Eye strain from LED backlighting in MacBook Pro

Being unable to say/think the right words is a symptom I often experience when I am about to have a migraine, I think it's fairly common.

The way I see it, I have a pre-existing problem with migraines, which is exacerbated by eye strain. Therefore, these screens that are giving me bad eyestrain are greatly increasing my chances of having a migraine.

It's a shame what you say about the EIZO screens because looking at their website, they seem to have really done their research and are trying to address all of the separate problems of flicker, brightness and blue light. Search for 'the latest on computer screens and eye strain' on their site.

Oddly enough, I have been having real problems with throat infections for the past few years, which is about the same amount of time since I got my first MacBook. I'm sceptical however about whether they can possibiy be linked to my computer use.

I think we have to make assumptions to rule out something or to prove it. The thing with flicker is that different frequencies should affect us differently. Most LCDs use 60 Hz as a refresh rate. And while I know the refresh rate doesn't cause flicker in LCDs, it limits the content on the screen from flickering faster than 60 Hz. So if there were to be a slight pixel or subpixel flicker in recent LED backlit screens, it would be limited by the set resresh rate. Theoretically, pixels can use any frequency from 1 to 60 Hz if they are instructed to flicker. 30 Hz is most likely being used in my opinion, as it wouldn't be visible to the naked eye if 60 were being used. And I read of people seeing it. A possibility is that different Hz are being used on different devices. Even if the PWM backlight flicker of approx. 120 - 200 Hz has no effect on a person, there is no certainty that 30 Hz wouldn't be highly troublesome.

Pixel flicker can be invisible when mixed correctly with non-flickering pixels or with pixels that are on while the neighboring ones are off. An example is these test pictures http://www.techmind.org/lcd/ . It is highly likely that one of these test pictures will flicker when only the offending pixels are displayed. The flicker seems invisible when mixed with non-flickering pixels. RGB sub-pixel dot-inversion flickers the most on my screen, while it isn't normally visible at all on my screen and doesn't show on attempts to measure the flicker with a camera.

Who is to say the entire screen isn't made out of such flickering pixels being turned off while the neighboring ones are temporarily on only to get turned off when their time comes resulting in a snow-effect often invisible to the naked eye and not detected by traditional attenpts of flicker detection?

Dovez,

What you are describing is "temporal dithering" and is ABSOLUTELY ONE HUNDRED PERCENT PRESENT on modern screens, specifically Macbook Retina displays. This isn't the entire answer, but it's a big part of it. To make a "PWM FREE" display, they have to do SOMETHING to dim things. LED's are either on or off. I've long suspected that supposedly PWM-free devices are in fact flickering in a way that our eyes can detect but which cameras can't. If you tell me an LED device is at half-bright, but "not flickering" I have to say that's simply not possible!

I've witnessed this up close on a Macbook Retina display.

It isn't the only problem, but it's ONE problem.

I want to clarify that I meant that besides the pixels on the test pictures flickering other pixels might be doing the same or similar thing, pixels that have no test pictures to show that flicker. It may seem unlikely, but we have to consider everything. The answer in my mind is almost certainly some sort of flicker, because I have seen flicker affect me the same way. Look at this http://www.youtube.com/watch?v=vdU_DbRp4Og multiple times on a comfortable display and see if it doesn't cause the same symptoms as an uncomfortable display does. Warning: epileptic people don't watch the video! You can try wanching it for as long as it takes for the symptoms normally to appear on a bad LCD.

With all due respect, what you say is simply not true. LEDs (and diodes in general) can indeed be "half on". Diodes are actually analog devices, and a diode's voltage-to-current response curve goes very sharply from almost no current (with reversed voltage or low forward voltage) to very large currents (with high voltages). It's true that it isn't _easy_ to get an LED to the "half on" (half-bright) state, because you would have to control voltages quite precisely to be able to do that, and given the usual variation in LED voltage/current response curves, this is tricky. But it is by no means impossible. Also note that dimming is pretty forgiving; when you dim a monitor to 50% you don't really care if it's _exactly_ one half of the illumination of the maximum; you just care that it dims smoothly from fully on to fully off (or to as dim as it can get).

Does Apple actually do that in their monitors? I have no idea; ask an Apple engineer. (They probably won't tell you, though, because of NDAs). I do know that Apple filed a patent recently for a way to control the intensity of an LED light source using purely analog means (i.e. without dithering). This suggests to me that they may indeed be doing analog dimming (or at least are experimenting with it). It makes sense; a lot of people are sensitive to flicker, and if they perfect it they can use it in their advertising ("now with Ultra-Flicker-Free technology!") and then charge an arm and a leg for it.

Just to be fair, the comments i read about EIZO and HP (advertised as flicker free) were posted by two people. Maybe other sensitive people find these displays comfortable. For example EIZO 2436 is flicker free when the light level is above 20% or so according to some sites and the user has to disable the eco mode from what i read. I am not sure if the user complaining actually made these adjustments that could help.

But on the other hand even if i buy a display like this, if i connect eg to mac mini, will the screen be comfortable to use for long hour use? I read of people using a screen before with no problem and after connecting to mac they have trouble.

Clues to solving the mistery.

1) A change in processor can cause the screen to become uncomfortable.

2) Changing the screen resolution may make the screen usable.

3) Updrading the OS can make the screen unusable.

4) PWM measurements showing it isn't used

5) Benq and Eizo (partially) making PWM-free LCDs

6) Some CCFL LCDs are fine and some LED LCDs are fine.

Lets examine the screen resolution working clue. No spectrum changes (the same blue light), no polarization changes, no glare changes, brightness changes etc. The same computer withe same screen with a different resolution working.

OS: same hardware, but different OS may introduce the "eye strain" plague.

Changing processors: was it the actual processor change or was it that when a new processor was introduced a new driver for him was installed?

Only LEDs are affected? No. Any screen that is usable might become unusable with the right software. What can software do to the screen? It mainly controls its content, its pixels. What can pixels do? Change color, color intensity, turn on and off. Color and color intensity would be noticeable, but the right kind of flicker not.

PWM-free monitor is one of the essential requirements (but not for all apparently), but to solve the second problem one might need some very old PC or mac with some very old OS and drivers to use it with the monitor. This seems like the only reliable combination for people with "eye strain".

All of this is very interesting and there could definitly be some answers in here somewhere on how to make screens more tolerable. I would like to throw a curveball out there for anyone to respond to as well. Let's start with this baseline Hypothesis: The central question still remains is it Flicker ( Whether created by OS or anything else), Spectrum (Whether created by pixel color or back light) or both of these regardless of what is creating the issue to our eyes/Brain, right?

Here's the curveball:

Overhead LED bay lighting, or any LED light bulbs completely unrelated to computer OS systems. How many of you have been around LED lights, do you have the same symptoms with LED room lightning you do with your LED monitors? If your looking for some LED lighting to test because you are not sure here is a short list of places you can find them:

Car tail lights ( They have an appearance of many red dots instead of one red light)-Some head lights but mostly just those Audi "Eyebrow lights" so far

Overhead lighting at gas station pumps

Some grocery stores, they seem to focus on produce areas at the moment

Airports

Southwest Airplane's

So my question to those who are looking at technical reasons for monitor discomfort is what is the same about LED lighting as with computer monitors and TV's? We'll we know for sure spectrum of the LED light is the same with a range of 3500K-5000K the most typical white/blue spectrum created by LED lights, even those light bulbs that have yellow or white covers are still blue/white LED diodes beneath the covers. And we know they are using PWM in the lighting to control power consumption, so do the monitor setting's/OS/Pixels have much more to do with the issue? I can tell you for certain I have major Vertigo, migraine, eyestrain and nausea that is exactly the same from both the overhead LED lighting and any monitors. I would like to test a "flicker free" monitor because that would at least help understand spectrum Vs. Flicker somewhat more🙂

There have been two changes in recent years, new energy eficiency with computers and new lighting however it was not until backlighting went from CCFL to LED did I start to have any major issues......

I would also add this little update from my personal medical journey:

Visited a Neuro-opthamologist and was told she has seen over 100 patients in our area with very similar issues and symptoms. She was doing a study with her student Dr. and when her student moved on she let the study laps.....

I still think we're dealing with a combination of symptoms. I know that LED signage (outside local businesses) is nearly impossible for me to read, and that depending on the day and how my eyes are doing the LED tail lights of modern cars can be problematic for me.

One local business was forced to remove their LED sign after it literally caused an accident by blinding oncoming drivers in a busy intersection... so this problem is becoming more widespread.

However, I know for a fact that we are dealing with multiple issues.

LED's changed sometime in the past 2 years - early LED monitors cause me no problems but more recent ones do. The same laptop with the "newer screen" (exact same electronics otherwise, same model number and part number) causes issues. The screen is visibly BLUER (they call it "whiter").

LED lighting has also become more prevalent... this in turn causes problems.

Lastly, this monitor strobing or flickering or snow or temporal dithering - I believe this has multiple sources, as well. Some are from the monitor itself (PWM) and others from the OS or drivers.

All of these issues combined cause real problems for viewing modern monitors on modern OS'es.

Want to add that many LEDs can and probably are flickering because of various reasons. The screens may be flickering because of even a defferent reason than LED lighting. Flicker could very easily be the common denominator. There are just many different reasons for it to be present and there are also different forms of it. To convince yourself how strong flicker can be watch 10 minutes of the youtube video I left a link to earlier. It causes the same symptoms in me on a problem-free screen that watching a problematic screen for a too long time does - eye pain and headaches.

Anyone have a good camera to film the LCD pixels/subpixels from the close as it is seen on this video: http://www.youtube.com/watch?v=MFCeBsBz9Qs ?

The person filming needs to have a pwp-free apple laptop that is causing the person strong eye strain. White content needs to be filmed so all 3 subpixels are displaying a color. The right shutter speed is required or else no flicker may be visible even if there is some. Would be nice to see if flickering is there or not. I guess if some people can see it with the naked eye, then a camera should show it more clearly. But perhaps it is not as common as I think and very few people have actual pixels flickering on their screens. I'm also not sure if this method proves anything if nothing is seen. It theoretically should, but who knows. Also, to be sure, a few people would need to repeat the test on their screens. But most likely noone will even try it.

Gurm42 (..)LED's are either on or off(...)

That is not true. LED's could be dimmed by adjusting current. (the same is CCFL and EEFL etc.)

http://powerelectronics.com/site-files/powerelectronics.com/files/archive/powere lectronics.com/lighting_systems/led_drivers/analog-dimming-fig1.jpg

http://powerelectronics.com/lighting/how-add-analog-dimming-virtually-any-led-dr iver

LCD with LED or CCFL is flickering becouse:

- backlight PWM (could be removed in most backlight inverters (ex. http://www.maximintegrated.com/datasheet/index.mvp/id/5541, http://www.electronica-pt.com/images/fbfiles/files/OZ9938GN.pdf etc.), becouse most drivers have analog dimming bult in, but not used by producer). I did it myself in cost of 1 kilo resistor.

- in 6 BIT TN panels becouse FRC (temporal dithering called Frame Rate Control, http://www.tftcentral.co.uk/featurescontent.htm#dithering)

- V-com (http://www.intersil.com/content/dam/Intersil/documents/an12/an1208.pdf) pixels are flickering less or more becouse can't be always on from technical reasons

other unwanted effects are:

- snowing effect (cristal effect) becouse anti-glare coating (glare screens are better) and from my observations from additional reflective polarizer inside LCD panel (http://www.laserfocusworld.com/articles/2011/10/reflective-polarizer-for-lcds-in creases-screen-brightness.html)

- backlight leakage when matrix is too heavily compressed in some places by bezel (could be removed easily in ex. TV by loosening the bezel screws)

What I done to my 32 AKAI TV/monitor with LG IPS 8 BIT panel with EEFL backlight, 1920x1200, (http://wiki.answers.com/Q/What_is_the_difference_between_ccfl_and_eefl)

1. It had: 14 CCFL lamps (about 100 Watt) lighting from behind screen, (not removable PWM (even with 100% brightness , no sign of FRC becouse 8 BIT, V-com small flicker, snowing effect and huge backligth bleedings in every corner of screen). Effects: tired eyes, eye pain and burn ....

2. I removed PWM first. Backlight inverter had OZ9966SN driver (which works with external PWM, internal PWM and analog driving. It has two legs: PDIM and ADIM) manufactured by 02 Micro. There was no accurate datasheets for it, so I read OZ9906 LED controller manual with the same brightness control plan.

I disconnected brightness wire entering inverter board from main board (it had external PWM). PWM was gone so the EEFL lamps worked on 100% duty. PWM 200 hz flicker was gone (observed by camera and eyes). Effects: nicer screen and no more tired eyes (i tested it many hours in 2 weeks), huge heat from infrared radiation emitted by 14 lamps. So second step was turning on analog dimming on OZ9966SN. ADIM had 5 Volt on his leg and was connected with VREF leg (but analog driving is in 0,5 V - 2,5 V interval). So I short this leg by 1kOhm resistor to ground. With less then 0,1 V on this leg, driver chip dimmed lamps by changing it's electrical current to minimal. CCFL and EEFL could be dimmed by analog only 2,5 times with current from

Effect: before with 100% backlight turned on and 100% contrast LCD panel shined like a sun. After dimming you can look at without fear of blindness. TV/monitor is cold now. But my eyes was still in pain. Not such severe but still.

3. I dissasembled LCD panel and added 2 cm opaque PMMA opal colour plate (with UV blocker) to block excessive light which can't be more dimmed by analog driving. Better but still eye burn in right eyes.

4. I removed all LCD corner backlight bleeding by loosening the bezel and LCD panel screew.

5. Mayby some snowing on LCD panel ? There was 2 polarizers sheets in LCD panel. One was reflective. After removal snowing was gone. Screen is super clear viewed by naked eye. But no change to eye burn. I cant look this one from many meters without pain.

Other things with no effect: 2 cm yellow PMMA (with UV blocker) inside panel which blocked all blue spectrum of EEFL maps (screen was yellow and the light bleeding was yellow), F.lux program, yellow tinted glasses, sunglasses, black screen by showing a black colour .jpg (there is always some light bleeding by all the matrix, not like in totally dark CRT).

In the same time I can watch another 24' CCFL ACER monitor with 4 CCFL on edges of screen and old CRT very well.

my applications:

1. PWM: tired eyes

2. Lower intensity of backlight light in not the same as dimming by panel contrast to eyes. Eye burn effect and pain comes after a longer time.

3. Snowing effect, backlight bleeding, small v-com flicker:only visual pollution without consequenceswith high probability

Nothing I can do next with this model. My eyes can't tolerate probably or EEFL lamps or rear backlighting.

Previous LED 27' LG was much worse to me. I damaged the lenses of his matrix during careless dissasembly so beware. The panel has vertical stripes, but as crap could be tested now 🙂 Plan: removing PWM and add analog driving to LEDs.

Edit: After 1 month of testing I feel uncofortable in supermarkets with fluorescent light and feel burning in my eyes,I feel anxiety to stores with TV sets and I went back to my old CRT for now.

Kxtr73, Very interesting report on your findings and much more in depth experimentation than the average Joe could complete, thank you for the information. I feel your pain in supermarkets and stores, I get Vertigo and migraine issues from both fluorescent lights and much worse from any LED lighting. I visited a Neuo-Pthamologist Monday morning who explained she has seen at least 100 patients with our exact issues so we are not alone. One question I have for you: Did you have issues with lighting before all of your testing or do you think they became much worse afterwards? I ask because there are scientists who believe LED light may permanently damage the retina, I would be very careful, you could be causing damage by looking at these LED screens during your tests....

I believe my sensitivity is increasing, it is either that or the prevalence of LED lighting in everything is increasing and I am not recovering anymore...

Take a look at this slow motion video of an LED light bulb and share your thought's on that as well!

WARNING: If you are very sensitive this will have an effect on you, I can only watch it for 10 seconds!

http://www.youtube.com/watch?v=zAfWKcg8Bq0

Wow, nice Post Kxtr73. As CoreLinker explained to me prviously, the FRC dithering is nowadays always there, even for 8bit panels in order to better cover the color space.

So maybe that’s why you couldn’t completely eliminate your symptoms.

As an update from my side, I have ordered (again) the surface rt and probably will also keep it. It is usable for an hour or two a day without big issues. However what I also noticed is that my symptoms appear much sooner if I watch pictures or some other color rich content on the surface rt. So looks like the Tegra3 reduces the dithering if less colours are needed.

Stefan