Ranked on a 1-10 scale based on symptom severity. All laptops running Ventura unless otherwise specified. All using default display settings. MBA 13" M1 -- Instant eyestrain and nausea. (10/10 I only used this for a few minutes. Brutal) MBA 13" M2 -- Delayed eyestrain, no nausea. (6/10 I used this for a week. The eyestrain starts off mild and increases over time. I can only use it for 15-30 minutes before it's too painful. Not sustainable for long-term use.) MBA 15" M2 -- Instant eyestrain and nausea. (9/10 I only used this for a few minutes. Much worse than the 13" version.) MBP 14" M3 -- Instant eyestrain and nausea. (9/10 I only used this for a few minutes. Very uncomfortable.) MBP 13" M2 -- Mild eyestrain, no nausea. (3/10 I used this 1.5 hours daily for 6 days. No major issues using the preinstalled Ventura. HOWEVER -- when I upgraded to Sonoma, it was immediately unusable. I can only use this machine running Ventura.) MBP 14" M1 Pro -- Mild eyestrain, no nausea. (2/10 I used this 2-3 hrs daily for 10 days. Used for up to 5 hours in one day with no major issues using the preinstalled Sonoma. HOWEVER - when I downgraded to Ventura, the eyestrain kept getting worse. I can only use this machine running Sonoma.) MBP 16" M2 Pro-- Instant eyestrain, mild nausea. (8/10 I only used this for a few minutes. Could be a 9/10 or 10/10 but I'm unwilling to find out.) MBP 16" M1 Pro -- Mild eyestrain, no nausea. (2/10 I have only used this for 1 hour so far, seems not too bad. Running Sonoma.) Most recent update: 12/28/23

Sonnis If you are desperate, I can use the MacBook Air 15 M2 by taking a drug called fluxarten. In my case 5mg per day is enough and I can work.

Your post matches up with my experience with these models as well. I haven't tried the M3 Pro yet or the M2 15" Air but otherwise spot on.

I've tried most of these and found them to match up to my findings as well. The worst for me were 14 and 16" MBP, then m1 air, then m2 air 15", then the best was m2 13" air (still not great as a daily driver). I have not tried m2 pro 13". I eventually gave up on macbooks, sold my units and bought a Thinkpad x1 carbon and it is 10x better than any Mac I've used in the last 5 years. I really did not want to give up MacOS but I was getting severe symptoms and now I just wish I made the jump sooner because I'm pretty sure I suffered some permanent damage over the last few years without being away of what was happening. Thanks for sharing your results!

Would you mind sharing the exact specs of the M2 Pro that you have? I'm on an Intel Mac right now but I've been thinking recently that I'm gonna have to upgrade at some point…

Great post thank you! +1 for sharing the specs of your M2 Pro 13". I too am still on an Intel Mac (2015 pro) and badly need to upgrade.

devilgrove It's the base model 2022 13" MacBook Pro M2, 8GB, 512GB running Ventura. The other machine I'm having success with so far is the 2021 14" MacBook Pro M1, 16GB 512GB running Sonoma. They both have a different "feel" to my eyes and brain, but so far I have been able to adjust to each one. Oddly enough I can't use the 13" with Sonoma or the 14" with Ventura.

Did you attempt to run the with forced low power / high power mode? Or simply with power supply vs low battery, With ProMotion the framerate it gets capped at 60 hz in low power. What application you run also makes a difference on what framerate it runs at. For example if you have some constant overlay that changes it uses a stable 120hz instead of going up and down. This can be observed with Quartz Debug from Additional Tools for Xcode.

async No I didn't. All my testing is just using the OS and Safari, letting the battery run down with default settings.

I'm testing as many MacBooks as possible. Here are the results.

macsforme

DigDeal Sorry to hear the firmware downgrade did not work for you. 😦 I wish there were better methods for empirical measurements of these problems (I see there are some hopeful developments in other threads here recently). Because of this, it can be hard to determine whether some irreversible changes are occurring during software updates (not limited to the main firmware), or if our brains now associate the whole device with being a threat, even after reverting to the previous, tolerable state. I hope our continued research here will ultimately bring some clarity.

@GBowler noted different levels of eye strain between two identical MacBook models with different installed firmware versions. I also have several MacBooks of the same year (2015), which all had recent firmware versions, but among them I find certain screens more tolerable than others. So, while firmware remains a possible factor, I am currently leaning more toward attributing my personal issues to panel variations.

RobC

macsforme you could check the display panel of your Mac with the following command:

ioreg -lw0 | grep \"EDID\" | sed "/[^<]*</s///" | xxd -p -r | strings -6

DigDeal

macsforme yeah, pretty sure at that stage it's at least partially psychological. I'm literally scared to update any usable device at this point.

Interesting note, again proving my point that it's 100% software related, is that using "safe mode" makes it tolerable. Not ideal, but not as harsh as in regular mode. Most likely some sort of hardware acceleration is disabled.

vegeta97

i think laptops with ips lcd screens would cause the least eyestrain?

macsforme

RobC Unfortunately, all display EDID values for this MacBook model/year appear to be identical. 😕 My understanding is that Apple removed the ability to see the display model/vendor after people were returning earlier retina MacBook models which did not come with a display from their preferred vendor.

RobC

macsforme . Sorry, the right command is:

ioreg -lw0 | grep IODisplayEDID | sed “/[^<]*</s///” | xxd -p -r | strings -6

I used this commands with several MBP 16 2019 and all three panels were different.

Donux

RobC Here you go, under the marketing layer, there is the same or maybe even worse situation than in PC market. Does Dell have the same problem ?

macsforme

RobC I currently have four of these machines. All four return only "Color LCD" for that command. Three of the four (including the one I subjectively find most comfortable) return the exact same entire EDID string, while the fourth is not a byte-for-byte match but when parsed by edid-decode all values still appear to be the same. If the ability to see more panel info returned on later Macs, that is good news and may be helpful to us here.

JTL I suspect you are right about VBIOS being stored on the main firmware chip (especially the Intel integrated VBIOS, but likely the discrete AMD VBIOS as well). My initial tests consisted of reconstructing an old firmware image (with some help from some guys on the MacRumors forums) from a combination of donor dumps found online and Apple's own .fd firmware images from their macOS installers (specifically an old version from the El Capitan era). While this seems like it should include an older GPU VBIOS, my preliminary testing did not seem to show any differences in GPU info listed in the OS between old and new firmware versions on this model (although there is still more testing to do). Also interestingly, someone on this board (maybe you? 😆) found an "AMDFirmware" update script in macOS installers, which seemingly suggests an AMD VBIOS update outside of the main firmware, but that could also be for different GPU models (such as those in desktop machines).

JTL

macsforme I have been investigating reports of eye strain on MacBooks being triggered by a firmware/OS update. I also bought a few tools to dump and flash different firmware versions on MacBooks (specifically testing the MacBooks I have, which are mid 2015 15-inch Pro models, both with and without the discrete AMD GPU). So far I cannot pin down the issue, but there have been several reports like yours and I plan to keep investigating. A GPU VBIOS update seems the most plausible.

This might not be possible, because it requires a "known good" system of the same model but in that case it might be possible to compare firmware dumps between a "bad" versus "good" system and potentially identify what's different.

From what I remember (almost) all laptops integrate the GPU VBIOS into the UEFI ROM image, as opposed to desktop GPUs where the GPU VBIOS is flashed onto a separate chip, so identifying or changing the VBIOS in use can be done with external tools.

Donux

DigDeal Maybe this is how they increase sales? Got eye strain, aha, need better product? I mean this sounds ridiculous but in this world I am surprised day by day as to what is possible when you have weak CEO and very pushy board.

DigDeal

Donux yeah well this plan backfired lol. I'm not buying another Mac, thank you. At least on Windows I could make a backup and roll back drivers and software on my own if needed.

Hunter20

NewDwarf

Link is broken, can you upload the kernel extension file again?

Is every MBP 2019 15" model pwm free, or are there panel variances?

MAS-76

Hunter20

Here's AHKinject_SCN078.dmg

https://mas-fx.filemail.com/d/sofofeusdcicyvs

Really interested if this works for you. Please report back! I tried on a 2019 16" MBP with Sonoma and didn't seem to work. See above. Let us know if this works on your 15"!

Hunter20

MAS-76 Thank you for sharing!

I found this documentation (see link) for AHKinject_SCN078.dmg on the internet. The document is from 2016. So I maybe the DMG was build for older MacOs versions, perhaps you can try to downgrade to MacOS (Catalina?) and try again. I currently don't have a macbook with amd card, but I am going to look if I can find one to try.

https://docplayer.net/28661617-Temporal-dithering-is-enabled-by-default-on-apple-macs-with-no-user-visible-option-to-disable-it.html

daniel_mate

MAS-76 @Hunter20
I have 2019 16'' with Radeon Pro 5300M

Running Ventura 13.6.1

Was able to load the kext fine

$ kextstat | grep -v apple

Executing: /usr/bin/kmutil showloaded

No variant specified, falling back to release

Index Refs Address Size Wired Name (Version) UUID <Linked Against>

178 0 0xffffff7f971eb000 0xff6 0xff6 com.amulethotkey.driver.ahkinject (1.1.1d1) C89AB321-97E6-3C86-ACBA-6498B6DCC100 <6 3>

tbh I'm not sure if I see/feel any difference.
btw I'm also using BetterDisplay, with the virtual screen the screen already felt more "static", thus not sure if that was already disabling dithering, so I do not see further improvements.

RobC

macsforme Did you tried with the following version of the command?

ioreg -lw0 | grep IODisplayEDID | sed "/[^<]*</s///"

If you put every answer line into an edid-decoder like this one: https://hverkuil.home.xs4all.nl/edid-decode/edid-decode.html you could get much more details of your panels, like the vendor and date in which they were manufactured among others.

JTL

macsforme I suspect you are right about VBIOS being stored on the main firmware chip (especially the Intel integrated VBIOS, but likely the discrete AMD VBIOS as well). My initial tests consisted of reconstructing an old firmware image (with some help from some guys on the MacRumors forums) from a combination of donor dumps found online and Apple's own .fd firmware images from their macOS installers (specifically an old version from the El Capitan era).

If you have any forum threads or information from elsewhere I might be of interest.

If you sent me the firmware dumps/images in question I might be able to take a look as well.

Also interestingly, someone on this board (maybe you? 😆) found an "AMDFirmware" update script in macOS installers, which seemingly suggests an AMD VBIOS update outside of the main firmware, but that could also be for different GPU models (such as those in desktop machines).

Not sure I recall that exactly, but what I do remember is some desktop Mac systems used GPUs similar to the same model as commodity PCs but had "Mac Edition" VBIOS firmware for them and it's a thing to flash cards sold for PC use with "Mac Edition" VBIOS's to improve Mac compatibility. So that's some prior art for differences between Mac and PC VBIOS's, but considering what else I said I wouldn't be surprised if the situation with Macbook's is different enough that might not be relevant.

macsforme

macsforme I can also confirm that one of my 2015 machines has a harsh panel while another has a calm panel (subjectively). The other two have not been tested extensively yet.

macsforme Three of the four (including the one I subjectively find most comfortable) return the exact same entire EDID string, while the fourth is not a byte-for-byte match but when parsed by edid-decode all values still appear to be the same.

Just in case anyone might rely on it, I wanted to officially recant my claim that one of my four 2015 15-inch Retina MacBook Pro screens was more comfortable than the others. In fact, all four seem uncomfortable in roughly equal degree. There were some differences depending on the operating system version and configuration running on the machine each screen was connected to, which initially led me to mistakenly believe there was a variation. At the end of the day, after a day or two of use (with a delayed onset), all of these 2015 15-inch Retina MacBook Pro screens give me headaches and vision problems.

I plan to make another thread sharing more research and overall findings regarding this MacBook model with regards to eye strain.

macsforme

RobC Yes, I tried the command as you posted it. As I stated, all the info is the same when parsed. Manufacturer: APP; Model: a02e; Made In: week 4 of 2015; Serial Number 0; Monitor Name: Color LCD.

Donux

RobC This command does not work on Macbook Air m2 13. If I remove EDID part I would bunch of variables and then last two IODisplayParameters and IODisplayWrangler. Otherwise there is no EDID information available. Any ideas how I could check out my panel manufacturer? This could potentially be a game changer, since I could buy used macbook and ask to check this information before buying to make sure panel is not some mickey mouse design.

RobC

Donux In your case try:

ioreg -lw0 | grep IODisplay | sed "/[^<]*</s///"

If you put every answer with the "00ff…>" format into an edid-decoder like this one:

https://hverkuil.home.xs4all.nl/edid-decode/edid-decode.html

you could get much more details of your panels,

Donux

Sonnis Could you please tell what is the panel manufacturer of this laptop you have used? - "MBP 14" M1 Pro" which caused very minor symptoms for you. This could be a really good purchase once the panel details are known, since on the second hand market these laptops already going quite cheap, yet still very powerful.

macsforme

JTL This is the post from you I remembered. After a cursory look, those firmware payloads look like they're possibly for later Macs (iBridge references, web searches returned AMD Vega references, etc.), so not affecting my 2015 machines. I also found a few posts here suggesting that later 2015 MacBook Pro display panels replaced by Apple due to "staingate" may be harsher on the eyes. I can also confirm that one of my 2015 machines has a harsh panel while another has a calm panel (subjectively). The other two have not been tested extensively yet.

So far I have not found good/bad machine differences that I can confidently attribute to firmware (my research continues). If I discover anything, I would be happy to collaborate over analyzing the differences.

« Previous Page Next Page »