OLED yellow vertical edge artifacts

raymod2

n00b
Joined
Jan 21, 2023
Messages
5
I own the following monitors which have similar dot pitch (109 ppi vs 105 ppi) so the optimal viewing distance is roughly the same. I typically view them from around 30 inches away.

Asus ROG Swift PG278Q (27" 2560x1440 LCD)
Asus ROG Swift PG42UQ (42" 3840x2160 OLED)

The OLED monitor displays an annoying visual artifact that I noticed right away. It is most obvious when a yellow box is displayed on a white background (for example when highlighting text). There is a bright red vertical line on the left side of the box and a bright green vertical line on the right side. The graphic below demonstrates the problem:

oled_vs_lcd.png


The vertical lines are clearly visible in the photograph of the OLED on the left. They are not present in the photograph of the LCD on the right. The subpixel layouts of the monitors can also be observed in these photographs (WBGR for the OLED monitor and RGB for the LCD monitor). You can see that the WBGR layout results in vertical misalignment between the white pixels and the yellow pixels. Based on that observation I tried another test and displayed a white/yellow/white vertical line on a black background:

vertical_lines.png


The line of the left is one pixel wide and the line on the right is two pixels wide. Here the artifact is even more obvious. The center of the line is too wide AND it's the wrong color. Due to the geometry of the WBGR subpixel layout there is no perfect way to display these images. If the yellow pixels are displayed correctly then the center of the line will be slightly offset from the top and the bottom. Nevertheless, it seems to me that would be far less noticeable than the artifacts that are caused by attempts to align the pixels.

Can someone else with an OLED monitor try to reproduce these results? I'm curious if other Asus PG42UQ monitors have this problem (I am running firmware V033). I am even more curious if the LG C2 has this problem. Just draw a yellow box on a white background in Windows Paint and look for the red and green vertical edges.
 

criccio

Fully Equipped
Joined
Mar 26, 2008
Messages
13,979
I see this on my 42" C2.

Thankfully if I wasn't looking for it I wouldn't notice and I certainly don't notice in normal use.

buEBvPDShUo_CEme7lxUKCyA=w1918-h1439-no?authuser=0.jpg

Ap_7zEo3y9Hr3Bo22FmfE31w=w1918-h1439-no?authuser=0.jpg
 

Zepher

[H]ipster Replacement
Joined
Sep 29, 2001
Messages
20,316
I don’t see any red or green lines that I can notice on my Alienware.
E8CB7DC2-B79F-4AD2-9565-426CDB7DD355.jpeg


A778F54D-B23A-46A6-A5AB-8C1386DD8D50.jpeg
 

vick1000

2[H]4U
Joined
Sep 15, 2007
Messages
2,344
WOLED pixel layout, Big white subpixel. QDOLED pixel layout, triangle. Not good for text and the like.
 

raymod2

n00b
Joined
Jan 21, 2023
Messages
5
Thanks for the photos. It looks like the LG C2 exhibits the same problem as the Asus PG42UQ (both using the same WBGR panel). This implies they are using the same (flawed?) algorithm to map pixels to WBGR subpixels. criccio: Another place where I see this regularly is in the Windows File Explorer. The folder icons are yellow and they have a red line along the left edge. It's easy to assume that this was intended (maybe to give the icons sharper edges) but you don't see this on other monitors.

The Alienware AW3423DW doesn't have this problem as the photo above demonstrates. I suppose that is expected because that monitor only has 3 subpixels (it omits the white subpixel) so mapping pixels to subpixels is straightforward. I'm a little confused about the link that Errok23 shared. It claims there is color fringing on that monitor due to the triangular layout of the subpixels. But that shouldn't matter. As long as the subpixels are close enough together their combined output should appear as one color. That is evident in my photos above where the red and green subpixels appear as one yellow pixel when they are both illuminated at the same time.

I don't necessarily agree with vick1000 that the subpixel layout of these monitors makes them bad for displaying fine details such as text. Subpixel rendering (such as ClearType) is a mess with the WBGR layout but that could theoretically be fixed in software. The color fringing could theoretically be fixed in the monitor firmware.
 

Lifelite

Supreme [H]ardness
Joined
Aug 26, 2015
Messages
6,299
hmm I have the AW3423DW and I can't notice it at all.

Hard to get a picture with all the phone camera BS that filters light and such:

HptvsPH5QY_le302JSvCkEFQ=w1002-h1335-no?authuser=0.jpg


heFIK6tgOoUi4lGgleho3gXA=w1002-h1335-no?authuser=0.jpg

kGqGtv-jXhgsfxVCROI9ML7Q=w1002-h1335-no?authuser=0.jpg


Looks like a bunch of pixels but zooming out looks rather normal lol

1674601028359.png
 

raymod2

n00b
Joined
Jan 21, 2023
Messages
5
Lifelite: What are you showing us? Some of the photos look like vertical edges between white and yellow. I agree there are no artifacts there on the AW3423DW (as Zephyr showed in his photo). Your photo with the text looks like white on a dark background. What does it look like with white on a *black* background?
 

Lifelite

Supreme [H]ardness
Joined
Aug 26, 2015
Messages
6,299
Lifelite: What are you showing us? Some of the photos look like vertical edges between white and yellow. I agree there are no artifacts there on the AW3423DW (as Zephyr showed in his photo). Your photo with the text looks like white on a dark background. What does it look like with white on a *black* background?
Oh sorry never described. The text was a forum post here, text with black background. For white and black, I just cannot get my camera to properly focus because of the insane contrast lol


dMExWcNFVXgqgxNxDzzIJ3Ow=w1003-h1337-no?authuser=0.jpg
 

deaedius

[H]ard|Gawd
Joined
Jun 18, 2014
Messages
1,127

Chief Blur Buster

Owner of BlurBusters
Joined
Aug 18, 2017
Messages
314
While it's been already publicly known for several weeks that I'm working on OLEDs, I can't say much yet (NDA), but monitor manufacturers are currently working to improve this further. For the specific red/green fringing, it's not because of white subpixel, but by other considerations.

However, many lurkers, myself included, are appreciating the photography occuring in this thread, and welcome people to post as many photographs as possible from as many OLEDs as possible.

Subpixel Photography Tip:

To get clearer smartphone photos with better focus:
  1. Use maximum smartphone zoom (pinchzoom all the way in).
  2. Move phone slowly FURTHER from screen (~3 to 10" depending on phone).
  3. Take a photo when photo preview when becomes darkest/moire-artifacted (this means focus became best).
Post photo without downscaling (keep original resolution).

This macro zoom trick from a distance produces better subpixel photography for most smartphones. Try approximately 8 inch distance, plus maximum zoom, plus tap-n-hold to automatically lock focus. You know that you focussed perfectly when the photograph suddenly becomes dark (showing black gaps between pixels), rather than a bright photograph (blurry photo). The temptation is to try to choose the brightest photos, but those are the out-of-focus ones. The dark photos are the ones that focussed so well that black-color between pixels are fully focussed. So zoom in to maximum macro, then physically move your smartphone slowly further away from the monitor to adjust smartphone distance -- until you suddenly see darkness/moire artifacts in the photo preview (telltell hint you nailed the subpixel focus) and then press the shutter button.

3rd party apps (DLSRcamera or ProCam) is useful in that you can use manual focus as slider, since autofocus sometimes has problems in this situation. But most phones should eventually usably focus once phone is sufficiently far. You ideally want to slightly underexpose the photo too, because overexposed photos can make the fringing less visible in the photo than in real life.

EDIT: You can follow the above steps to focus on bright objects, tap-and-hold on the smartphone screen to lock the smartphone focus. That temporally disables smartphone autofocus on both iPhone and Android. Then you can move your smartphone to photograph a darker part of the screen without this focus-issue / overexposure-issue.

The vertical lines are clearly visible in the photograph of the OLED on the left. They are not present in the photograph of the LCD on the right. The subpixel layouts of the monitors can also be observed in these photographs (WBGR for the OLED monitor and RGB for the LCD monitor). You can see that the WBGR layout results in vertical misalignment between the white pixels and the yellow pixels.
Correction: It's not WBGR. It's RWBG on all LG panel with that fringing.

Samsung is not immune to fringing either, it's just different colors and different fringing direction that has the problem.
- Samsung has horizontal fringing (e.g. green fringes at top or bottom edges of objects) on certain colors OTHER than yellows and whites
- LG has vertical fringing between yellows and whites (e.g. this thread).

Some people are more bothered by Samsung's fringing effect for their specific video gaming colors, and others are more bothered by LG's fringing effect for their other specific video gaming colors. The 100% out-of-focus Samsung photos does not show this very well, especially since Samsung fringing doesn't occur with yellow-and-white. So not apples-vs-apples anyway.

Many sources including RTINGS confirms RWBG pixel groups per pixel. There's already a public peer reviewed research paper describing why they chose RWBG pixel structure way back in the television era.

1674689863314.png


I have more to say, when the information is no longer embargoed. Keep tuned.

Just only commenting/correcting publicly known info with other publicly known info.
 
Last edited:

pendragon1

Extremely [H]
Joined
Oct 7, 2000
Messages
43,710
isnt this just the nature of the new pixel layouts? i see then on my hisense tv sometimes but just carry on...
 

Chief Blur Buster

Owner of BlurBusters
Joined
Aug 18, 2017
Messages
314
isnt this just the nature of the new pixel layouts? i see then on my hisense tv sometimes but just carry on...
Some of us are massively more bothered by it, especially with desktop use at close viewing distances, unlike televisions at further view distances.

However, there's a fix for your TV. People can fix ClearType on any BGR panel of any technology (including HiSense) via downloading Better Cleartype Tuner and switching ClearType to BGR instead of RGB.

1674693797492.png


The software is great for all televisions including odd OLED and LCD pixel layouts, to make text look better.

Then the HiSense television text became sharper, and looked closer to a computer monitor, much better text more comfortable at close viewing distances (like using a television as a desktop computer monitor, which I also do too sometimes). This is because many HiSense LCDs use BGR pixel layout, and ClearType has a BGR mode.

RTINGS HiSense U7G television at www.rtings.com/tv/reviews/hisense/u7g

1674693670179.png


For television-targeted panels, there are some reasons why some things are done this way -- like relocating television electronics to the bottom edge of the screen, or adjusting a pixel structure for better pixel wear balance between color channels (e.g. mitigate OLED pixel wear and tear more equally on all colors). Some LCD panels are fabricated in a way that makes it easier to design a television around BGR instead of RGB, since panel vendors and television vendors are often left-hand, right-hand. Netflix users don't really worry about computer text rendering as much, so there's been more freedom to go away from RGB on televisions.

There's lots of public papers that talk all about these pixel structure oddities for plain old fashioned video use, but it is quite true that many television manufacturers aren't really testing for computer text clarity, industry-wide. RTINGS talk about this quite a bit.

BTW, we were once using CRTs, where subpixels had no correlation to pixels...
___

Anyway, to get back to OLED color fringing topic, Better ClearType Tuner can help reduce fringing with computer text quite a bit on any OLED panel ever released. So definitely download it for your OLED to improve your text. However, this does not affect color fringing on things like yellow objects on black or white backgrounds. (Continued photography is welcome on all models and brands of OLED, I'd love to see more.)
 
Last edited:

pendragon1

Extremely [H]
Joined
Oct 7, 2000
Messages
43,710
I fixed ClearType on a HiSense TV via downloading Better Cleartype Tuner and switching ClearType to BGR instead of RGB.

Then it looked perfectly like a computer monitor.

This is because many HiSense LCDs use BGR pixel layout, and ClearType has a BGR mode.]

RTINGS HiSense U7G television at www.rtings.com/tv/reviews/hisense/u7g

View attachment 544370
huh, ill take a look at that. i have the 55" U78G, which is think is just the canuckistani version. one of the last firmware updates also added a "hdmi input optimization" option with picture and text modes. havent tried it though...
 

Chief Blur Buster

Owner of BlurBusters
Joined
Aug 18, 2017
Messages
314
Oh sorry never described. The text was a forum post here, text with black background. For white and black, I just cannot get my camera to properly focus because of the insane contrast lol

1674694568915.png


Even in this blurry photo, you can see the publicly known Samsung QD-OLED horizontal fringing artifact. So it's not just LG.
  • GREEN fringe at top of WHITE objects on dark backgrounds.
  • MAGENTA fringe at bottom of WHITE objects on dark backgrounds.
Just that you have to test different color pairs when testing on LG, versus testing on Samsung. And fringing direction is different, since Samsung is horizontal fringing and LG is vertical fringing. There are other color pairs that have fringing too, but these cherrypicked color pairs (LG white/yellow and Samsung black/white) is just among the more noticeable, especially for pixel-sharp edges more common with computer use than with video.

This is less visible at television viewing distances (and television content), but more visible at desktop viewing distances (and computer content). So make sure you test apples-vs-apples angular pixel densities. Also the addition factor that human vision varies a lot, some of us blatantly see the color fringing and are annoyed by it, while others just don't care at all. RTINGS and other reviewers talk about OLED fringing already, so this is all public information already, which I'm just parrotting here from my POV.

Camera Tip Addendum for Samsung Black-and-White: Follow the camera instructions in my earlier post, but focus on a bright part of the screen. THEN tap-n-hold to lock the focus (works on Android and iPhone), and THEN move the camera to the dark part of your screen. The pixels will stay sharp and avoid becoming overexposed.

Bottom line: Let's be fair; neither Samsung nor LG is immune from color fringing in computer use.

Keep taking more photographs; More the merrier. Few of us, myself included, have concurrent access to all the dozens of OLED televisions and monitors ever made -- so this sharing comparison photographs of OLED is quite useful.
 
Last edited:

raymod2

n00b
Joined
Jan 21, 2023
Messages
5
@Chief Blur Buster: At first I thought you were being pedantic and the names WBGR and RWBG were equivalent. Now that I think about it the distinction is important and you are correct. There is no algorithm mapping pixels to subpixels that attempts to prioritize vertical alignment. They are simply lighting up the appropriate subpixels in each RWBG group. The reason yellow is such a problem is because it is comprised of red and green and those subpixels are so far apart that they don't blend into one color. My 1-pixel vertical line uses R__G and my 2-pixel vertical line uses R__GR__G. The yellow "pixel" in the center is actually comprised of subpixels from two neighboring pixels:

797652_vertical_lines.png


The photograph below uses all the primary colors (red, green, blue) and all the secondary colors (yellow, cyan, magenta). The primary colors don't need any blending but they exhibit black vertical lines due to uneven gaps between pixels:

The red box has a 4-subpixel gap on the right (R____W__).
The green box has a 5-subpixel gap on the left (_W_____G).
The blue box has a 4-subpixel gap on the left (_W____B_).

The secondary colors blend differently depending on the proximity of the required subpixels:

Yellow blends the worst because the subpixels are very far apart (R__G).
Cyan blends the best because the subpixels are right next to each other (__BG).
Magenta has moderate blending (R_B_).

1674704973154.png
 

Chief Blur Buster

Owner of BlurBusters
Joined
Aug 18, 2017
Messages
314
@Chief Blur Buster: At first I thought you were being pedantic
Definitely not being pedantic.
Dryly, this is deadly important scholarly research.
Onwards:

and the names WBGR and RWBG were equivalent. Now that I think about it the distinction is important and you are correct. There is no algorithm mapping pixels to subpixels that attempts to prioritize vertical alignment. They are simply lighting up the appropriate subpixels in each RWBG group. The reason yellow is such a problem is because it is comprised of red and green and those subpixels are so far apart that they don't blend into one color. My 1-pixel vertical line uses R__G and my 2-pixel vertical line uses R__GR__G. The yellow "pixel" in the center is actually comprised of subpixels from two neighboring pixels:

797652_vertical_lines.png


The photograph below uses all the primary colors (red, green, blue) and all the secondary colors (yellow, cyan, magenta). The primary colors don't need any blending but they exhibit black vertical lines due to uneven gaps between pixels:

The red box has a 4-subpixel gap on the right (R____W__).
The green box has a 5-subpixel gap on the left (_W_____G).
The blue box has a 4-subpixel gap on the left (_W____B_).

The secondary colors blend differently depending on the proximity of the required subpixels:

Yellow blends the worst because the subpixels are very far apart (R__G).
Cyan blends the best because the subpixels are right next to each other (__BG).
Magenta has moderate blending (R_B_).

View attachment 544401

Your post is very correct.

I knew about all this already too; It is correct R and G is spatially far apart in RWBG, unlike for RGB and BGR.

Conversely in fair defense of LG -- this line of thinking allows you to easily explain why Samsung QD-OLED has the green/magenta fringing behavior. See this image from RTINGS is very easily explained in the very exact same way:

pixels-small.jpg

(individual subpixel shapes can vary, the Alienware OLED is slightly different but still triangular)

The green subpixels at top of a "pixel", the red/blue subpixels at bottom of a "pixel" creates the green/magenta fringing. White objects with sharp edges (computer graphics) on black screens. You have a green fringe at top edges of white objects on black backgrounds. And purple fringe at bottom edges of white objects on black. Because those are the last subpixels illuminated against a totally black background.

Regardless of LG or Samsung, the choice of pixel structure can create side effects. Even plain LCD RGB can have color fringing, just less visible. Like the red left edge / blue right edge white boxes on black backgrounds LCDs. So even RGB and BGR is not immune. But much less so than more unconventional pixel structures that increase relative spatial distance between bright subpixels (bright color primaries).

At higher pixel densities and further view distances, this all can become a nonissue, but low pixel densities at close viewing distances, this can be annoying to some, without mitigations/tweaks.

____

...So explanation needed for these subpixel oddities that leads to color fringing...

Why these odd pixel structures on all OLED computer monitors on the market?

TL;DR: One of the many reasons is to make them burn-in resistant.

From the publicly available papers, e.g. several by SID.org (Journal of Society for Information Display), etc. -- there's been a problem with direct-spectrum RGB OLEDs because they have far faster burn-in -- e.g. some of the RGB OLED smartphones have burn in far earlier than recent LG OLED TVs for example (e.g. LG C1+). Samsung and LG went different directions on all their mitigation measures in longevity per nit, with all the interacting features (economics, defects, etc), whether by using fat subpixels (Samsung) or by using brightness-enhancing white subpixels (LG).

By writing this, I'm just really summarizing key points of papers by writing this explainer, the publicly released papers explain various reasons, including W pixels to increase brightness without premature wear on overdriving R, G, B. Plus, additionally, the unification of primary light (e.g. blue light or white light) and generating the primary colors via quantum dots (Samsung route) or color filters (LG route) to create the different OLED colors.

For LG WOLED, some of the publicly published science papers are literally roundabout ways of saying W subpixel also doubles as a preventative measure to prevent overdriving R/G/B to burn-in-risky voltages and brightnesses. Common sense compromise. So currently, the television and monitor world's obviously currently gone away from direct R-emissive, G-emissive, B-emissive OLED in order to make OLED last long enough without permanent burn in. So that you're using the same light spectrum for all primary colors, to allow equallized OLED-color pixel wear. It's a fascinating read on all the published science papers -- but the short story of one major purpose of the really odd shapes of pixel structures on all the various vendor OLEDs: Make OLEDs last longer without permanent burn in.

(...There's other reasons, such as panel manufacturing defect rate, but panel longevity per nit is the biggie...)

The current switch to these odd pixel structures were engineering considerations that also allowed OLEDs to last long enough to be used for computer use. There's the usual temporary image retention stuff that fades (phosphorescent style effect), but the RTINGS-scare-story of permanent burnin is a nonissue for me post-2021 on the OLED currently sitting on my desk. I just merrily Visual Studio away on my OLED for hours, static elements included, Dark Mode, default out-of-box 75% brightness, orbiting enabled, Pixel Refresh overnightly...

And the trailblazers of "LG OLED TV using them like a computer on a desk" paved the way to show that the newer OLEDs (e.g. C9, C1, C2) didn't burn in as fast as the early C6 or C7 -- IIRC, that's the first OLEDs that RTINGS burnin tested and sounded the burn-in alarm for. Kudos that these current post 2020s OLEDs now last much longer than a few years ago. Remember, even LCD backlights can slowly dim over tens of thousands of hours of wear and tear -- my Benq XL2411Z is down to 100-150 nits due to wear on its LED backlight. Or that LCD becomes very splotchy/dirty looking (see 10-year-old airport arrival/departure LCD screens that's been on 24/7). While OLED can still have burn in risk if too bright, the current modern OLED lifetime can outlast plasma displays, thankfully.

Current all-white-primary-source OLED (four white subpixels with three color filters by LG) or all-blue-primary-source OLED (QD wavelength-converting phosphor for Samsung) combined with recent materials and formulations can now outlast a lot of historic CRT/plasma phosphors.

If you research it out, you've already noticed users are happier with the longevity of newer OLED technologies. Here. I'll cherrypick one -- YouTube: 2-year no-burnin computer test of LG C9 and also read the YouTube comments too. You can see the YouTube comments that are saying they've used their LG OLED for years without permanent burnin. You get the gist. They've really improved OLED lifetimes since the early days.

And yes, I'm casually using OLED like a regular LCD pretending to be a burnin-unaware user -- semi-intentionally trying and failing to permanently burn in my OLED -- for science's sake :D

We should all finally welcome OLED overlords to the computer desk.

Back to OLED photography, would love to see more!
 
Last edited:

kasakka

2[H]4U
Joined
Aug 25, 2008
Messages
2,899
I remember noticing the yellow background fringing issue on my LG CX 48" when I got it but then I just shrugged my shoulders and didn't think of it more than that. I think these subpixel problems will pop up so rarely that they wouldn't be a dealbreaker for anyone compared to the text rendering related problems.

As for burn in, that CX 48" at 2.5 years is still going strong, even though it has seen 2 whole years of 100% working from home ~8h a day + personal use. I don't know if it's gotten dimmer in the process but I really can't tell.

For me the disappointing part about OLEDs is that there has been very little advance in these smaller 4K sizes, barely any progress over the CX besides the 42" size. This year the larger TVs from both LG and Samsung got brighter, but it might take a few more years for that to trickle down to smaller sizes. Meanwhile I have no interest going back to 1440p or its UW equivalents. So for my next monitor I'm looking at mini-LED models instead.

PS. Really appreciate your posts Chief Blur Buster . Always educational!
 
Last edited:

Lifelite

Supreme [H]ardness
Joined
Aug 26, 2015
Messages
6,299
View attachment 544374

Even in this blurry photo, you can see the publicly known Samsung QD-OLED horizontal fringing artifact. So it's not just LG.
  • GREEN fringe at top of WHITE objects on dark backgrounds.
  • MAGENTA fringe at bottom of WHITE objects on dark backgrounds.
Just that you have to test different color pairs when testing on LG, versus testing on Samsung. And fringing direction is different, since Samsung is horizontal fringing and LG is vertical fringing. There are other color pairs that have fringing too, but these cherrypicked color pairs (LG white/yellow and Samsung black/white) is just among the more noticeable, especially for pixel-sharp edges more common with computer use than with video.

This is less visible at television viewing distances (and television content), but more visible at desktop viewing distances (and computer content). So make sure you test apples-vs-apples angular pixel densities. Also the addition factor that human vision varies a lot, some of us blatantly see the color fringing and are annoyed by it, while others just don't care at all. RTINGS and other reviewers talk about OLED fringing already, so this is all public information already, which I'm just parrotting here from my POV.

Camera Tip Addendum for Samsung Black-and-White: Follow the camera instructions in my earlier post, but focus on a bright part of the screen. THEN tap-n-hold to lock the focus (works on Android and iPhone), and THEN move the camera to the dark part of your screen. The pixels will stay sharp and avoid becoming overexposed.

Bottom line: Let's be fair; neither Samsung nor LG is immune from color fringing in computer use.

Keep taking more photographs; More the merrier. Few of us, myself included, have concurrent access to all the dozens of OLED televisions and monitors ever made -- so this sharing comparison photographs of OLED is quite useful.
That's a very fair assessment. My typically sitting distance from my monitor is more than arms length away. Combine that with the 1440p resolution, my mid 30s eyesight, and the fact that I typically use my second IPS monitor for web browsing and things that have text the majority of the time....it doesn't bother me. The gaming fidelity and experience is 100% worth it to me.

Though worth noting even when I get REALLY close to my monitor, I still cannot see it. To take the picture I basically have my phone planted against the screen with x3 zoom (much to my discomfort, but [H] has needs!)
 

raymod2

n00b
Joined
Jan 21, 2023
Messages
5
@Chief Blur Buster: I don't agree with your assessment that the LG and Samsung panels should both suffer from color fringing due to their subpixel layouts. The Samsung triangle layout seems perfectly suited to avoid this problem. All subpixels are equidistant from each other so all secondary colors should blend equally well. Also, according to your argument the Samsung panel should show color fringing on all 4 sides not just 2 (left=red, right=blue, top=green, bottom=magenta). I have never seen an Alienware monitor in person but I suspect if there is any color fringing it is far less severe than the yellow vertical edges on my Asus monitor. The photos that @Lifelite provided are inconclusive. Maybe he could reproduce my 6 color boxes and photograph them from around 6 inches away using 2x zoom in portrait mode (for a fairer comparison to my photos).

Regarding the problem of burn-in: I don't see how adding a white subpixel can prevent this. If I display a red box at full brightness on a black background then the red subpixels will wear out faster than all the other subpixels in that square and eventually that square will remain visible when displaying other content. The white subpixel doesn't even get used in that scenario. From my understanding the purpose of the white subpixel is to increase brightness. All the subpixels are actually white with color filters in front of the red, green, and blue ones. Since there is no color filter in front of the white subpixel all the light gets through. The Samsung panel doesn't need a white subpixel because it doesn't use color filters. All the subpixels are actually blue and It uses quantum dots to absorb all the blue light and emit the color for which they were tuned (and they do this with nearly no loss of brightness).

From where I am standing it seems like Samsung has the better technology (too bad it is not available in a 42-inch 4k format). However, not all is lost with the LG panel. Unless there is some hardware reason why the subpixels need to be grouped as RWBG I think we would get a much better picture by changing the grouping to WBGR (the way I originally assumed they were grouped). The red and green subpixels would be neighbors again like they are in almost every other display. Also, the white subpixel wouldn't interfere with subpixel rendering techniques such as ClearType. I'm a little confused why they chose the RWBG grouping in the first place.

You can almost do this in software as I have demonstrated below. The right image looks better when rendered on my Asus OLED panel. The only problem is a missing column of yellow pixels on the right side. I can't coax my monitor to display them because that would require lighting red subpixels next to white subpixels. The monitor firmware could presumably improve on what I have done.

1674759595040.png
 

Chief Blur Buster

Owner of BlurBusters
Joined
Aug 18, 2017
Messages
314
lol do you understand who you are doubting?
Em, the ONLY thing I disagree is "I don't agree with your assessment that" ... "panels should both suffer"

I never said that the panels should suffer. The forum member 'raymod2' simply only incorrectly assuming because of the restrained tone of my writings.

I know that they should NOT suffer from color fringing. I simply explained WHY they currently suffer from fringing.

That has nothing to do with me accepting the fringing which I don't like either... There are people who don't mind and don't care. Much like people who don't mind and don't care about 60Hz or tearing or TN-vs-IPS color. But there are people who do care deeply, myself and raymod2 and many others included.

The rest of raymod2 post is fine and accurate.

I just can't yet say much except confirming publicly available info. I will soon be able to say more in future, so keep tuned.

From where I am standing it seems like Samsung has the better technology (too bad it is not available in a 42-inch 4k format). However, not all is lost with the LG panel. Unless there is some hardware reason why the subpixels need to be grouped as RWBG I think we would get a much better picture by changing the grouping to WBGR (the way I originally assumed they were grouped). The red and green subpixels would be neighbors again like they are in almost every other display. Also, the white subpixel wouldn't interfere with subpixel rendering techniques such as ClearType. I'm a little confused why they chose the RWBG grouping in the first place.

You can almost do this in software as I have demonstrated below. The right image looks better when rendered on my Asus OLED panel. The only problem is a missing column of yellow pixels on the right side. I can't coax my monitor to display them because that would require lighting red subpixels next to white subpixels. The monitor firmware could presumably improve on what I have done.
Correct, nothing new to me -- I knew that in 2022;

Hopefully I have good news to say. But given current agreements I can't say anymore about these possible solutions, firmwares or timelines yet.

The bottom line is that smartphone-style subpixel awareness needs to be ported to PC monitors and PC operating systems. Basically smartphones were the trailblazers here, doing a "Pentile-aware ClearType" algorithm for everything on screen (text and graphics). Windows doesn't have a custom subpixel-awareness system to eliminate fringing on custom subpixel systems. That eliminates a lot of fringing with custom subpixel layouts. I already know that.

Also, I've also written about shader-based (or ASIC/FPGA based) CRT electron beam simulators as a custom scaling algorithm that decouples pixels from subpixels, and I've written those posts for years (e.g. talking about MAME HLSL, and similar algorithms used simultaneously with my RetroArch BFIv3 suggestion on GitHub etc.). CRT subpixels (phosphor dots) were completely unliked from computer pixels, and everything looked fine! It is possible to software-simulate that in realtime at current subpixel counts in a shader or FPGA, even for non-integer-subpixel-count mappings per pixel. MAME HLSL is an example of publicly available source code that has no integer-subpixel-count per pixel, as an advanced CRT filter used for arcade emulation. In the OLED case we've got at least exact subpixel count per pixel, so the job is much easier, but spatially-aware subpixel mapping per pixel is very old science, with lots of existing masterpiece art that is not yet milked for the desktop OLED market quite just yet.

Let's face it: For a long time it was the television market, and suddenly lots of computer use and closer viewing distances by 20/20-vision young esports eyes...

Just give the OLED consumer desktop monitor market some time to mature. It's bluntly so bleepin' shiny brand new.

Keep the photographs coming from everybody. They're useful. More the merrier.
 
Last edited:

Chief Blur Buster

Owner of BlurBusters
Joined
Aug 18, 2017
Messages
314
That's a very fair assessment. My typically sitting distance from my monitor is more than arms length away. Combine that with the 1440p resolution, my mid 30s eyesight, and the fact that I typically use my second IPS monitor for web browsing and things that have text the majority of the time....it doesn't bother me. The gaming fidelity and experience is 100% worth it to me.

Though worth noting even when I get REALLY close to my monitor, I still cannot see it. To take the picture I basically have my phone planted against the screen with x3 zoom (much to my discomfort, but [H] has needs!)
Same here, the gaming fidelity of the Corsair Xeneon Flex is 100% worth it too.

OLED is a huge advancement in motion quality for gaming, even if BFI isn't arrived yet. Brute framerate-based motion blur reduction works absolutely beautifully on OLEDs, with a 240Hz OLED beating a 360Hz LCD in motion clarity at maxed framerate=Hz, thanks to the nearly "GtG=0" behavior of OLEDs. LCD strobing still outperforms if you don't mind disadvantages (flicker, brightness loss), but 240fps 240Hz OLED is much brighter and more colorful while still having motion clarity that resembles older strobing implementations such as LightBoost (but without the disadvantages). While still achieving better motion clarity than a non-strobed 360fps 360Hz LCD, with the delightful inky blacks. LCD has a place, but OLED is going to practically muscle in rather quickly over the next two years.

Practically all the first 240Hz OLEDs already have fewer disadvantages (and less lag) than the first 240Hz LCD in 2016 (which wasn't as good as today's 240+ Hz LCDs), even if the newest LCDs still has less lag -- but this is a very mighty strong debut for the 240Hz and 3-figure priced era of OLEDs. Undoubtedly, there's the ...em... fringing that bothers some but not others. It bothers many reviewers such as RTINGS, so all the companies have an economic interest in working on this matter. Give it time.

Now that being said, these are desktop computer monitors finally hitting more mass-market three-digit pricing and I also want to Visual Studio and PhotoShop on them, and I have a big interest in reducing color fringing and improving text clarity, too.
 
Last edited:
Top