Eye strain from LED backlighting in MacBook Pro

Maybe Recovery Mode disables GPU hardware acceleration. You could try to disable it (if possible on Macs) within the regular environment to see if it helps. Judging by Windows 10, Android, iOS, and modern Linux distros, I bet the whole Desktop is GPU accelerated, and maybe it feels more sluggish or less smooth in Recovery Mode. That'd be a clear hint that it's turned off.

Yes it does feel less smooth in Recovery mode, all the lags of page cascading.

Can't find a way to cancel GPU acceleration in OS anywhere I can think of, I don't think it's possible?

My friend has his "trick" too: use a higher than native resolution, that's it. , he's been sticking with 1440 x 900 ever since he found its alleviation effect, for strange reasons. (a "diffused" strain rather than a "focused" strain in lower Res),

I didn't see any dramatic differences right away, maybe the contrast is too low already on my macbook, going to give it more time.

By the way it just seems less and less people are posting on this issue? Maybe human eyes adapted to it gradually and eventually ? I remember in those early days I could not stand WiFi for too long without feeling sickness or dizziness in the head, no kidding, within half an hour at most I had to put my tablet away, but now I literally feel nothing at all glued to it all day!

Maybe when I finally get used to it I can't tell if it is my new "tricks" that just work or it is my eyes that are "compromised" in order to grow on new products.

Don't know if that is good or bad thing, but I have to say to live in this era I feel like equally blessed and cursed. There are times I look around at all these gadgets, with all this penetrating technologies like 4g, 5g, wifi, AR, VR, AI as a intertwined big big web entangling the world's every last inch no escape no end, that makes the world more and more "unreal" to me and this scares me no end. I'm blessed I have a loving family and a loving girlfriend I feel more real when I'm with them be honest yet it also so happens at more and more occasions I find myself to be forced into "unreal" alternatives for lack of real alternative. Sentimental? Maybe, last time I put on a VR I easily went motion sickness in minutes if not seconds, I can see a day I finally feel like home at it and this scares me and saddens me even more. This is an "artificial" evolution some part of me cheers for it while the other part chills. Over reacted? Maybe, but suddenly at this point is it not a concern anymore whether this macbook eye strain can or can not be resolved by mighty fruit which you think is not a religion in disguise already?, either way, in the long run, it does not really make that much of a difference. Nite.

Ok so this is the solution for me: Avoid Intel Iris cards!

I was going crazy from the nausea the late 2013 mac pro was causing me so I decided to test several models to see what can be.

Late 2013 mac pro whit intel iris card: after 1h of use extremely nauseos whit very red eyes. Installed matte filter for display , changed fonts , resolution , external display nothing worked.

2015 mac air whit intel hd: even after 5h whit no pause i was feeling fresh.

2015 12' macbook whit intel hd: daily use for few months without issue

late 2016 13' mac pro touchbar whit iris pro: same nausea and headache after 1-2h of use, a little better than 2013 but still serious enough that i had to stop using it.

late 2016 15 pro whit intel hd 530 (amd 450 model) : daily use up to 6h/day without any problem.

So as you can see I was very surprised when I realised the only problem is the Intel Iris pro card (probably drivers).

I am sure many users that have problems use a mac whit iris pro card, try to switch to a HD model and I bet you will see the difference.

"What good is an eye doctor in this case if he cant help with graphics coding?"

He may be able to tell you why your eyes are behaving in that manner while millions of other users do not have your symptoms. Guessing about possible medical issues is never a good idea.

Hi, Josh

Sorry to hear your story!

How about first invite you to read a great article which spooked me the other night?

To Quote just two paragraph::

Unlike some displays that have 10-bit color built into the hardware of their panels, the new iMac screens will have 8-bit panels and will use a processing trick called dithering to present 1 billion colors to the user.

An Apple spokesperson said the new iMac will use an algorithm that employs both temporal and spatial dithering. The former takes advantage of the human eye's tendency to mix two colors in close proximity to create a blend of the two, while the latter achieves the same effect by having a pixel flash between two colors very rapidly.

You have to open that link to have the full context, you may then very quickly, like me, ask Apple a very simple question::

How about just give us that 10 bit panel, instead of that 8 bit panel along with all the dithering pain in Apples ***?

So why do I bring this up, because as many other members have suggested, that something called temporal dithering (the term spooks me forever) is very likely the cause, it is actualized either through GPU graphics driver or, on a hardware level, by the built-in macbook/iMac display through FRC method, !

In short, it is to cheat million of colors (as Apple advertised on the official page) on a natively incapable panel.

Now, I have to say that somewhere buried in the middle of this mega thread are no less than ten instances as I recall that mentioned either of these things::

✅ the same external monitor that never had any problem connected with Windows computer, started to give the user eye strain when connected with specific macbook model

✅ the same macbook that induces a feeling of nausea, after installing Windows using bootcamp or Linux using USB drive, feels like a different computer, the eye strain also dissapeared or or alleviated to the point of almost nonexistent.

✅ the same macbook of no eye strain, after major annual update, started to cause such.

Posts like that indicate the graphics driver is undeniably involved.

Apple also made dithering mandatory in OS X/macOS, no options to turn if off.

( PS. I suppose when the original poster created this thread back in 2008 he was referring to PWM flickering, because when it first came out it was a far less refined method, and through these years things have been improved and Apple product is doing reasonably well in area of PWM implementation? Check this. So while some older macs of specific model might have this PWM issue the newer ones not likely)

Anyone still have suspected problems with Temporal Dithering and Mac hardware?

I managed to create a KEXT injector that should (in theory) work to disable the dithering.

If interested reply back with what hardware (Mac model) and OS X version.

My email is *******

Thanks

<Email Edited by Host>

If changing the LED screen alone has fixed the problem, is it possible that a third source of flickering, other than PWM and dithering, which is called inversion, is in play here? 😐😮

The following is from that link,

In a pixel on an LCD monitor, the amount of light that is transmitted from the backlight depends on the voltage applied to the pixel. For the amount of light, it doesn't matter whether that voltage is negative or positive. However, applying the same voltage for a long period would damage the pixel. Do you remember how electricity decomposes water into oxygen and hydrogen gas in chemistry class? Similar things could happen inside the liquid crystals that are in the pixels.In order to prevent damage, LCD displays quickly alternate the voltage between positive and negative for each pixel, which is called 'polarity inversion'. Ideally, the rapid polarity inversion wouldn't be noticeable because every pixel has the same brightness whether it a positive or a negative voltage is applied. However, in practice, there is a small difference, which means that every pixel flickers at about 30 hertz. In order to make this less noticeable, pixels with positive and negative voltages are interleaved, such that on average the screen as a whole keeps the same brightness

Does it "click" on you reading what's on that site? Does it make sense?

The "pixel walking" (inversion flickering) won't appear like "walking" on retina screens since individual pixels too small to spot, overall it looks more like flickering!

Normally this inversion is not an issue at all, but yes manufacturers can screw it up big time if such is unfortunately the case.

I am not a fan of sharp blacks too!! They look so artificial , so "out of place"! we rarely, if at all, encounter such degree of blacks, such degree of burning contrast in every day life, I once thought it too that it is this blacker than black that was causing my eyes hurt, but I assure you it's not.

If the illumination is stable, a black is powerless to hurt you, but armed with "flash", it becomes a different story 😐

Here's a new episode from this flicker-hunting season::

A friend has an Acer notebook which has PWM setting in BIOS!!! Sounds too good to be true!! But that's after he suffered a good six months of eye strain until he bumped into a post that teaches how to unlock hidden advanced BIOS features for Acer notebooks. When unlocked, the specific item is under the Integrated Graphics sub menu, it's called Backlight Control which lets you select one of the following::

PWM inverted

PWM normal

GMBUS inverted

GMBUS normal

The default value is PWM inverted, he figured better get rid of anything that has the cursing letter "PWM" in it so he changed it to GMBUS inverted without consulting what the heck that is,…. and poof! The eye strain is no more.

Now some theory ::

1. Maybe how wrong I was, I might have underestimated the PWM factor. Very probable that most panel/motherboard OEMs have certain kind of "PWM Interface"* embedded in BIOS/firmware, I believe this is universal to almost all laptop PC's and Mac's, and it is actually an "exception" that Acer allows user to gain such access to switch it off. (through "hack")
2. If OEMs have this PWM Interface embedded in BIOS/firmware, the graphics vendor can write codes targeting it.
3. The friend's Acer notebook is listed on notebookcheck site as PWM-free, obviously It doesn't mean PWM Interface is not onboard.
4. As long as the Interface stays open, it matters not the panel is PWM-free or it has a fixed high frequency of PWM, because the software can easily take over and take control.
5. Intel graphics driver, and/or Intel power management driver, should have full privileges to this PWM Interface, Intel drivers may deem it "handy" to use PWM for dynamic control of brightness and panel power if it has detected such interface, thus completely ignoring other methods.
6. The frequency of PWM is probably dynamically adjusted too via the driver, when you watch full screen movie, the driver "sees" to it that one frequency band should be used, when you edit text or browse web pages (static content) another frequency band is used.
7. Because the driver can communicate with the LED backlight through PWM Interface, a driver update can completely change how the lighting behaves if it deems such change convenient or energy efficient.
8. External monitor has independent power cord so power saving is not a concern, it also has its own brightness mechanism, all of these data should be recognizable and later read by the graphics driver as the monitor plugged in, the driver shall then decide it's not its business to force PWM on it. (That said, I don't know if there is watertight evidence that the driver absolutely can NOT force PWM on external displays).
9. With that being said, maybe it can be concluded if the monitor is eye-straining there must be a dithering source, unless the monitor on itself uses a very bad PWM for dimming.( I think that's a rarity)
10. Apple should seriously consider adding PWM and Dithering checkbox to the display settings in a future update, it's not that hard to do. Night Shift is a good move, it shows Apple's concern over the health impact of excessive blue light from LED display is real. So why stop there, Apple? You have enough cash, please continue doing what's right for the customer!! This Eye Strain thing is real!

* "PWM Interface" (or PWM Programming Interface) is just a made up term, some or all of the above may be proven wrong, I'm evolving.

You may not get anywhere with Apple here. I've had lengthy discussions with them about my first reaction- very severe, and they tacitly recognized the problem, but officially they'd tell me nothing.

Later my contact there called me to set up an interview in order to get the info "directly to the engineers" at Apple.

It was a frustrating and lengthy call, since he was obviously being coached as to what to ask, how to respond, what specific language to use, etc. Many long pauses, none answers, evasions. It was clear Apple was interested in getting details of my experiences. They know it's a problem. They know it only affects a percentage of the population, from mild irritation to severe migraines and vomiting. Who knows what they will do? If anything. They're the richest corporation on the planet for a reason.

Some people can't eat peanut butter, but this is fixable. I'd gladly pay a premium to get my professional life back. By refusing to officially acknowledging the problem and sharing what they know, Apple is hurting people who depend on their products.

That's their choice, but you wont get their attention here.

When Bob the Bot makes a snark: Yes, many of us have been to eye specialists. There's nothing wrong with our eyes. For most of us they're better than most. Whatever causes this - it's not known to the broader medical community. Little research is being done publicly. Most opthamologists aren't computer specialists.

This forum's purpose is to let people share info and realize it's not just them. So it's not for you is it? Bot.

Continued...

(Pics and quotes) From this 2017 review by Mehdi Azzabi

https://www.rtings.com/monitor/tests/motion/image-flicker

Surprise!

1. BFI frequency can be independent from that of PWM, and it tends to be lower than that of PWM.

2. EVEN On flicker-free PWM-free monitors BFI can still be present.

Image flicker is a behavior commonly found on monitors where images shown on-screen will appear as a series of short duration impulses instead of staying on-screen constantly until they have to be replaced by the next image. Flicker has a large impact on the appearance of motion. Its appearance can either be an intentional method to improve motion clarity (usually referred to as black frame insertion or backlight strobing) or simply a side effect of the screen's brightness adjustment system (PWM Flicker).

Side effect....!

So aside from dimming, PWM can concurrently work as a means to cope with motion blur, taking full advantage of this "side effect"?

IS that why some devices to intentionally use a low freq band for PWM?

You figure out!!!

Following are some of the dedicated methods dealing with motion blur::

(Quotes) from this wikipedia page

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

M e t h o d :: S t r o b e d b a c k l i g h t s

Different manufacturers use many names for their strobed backlight technologies for reducing motion blur on sample-and-hold LCD displays. Generic names include black frame insertion and scanning backlight.

Reading through this, i keep ask myself why somebody thinks sample-and-hold must go?

In some cases it's bad I can see it. (games?, movies?)

But too many scenarios out there in which sample-and-hold makes more sense::

Reading, drawing, texting, coding, writing, surfing, modeling, composing, idling, mind drifting, laughing at your screen,....

Not much "motions" involved in all those scenarios!

So toss that beautiful baby out with bath water??

Genius!!

Philips created Aptura, also known as ClearLCD, to strobe the backlight in order to reduce the sample time and thus the retinal blurring due to sample-and-hold.

The word strobe immediately alarms!

What's the Frequency of this strobing then?

You figure out!!

Samsung uses strobed backlighting as part of their "Clear Motion Rate" technology.[9] This was also called "LED Motion Plus" in some previous Samsung displays

Ditto!

BenQ developed SPD (Simulated Pulse Drive), also more commonly known as "black frame insertion", and claim that their images are as stable and clear as CRTs.[11][12] This is conceptually similar to a strobing backlight.

Stable? what does that mean?

CRT's have phosphor screens to soften the flicker graph because of decaying effect of it between refreshes.

Is that decay being simulated too?

Sharp Corporation use a "scanning backlight" which rapidly flashes the backlight in a sequence from the top to the bottom of the screen, during every frame.

One scanning cycle every frame? So..., 60hz?

Can I say bad?

nVidia has licensed a strobe backlight[15] technology called LightBoost to display manufacturers. This is normally used to reduce crosstalk during 3D Vision, which utilize shutter glasses, however, it also eliminates motion blur due to its ability to keep pixel transitions in the dark between LCD refreshes.A 'hack' method or utility tool is needed to take advantage of LightBoost backlights for blur reduction benefits.

No idea on that one, more data needed about LightBoost.

BenQ later developed their own native "BenQ Blur Reduction" technology, integrated into several of their gaming monitors. This offers a strobe backlight which can be easily turned on and off by the user. There is no control over the strobe timing or strobe length for the user, although third party utilities have been produced for this purpose. Newer firmware for the BenQ Blur Reduction monitors allow direct user control over the strobe pulse (timing) and strobe length (persistence) directly from the Service Menu.

Stand up and applause folks! Common Sense finally arrived at the door of BenQ so as to GIVE BACK TO US USERS THE MAGIC SWITCH TO TURN STROBES OFF!!!!

And Interestingly, the monitor has its own firmware driver for strobes !!?.

Eizo have also introduced their 'Turbo 240' option used so far on their Eizo Foris FG2421 gaming display. This allows the user to control the strobe backlight on/off easily to reduced perceived motion blur

Applause!! Common Sense is not exclusive!

LG introduced a similar 'Motion 240' option on their 24GM77 gaming monitor

ULMB is a technique provided alongside NVIDIA's G-sync technology, and linked to the G-sync monitor module. It is an alternative option to using G-sync (and cannot be used at the same time), offering the user instead an "Ultra Low Motion Blur" mode. This has been provided on various monitors already featuring G-sync (e.g. Asus ROG Swift PG278Q, Acer Predator XB270HU). For newer games with a higher demand for graphical power, G-Sync is preferable over ULMB.[17]

NO idea ....

To be continued....

Because manufacturers have their own practices pairing and calibrating shades, Ill do it the easy way, leave out unnecessary complications.

First lets see how to make new shades:

Map 64 shades to RGB value, multiply them by 4:

[0, 1, 2, 3, 4, .... 61, 62, 63] ==> [0, 4, 8, 12, 16, ... ... 244, 248, 252]

They would be displayed natively, no dithers required.

For the missing values ==> [1,2,3; 5,6,7; 9,10,11; ... ... 245,246,247; 249,250,251],

they will all be dithered in!!!

Lets see how to fill in gap between shades [4] & [8], that is, [5, 6, 7]!!!

{other triples would follow the same rule}

[5] is at the 1/4 point across, so 75% of the time the sub pixel would be put on shade [4] and 25% of the time on [8]!

[6] is the midpoint, a 50/50!

[7] is at the 3/4 point, so 25% of [4] and 75% of [8]!

The break points [MARK] would be precisely controlled, hence the term Frame Rate Control.

each cycle would be four, or two, or even one frame length if response time can afford.

everything's ready now, time to start dither simulation!

my first pick would be rgb(40, 125, 210)!

Values sent to each sub pixel would be cycled in this sequence::

R ==> .... || 40 . 40 . 40 . 40 || 40 . 40 . 40 . 40 || ....

G ==> .... || 124 . 124 . 124 . 128 || 124 . 124 . 124 . 128 || ....

B ==> .... || 208 . 208 . 212 . 212 || 208 . 208 . 212 . 212 || ....

that's like looping through 4 colors, extracted by column

Not much differences among them I must say,

I have a bad feeling the flicker will be weak.

Animate it anyway!

Flickers?? Yes it's weak!!

That's a huge let down!! I was expecting much more than that!

If you can even sense it that's because I've slowed way way down that already "sluggish" gif!!!

maybe I would rate it 3 out of 10; 2 for the flicker, 1 for the aftershock of disbelief.

First impressions? The space in adjacent shades are too cramped to embody a serious flicker. That's not to say human eyestrain gene won't pick it up though, just compared with BFI and PWM it's definitely dwarfed (me opinion, of course)!

Other colors at different positions on the rgb scale have been tried for dither too but Ill not bore you with more gifs!!

Seems unable to do enough harm?

Or maybe in other ways likely overlooked, remember the [MARK]? i ll return to that later 👿

As for what dynamic PWM is, Goog

Very sophisticated now, not single backlight piece, but backlight array/matrix.

note the spatial/temporal control of these units on the fly, all programmable

at one moment it could be a double scan 3 phase

next moment maybe a n-scan x-phase in relation to brightness/contrast/battery..etc

(thus may explain why less or more strains at different settings)

that's just 1D!

and don't forget OLED, where every pixel can be turned on/off individually!

Translation: for a 2000 x 1500 pixel dimensions OLED screen, thats 2000 x 1500 = 3,000,000 or 3 million micro backlights alike!!!

No rules against that it can not be used as strobes to do dimming or for anti-motion blur purpose

that's a 2-phase, chessboard pattern.

on and off squares are among each others, so the flickers "blend in".

All patterns programmable, where and when to be on/off.

in 2 yrs we might have display A.I. decides for us about what is the best for our eyes to ingest or interpret

idlike to decide myself thank you.

Opt out!!

It took some very long time before we got the lighting right for the best results, if you see "noise" dancing in the video that is what it is on this monitor, live! We made sure that no amount of grain by the camera itsefl got in there alive!

click the gear icon hit 720p [MUST!]

see how the "dithering" is not continuous

e-ink test part 1: dots resume/stop moving as page scrolls - YouTube

seeing news on youtube, the dots update very quickly, hardly any room for temporal dithering to step in i'm afraid

e-ink test part 2 : video test - YouTube

No "settings" gonna work, other "workarounds" aren't recommended because, you may unwittingly trap your eyes into an irreversible sub health state.

This WWDC Apple introduced the system level dark mode(finally!!) which will remove/minimize OS whitespace, whitespace allows even bigger amplitude, the dark mode contains and has it flattened, so it should help.

You may want to visit ledstrain dot org, a lot of info! JTL is very helpful there.🙂