a white wall in front of them, and from behind them, you randomly project the DC lamp light through different LCD layers to see if they have the same reactions that you do to the projected light on the wall. If they do (in this single-blind test), that could be pretty significant. (And to be complete, you should be one of the subjects in this test as well, with somebody else randomly switching out the layers.)

 

[informerkh]

Since you know others with similar eye pain problems with LCDs, do you know some who would be willing to go through a couple simple tests with you to start gathering more data?

If so, I'd suggest setting up a test where the subject is looking at a white wall in front of them, and from behind them, you randomly project the DC lamp light through different LCD layers to see if they have the same reactions that you do to the projected light on the wall. If they do (in this single-blind test), that could be pretty significant. (And to be complete, you should be one of the subjects in this test as well, with somebody else randomly switching out the layers.)

Yes, that is a great idea and I want to do something similar!
There are three problems with the execution:

1. People are scattered around the world and even within one country. As I need people who are willing to participate, have free time, have these layers... It is super hard to organize. And only two people are ready to disassemble screens right now, other simply do not have skill or waiting curiously for the results. But I hope things would improve.

2. Screen layers have to be found, so it should be exactly a bad screen, preferably the worst one for clear conclusions. Because of the next point in this list, the best thing to do is find some expensive modern smartphones with IPS screens and that takes some time and money (or less money and more time if service center can give some).

3. People have different sensitivity levels. I have high sensitivity but others have less. This problem is the most prominent on modern smartphones and most of them are OLEDs, so there's nothing to disassemble.

The lady I mentioned took the slightly bad IPS screen, disassembled it and found out that:
- she has way less problems with the exact same prism layer
- she has problems with a layer that comes above it, it is sort of another prism but probably multifunctional, we couldn't figure out its name. It looks very similar to prism film and may be some sort of reflective polarizer combined with prism film... I am not sure.
- she has pressing sensation that resembles phone when she takes lover glass substrate with transistors, that is glued with rear polarizer, and that undefined multifunctional layer together. She feels pressure when she applies them above the safe screen (with no image, just white background).

As I said, she is at the very beginning and she doesn't want to jump to conclusions.
This is a tricky problem and it affects people randomly, so there is no critical number of EEs to do all the tests yet. I hope it will change over time. Maybe if I have some clear findings or simple tests that everyone can do, there will be more data.

 

[Tom.G]

One more WAG (Wild Ass Guess):

Sensitivity to differential polarization. If you can get two polarizers (perhaps sacrifice a pair of plastic polarized sunglasses) to hold infront of your eyes.

Try viewing a screen thru the polarizers and see if you get any improvement/worsening when you independently rotate the polarizers. Try it with both 'good' and 'bad' screens.

Oh, and another WAG:
Do the screens give you a problem if you view them with only one eye?

Cheers,
Tom

 

[informerkh]

Hi Tom,
Those two ideas are also interesting!
I personally haven't tried polarizers independently yet, but I may try and see what happens. I asked the lady I mentioned (she uses nickname Mrak0020 on the forums I mentioned) and she also haven't tried. But both me and her tried different polarizers on top of bad screens, it changes nothing. I suppose polarization should not be differential if a screen is covered with a "good" polarizer above it.

I tried to look both eyes and one eye. Here what I've got: when I look with both eyes it hurts. When I look with one eye it also hurts. And what is interesting, it seems it still have some connection to accommodation or convergence despite it is one eye only. And here is why.
Mrak0020 also tried it and she also have more severe reaction (less sensitivity level, so less bad screens, but more severe reaction, her eyes hurt more). She has new visible red blood vessels in her eyes immediately after looking on a bad screen.
She was playing with stereograms and a simple test for binocular vision dysfunction: Brock string. Using those she understood that bad screen influences her convergence (it is the process of eye muscles rotating eyeballs so that they look to the same point), she had never any problems with it but after looking on the screen she can't focus far, eyes converge at the wrong point. That can be fixed with exercises for convergence, or looking on parallel-view stereograms. Looking on the bad screen influences her convergence immediately again, exercises do not prevent this. Screen does something to the eye rotation or focus.
Now when she looks with one eye only, most symptoms remain, and she sees those new visible blood vessels in the mirror on the surface of that one eye. So it is connected to binocular vision or eye muscles or eye focus, and a single eye exposure is somehow enough for the pain.
I haven't tried Brock string yet, we've met with Mrak0020 not so long ago.

 

[informerkh]

Just wanted to clarify if maybe something was misleading. I do not represent any organized group of people, and due to the different background, skills and countries it is hard to do anything organized. I've seen a number of similar stories and Mrak0020 spoke to different people with the same problem. They always are easy to recognize because people are not sensitive to flicker or ruled all the flickers out (it is not so hard to do, typical screen flickers can be measured or avoided and there are guides how to do it). People complain about the pressure, pain, inability to focus on image and weirdly "piercing colors" of the screen, especially white (no matter what the brightness is, it could be minimum brightness). Just like in this review of a monitor: https://libredd.it/r/Monitors/comments/pmmt2l/trying_to_reduce_eye_strain_and_headaches_from/
It is about a flicker-free monitor and it says "I still get bad eye-strain and headaches from looking at the monitor even just for a few minutes. It feels like the colors, especially the white color pierce through my eyes even at lowest brightness if that makes any sense." There are two similar reviews in Russian for the exact same model.
I also checked spectrum by simply replacing backlight and it changed nothing. And Mrak0020 tried about 50 different bad phones, they couldn't use absolutely same diodes. So it seems to be neither flicker nor spectrum related.

We've teamed up with Mrak0020 and we are eager to try different tests, that is the best we can do now.