I disabled dithering on Apple silicon + Introducing Stillcolor macOS M1/M2/M3
I can confirm on sRGB and Colour LCD profile macbook air m2 13 inch when you check and uncheck "Disable dithering" something changes within that black to white colour gradient. However, banding remains consistent on both settings, rather some sort of noise artefacts that change when you check or uncheck. So I can only say yes - something is happening, but weather that something is something intended, I am not sure.
I am wondering is there any possibility to add to the mac app similar to AMD GPU options dropdown for bit output?
Donux are you viewing the gradient on chrome? try firefox, chrome adds it's own (non-moving) layer of dithering to all gradients that makes it harder to tell where banding appears
- Edited
aiaf I feel like this explains what was happening on my monitor! I've recently realized that when Stillcolor is enabled, even though that "checkerboard and obvious visible flickering" was finally fixed, there is still a very slight amount of dithering on certain colors that I can see flicker if I'm moving my head, even though this happens on wayyy less colors than it does before I installed Stillcolor.
So even though I thought that my monitor was just 6-bit at first, I was probably wrong and I'm pretty sure it's actually 6-bit + FRC 8-bit, and the dithering that I'm still seeing on a few colors is monitor FRC — and the checkerboard pattern and "banding" issues I got even while Mac dithering was still enabled was due to the monitor FRC conflicting with Apple's GPU dithering.
Eye strain is still a good amount less with Stillcolor disabled, but I was still noticing a little bit of annoying flicker, and FRC on the monitor itself explains it. I will try soon with a newer 2K monitor (that I think is true 8-bit? but I'm not sure… we need some easier way to tell this) to see if I can reduce flicker even more
BTW yes a way to finally change color depth on an M1 Mac would be awesome!!
- Edited
aiaf I'm starting to suspect that even if it's not a second layer of dithering, maybe the reason why even a very dark gamma doesn't seem to affect banding on specifically the internal XDR screen is because of the individual control of mini-LED brightness zones.
If that's the case, that would also mean the screen might be using different levels of PWM in each backlight zone depending on which color is displayed there.
(This would be similar to what I've noticed in slow motion recordings of some OLED iPhones and the Apple Watch, which have backlight control per pixel. Sometimes certain shades of gray flicker at different rates than others.)
There's also the chance that the internal screen is 8-bit + FRC to achieve 10-bit compatibility (and thus achieve more precise shades of gray)… and this stage of FRC happens at the panel level after it's already received any 10-bit signal, similar to a lot of ""10-bit"" external monitors?
I wonder if there's any way to force all mini-LED zones on an XDR MacBook Pro to the same exact brightness?
For example, get black to show up as that classic "blue-ish IPS glow" instead of completely turning off those zones as the Mac usually does, which would also solve the infamous and widely-discussed "blooming" issue these Macs have with white text on a pure black background.
(Of course, at the obvious cost of reducing contrast ratio, but I've never cared about black levels honestly. I actually prefer displays with just one uniform backlight zone... after all, you don't light up a room or a sheet of paper with "10,000 invididual ultra bright mini-LEDs" blasting in your face 
)
- Edited
Hey @aiaf I tried your app today with MBP 14'' M3 connected with display port to my Asus VG27AQ 27'' 2K monitor. So first I tried app just on a laptop, can't say if something changed because that MBP is using PWM, so I was affected by that. After I tried just on my monitor, and I think after 5-10 mins I started to have slight nausea, I think the same as it was before. So I don't know if its working with external monitors or not. I need to grab MBA wtih M2 without PWM and test your app properly. Will try to do it during this week. Thanks for your hardwork btw, would be great if you could share your buymeacoffee or something similar
My monitor is true 8bit, without FRC, so all new MACs are forcing it to use dithering somehow.
Also another one interesting thing I noticed, that if i'm using different version of display port cable to connect to 4K monitor for example 1.2 or 1.4 (don't remember exactly, need to check specifications for it), mac will not be able to send 10 bit 444 because of the display port limitation and bandwidth. I saw it on some asus pro art monitor, it was 4K 8bit+FRC. So in the monitor OSD I was able to see 8bit 422 or 444 with my old display port cable, and I saw 10bit 444 with newer cable version, I think it was 2.1 display port.
Just throwing this out there for those still having issues.. could also text rendering be an issue? Especially on either rgb/bgr pixel layout of your monitor? With one being better than the other with how macos renders text ?
I don't own a Mac currently btw. Looking to hear more feedback first before trying
Does anyone have any contacts to Apple, so that this could be shared to Apple and maybe, just maybe they could implement this as a feature so that those sensitive could disable dithering.
Also, we should encourage them to do PWM free displays. Even high HZ will cause problems. It just needs to be completely flicker free.
Maxx Does anyone have any contacts to Apple, so that this could be shared to Apple and maybe, just maybe they could implement this as a feature so that those sensitive could disable dithering.
If it works the way I think it does, it works with controls already built into the driver by Apple, just hidden from the user.
Maxx Also, we should encourage them to do PWM free displays. Even high HZ will cause problems. It just needs to be completely flicker free.
I think the M2 Air is a "classic" IPS display without PWM, so in that sense you're already covered.
But the others are miniLED based displays, which have a different backlight compromised of many individuals LEDs behind the display which can be individually controlled. From what I've seen with other implementations of miniLED none of them seem to be exactly PWM free unfortunately.
One thing though - I've never had any relief from Ditherig in Windows. I do see the banding in lagom.nl, but the display still irritated, though it is not PWM.
My Lenovo x280 does not irritate with Ditherig or without. If I don't enable ditherig, it does not show banding, but still it is irritation free.
Would anyone understand why this is?
jordan For everybody asking for a donation link, I appreciate your support! But please donate to LEDStrain or any other awareness raising campaigns, perhaps even the Asahi Linux project, or a charity of your choice. The community needs more momentum to be able to reach device manufacturers and compel them to make changes.
JTL exactly, it's the same property that NewDwarf tried to disable in his boot-args attempts. Apple engineers know every inch of their display pipeline, we're not inventing new things here, but asking Apple to make these options accessible to the end user is another matter entirely. Has anyone filed a radar with Apple re. dithering and color depth?