[PCM2]

It is and of course shouldn’t happen regardless of system. Just to add a datapoint, I didn’t have that issue on my system (RTX 3090, Windows 11).

[Weissrolf]

Acer X32 FP unit A vs. B.

[sblantipodi]

is there someone who asked Acer to release a new firmware that “remembers” the adaptive dimming setting in SDR and HDR without the needs to manually switch between adaptive dimming on and off every time we switch from HDR to SDR and vice versa?

PS: I tried AC Mirage on this monitor and now I’m blind. Mirage menu is pretty dark but sometimes when it loads contents or benchamark if flashes the screen completely white at max brightness.
Switching from dark to max brightness is crazy, I’m blind…

[Weissrolf]

Acer X32 FP, fullscreen black, local dimming black-level increase per Brightness setting. LED zones are only completely disabled in SDR mode below Brightness value 21 and never completely disabled in HDR mode.

SDR: (0-20 off), 21, 40, 59, 80, 100
HDR: (0-6 still on), 7, 16, 24, 32, 40, 48, 56, 64, 72, 80, 88, 97

Every time one of the above Brightness steps is reached LED brightness for full black increases by a step. Mixed content may use different steps, because the local dimming algorithm decides this. But overall this should mean that you want to use the highest brightness value before the next step to get higher dynamic range (brightness 0 uses the same black-level as brightness 20 while using higher levels for everything else).

[Weissrolf]

I received a third unit (C) of the X32 FP. Most things seem to be similar to the second unit (B) with these exceptions:

– On a fullscreen black image unit C show more backlight bleeding in the upper right and lower left corners, but a bit less light clouding in the other corners. With Windows 11’s dark theme (dark blue) backlight bleed is visible in the upper left corner of unit B and considerably more visible in the upper right corner of unit C.
– White-balance is more uniform across the screen on unit C, whereas unit B has more of a slight magenta tint on its right side. Overall unit B seems to be a tad warmer with a tendency for more magenta at same settings as unit C. Since the X32 FP seems to generally shift more towards magenta in HDR mode this is a welcome improvement on unit C.
– Unit B seems to be about 1/6 to 1/3 EV brighter than unit B at same brightness settings (about 5-10 steps of the brightness setting) in both SDR and HDR mode. This seems rather surprising, but ion parts may be attributed to the better color uniformity.
– I discovered a stuck (full brightness) blue pixel on unit B today, mostly visible with black/dark colors. If it was present before then I didn’t notice it, but I currently assume that it only just developed. Overall the panel of the X32 FP seems very sensible to touch, reacting even to very light pressure (also affecting backlight bleeding).

Vertical “scanlines” are present in both units, but still much more subtle than in my first unit A, especially in the center. Most people likely wouldn’t notice them at all. I suspect that Rtings’ “text clarity” score was affected by this, though, as the supposedly black text lines show red sub-pixels lighting up where they should be off.

[PCM2]

Thanks for keeping us updated with your findings and the consistencies and inconsistencies between units. 🙂

[Weissrolf]

I would have to send in unit B for repair of the joystick, so I thought I give another B-stock unit (C) a chance first. It‘s good to see that one being more similar to the better B unit than the worse A unit.

But backlight bleeding seems to be quite over the place with these panels. I also noticed that they are very sensible to even slight touch/pressure. So different pressure points when the panel was built into the chassis might play a role with backlight variance.

[sblantipodi]

I bought 7 units before getting a nearly perfect one.
But what are the scanlines you are talking about? I never seen it on my old units.

Can you take a picture of them?

[PCM2]

You can’t accurately capture them with a camera and not everyone will notice them. But they’re covered here.

[Weissrolf]

I suspect that Rtings might have captured them in their text rating of the X32. Their photo shows red subpixels lighting up in the supposedly black lines of text.

Often these vertical lines are only temporarily visible, after changing on-screen content. They happen more often in HDR mode (maybe heat related) and at lower refresh rates.

[PCM2]

I’m not really sure that’s what they captured, but whatever it is it’s odd that it only appears for some vertical lettering elements on the Windows 10 screenshot. I wonder if they took multiple photos and observed this same thing? A better representation of how they would appear to the eye is found in our review of the Gigabyte AORUS FI27Q-X. These could be captured quite consistently (albeit with some camera moiré – the wavy lines) at 60Hz on that model due to the pixel density and nature of the dynamic interlace patterns, but on the X32 FP it wasn’t possible. To the eye the effect is quite comparable.

[Weissrolf]

At this point I am pretty sure that the vertical lines that sometimes appear on the Acer X32 FP are FRC related. This is why they get worse at lower refresh-rates. Additionally FRC seems to employ averaging over multiple consecutive frames, which is why the lines vanish after a short time with static images. Both can be made visible by using very low refresh-rates and even more so by making the monitor use extreme overdrive for its FRC (WARNING: extreme flickering!).

Use the following test for everything below:

https://arnowelzel.de/tools/monitor-test

FRC artifacts (lines and others) without flickering:
Disable DSC and HDR to enable custom timing configurations in your GPU driver. Create a custom timing of 30 Hz. At least with NVidia this custom timing can then be used by both SDR and HDR as long as DSC is disabled. Overdrive settings do not seem to influence FRC this way.

Using HDR increases the effects described below, bit depth doesn’t seem to make a difference.

FRC artifacts *with* flickering: WARNING, extreme flickering!!!!
Do the same as above, but create a custom timing between 7 to 28 Hz, the lower the better (aka 7 is best). This will cause the screen to display a moving popup, telling you that the input is invalid and disabling the OSD (other than input select). BUT it will still display an image, but use *extreme* overdrive for *both* content *and* FRC. As a result the whole image will flicker, usually the darker the color the more it will flicker.

FRC/image averaging over multiple frames:
This can be made visible even at 160 Hz by switching between fullscreen purple and the pixel-shift (black/white pixels) image right before purple (press cursor left/right to switch back/forth). If you look close you will see a short flash of green before the image settles to gray. The lower the refresh-rate the more this becomes visible, at some point you will see the bottom of the screen going rather red than green and you can also observe the same when you switch from fullscreen blue to the pixel-shift pattern then.

You may notice pronounced clouding, especially with darkest gray. This seems to be LCD pixel based rather than LED lighting based and happens on my third unit, but not on the second unit. Curiously the 2nd unit has a generally *less* uniform color and brightness balance over its screen area (while being brighter/more contrasty), despite being free of clouding in this test. This may also be observable with the + symbols test of the test suite without producing flickering (viewing angles and contrast of the + symbols being more unevenly distributed).

Sometimes the averaging seems to get stuck with its buffering, which then leads to image retention over several seconds (like half a minute or even longer). Personally I only observed this clearly in these special test-conditions, but at least one user on Amazon posted pictures of strong image retention with his unit.

I also had a few cases where moving a window from one screen to the other resulted in the window being shown on both screens (aka being stuck on the screen it was moved away from). It’s not clear, though, if this is the monitor getting stuck with the image in its buffer or rather a graphic-card/driver error.

You can trick custom timings for both HDR and DSC modes by creating them in non DSC SDR mode. In HDR mode they stay accessible via both Windows display prefs and NVcontol panel, in DSC mode they stay available in Windows display prefs only.

And for whatever reason the combination of my 2070S and Acer X32 FP doesn’t dropout using DSC + VRR at 120 Hz refresh-rate. So that works better than trying to get the stupid NVidia driver to run 120 Hz without DSC (121 works, 119 works, 120 only until the next switch off/on).

PS: running the monitor at 6 Hz results in a black image and running it at 5 Hz results in overdrive being disabled. Less than 5 Hz is not possible.

[PCM2]

In theory at least, the monitor should only be employing FRC to achieve a 10-bit signal. Using Nvidia graphics driver defaults should disable the FRC stage. I say in theory, because some FRC behaviour may still remain. I observed these interlace pattern artifacts for 8-bit content and with the monitor fed an 8-bit signal all the same.

[Weissrolf]

In 2 decades of using LCD displays only one of my monitors (Viewsonic VP920) used true 8-bit output without any kind of FRC. Everything else employs FRC constantly to allow OSD contrast, color temp, gamma and all the other settings without throwing away colors/gradations, regardless of whether the GPU output is 8-bit or 10-bit.

You can see this when you display an image of 256 gradations (8 bit) of gray and then lower contrast from its default 50 towards 0. Instead of crushing shadow tones to pure black all 256 tones get gradually darker.

[PCM2]

That makes sense and I did suspect as much, that there’s almost always a ‘reserved’ FRC stage regardless of whether it’s used to enhance the signal bit depth. 🙂

[Weissrolf]

I am pretty sure that contrast on unit 3 is a bit lower than on unit 2, so not just brightness. The joystick is a bit spongy downwards, likely the first direction to fail.

The images below show flickering FRC clouding at 7 Hz. Disregard the white-point differences, but unit 3 (below) clearly has some clouding problems there where unit 2 (above) does not.

[sblantipodi]

I’m using the monitor at 60Hz and 120Hz, and never seen scan lines on my unit…

The most annoying thing of this monitor is the bad overdrive…
It is really annoying even when surfing the internet on this page for example.

If I scroll this page I can clearly see the text overshoot while scrolling, the text has a red helo during the scroll…

Another problem I have is that on Linux I can’t use 160Hz/144Hz, after a while on the high refresh rate it starts flickering…
this does not happen on windows…

I’m pretty bothered by the text overshoot and I’m looking forward for an upgrade…
I don’t like OLEDs as I don’t consider them “general purpose monitors” and I don’t know what to buy now…

[PCM2]

Why not set overdrive to ‘Off’ and see if you can get used to it like that? It’s natively a pretty fast panel even without overdrive, you shouldn’t feel too hard done by if you prefer just using the ‘Off’ setting. I agree a setting between ‘Off’ and ‘Normal’ would’ve been nice but it isn’t an option and you may find it easier to adapt to the slower pixel responses (which are still a reasonable speed) than to the overshoot which clearly bothers you for some transitions.

[sblantipodi]

thanks for the suggestion PCM2 and sorry for the late reply.

I don’t have the PG32UQX anymore…
Do you consider X32FP as slow as PG32UQX when overdrive is off?

just a curiosity for comparison…

[PCM2]

See for yourself from the pursuit photos. It’s pretty clear the X32 FP with overdrive ‘Off’ (left) is faster than the PG32UQX with ‘OD = Normal’ (right). Or more to the point, the ASUS has distinct weaknesses for the dark background in particular which the Acer doesn’t. And it’s even worse for some transitions not shown here, it simply has some clear stand out weaknesses that the Acer doesn’t.

[Author]

Posts