DisplaysShouldNotBeTVs @aiaf I can definitely still vouch for the fact that there's something weird going on with macOS (or, specifically newer macOS DCP firmware, still haven't tried a DFU restore yet) aside from dithering, both in terms of backlight control and flicker

Installed a matte screen protector to get glare out of the equation entirely and I can notice this switching between macOS Ventura 13.6.6 (but on top of Sonoma 14.4.1 firmware) vs. Asahi Linux 40 (but on top of "temporarily loaded in" Ventura 13.2 DCP firmware)

-

Let's split flicker into two separate beasts, the classic type of "FRC" temporal dithering that is definitely distracting and a big source of strain but doesn't really "mess with the position of things" - text and colors are obviously flickering, but every pixel despite this flicker still stays within it's correct place and it doesn't feel like the entire UI is twitching erratically in all four directions like you're trying to look at it on a moving boat.

Then, there's the really weird stuff going on inside macOS which I can only describe as a "disorientation pattern". This is the type of flicker that actually makes it feel like everything is moving around, no pixel ever feels "locked" to one location to my brain, your attention randomly jumps to small items on the screen, edges seem to "glow", and seems to be responsible for the really weird "false 3D depth" effect many people such as @async have discussed here. "When moving one item around it also feels like something on the entire other side of the screen is slightly changing too."

-

On macOS, it feels like I can successfully disable "typical FRC" flickering with Stillcolor, but the weird "disorientation pattern" flicker remains. On the M1 Air it's tolerable enough to use outdoors, but indoors even with the matte screen protector on I get tired so fast looking at macOS even with Stillcolor and disabling as many IOMFB properties as possible.

The yellow glow that @Ruoma mentioned is here and even feels like it's shining through my matte screen protector. here are obvious shenanigans going on that are either coming from macOS or the newer Sonoma firmware.

-

Meanwhile, while booted on Asahi, the "typical FRC" is currently unsolvable since there is no Stillcolor equivalent in Linux to interact with the IOMFB registry. I can notice flicker that feels similar to the type of FRC used on external monitors. The same "subtype" of flicker that is also present in macOS but can be disabled with Stillcolor.

However, the "disorientation pattern" flicker is not here at all. Even though I can still notice FRC which is definitely still a big issue… on the other hand, everything feels "locked into one location", images feel much flatter, the "false 3D depth effect" is significantly minimized. And very interestingly, the feeling of "backlight glowing" essentially stops.

When scrolling within one small scroll area, unlike macOS, it doesn't feel like "the rest of the screen is moving as well". Only the small scroll area feels like it's moving.

With the matte screen protector, I can actually use Asahi Linux indoors without the instant "tiredness and nausea symptoms" I get from my macOS install. I still get external monitor-esque FRC strain symptoms (which probably could be fixed if Stillcolor was ported to Asahi Linux 👀👀) causing the screen to still feel harsher than a FRC-free display, but it's so much better than macOS.

Web browsing on Asahi vs macOS feels like using two entirely different computers, and this is true even if I only focus on the solid colors and not the text rendering.

"FRC" flickering is still present in Asahi because I can't use Stillcolor, but the really weird "instantly recognizable Apple disorientation effect" type of flickering that Stillcolor can't disable, for lack of a better phrase, is temporarily gone while using Asahi until booting back into macOS.

-

There is definitely a lot more to investigate aside from enableDither. If Asahi can still have "totally recognizable FRC flickering going on" to me while so many other symptoms are gone, there are definitely other types of flicker that macOS is adding in that make it much worse than just regular FRC.

    DisplaysShouldNotBeTVs I’m upgrading from a 2008 Mac Pro with a Cinema Display 🙂 Was hoping to connect the air to a studio display but maybe I should revert to a Mac Studio as I assume with an external display, I’ll have more options to get a doable solution.

    I’ll try the betterdisplay setting. This has nothing to do with a dummy mirror right? Cause I’m confused on how to do that. Just a setting on the internal display right?

    thanks

      DisplaysShouldNotBeTVs

      When Asahi Linux was in early stage people were reporting reduced eye strain compared to MacOS. BUT earlier versions had no GPU drivers included (Asahi team had to reverse engineer them). Possible explanation is that dithering functions do not work without proper GPU drivers.

      The Apple Silicon gpu drivers were added to Asahi in 2022-december (see here)

      Maybe you can try Asahi versions before 2022-december to check in difference for eye comfort.

        Ruoma I guess nowadays you need to pay a bribe at apple store in order to get not rotten apple 🙂

        Hunter20 Maybe you can try Asahi versions before 2022-december to check in difference for eye comfort

        How do you do that? I was trying to search how to install earlier versions and there was literally no results, like no one has ever wanted to do that so far and I couldn't find any way to force the installer to download an earlier version etc

        Also, the main point of what I've already been saying in my previous post is even though FRC is present in the newer GPU accelerated versions of Asahi, it still feels sooooo much better than macOS. Asahi only has regular FRC temporal dithering (which is still a big problem of course, but far less bad that whatever macOS is doing) BUT seems to totally not have the weird disorienting types of additional twitching false 3D effect flicker that on the other hand macOS still has at all times (even after using Stillcolor)

        Asahi on "temporarily loaded in" 13.2 DCP firmware: FRC flicker cannot be disabled, but the "Weird Flicker" (AKA any additional disorienting false 3D type of flicker) is NOT here at all!

        Ventura 13.6.6 on 14.4.1 DCP firmware: FRC flicker CAN be disabled with Stillcolor, but the "Weird Flicker" still happens no matter what, even after using Stillcolor

        Just to clarify, my brain is actually able to tell these two types of flicker apart which is why I feel comfortable saying this, lol

        The main thing I haven't tried yet is running Ventura with real Ventura DCP firmware instead (instead of the weird "Ventura OS on Sonoma firmware" limbo state I'm currently in) which I need to DFU downgrade and wipe to do. Which means that as of now I'm not sure if the Weird Flicker is a macOS thing in general or more specifically a Sonoma DCP firmware thing.

          Regarding my issue with the yellowish white, I don’t understand why a white background with black text would need to flicker or dither or frc or whatever is going on? I mean it’s just white no? Why do I get slightly nauseous with just white? Or is it a magic Apple white that only true 10 bit panels can properly display??

            Hunter20 It's not actually white, notice how if you mess with turning off the brightness compensation property it will switch from a yellow white to a reddish white implying that the yellowish white you were just seeing was not actually true white. Who knows if the reddish one is true either…

            One of the reasons why they do this is probably relating to something that I observed on an earlier 2016 Intel MacBook Pro: in Boot Camp Windows, where there is much less color calibration going on, I noticed that when turning the brightness up and down the screen will alternate around every 10% between a bluish white and a yellowish white. Probably as a side effect of however the hardware backlight works. I've also noticed a similar thing to a milder amount on Asahi Linux on M1 Air. Yet while booted into macOS on either Mac, this "doesn't happen".

            So, to "compensate" for this variance in color accuracy, I've noticed while booted into macOS, even on (some) Intel Macs in addition to M1 ones, the white is changing very slightly (which essentially needs flicker to achieve in such a subtle amount) to prevent it from "becoming inaccurate" and looking more similar at all brightness levels… and then causing tons of strain from doing so

            In addition to this brightness compensation there's probably tons of other post processing layers we don't know about that are ALSO affecting the white even further

            (to be clear, this is in regards to the internal display, most of this doesn't happen with external monitors. Stillcolor seems to actually fix external monitor issues for me entirely, the remaining problem in my case though is that all of the external monitors I own have their own internal FRC when provided an 8 bit signal even with a totally clean source like an old PC, and that's still a big issue for me.)

            So yes, you're actually semi correct in that "it's a special white that only Apple displays can provide".

            Also, to make matters worse, Apple will still dither even if they switch to true 10 bit internal displays in the future, because they actually made any Mac that supports Display P3 use a 16 bit color space on the GPU. I'll make another post soon about this. They mentioned / confirmed this detail in the WWDC 2017 talk about introducing Display P3.

            Finally, about the "black text" part - remember that (unless you're using a retro pixelated style font that is perfectly aligned to the grid) all the edges and curves of black text are actually made up of shades of gray and will thus flicker on any temporally dithered display.

            P.S. a quote from a blog post when Apple introduced EDR:

            https://prolost.com/blog/edr

            Apple has remapped “white” to something less than 255-255-255, leaving headroom for HDR values, should they be called for. The operating system is complicit in this trickery, so the Digital Color Meter eyedropper shows “white” as 255, as do screenshots.

            With Catalina, Apple quietly changed what “white” means for millions of Macs, and none of us noticed.

            Between high-bit-depth P3 displays, display color management, software calibration, and True Tone, it’s probably been a very long time since “white” in a Mac UI was truly 255-255-255.

            So yes, you're actually semi correct in that "it's a special white that only Apple displays can provide".

            Also, to make matters worse, Apple will still dither even if they switch to true 10 bit internal displays in the future, because they actually made any Mac that supports Display P3 use a 16 bit color space on the GPU. I'll make another post soon about this. They mentioned / confirmed this detail in the WWDC 2017 talk about introducing Display P3.

            Finally, about the "black text" part - remember that (unless you're using a retro pixelated style font that is perfectly aligned to the grid) all the edges and curves of black text are actually made up of shades of gray and will thus flicker on any temporally dithered display.

            P.S. a quote from a blog post when Apple introduced EDR:

            https://prolost.com/blog/edr

            Apple has remapped “white” to something less than 255-255-255, leaving headroom for HDR values, should they be called for. The operating system is complicit in this trickery, so the Digital Color Meter eyedropper shows “white” as 255, as do screenshots.

            With Catalina, Apple quietly changed what “white” means for millions of Macs, and none of us noticed.

            Between high-bit-depth P3 displays, display color management, software calibration, and True Tone, it’s probably been a very long time since “white” in a Mac UI was truly 255-255-255.

            Due to all of this I have given up on "strain-free MacBooks / iMacs / iPads / iPhones / Apple Watches" at this point.

            Using a Mac as a desktop with Stillcolor and a safe true 8-bit zero FRC external monitor is the only way to fully salvage Apple technology

            (aside from hacks like streaming macOS through remote desktop to an old Windows laptop, which is my main and only strain-free setup at the moment. If you're wondering why I'm not using an external monitor myself it's because personally I don't want to spend time getting external monitors working as I have ergonomic issues using desktops anyway, I strongly prefer laptops and the ability to not have to sit in one place all day.)

            My goal is to fully switch my portable workflow over to Windows and Android this year, because, even though there are still many bad devices there, at least I have options such as different brands to choose from and easier ways to fully swap the LCD inside of a laptop

              @DisplaysShouldNotBeTVs What I don't get is, if stillcolour is enabled and that removes dithering, what is it that's bugging me about the white? I thought dithering was the color managing flicker?

              Been playing around with screen settings and better display. Also just downloaded iris tech, seems to be working better for me than night shift and apple settings. HW brightness is up full, and lowered with iris.

              axu2

              I'm not an apple developer nor a well-connected business person but I do occasionally read this well-regarded blog called Stratechery. The blog author has interviews with the CEOs of these tech giants. It would be amazing if he brought up these issues with someone in a relevant position seeing as he talks about Apple's "strength in design and user experience". Seems like they're neglecting the basics of user experience.

                Ruoma I have issues with white backgrounds on recent MacOS versions also. Its the worst in Safari. Other built in apps are better. Not sure why.

                DisplaysShouldNotBeTVs Apple has remapped “white” to something less than 255-255-255, leaving headroom for HDR values, should they be called for. The operating system is complicit in this trickery, so the Digital Color Meter eyedropper shows “white” as 255, as do screenshots.

                With Catalina, Apple quietly changed what “white” means for millions of Macs, and none of us noticed.

                By the way, another unsolved issue with e.g. the M1 Air is that there is currently no way to disable HDR on the internal display and prevent the sudden shifting in brightness and color profile/tone mapping whenever there's something resembling HDR content on screen.

                In some Intel Macs like the 2018 MacBook Air you could Option-click display preferences to uncheck Use extended dynamic range, but no such option appears on the M1 Air at least in Ventura.

                When I Quick Look an HDR image, the brightness still increases and shows extremely bright HDR areas on that image much brighter than the white of the desktop (in cases where I'm not at maximum hardware backlight). And if I open some apps such as Blender, they very slightly activate HDR mode as well and shift around the entire color profile. With dithering disabled, this is super obvious as if I have a banding test image open at the same time the banding will shift a ton during the HDR transition.

                I've messed with about every possible IOMFB numeric/boolean property that can easily be modified and cannot get internal display HDR to disable.

                Unlike iOS, the "double invert" trick does not work — if you combine Invert Colors with the Option-click "Inverted Framebuffer" option in BetterDisplay, HDR will still be present. (This is different to iOS where in that case if you combine two Invert Colors, Zoom and Classic Invert, HDR can be disabled)

                You can also search the Console app's log for "SkyLight" or "headroom" and see messages relating to the HDR changes taking place.

                It's also really deeply embedded into the color management system as well — so even if you use max brightness (to "prevent" there from being any headroom for HDR) but then apply color profile dimming or even overlay dimming, the HDR image will still "shine through it".

                One way to semi-bypass HDR is to activate full screen "Ctrl-Scroll" accessibility zoom, but you can still notice the hardware backlight behind the LCD changing whenever e.g. the Quick Look window is hidden and shown even if the colors are now "clamped"

                It also points to a flaw of the "quantization" feature of BetterDisplay @waydabber, as even if quantization is set to 16, and I have a gradient displaying 16 shades on screen, there can still be an HDR white showing on screen that is brighter than that which is implying that the video card is not actually being limited to 16 shades. (I understand this may be a limitation of achieving the quantization directly through the video card LUT, as the video card is probably not taking the values given to it exactly as specified. There's also another issue with quantization I noticed where macOS will sometimes display in-between "smoothing" shades that are more precise than the set quantization level, the video card is definitely doing some kind of smoothing out of the LUT and not using it directly.)

                This "fake HDR" is something that is not present in say, Asahi Linux, at all, so it's another piece of the puzzle to investigate in regards to macOS

                @aiaf

                  Hello, As you probably know, the MacBook Pro 14" A2442 and A2779 has a problem with the display being somehow paired with the logicboard. If one replaces the display with an original display from another Macbook, backlight artifacts are displayed. I am interested in this thread because it deals with various hidden settings of the t-con board. I'm currently trying to research as much as I can about how this t-con board works and figure out if there isn't a software way to allow replacement of displays on these macbooks without backlight issues. It seems to me that the problem with the calibration could somehow be related to the uniformity correction property, because the replaced displays have artifacts in the upper part (they are the most pronounced), but the backlight is also weaker around the edges of the display. Can anyone guide me where to direct this research? Thank you

                    Andrejmm1234 If you have a 14" MBP with a replaced display with you, if you disable the uniformity2D setting in either Stillcolor or BetterDisplay Image Adjustments does it affect the replaced display? I'm curious

                      DisplaysShouldNotBeTVs That is quite interesting. When I turn the uniformity correction feature on and off in better display or Stillcolor, nothing happens - the display looks exactly the same with it being turned on or off.

                      axu2 i was last year, asked a couple teams, they never heard about our problems. I shared with them our community. Nothing changed as you can see..

                      Apologies for not regularly following up with this thread.

                      I appreciate all the investigations re. the various properties. Is IOMFBContrastEnhancerStrength verifiably problematic re. eyestrain? Any before/after pictures or videos? What does it do exactly?

                      I still firmly believe that the community should spend its time and effort on reverse engineering the TCON's firmware, understanding it, or at least doing some comparative analysis between the various versions. In particular, comparing the payloads/configs of the M1/3 Air TCON against an M1/3 MBP, for a start.

                      @DisplaysShouldNotBeTVs re. Asahi Linux, IIRC, they do not expose any methods from user space to change these properties. They sort of go with some default values and rely on mostly on default driver behavior. Maybe you can raise it up as in issue their Github to explicitly disableenableDither and see how that goes.

                        Andrejmm1234 I believe this is a TCON firmware issue, upon replacing the display, the TCON must be accordingly updated by Apple to drive the display correctly. We've uncovered a few frameworks and code that deal with updating TCON firmware.

                          Applesexual this should never happen unless you restart your Mac. What macOS version are you running?

                          aiaf Is there any way how to update tcon board right now or any sign that it could be done without Apple software? Because Apple refuses to pair displays that were not purchased through Apple.

                          dev