Using Eink Monitors To Detect Temporal Dithering

jordan

Update - 11/01/2025 - Original post has been replaced with corrected information.

I initially reported that Windows builds 1809+ were using forced temporal dithering, but my testing methodology was flawed due to a software issue I wasn't aware of.

What Went Wrong:

The display mode I was using on the Boox Mira Pro has a bug that triggers the monitor to change its own dithering even when displaying static content, which shouldn't happen. This is why I thought it was an OS issue—e-ink displays are completely static and retain whatever image is shown until the video output changes (you can even unplug the monitor and it keeps the image).

The Real Issue:

The spatial dithering modes on e-ink displays are used to show gray shades. After more testing, one of the preset modes on the Boox Mira Pro is definitely bugged and causes movement/dithering on certain content, such as specific wallpapers. When I switch to video mode (which applies full-screen spatial dithering) or high quality mode (which uses zero dithering), I don't see any movement. In fact, when I enable GPU dithering, I can actually see temporal dithering activate in the non-bugged modes—it goes from completely static to chaotic movement/temporal dithering.

Video Demonstration:

I recorded a demonstration showing Windows 10 1809 with a GTX 1660 Super, where I use a color control app to disable and enable GPU dithering. You can clearly see the e-ink screen expose GPU temporal dithering when I turn it on using a non-bugged display mode: https://www.youtube.com/watch?v=Moq-LLhw3Nc

Corrected Findings:

After proper testing, I can confirm that the RX 6600 and 1660 Super do NOT use temporal dithering in Windows 1809-22H2 and Windows 11 25H2 while using a true 8-bit display. The Boox Mira Pro is recognized as a true 8-bit display by the system. What I was seeing was the same type of dither logic through each configuration, which turned out to be entirely due to that bugged display mode.

E-ink Testing Still Viable:

E-ink can still be used as a test tool for detecting temporal dithering, but you need to:

Test each display mode thoroughly
Ensure none show movement when content is static
Use high quality mode (zero dithering) for most reliable testing—it makes the screen flash with any movement (e-ink's slowest refresh mode), which actually works well for detection. As far as I know Dasung is the only Eink monitor brand without high quality mode.
Be aware that different displays may behave differently with 6-bit or FRC panels

I also tested with a Dasung 253 monitor and didn't see any bugged modes, so this appears to be a Boox software/firmware issue. The other modes on the Boox appear to work fine.

I apologize for the initial misinformation.

JTL

jordan I have heard people say Radeon cards dither on Linux unless you tweak settings, but that wasn’t true at all here — CachyOS handled it perfectly by default, zero Temporal dithering.

I have reason some of it might be a VBIOS control.

AshX

jordan

This is really cool (well, it would be cool if dithering didn’t cause such severe symptoms in people). Fascinating that 2018 seems to be the cutoff point because it lines up with a lot of stuff.

Makes me wonder if I downgrade the 2015 15” MBP (that is basically falling apart) from Catalina if it’ll be tolerable. I doubt it.

Still curious what changed after 2018 industry wide to suddenly push dithering on everything (is it P3 becoming more standard?).

WhisperingWind

jordan

Further research may be helpful, as in my experience Windows 10 22H2 does not appear to perform temporal dithering by default. I’m currently using a Sapphire Radeon RX 6600 Pulse, and testing was carried out with a Blackmagic UltraStudio Recorder 3G. Driver settings: RGB 4:4:4, 8-bit. Driver version 24.12.1 (Adrenalin Edition).

The testing methodology is straightforward: a Python script compares frame N with frame N+1 on a static image. If any frame shows pixel changes, it flags the presence of temporal dithering. The testing is done with a custom Python script, with an additional check described here.

I’m quite sensitive to temporal dithering, but I’ve been using this setup for almost 10 months with great comfort for my eyes.

I plan to try other driver versions in the future. However, my impression is that the issue is not the OS, but the hardware or a driver setting that may depend on the hardware used. But I don't yet know exactly what the issue might be.

jordan

JTL Interesting. The card I tested is the sapphire pulse variant.

JTL

jordan The specifics I haven't investigated in a while 😅

jordan

JTL I think there was even some stuff in intel vbios too that someone dug up.

AshX Agreed seems 2018 was definitely when this all started.
hmm it might be! I think people just want vibrant colors and no banding which of course costs us our health. soooo unnecessary. At least there should be a toggle to revert to basic video output with no flicker.

photon78s

There is the Boox Mira (non pro) which is 13in but probably out of stock too. I wonder if boox stopped listing them due to tariffs? I think it should work on other monitors too, I know dasung is seen as 8 bit too which I think is intentional so they are less likely to dither. You can play with different display modes and see how the dithering visibility changes. For example if I cap the boox to pure b&w most of it is hidden so if I need to tweak things ill enter that mode and then back to normal grayscale when I need to record it in action. I have a color mira pro too but havent tried that. I wonder if that can show any color fringing?(I think thats what its called)

I think its possible yes but I just dont want to look at the dithering long since it has bothered me on normal screens. I know the grayscale mode on eink uses its on spatial dithering which isnt triggering so if its dithering it should change how it looks when things are static. I also wonder if dithering would be easy to capture on lcds if you set the refresh rate really low? Doesnt it dither at half the refresh rate?

photon78s

So far the Boox Mira Pro is out of stock. Hope this technique works with other e-ink displays. In theory it should unless not recognized as 8 bit display, etc.

I think this is very helpful. However out of curiosity I asked an LLM (ignoring potential inaccuracies):

Can you use an e-ink monitor connected to a computer / gpu to visually detect different temporal dithering strategies?

Here is what it said:

Short answer: Yes — but with big caveats. An e-ink display can reveal some temporal-dithering behaviors if you control what frames the GPU sends and the temporal frequency is slow enough (or you deliberately slow it), but e-ink’s slow, firmware-driven updates, ghosting compensation and internal waveforms make it poor for detecting most modern high-frequency temporal dithering (FRC, subframe PWM, high-rate multiplexing). It’s much better for testing spatial dithering and static patterns than for catching millisecond-scale temporal schemes.

You can reliably use an e-ink monitor to detect or study:

Low-frequency temporal alternation (e.g., 0.2–2 Hz): if you force the host to alternate frames slowly, the e-ink will show visible successive states and you can see time-averaging effects (if you photograph with long exposure).

Time-averaged result of temporal dithering: because e-ink integrates slowly, the long-exposure photo or the human eye over seconds might show an average brightness — useful to validate that a sequence of alternating frames will average to a target gray.

Spatial dithering & spatio-temporal patterns: static spatial patterns (error-diffusion, blue-noise masks) remain visible without flicker; you can compare static vs temporally averaged results.

Host-side algorithms issuing frame sequences at low rates (e.g., software FRC using long T): if the dithering algorithm uses long periods (seconds or many 100s ms), e-ink will show the sequence.

Firmware update artifacts: e-ink’s own temporal compensation patterns (often visible as ghost frames or inversion artifacts) — useful if you’re trying to separate GPU temporal dithering vs display-internal sequences.

You will not reliably detect: 60 Hz / 120 Hz FRC, subframe binary multiplexing at kHz (DLP style), or other high-frequency schemes.

JTL

jordan I think there was even some stuff in intel vbios too that someone dug up.

In any case it wouldn't be the same but that also does ring a bell.

photon78s

jordan

I've read that pixel inversion can be perceived as half refresh rate.

https://display-corner.epfl.ch/index.php/LCD_dynamics

The "high-frequency schemes" get more complicated but hopefully less of a concern.

All this to say this approach is worth exploring more and also e-ink.

AshX

photon78s

I wonder if this is pixel walk or temporal dithering: https://streamable.com/tkohgw

photon78s

AshX

This is pixel inversion:

https://www.youtube.com/watch?v=4H6bETUx2Lc

Looks very different from your video.

Have you tested the screen using pixelinversion.com? I must warn that you may see visible flickering on these test as intended.

AshX

photon78s

Wow. This is very different. It is interesting that the subpixels also appear “chipped” inconsistently. The video doesn’t indicate what device is being filmed. My video is at 240 fps vs that video’s 120 fps.

My iPhone 5 also looks similar: https://streamable.com/5saleu

I will say that this form of flicker is much less intense than that of newer Macs. But it looks differently from the videos I’ve seen of temporal dithering on newer LCD devices.

This flickering is “faster” so am I to assume the frequency is higher or lower? Not quite sure how that works. Both videos I took are on gray colors.

Aquila

These are the MUST;
📌 Win 10 1809 (base or LTSC)
📌KSF (phosphor free) monitor
📌A safe GPU

In fact we have a OS dithering, (software) and hardware dithering (monitor and GPU).

A true 8bit panel sadly isn't the solution.
Many times a 6bit+frc or 8bit+frc are better than a true 8 bit panel with harsh blue or modern panel

Dithering is absolutely a problem as the PWM but the nature of the panel is probably more important.

markdotpeters5

Did you try this setting? I don't have the tools to measure dithering on desktop OS but have no issues at all with Windows 10 22H2 so would be surprised if it does with this setting enabled.

(not sure if you can see that pic, it says 'not available in your region' for me!)

jordan

markdotpeters5 I didnt see that option at all. which driver version are you on? I am going to load 22h2 again and try a couple other things so I can check for that again

markdotpeters5

jordan are you able to see the picture? It's in the Nvidia app.. I'm on driver version 581.57 (The Game Ready drive if that matters)

jordan

markdotpeters5 Yep! Thanks I will load that driver when I load 22h2 again and see if I can find that option.

Aquila someone actually posted last year about the studio driver helping their eyestrain. Ill test it on 22h2

Aquila

What's your opinion about Nvidia Studio drivers?

jordan

WhisperingWind But does the system see the ultrastudio capture device as a monitor with a proper monitor edid? My worry with capture cards is that they arent treated the same as a monitor. I can visibly see the temporal dithering on my rx6600 pulse on every build after 1809 even before installing the radeon driver. 1809 its self has zero TD upon my testing so for my setup is appears to be the OS. I hope more people can jump in with testing this.

markdotpeters5

jordan I just did a test using OBS to capture my desktop, then the Ffmpeg script to highlight the changes between frames as linked earlier..

As a test I played a video on one half of the screen and had the lagom.nl dithering test picture on the other half…

Absolutely no movement on anything but the video, so looks like my 22H2 isn't dithering.

WhisperingWind

jordan

But does the system see the ultrastudio capture device as a monitor with a proper monitor edid?

The recorder is recognized as a TV. But the “monitor” I use with Windows 10 22H2 is also a TV 🙂

I hope more people can jump in with testing this.

I hope so, too, since it’s still unclear what the issue might be. I don’t think Windows 10 handles TVs and monitors separately; however, since my only true 8‑bit display is the TV, I can’t confirm or refute that.