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

waydabber

DisplaysShouldNotBeTVs Not sure what's wrong with it. Sometimes similar things happened of course while experimenting - usually booting to the Recovery environment, getting rid of the windowserver configuration file under /Library/Preferences, starting in safe mode etc, forcing a boot in clamshell mode with external displays connected, etc solved the issue (at least to gain back control). None of these seem to work and the Mac is fully dead after a brief loading episode, even keyboard backlight is gone. Did not have this issue on my M3 MBP though. And until turning local-dimming back on, things did work on the M1 MBP as well. Anyway, will figure something out, no worries. I'll wipe the whole thing and reinstall I guess. 🙂

UPDATE: I had to decrypt the volume in Recovery Terminal to bypass a (for some reason?) broken filevault login apparently, this fixed things (+clearing nvram, deleting windowserver config). 🙂 All good now.

I tried again, now it did not freeze like last time on the M1 MBP so it might have been a one off thing. But the entire process is somewhat unstable.

waydabber

DisplaysShouldNotBeTVs Nothing fancy, just show a full white screen briefly to create uniform max LED output. If you tried it you'll see the flash.

Donux

karut Could be, I will get in touch with folks from github, but I found this forum post, https://ledstrain.org/d/1797-switchresx-developer-m1-mac-does-not-allow-8-bit-color for switchRresX, so apparently this dev knows authoritative source and its fixed 10 bit signal.

waydabber

Donux Hmm. But if you guys want banding and limited color depth, I can do that, I'll simply add a quantization filter to the app - it should effectively downgrade the color depth to any arbitrary level and everybody can tweak that until it looks best in his/her eyes.

(ok, maybe not perfectly as if I use gamma table manipulation, as the table's max resolution might be an issue - a max of 9 bpc seems to work in practice on Apple Silicon - combined with the inherent tendency of the relevant API to average on boundaries - but the result still looks cool especially with GPU dithering off at the same time).

Donux

waydabber I will use and try everything that could potentially make MacBook great again 🙂 Please do not be discouraged, I think we should try everything that is within our agency to do until we bump into total dead end. Note that folks who said Asahi linux solves an issue, backed off in the same forum, https://ledstrain.org/d/1854-mac-eyestrain-not-the-panel-macos-asahi-linux-great-please-try/5, so … its hard to say that Asahi is a proof that OS software is an issue.

DisplaysShouldNotBeTVs

waydabber it would also be really nice to also be able to choose to force a null profile / linear color profile through modifying the gamma table, to remove any additional color calibration or gamma curve that would typically be applied by macOS

e.g. a website with a rgb(128, 0, 255) square would send as close as possible of those exact values to the display, even if this made things appear very uncalibrated or "not correct"

even though this would reduce color accuracy, it would also help reduce the likelihood that something is being mapped "between bits" / reduce the frequency of floating point colors occurring that might be temporally dithered at some point in the pipeline (because a non-linear gamma curve was applied to the color, for example)

to give an example:

i always leave my Windows devices completely color unmanaged with the "windows calibration loader" service disabled —

to my eyes, despite the color inaccuracy in photos this entails, this results in the best and most comfortable to read appearance of text and website content… but when i've tried to apply a color profile on Windows, text immediately feels harder to read. this is why i'd like to recreate the same "unmanaged" setup on macOS 🙂

Ruoma

This thread is getting unreadable for non technical people. Is there maybe a way to summarize the latest consensus and add it to the first post?

I just want to know which laptop is “fixed” and which isn’t. And how to do it if still color isn’t enough.
Also would like to know if a Mac Studio with an external display would be affected.
Or an Apple studio display with a Mac mini.

Would be much appreciated.

Donux

Ruoma My vote so far as to what is in still color and betterDisplay mirroring as - not fixed, but it does have visual improvement (Macbook Air M2 13).

SDR brightness related stuff - I assume only applies to macbook models that use mini LEDs. Please correct me if I am wrong.

waydabber

Donux Here is a version with the quantization filter. Simply use the slider to set the number of colors/channel.

https://github.com/waydabber/BetterDisplay/releases/download/v2.0.0-pre-release/BetterDisplay-v2.3.0-b27847-pre.zip

It's available under Image Adjustments. Please note that quantization is a Pro feature.

This example will result in 16 color per channel (red, green, blue), which is 16x16x16 = 4096 logical colors on screen. However due to API limitations the quantization is not perfect, there are some interim colors present at step boundaries.

@DisplaysShouldNotBeTVs - it is possible to create a custom ICC profile or simply edit the display's EDID (you can use BetterDisplay for override) to alter values based on which the factory ICC profile is created by the OS. This method does not work for the internal display. ICC profiles can't be easily applied to the XDR screen (you can force enabling color profiles with BetterDisplay for XDR displays - by enabling Unlock Color Profile management for XDR displays under Settings/Displays/Overview/Advanced…) but the effect will be partial only, proper calibration can be done via custom XDR presets).

Ruoma

@Donux I need to upgrade my 2008 Mac Pro with Cinema Display and was thinking a new studio display with a new mac, preferably a spec’d out mba, which together is a big purchase. Reasoning being if the mba didn’t work out, I would still have the ‘better’ studio display to connect to for long sessions. But if the mba would ‘infect’ the studio display, I’d rather get a Mac Studio, unless Mac OS is the l issue real issue here, which seems to be the case?

Donux

Ruoma Personally, I have not seen correlation between more premium product and less eye strain. There are a lot of posts about people getting top of a range product from apple, and complain about eye strain:

https://discussions.apple.com/thread/254774275?sortBy=best

https://discussions.apple.com/thread/255372278?sortBy=best

Only iMacs I know had a decent record, but I lost track of current situation. Not sure about studio display either, but I have read few posts about returning it due to eye strain (could be GPU cause). So obviously if you have a great return policy in your location, I would say there is no risk as long as you can use it for few weeks.

At the same time - its all about convergence, so all apple products are going into a direction where pretty much all models are same in some way. So if they make mistakes or unexpected outcomes with their patented technologies, it would probably affect every single product.

DisplaysShouldNotBeTVs

Ruoma was thinking a new studio display with a new mac

i'm not so sure about the studio display, despite it just being an LCD and not an XDR mini-LED, it uses the new "reference modes" system typically see on the XDR displays in place of the standard color profile menu

the reference modes system makes it actively hard to switch between color profiles as you usually would, and also implies that the display is doing its own hardware-level color management on top of what macOS is already doing — making it more likely the Studio Display is also performing it's own temporal dithering itself

in comparison, the M1/M2/M3 Airs and low end "Touch Bar" 13" MacBook Pros just use the standard color profile system for their internal screens, and seem to work the best with Stillcolor.

i'd say in the case of which laptop to get, try an Air first with Stillcolor…

if it doesn't feel comfortable to use, return it and try Stillcolor on one of the Touch Bar 13" Pro models instead. (but make sure to disable the Touch Bar, as it is OLED with strong PWM, there are some third party apps that allow you to black it out)

avoid all Pros with mini-LED XDR displays (aka all the MBPs with SD card slots), i got burned so hard from my M1 Max 14" MBP's screen, it's why i even learned i had issues with screens in the first place

Ruoma

DisplaysShouldNotBeTVs

Many thanks. To make sure I understand:

the airs should be completely ‘fixed’ with still color? I’ve read numerous threads from people also having issues with the airs, but all pre-stillcolor…
Buying an M series desktop Mac connected with the ‘right’ monitor should be problem free? Or would it still need something like still colour?

Any thoughts on wether the upcoming oled iPad Pro with dual oled layers would perhaps magically be better than one oled layer?

DisplaysShouldNotBeTVs

Ruoma Buying an M series desktop Mac connected with the ‘right’ monitor should be problem free? Or would it still need something like still colour?

For external monitors, Stillcolor definitely needs to be used, otherwise all monitors will flicker when connected to an M1 Mac.

But when Stillcolor is on, it's been verified through a capture card that as long as the monitor itself truly flicker-free (as in true 8-bit with no FRC on the monitor's side either, not just PWM-free), the display output will be totally still.

Stillcolor is what has made M1 Macs great machines for external monitors for the first time

DisplaysShouldNotBeTVs

Ruoma the airs should be completely ‘fixed’ with still color? I’ve read numerous threads from people also having issues with the airs, but all pre-stillcolor…

yeah, weirdly i haven't seen many reports after stillcolor released, maybe it's because everyone here has already returned their Airs after trying them a few years ago etc?

i'd just say to try it with an M2 Air for example, but one thing to note is that the M2 and M3 Air have a "billions of colors" display according to their spec sheet, which might imply an additional layer of dithering flicker applied by the hardware after the signal has been produced by the OS. (We haven't figured out if this is true yet, for example this possibly may only be true on the mini-LED Pros)

So the M2/M3 Airs might work for you, or they might not. What's good is that Stillcolor is known to actually function correctly and improve the display to a degree on these Airs… (compared to mini-LED Pros where there's a LOT more going on that hasn't been figured out yet)

But if "the level of improvement it gives is enough to make it usable" can vary from person to person. Make sure to run Stillcolor, that is essential.

If the M2 or M3 Air doesn't work, return it and try a M1/M2 13" Touch Bar Pro or M1 Air, as those only say "millions of colors" on their spec sheet which could possibly lead to Stillcolor being even more effective.

And always make sure to disable any other potential flickering light sources on the laptops of course, like the keyboard backlight or Touch Bar where applicable.

DannyD2

Ruoma For me personally, Stillcolor has completely fixed the visual issues I was having with my Macbook Air 15-inch M2. I've seen reports from people who still say they have residual issues, but I wonder if that's due to something like excess blue light that affects all displays and is not necessarily germane to Macs. There's also one thread with obvious backlight flicker that might just be broken hardware.

jordan

DisplaysShouldNotBeTVs woah really ?! So your saying.. a m1 Mac mini will even have absolutely zero dithering and frc with a true 8 bit monitor ?

macsforme

DisplaysShouldNotBeTVs But when Stillcolor is on, it's been verified through a capture card that as long as the monitor itself truly flicker-free (as in true 8-bit with no FRC on the monitor's side either, not just PWM-free), the display output will be totally still.

I wonder if this conclusion may be premature, since the capture itself was reportedly at 10-bit color, whereas the vast majority of monitors are no more than native 8-bit. This seems to be a factor we cannot ignore. I would be willing (even happy) to be proven wrong.

On another note, the ability to force an arbitrary color depth on macOS would be huge. My desktop monitors are all 6-bit + FRC, but I would be perfectly happy with a stable 6-bit image with none of these games macOS is playing. The most comfortable screen I’ve looked at in a while was an older HP Windows 10 laptop with a 6-bit screen.

DisplaysShouldNotBeTVs

jordan That's what aiaf proved with their capture card test at least, if you look at the "video diff" recordings they took, the moment they enable Stillcolor, areas that used to be filled with really annoying moving static patterns suddenly become completely still.

The main challenge is finding a true 8 bit monitor without FRC, as a lot of manufacturers often report this incorrectly.

(Like, I have a BenQ that no one has ever mentioned is 6bit+FRC, but even when plugging in an ancient Windows laptop, there's super obvious (to me) moving static on backgrounds everywhere I notice every time I move my head, that is clearly being generated by the monitor itself because I see the same pattern no matter what PC or Mac is connected.)

Some people are saying "cables might affect output" but everyone talking about that seems to own 10-bit (as in 8+FRC) or HDR displays, I believe that the best + most consistent results would be on true 8-bit monitors

jordan

DisplaysShouldNotBeTVs oh wow that's incredible. I know the HP omen 27q 2023 monitor is true 8bit BUT the better one is the LG revision. They come random with either a boe or LG panel. Revision starting with L is LG and B is BOE, says it on the box. Someone on led eye strain telegram reports they can use it for 8+ hours. Some other people have said the boe version is decent still. Just the boe is bgr and LG panel revision is rgb

Blooey

DisplaysShouldNotBeTVs Some people are saying "cables might affect output" but everyone talking about that seems to own 10-bit (as in 8+FRC) or HDR displays, I believe that the best + most consistent results would be on true 8-bit monitors

We should only see dithering if the destination bit-depth is lower than the source. The dithering algorithm should do nothing to a signal if the destination bit-depth is equal or higher than the source. So it makes a lot sense that using equal or lower data-cabling than the destination will disable FRC dithering at the destination. The destination will receive a source that has less bit-depth than its own native LCD destination. So at the monitor end, with that incoming signal being equal bit-depth to the native panel, the FRC dithering algorithm built in to the monitor can do nothing to the signal. It passes through to the display unchanged.

This cabling trick works because the OS seems to prioritize resolution over frame rate, over bit-depth. So as we choose lower spec cables, the OS is forced to first lower the color bit-depth. And because we have enableDither=false, the OS won’t apply a temporal dithering step when dithering down to the lower bit-depth for the cable.

So that should be why the 8-bit+FRC displays are requiring lower spec cables for all of this to work. Just wanted to try and help clarify some of this. I see your point that it is simpler to choose a higher bit depth monitor, without built-in FRC. Then there is no need to disable FRC at the display using lower spec cables, or fiddling with the OS display settings to saturate the cable.

photon78s

jordan

Which is the preferred version? They both are listed as non FRC 8 bit. The qs is higher refresh rate at 240hz versus 165hz.

https://www.displaydb.com/monitor/hp-omen-27q

https://www.displaydb.com/monitor/hp-omen-27qs

Would it be safer to first test using a bandwidth limited cable (HSwE) or is that not necessary anymore for Stillcolor or BetterDisplay?

DisplaysShouldNotBeTVs

waydabber This definitely works to a degree and seems to improve the screen in some ways, and it's awesome to finally see that classic "IPS black" shade of blue instead of blooming on totally black backgrounds!

But it doesn't look like this fully disables local dimming, because if you then turn Software Brightness really far down after applying this, if you move around a Dark Mode window you'll still see a trail of brighter LEDs following it and lagging slightly behind. And if you use overlay dimming instead of colortable dimming, there's also a dark halo around the mouse cursor against some backgrounds, which implies the mini-LEDs are still changing dynamically.

jordan

photon78s

The top link should be the one. The exact Amazon link is:
https://a.co/d/4Ia7bN1

Like I said it's a panel lottery between boe and LG. The pic attached you can see starts with L for the revision which is LG. starting with B is boe panel.

I think 144Hz might be native refresh rate ? Unsure if 165 is OC

Unsure on the QS.

https://www.rtings.com/monitor/reviews/hp/omen-27q says 10 bit but they must have HDR enabled which should be disabled of course.

« Previous Page Next Page »