I disabled dithering on Apple silicon + Introducing Stillcolor macOS M1/M2/M3

Donux

tiffoknee I have checked it on Amazon, so there are quite a few people complaining about accuracy of measures. But its normal for products to improve over time, so it could be just some period where there was a big variance. I assume you are doing those tests on still image or desktop right ? If not, probably it would be best to use the same image at full screen and also to make sure HDMI is set to the same standard (if it is possible). And of course those monitor contrast, brightness knobs should stay the same, as well as distance and coordinates of measuring (including ambient light conditions)

tiffoknee

Donux yes I’ve tried to keep everything as similar as possible, but here’s definitely some weirdness around whether the phone is plugged in to main power - so now I’m wondering about the monitor as well. Is it picking up EM?? Not sure!

Donux

tiffoknee Well, we could go into nitty gritty details, what it seams like - your findings go side by side as to how people feel. Except, of course, those intel machines that clearly have PWM, and there are quite a few to watch for, not to mention batch issue. Generally speaking, if you suffer from eye strain on apple silicon, switch to intel. Intel might cause also eye strain due to PWM if you will have it, but at least, you could count on external monitor to be much better solution to apple silicons. I would really like to reach out to Apple, really to the very core team and say guys you are loosing customers here who could purchase those 2k-4k Macs, but only if those were solid flicker free devices. But if apple is not listening, maybe apple is not the same apple as it used to be. It has changed, grown into something else.

Clokwork

Donux Apple seems to have little to no interest in assisting us as we play a small roll in their overall revenue. It's sad but true. If Apple Silicon devices are truly only outputting 10 bit even when in sRGB mode, then yeah, many of us are in trouble until better screen technology comes along.

My issue is, lets say a true 10 bit monitor becomes the norm where P3 Wide color can be displayed without any tricks (FRC). Then the goal post will move. What else will marketing have to rely on? sarcasm incoming You will then see that the next big thing is 12 bit color… then 16 bit color… etc…

Clokwork

Is this app still effective as of today on newer hardware from Apple?

macsforme

Clokwork After upgrading from Sonoma to Sequoia, subjective strain felt worse but now I can’t be sure if I was just having a bad day or if there’s actually a difference. StillColor purports to work on Sequoia, in that it sets enableDither to false in ioreg output as expected. I have not verified capture card output on Sequoia specifically, but it did successfully disable dithering on external displays on previous versions so I assume that still works.

Addressing my work-issued M1 Max 16-inch mini-LED MacBook Pro specifically, I am unconvinced that StillColor disables dithering on the built-in display. When staring at the screen and using my peripheral vision, I would argue that I can clearly see temporal dithering (or FRC, or some kind of color flickering) on the built-in display regardless of StillColor being enabled or not. So in my opinion, an Apple silicon desktop with StillColor enabled and a known-comfortable external display is your best bet (although some on this forum still find this configuration to be uncomfortable).

Donux

macsforme Well I understand your point and feel I was there just like you before, but now I would disagree on this one. On BENQ PD2705Q, macbooks no matter what felt aweful, windows machines were OK (Especially windows 10). BENQ has confirmed that monitor is true 8 bit monitor. Also tests peprformed by @tiffoknee also support this, although I wish @tiffoknee did a test using usb-c port, as HDMI in my opinion is already an obsolete format and everything should be usb-c, but I assume his monitor does not have that option. Overall I think sometimes we get attached a bit to specific ecosystem and there is a bit of wishful thinking going on. The truth of a matter is - for eye strain sufferers apple silicon is not recommended machine, because once you get an eye strain, you are all of the sudden in a lots of medical problems - fatique, eye strain, headaches, brain fog, etc. When you are paying extra for apple, you actually buying much more than a machine. You are buying an experience, and a pretty aweful one 🙂

AshX

macsforme

I downloaded Stillcolor on the new MacBook Air computers and the M4 Pros all running Sequoia at the store and noticed no difference. Unless display units are somehow locking programs out (Stillcolor successfully installed) I wonder if the new OS applies other forms of dither or if it introduces new features that cause discomfort.

We need more data as to whether or not Stillcolor is only effective on certain MacOS versions. I wouldn’t be surprised since the trend for both MacOS and iOS since 2021/2022 seems to be more intense dithering and other yet-undiscovered software gimmicks that make screens more uncomfortable and difficult to view. If this is true, either @aiaf might be able to run some tests on the latest OS and MacBooks, or we just have to compile a list of hardware and corresponding MacOS versions that Stillcolor effectively tilts the scales in a positive manner.

This is such a difficult problem to solve because just when you think progress has been made, Apple introduces new software that seems to undo all the progress. It’s like playing whack-a-mole. I’ve read about serious quality control problems with the MBA line going back to the M1, and I’ve seen it myself with the latest M4 models. So clearly there is a hardware defect component that is affecting folks. Trying to determine where hardware or software begins and the other ends is a real challenge.

macsforme

macsforme StillColor purports to work on Sequoia, in that it sets enableDither to false in ioreg output as expected. I have not verified capture card output on Sequoia specifically, but it did successfully disable dithering on external displays on previous versions so I assume that still works.

I just confirmed via capture card and the ffmpeg time diff method that StillColor on Sequoia still disables dithering on external displays. Subjectively, I could not tell any difference on the built-in panel with StillColor on versus off.

Clokwork

Donux Yep. I will be moving to Linux where possible. I also can't wait for Valve to release their Steam OS. That way The only thing I might need windows for are a few programs and won't have to use it often.

xenu

Hello,

I updated a M2 Macbook Air to the latest MacOS 15 Sequoia, and I found that StillColor was able to briefly disable temporal dithering, but some minutes later I found it was enabled again.

Has anyone seen this issue? I didn't see it with the previous MacOS.

Thanks.

amusesmile

AshX Just chiming in that as much as I appreciate the project more than I can express, having StillColor on my Sonoma M1 Mac doesn't fix the issue, so +1 to what you're saying. Maybe it makes it a little better but I feel there's still dithering or something weird going on with text. I'm actually in this thread to read some of the updates and learn more about what it's calling on the system to know how it should work technically speaking. I'm planning to invest in a capture card to measure frame-by-frame differences in an objective way comparing this modern Mac to a 2013 Mojave MBP that I know doesn't have whatever this dither/color/pwm/black-magic effect is.

AshX

amusesmile

I tried Stillcolor on 3 MBA 13” M4 laptops at Best Buy and it actually made the screen worse. It was more uncomfortable, with both dithering and uniformity 2D disabled.

I would love for more people on M4 to see if Stillcolor is actually working. I’m still yet unable to pick up one myself to try, but the results were not encouraging. I wonder if Apple somehow changed the protocol with the M4 series chips.

AshX

macsforme

So to clarify, you confirmed via capture card that Stillcolor does not disable dithering on internal displays despite it saying so via terminal? Are you using an M4 machine? If not, then this means Apple changed something within Sequoia to apply dithering on internal screens. How is that even possible on the older silicon machines? Does BetterDisplay achieve a better result?

Donux

AshX Apple probably didn't like it, since it interferes with their built-in obsolescence logic.

macsforme

AshX No, I was just observing that I cannot tell any difference on the internal display with StillColor on versus off. It supposedly does disable dithering per ioreg output, but subjectively I cannot tell the difference. I have no way to capture the built-in display output in order to analyze it.

This machine is a MacBook Pro 16-inch with M1 Max.

AshX

Donux

Have we seen @aiaf around here in a while? Last I heard he was working on Stillcolor 2.0. Does anyone have a way to contact him outside of these forums?

Can anyone here who has an M3 or older machine running Sonoma or Sequoia compare to see if Sequoia caused Stillcolor to stop working? I will be receiving a MBA M4 this week so I’m happy to check on my end, but it would be great if someone with previous gen machines could also check. If I had a spare $700-800 lying around I’d pick up an M2 or M3 MBA to compare myself.

I’m new around these parts but I’m surprised we’re not on top of this. All I see is a lot of folks guessing and not a lot of empirical evidence - and Sequoia and M4 has been out for quite a while.

What about the BetterDisplay dev? Does he post here? Why do I see anecdotal posts saying BetterDisplay seems to work better on Sequoia to disable dithering than Stillcolor?

Can someone with more experience or permissions on this forum start a master post of some sort for us newbies? I’ve spent almost a month combing through replies in topics sometimes years old trying to glean any sort of information that has been proven to help. There really needs to be an organized master post or master list for new users and also just for organization’s sake.

Look what @aiaf was able to accomplish in a matter of weeks. I do think this is solvable but I get the sense a lot of good data is buried in the replies of old topics.

AshX

macsforme

Thanks for the additional details. I was re-reading early replies in this thread and there was the possibility of the Pro models applying hardware level dithering in addition to software. That would line up with what you’re experiencing - no dithering on external displays with Stillcolor enabled, but some level of dithering still enabled on the internal Pro display. It would be interesting if this could be captured and compared.

Donux

AshX But @tiffoknee did test what clearly shows apple silicons flicker more on external montitors too. Is it because he did not turn off dithering on the app ? Also, I have used macbook air m2 on BENQ PD2705Q, and it does flicker with macbook regardless what settings you do on betterDisplay, I have tried every possible combination. And eventually it gets brutal. And it is opposite situation with windows machines, even with those troublesome intel GPUs. I will certainly have another test with Asus Zenscreen soon for both machines, but if macbook will be strainy again, I mean then surely there must be something firmware, software related. I was wondering, is it possible to expose all of those parameters in the UI software so that we could play around and test it in various ways ? @aiaf I think did not like that his app was compied to betterDisplay, when he did most of the work for it I assume.

macsforme

Donux It is true there have been reports of Apple silicon flickering despite StillColor, and other concerning reports such as macOS having discomfort even on e-ink. On the Windows side (in the context of the ColorControl tool), some people alleged that there was another subtle kind of dithering/flickering with I believe the 30-series even with dithering turned off in the tool. Also, I tried the ffmpeg time diff detection method on my work desktop with an Intel UHD 630, where dithering was clearly visible to the naked eye on the unlock screen (on my main monitor), and this method did not indicate dithering in the capture. So, I believe either the capture device is receiving different output (detected as a TV?), or the ffmpeg time diff method has limitations that I’m unaware of (the conversion command provided in the OP appears specific to 10-bit color output, so perhaps it does not provide the whole story). I know others here have done frame-by-frame analysis using various tools, so perhaps there are other methods that can help provide a more complete analysis.

AshX

Donux But @tiffoknee did test what clearly shows apple silicons flicker more on external montitors too. Is it because he did not turn off dithering on the app ? Also, I have used macbook air m2 on BENQ PD2705Q, and it does flicker with macbook regardless what settings you do on betterDisplay, I have tried every possible combination.

I read through @tiffoknee’s posts and I wouldn’t be surprised if the Pro models because of their Liquid Retina XDR screens are applying a dither standard the hardware level.

I did a quick search for your BENQ monitor and a few sources claim it’s true 8 bit. The fact that it is so integrated with MacOS according to BENQ’s site and that it’s 2K resolution and millions of colors makes me wonder if MacOS is still dithering at some level.

A few questions:

By flicker what do you mean? Dithering? How is it visible?
What are the specs of your M2 MBA and what version of MacOS are you using?
If you are on Sequoia, have you tried downgrading to Sonoma or Ventura to see if an older OS rectifies the problem?

Does Stillcolor/BetterDisplay at least help the internal monitor of the MBA?

Donux

macsforme Yes, need to have a bit more structure in order to theorise a bit more. I personally started to think, if dithering was created not just to generate color where screen is incapable to display it natively, but also do it in a way that is reverse synced with PWM flicker, or one phase out. This way turning the actual perceivable flicker into noise pattern rather than high amplitute/modulation flicker. If manufacturers of panels got some of their panels slight off to the spec, software would produce even higher modulation/amplitude, instead of cancelling it out by producing more or less constant noise, instead of a flicker. But this is just a speculation.

ryans

macsforme

macsforme Also, I tried the ffmpeg time diff detection method on my work desktop with an Intel UHD 630, where dithering was clearly visible to the naked eye on the unlock screen (on my main monitor), and this method did not indicate dithering in the capture.

Just wondering how you captured this. Some capture cards let you specify the EDID. If you have the raw samples, VideoDiff is another option to ffmpeg. If you can share them I can also analyze.

« Previous Page Next Page »