Does BetterDisplay alter fixed XDR reference modes?

[gregbenz]

When using Apple's XDR presets for fixed reference modes (such as HDR video P3-ST 2084), the display should not be altered. It is meant to target very specific EOTFs and white points. However, I see that you can dim the display even when in such a mode. This seems like a bug and should at a minimum carry a clear user warning that you are reducing accuracy in a mode designed to be as accurate as possible.

On a likely related note, I am finding that my M1 MBP measures much lower than expected (79 nits vs 100 nits target via CR-250 spectro) even when BetterDisplay has been raised to max and then closed. It is possible there is a fault in my MBP, but it seems more likely that BetterDisplay is causing unwanted results. How can I ensure that BetterDisplay is not in any way altering Apple display behavior? Is there some uninstall process which might be used to wipe any LUTs or anything like that potentially used by BetterDisplay?

[waydabber]

The app does not alter the reference mode Presets themselves, it only adjusts the color table, when the user explicitly dims the display or uses Image Adjustments. You can of course disable these in the app, but there are times when the user uses fixed reference modes but temporarily wants to dim the display for any reason (obviously not when color sensitive work is being done). When the brightness is set to neutral (100%) and Image Adjustments are all neutral, the app will leave the color table untouched (meaning: it actively restores the default color table to make sure there is no adjustments to it whatsoever). In HDR mode using any adjustments will cause the EDR range (HDR range beyond peak SDR luminance) to be clipped, as Apple's color table SDK works in the SDR range only - so it is important to have the default color table applied (with no image adjustments or software dimming applied).

[waydabber]

Turn this OFF to make sure that the app does not alter the color table or use any soft of overlay, metal or compositor trickery to alter the output and let only the Preset (and native brightness control - if enabled by the preset) handle things:

[gregbenz]

@waydabber Thank you for the guidance and feedback on the intent here. I turned that "enable software-based video adjustments" setting off and see the same behavior as when I quit the app. So I then uninstalled BetterDisplay and searched for any traces of the application (per ChatGPT guidance for things like launch daemons, etc) and rebooted in safe mode. I still see the display is about 80% of expected brightness when in multiple different XDR reference modes.

I'm going to wipe the machine and reinstall Tahoe for my own education as I'd like to see if this is really a hardware fault. I can't think of a better explanation at this point, but it seems odd that the display would show very stable readings and good white point, while being roughly 20% below target values (100 nits shows around 78 measured, peak 1600 nits shows around 1290 measured in a ~10% window, etc). Is that consistent with any hardware / backlight / firmware fault you have ever encountered with an XDR display? (this is the 16" M1 MBP - so roughly 5 years old, seems in good shape but I only bought from the original owner earlier this year).

[waydabber]

Some level of LED wear is normal. 20% seems somewhat excessive at a first glance but not at all impossible for a 5 year old display if it was used a lot at high brightness (demo units running at high brightness continuously can easily get down to 80% in under 2 years). These mini-LEDs (especially first gen) were really pioneers at the time, they get rather hot when driven at peak performance and also draw a lot of power (that's why Apple normally limited them at 500 nits, mainly to limit battery impact - newer MBPs with smaller, more efficient mini-LEDs can get to 1000nits during normal use without any extra tweaking depending on lighting conditions). Heat causes extra luminance deprecation and aging. Most users don't actually notice this as one can compensate easily + a peak brightness difference between 1600 and 1300 nits doesn't look very dramatic actually. Of course if you want to use the display as a reference monitor, this can be an issue.

[gregbenz]

@waydabber Thanks, I appreciate it!

I have fully wiped the drive and reinstalled MacOS. I continue to see values ~20% below target, so this seems to confirm BetterDisplay is unrelated to my observed results. I bought this computer used earlier this year, so I cannot account for its history. The seller said it was very little used and the battery shows 88% capacity / normal, which would seem to back up what he told me. Perhaps it was stored by a vent, server (he's an IT guy), or some thermal concern. Or maybe just aging or some kind of fluke.

I didn't notice the issue until I put a spectro on it (though I do my photographer editing on other display and just use this one in variable brightness mode, so nearly impossible to tell). Very few people would notice that anything is off here, you really have to be using a reference mode and mastering content for it to matter at all. With Apple's "fine tune" calibration, I can close the gap a bit too (86 nits vs 100 target).

Even with this result, most photographers would probably not notice anything was off - and it still easily outperforms for EOTF tracking compared to most OLED displays (which have significant ABL - other than Apple's tandem stack XDR). I could potentially even create a custom reference mode with boosted SDR / HDR targets and end up with something closely approximating the HDR reference mode as the values I've tested seem to show a consistent % gap to target, the measurements are very consistent, and white balance is still great. Most of my editing uses the variable brightness mode and I'd probably just push it up a click and call it a day if this were my primary machine.

I did not see any error from MacOS "full" calibration, which surprises me here. I'm going to run it again and see if it outputs any kind of validation data. This would be rather obvious in the deltaE value if running CalMAN.

[radambe]

I have noticed that Apple's "calibrated" display modes are very rarely accurate out of the box - thus I'm sure the reason, at least in part, they added the option to measure and apply custom trim values for brightness and whitepoint to the display modes. You should use the "Fine Tune Calibration" option in System Settings --> Displays to measure the brightness and white point and enter in those values in. This will trim the calibrated display modes to measure more accurately.