Samsung QD OLED

High Hz on OLED produce excellent strobeless motion blur reduction with fast GtG pixel response. It is easier to tell apart 60Hz vs 120Hz vs 240Hz on OLED than LCD, and more visible to mainstream. Includes WOLED and QD-OLED displays.
DrR1pper
Posts: 61
Joined: 03 Aug 2015, 17:55

Re: Samsung QD OLED

Post by DrR1pper » 26 Mar 2022, 12:53

Inco^ wrote:
26 Mar 2022, 05:16
Thanks for your detailed answer. So if I understand correctly, you would still place the AW3423DW slightly behind the best fast-IPS monitors in the motion handling (clarity, smoothness) department.
Yes, in practice, I find it only slightly behind one of (if not the) best 240hz fast-IPS BFI capable monitors in BFI mode I've tried (Asus XG27AQM) and absolutely destroys it in non BFI mode. It is such a better all round user experience for me.* Can't be sure how it would stack against a 360hz of course but even if it felt up to 50% better than my 240hz (which I doubt), I'd still go with the AW3423DW.

Compared to the 240hz fast-IPS in non BFI mode, there's not a single shred of competition to be had with the AW3423DW. With BFI though, motion clarity is most definitely still going to be objectively miles better (as can be seen using e.g. PixPerAn reading in motion test...no way the Alienware can compete with BFI but it does seem like the Alienware can get a bit faster peak speed readability vs. non BFI IPS even though it's pixel persistent motion blur limited still and that's down to the complete absence of any ghosting. I will test this later and report back findings with pictures so you know what I mean better.) That being said, the brightness is so much higher on the Alienware vs. Asus BFI that it's easier/less strain to see clearly and when you combine the two together, the effect is strobe-like motion clarity. So much so, that it's more than good enough to compete with BFI for competitive edge and far more enjoyable in casual gaming. So for me it's a total annihilation win for the AW vs Asus in non-BFI and a good/great win over the Asus in non-BFI. But of course, caveat, YMMV.

Edit:

I just realised I think a good chunk of the reason for why it feels strobing-like is the post flick return to clarity of oled. So on IPS without BFI, when you flick the mouse around, it takes a good amount of time for the image to become unsmeared from ghosting again. Whereas it's the same instance as the flick is stopped. With BFI though of course, you are getting this exact effect even when in the flick which is why it's so amazing.

So in summary, the flick movements (sudden rapid change in where you're looking) is a total blur on the AW but it's only happening for a few milliseconds and resolves to clear the instance you stop the flick and this therefore feels like it's doing BFI. In medium flicks/movements, the motion clarity is at the AW's weakest but still pretty decent due to complete absence of ghosting contribution to motion blur leaving only pixel persistence motion blur as the cause of any motion blur. Then in slowest flicks/moments, it's pretty much as good as BFI because pixel persistence is not an issue at these slow speeds and ghosting is completely absent.
Last edited by DrR1pper on 27 Mar 2022, 00:21, edited 5 times in total.

DrR1pper
Posts: 61
Joined: 03 Aug 2015, 17:55

Re: Samsung QD OLED

Post by DrR1pper » 26 Mar 2022, 13:36

On the pixperan (at same 175hz) I was able to pass up to level 7 speed on Asus with non BFI and up to level 8 on Alienware. Asus BFI, I could just about successfully pass the test up to level 17 but if the screen were as long as the Alienware, I think I could do 18. The clarity is CRT clear of course but the dimness and contrast of the display makes it hard to lock your eyes on the text quickly enough to read it despite it being motion blur free. There is no contest that motion clarity is superior on IPS BFI...without a doubt. But the Alienware is more than good enough without BFI for competitive and casual gaming, especially for the brightness benefit, let all the perks of OLED. That being said, I look forward to the day BFI without the brightness compromise on OLED comes to market but I think that's a good ways off of course.

A future of true LED with BFI and 10,000+ nits peak for HDR is going to be glorious.

DrR1pper
Posts: 61
Joined: 03 Aug 2015, 17:55

Re: Samsung QD OLED

Post by DrR1pper » 26 Mar 2022, 13:45

Also, fyi, if you use PixPerAn readability test, you've got to make sure it's at the same refresh rate when comparing monitors because the speed of the text is dependent on the refresh rate. Took me a while to realise this when i first used it way back. So if you go from 120hz to 240hz, for the same readability test speed setting, not only will the text move across the screen at twice the speed for 240hz than it does for 120hz, the timer for how much time you have to attempt the test will also run down twice as fast. It's weird lol but it is what it is.

DrR1pper
Posts: 61
Joined: 03 Aug 2015, 17:55

Re: Samsung QD OLED

Post by DrR1pper » 26 Mar 2022, 19:09

I reread my reply from earlier today and realised it was poorly worded. Here’s the correction:
Inco^ wrote:
26 Mar 2022, 05:16
Thanks for your detailed answer. So if I understand correctly, you would still place the AW3423DW slightly behind the best fast-IPS monitors in the motion handling (clarity, smoothness) department.
Yes, in practice, I find it only slightly behind one of (if not the) best 240hz fast-IPS BFI capable monitors in BFI mode I've tried (Asus XG27AQM) and absolutely destroys it in non BFI mode. It is such a better all round user experience for me.*
Last edited by DrR1pper on 27 Mar 2022, 00:20, edited 1 time in total.

DrR1pper
Posts: 61
Joined: 03 Aug 2015, 17:55

Re: Samsung QD OLED

Post by DrR1pper » 26 Mar 2022, 22:16

To illustrate how I think the difference in severity of blur/smear from ghosting and pixel persistence vs. pixel motion speed causes the 175hz QD-OLED (and any other comparably high refresh rate OLED panel) to look strobe-like under high framerate gameplay.

Image

X-axis represents speed of pixels moving across screen. Y-axis represents severity of blur/smear.

There are two sources for total blur/smear of pixels in motion. The first is ghosting caused by pixel response time and the second is pixel persistence induced motion blur. The combination of the two is the resultant/total blur/smear you see.

It feels like the blur/smear contributed by ghosting scales in severity linearly with how quickly the pixels are moving (hence the linear blue lines). In which case, the slower the response time, the steeper the gradient of blur/smear from ghosting leading to earlier onset visual blur from ghosting on a display with slower response time (e.g. IPS vs OLED). The pixel persistence induced motion blur (orange lines) on the other hand are going to be identical in severity vs. speed of pixels in motion for both display types and it feels like the severity ramps up exponentially meaning little contributory blur effect in the low to medium end of motion. What this means for the total blur/smear you see (grey lines) on the QD-OLED vs fast-IPS non BFI is that it pushes the threshold of noticeable blur/smearing of pixels in motion to a higher pixel in motion speed. So the OLED feels strobe-like when casually/normally looking around because it is typically still below the threshold of noticeable blur.

Additional:

Total blur curves in the same chart.

Image

Zoomed in, if the green line represents the hypothetical threshold of noticeable blur/smear, you can see how it is pushed further up the x-axis (i.e. speed of pixels). Effect is like having BFI on as BFI flattens out the exponential persistence induced blur curve leading to the threshold of noticeable blur even further up the x-axis.

Image
Last edited by DrR1pper on 27 Mar 2022, 00:25, edited 8 times in total.

DrR1pper
Posts: 61
Joined: 03 Aug 2015, 17:55

Re: Samsung QD OLED

Post by DrR1pper » 26 Mar 2022, 22:33

Included hypothetical Total Blur curve for Fast-IPS with BFI. Threshold of noticeable blurring/smearing occurring at much higher pixel speeds.

Image

DrR1pper
Posts: 61
Joined: 03 Aug 2015, 17:55

Re: Samsung QD OLED

Post by DrR1pper » 26 Mar 2022, 23:51

phpBB [video]

Inco^
Posts: 101
Joined: 20 Apr 2020, 07:47
Location: France

Re: Samsung QD OLED

Post by Inco^ » 27 Mar 2022, 02:09

DrR1pper,

thanks for your time and explanations. So you actually feel like the AW3423DW has slightly better motion clarity than your previous 240Hz fast-IPS monitor (Asus XG27AQM) in non-BFI mode. But with BFI enabled, the fast-IPS, for you, looks more clear.
I understand your graphs and you have well explained the concepts: total motion blur comes from pixel persistence (driven by refresh rate) and ghosting (driven by pixel response time).

Regarding the graphs, which depict the blur versus the image speed:
1) you have assumed a linear relationship between blur from ghosting and image speed,
2) you have assumed an exponential relationship between blur from persistence and image speed.

Why the exponential assumption? I would have assumed linear as well. I understand that's an attempt to capture/fit your feeling. But right now with how you describe things, it seems like the most blur comes from persistence rather than from ghosting. And this way, enabling BFI is a game changer in your graphs. That's a bit counterintuitive to me, because I've used XL2546K (240Hz TN with BFI) and XV252QF (390Hz IPS with BFI) and for both monitors, enabling BFI helps but not by a big margin. I'm using XV252QF at 390Hz with BFI=extreme right now and I really feel like the motion clarity is hindered by ghosting from slow pixel transitions (not so much by persistence). The overall motion handling performance of the XV252QF is not that much reduced if BFI is disabled. And in parallel, most pros don't use DyAc on their BenQ's, do they?

I wonder if you could still fit your overall feeling of the AW3423DW performance vs XG27AQM performance, by using a linear model for the persistence blur instead of an exponential one. And by increasing in your graphs the importance of ghosting in comparison to that of persistence.

Thanks for the link to the video review of the AW3423DW. Unfortunately, the analysis is still not really focused on competitive shooters. Interesting to note, is that they claim they measured a 10%-90% pixel transition time of about 1-2 ms. It's faster than TN or fast-IPS, but not by as much as one would have thought, no? I would have thought OLED was 0.1 ms or something.
PG248QP - XL2566K - XV252Q F - PG259QN - XL2546K - Y27gq-25 - AG251FZ - LaCie Electron 22 Blue IV

DrR1pper
Posts: 61
Joined: 03 Aug 2015, 17:55

Re: Samsung QD OLED

Post by DrR1pper » 27 Mar 2022, 06:13

Inco^ wrote:
27 Mar 2022, 02:09
DrR1pper,

thanks for your time and explanations. So you actually feel like the AW3423DW has slightly better motion clarity than your previous 240Hz fast-IPS monitor (Asus XG27AQM) in non-BFI mode. But with BFI enabled, the fast-IPS, for you, looks more clear.
I understand your graphs and you have well explained the concepts: total motion blur comes from pixel persistence (driven by refresh rate) and ghosting (driven by pixel response time).
Most welcome! :)

No, I feel like the AW3423DW is a LOT better in motion clarity than the 240Hz fast-IPS monitor (Asus XG27AQM) in non-BFI mode (which is the mode I used that monitor in most of the time as I already found the non BFI mode's level of motion clarity paired with the significantly brighter and higher contrast of non BFI mode, good enough to sway my preference to it) in general/casual and competitive gaming. I was rather content using non BFI on the 240hz fast-IPS (coming from a 144hz/165hz-OC IPS just before that) so I was very pleasantly surprised and over the moon levels of extra happy when I tried the AW3423DW and discovered that for a non BFI monitor, it was significantly noticeably more clear in motion than the fast-IPS in non BFI. And the improvement in non-BFI motion clarity in competitive and casual gaming going from the 240hz fast-IPS to 175hz QD-OLED felt like it was at least 5x larger than going from 144hz/165hz-OC IPS to 240hz fast-IPS. To a level comparable to that of the fast-IPS in BFI mode in practice but without the significant brightness and contrast loss. But when you test the motion clarity of the fast-IPS and AW3423DW to the upper bounds of motion clarity, it of course becomes extremely clear that the fast-IPS in BFI has a much higher pixel-in-motion threshold before motion blur sets in. It's somewhere in the order of at least 5x (if not a good bit higher) for the fast-IPS in BFI than the AW3423DW from my testing with PixPerAn's readability test. But, if we ignore the fact that I was already happy to use the fast-IPS in non BFI the majority of the time anyway (meaning of course I'd be even more persuaded to shift my main display to the AW3423DW given the perception of near strobing like clarity on it whilst gaming)....the benefits of low motion blur for both competitive and casual gaming (at least for me) do not extend all the way up to the threshold of motion clairty that the 240hz fast-IPS in BFI mode is capable of. What the AW3423DW has actually shown/taught me is that actually, out of the two contributory factors that determine overall motion blur/smear on a display, provided the response time is low enough and refresh rate is high enough is that it's actually the ghosting portion of blur that is of greater contributor of perceived motion blur whilst gaming. E.g. although BFI removes reduces pixel persistence greatly that it raises the blur threshold so much higher than is the case for non-BFI, good BFI also removes a lot of the ghosting contributor of blur by making the BFI occur whilst a pixel is transitioning, thereby hiding the ghost. And so in actually, whilst in game, the motion clairty you get from BFI is I think coming more from the fact the BFI is hiding ghosting rather than removing pixel persistance (with the exception of flick shots that allow you to still see everything with zero motion blur whilst in the flick but this benefit is even overblown since flick shocks are muscle memory reactions rather than consciously calculated reactionary actions so as long as the frame resolves to clear by the time you're flick is over, the feeling you get is that it was almost as if there was no blur in the flick too if that makes sense). In hindsight, I realise this is why I become happy to stop using BFI as much after switching to the 240hz fast-IPS. Back when I had a 120hz BFI monitor that I would always use in BFI mode for competitive fps and most of the time for casual games (though usually at great discomfort due to even greater loss of brightness than on todays best BFI LCD displays...e.g. playing games like Metro 2033, i'd often walk into enemy NPC's because I just couldn't see them a game that was mostly took place in the darkness of the metro because the monitor in BFI was that bloody dim) because the response time of the panel (a 2012 vg278h 3d vision 2.0 monitor with 2d lightboost hack to enable BFI in 2d) had such long frametimes (being a 120hz display vs. todays 240-360hz) and much slower response times that I would frequently find myself above the noticeable blur threshold for that monitor in even slow mouse movements. BFI improved that by at least 10x fold such that my consistent long term average kdr in battlefield 3 literally jumped from 2.6 to 6 very quickly afterward that. Back then there was a bf3 stat site that recorded every game everyone ever played and you could see your performance stats recorded over time on charts and you could see extremely clearly the immediate cause and effect from shifting to that monitor, giving me an insane advantage literally overnight in that game. So much so that I was subsequently banned from most of the servers over the following year, believed to have been using aimbot or wallhack by admins to be able to consistently rack up such KDR's and total number of kills per game. I could understood though as it was typical of cheaters to be getting kdr's and kill numbers in that range. Thankfully, some admins were receptive to video recordings of me playing to prove I was not and allowed me back on whilst others not as people would tend to leave as a result of the absolute carnage that monitor permitted me to unleash on the enemy lol. Here's one of the videos I made at the request of an admin that banned me for suspected cheating so they could watch from my perspective me getting up to a kill streak of 21 kills and then an average kdr by the end of the game of at least 6 effortlessly....if interested: https://youtu.be/xy6zHQZV8pI?t=115

Sorry, segue over......with BFI enabled, the fast-IPS, stays crystal clear up to a significantly higher pixel motion speed threshold (as stated earlier by at least 5x if not a good bit more). However, the benefits of motion clarity for the majority of this much wider range of pixel motion is lost on me when I play both competitive (fps) and casual gaming. Couple that with the fact the brightness is considerable brighter without BFI which helps with ability to easily see and identify what may or may not be foe and there is a reactionary reflex benefit to be had there too. Combine all those factors and I feel I get a much better deal/benefit using the AW3423DW vs the 240hz fast-IPS in BFI mode.

Inco^ wrote:
27 Mar 2022, 02:09
Regarding the graphs, which depict the blur versus the image speed:
1) you have assumed a linear relationship between blur from ghosting and image speed,
2) you have assumed an exponential relationship between blur from persistence and image speed.

Why the exponential assumption? I would have assumed linear as well. I understand that's an attempt to capture/fit your feeling. But right now with how you describe things, it seems like the most blur comes from persistence rather than from ghosting. And this way, enabling BFI is a game changer in your graphs. That's a bit counterintuitive to me, because I've used XL2546K (240Hz TN with BFI) and XV252QF (390Hz IPS with BFI) and for both monitors, enabling BFI helps but not by a big margin. I'm using XV252QF at 390Hz with BFI=extreme right now and I really feel like the motion clarity is hindered by ghosting from slow pixel transitions (not so much by persistence). The overall motion handling performance of the XV252QF is not that much reduced if BFI is disabled. And in parallel, most pros don't use DyAc on their BenQ's, do they?
So that's a very interesting observation you've made on your 390hz monitor. Also, I may have answered your question with my prior paragraph why not the linear for both but if not, let me know and I'll give it a crack to explain my reasoning. That being said, you could be right. So, back to your observation. I think what's happening is that at such a high refresh rate, the response time is as long (if not longer) than a frametime. So BFI is unable to remove the ghosting from creeping into the present and/or previous frame. If this is the case, you need to lower the refresh rate enough such that now the response time (and thus ghosting) is less than the frametime and able to be blacked out by the BFI. You know, thinking about this, I think this is why I found my previous previous monitor that was a native 144hz able to OC to 165hz, felt less blurry at 144hz than 165hz. If the pixel has still not completed a transition by the time it is time to transition to the next state for the next frame, then that too would cause a blur (secondary to the ghosting itself). Very interesting.

If I'm right, higher refresh rate monitors usually also having higher response times, will mean setting your max capable 390hz BFI monitor (XV252QF) to the same 240hz refresh rate as your max capable 240hz BFI monitor (XL2546K) and turning on BFI on both, should yield a better motion clarity on the max capable 390hz BFI monitor (XV252QF) than the max capable 240hz BFI monitor (XL2546K).

Inco^ wrote:
27 Mar 2022, 02:09
I wonder if you could still fit your overall feeling of the AW3423DW performance vs XG27AQM performance, by using a linear model for the persistence blur instead of an exponential one. And by increasing in your graphs the importance of ghosting in comparison to that of persistence.

Thanks for the link to the video review of the AW3423DW. Unfortunately, the analysis is still not really focused on competitive shooters. Interesting to note, is that they claim they measured a 10%-90% pixel transition time of about 1-2 ms. It's faster than TN or fast-IPS, but not by as much as one would have thought, no? I would have thought OLED was 0.1 ms or something.
Yeah I found the quoted transition time perplexing and then I remembered it's not really the right metric to gauge relative response time. Should really be using metrics like what Hardware Unboxed uses now in addition to 10-90% rise/fall time and that is cumulative deviation. See, like how Vincent (of HDTVtest's) review of the AW3423DW measured an average response time of 1-2ms, Hardware Unboxed measured 1.28ms for the LG C1 OLED compared to a 3.57ms for the Alienware AW2721D (240hz fast-IPS). This is just a 2.8x better/faster response time for the OLED vs. a fast-IPS panel. However, the cumulative deviation was 67 for the LG C1 and 532 for the AW2721D. That's an 8x difference and much closer to the ~10x faster response time we often hear claimed OLED's are vs. top-end performing IPS. Point being, the cumulative deviation is a much closer measure for what the difference is which is why HW started using it.

Image

Image

You could be right that both ghosting and pixel persistence should be linear but the reason I thought it's not is because when I run the UFO test and PixPerAn readability test, there is actually an initial range of speed that is as sharp as BFI on both monitors. However the point at which blur starts to set in if the speed goes higher is higher for the AW3423DW than the 240hz fast-IPS (XG27AQM). So that made me think there might be a difference in contributory power to total blur from ghosting and pixel persistence across different pixel motion speeds. What got me thinking along those lines was my uni days of aerodynamics and how total drag is a combined effect of parasitic and induced drag and how those two change based on speed of airflow, lol.

Image

It's not exact the same combined curve behaviour but still, it's what got me thinking alone these lines. But thinking about it some more, maybe I'm wrong and it doesn't have to be linear for one and expotential for another. Both could be linear (or exponential) and that it's just the case that the human perception is as good as a flatness on initial portion of the total blur line (or curve) and that because there is no ghosting for the OLED, the threshold of when total blur becomes noticeable is of course going to be higher up the x-axis than for the fast-IPS. This seems more likely now I think about it more. Good stuff! :)

User avatar
Discorz
VIP Member
Posts: 999
Joined: 06 Sep 2019, 02:39
Location: Europe, Croatia
Contact:

Re: Samsung QD OLED

Post by Discorz » 28 Mar 2022, 03:38

Inside look of AW3423DW, panel model (Samsung QTM340YA01-C01) and more:
https://youtu.be/pqFCg0PZwOk?t=498

I hear people are complaining about 'PWM dim like' flicker. That is a big no-no!
https://youtu.be/pqFCg0PZwOk?t=350
https://youtu.be/6FIzbUsIK7Q?t=312
Text clarity/pixel layout is also not great.
Last edited by Discorz on 01 Apr 2022, 01:01, edited 1 time in total.
Compare UFOs | Do you use Blur Reduction? | Smooth Frog | Latency Split Test
Alienware AW2521H, Gigabyte M32Q, Asus VG279QM, Alienware AW2518HF, AOC C24G1, AOC G2790PX, Setup

Post Reply