Asus PG27AQN vs. Zowie XL2566K vs. 240 HZ OLED

[Nebulusedge]

Hey guys,

I recently build a new PC with a 7800X3D and a RTX 4080. With it I bought a Samsung G7 Odessy 27". I can't say that I'm totally dissatisfied with it but it feels like there is a lot of room for improvement. My current monitor doesnt feel fluid in motion although I achvieve high fps. Even with G-Sync the overall experience while playing competitive shooters like CSGO and PUBG is kinda meh (same for G-Sync off).

As I'm stilll in the return window of my G7 I informed myself about other possibilities. I found out about Dyac+ and ULMB 2 which greatly improve spray control and motion clarity for a smooth gameplay. I would buy the Zowie in a heartbeat but having a gaming monitor that is limited to 1080p in 2023 feels not futureproof and limiting. It's not like I don't want to play on 1080p at the moment which for PUBG is actually nessecary to get somehow in the ballpark of stable 360FPS.

Personally I'm interested in a very fluid and smooth display of the game. I like good colors and a good contrast but it's not mandatory for me. The main problem in my decision is the following: I don't think a monitor with 1080p and no g-sync will be a good decision in the long run as why I dont want to buy the zowie. The problem with the PG27AQN is the fact that I don't think its possible to achvieve stable 360FPS at 1440p for PUBG. I read about the E-Sports-Mode where you can simulate a 25" monitor at 1080p with it which would be optimal as I could use this for PUBG and the normal 27" resolution for things like CSGO where I'm capable of achieving 360FPS. I also thought about just setting the display to 240hz when I'm playing PUBG but I couldn't find any information out there how good this display would perform ULMB2 at lower frequencies.

I also read about OLED monitors and their qualities but it also seems like they lack in motion clarity for competitive games. I'm kinda sweaty and most of the games I play are FPS shooters. But there also seems many people that recommend G-Sync in combination with a good LED gaming monitor for the smoothest gameplay.

My main questions hover around the usage of ULMB2 at lower frequencies like 240Hz and what you guys thinks would be the best solution as I'm currently torn apart between all the different technologies.

Thanks!

[bumbeen]

ULMB2 sucks, edit, well, it doesn't "suck" that's a bit strong, but there's certainly a lot of room for improvement. At 360hz the strobe phase is off, has no adjustment. The crosstalk is worse than it should be, the bottom 20% of the display is still showing the previous frame more clearly than the current frame, while there's zero crosstalk for the next frame anywhere on the screen. The pulse width setting actually makes the crosstalk way worse because the start point of the strobe is fixed and and reducing the pulse width only moves the end point of the strobe closer to the start-point. This would eliminate crosstalk from the next frame, but there isn't any! So instead it reduces the visibility of the current frame, the one you should be seeing, in favor of the previous frame crosstalk.

For all their hype around it they could have done better with this, a strobe phase timing adjustment would fix all of this.

ULMB2 is better at 240hz, although still not ideal without phase adjustment, it's again strobing too early in the refresh cycle, but at least the bottom of the screen is showing the current frame instead of the previous frame.

15:11 GMT - couple edits

[tomjerry52]

ULMB2 sucks, edit, well, it doesn't "suck" that's a bit strong, but there's certainly a lot of room for improvement. At 360hz the strobe phase is off, has no adjustment. The crosstalk is worse than it should be, the bottom 20% of the display is still showing the previous frame more clearly than the current frame, while there's zero crosstalk for the next frame anywhere on the screen. The pulse width setting actually makes the crosstalk way worse because the start point of the strobe is fixed and and reducing the pulse width only moves the end point of the strobe closer to the start-point. This would eliminate crosstalk from the next frame, but there isn't any! So instead it reduces the visibility of the current frame, the one you should be seeing, in favor of the previous frame crosstalk.

For all their hype around it they could have done better with this, a strobe phase timing adjustment would fix all of this.

ULMB2 is better at 240hz, although still not ideal without phase adjustment, it's again strobing too early in the refresh cycle, but at least the bottom of the screen is showing the current frame instead of the previous frame.

15:11 GMT - couple edits

Do you think the delayed frame in the bottom 20% of the screen causes games like valorant to be unplayable?

[Dalek]

The PG27AQN colour accuracy is perfect, the colours look 'normal' and balanced, like what you would get with a standard monitor. It only took me 10-15 minutes to setup, afterwards I didn't mess with the OSD menu again (other than testing 25" mode)

I personally didn't like 25" mode since I was so used to using 24" 1080p, but you won't recover that much FPS from it. Not only that, I wouldn't use 25" mode given you pay so much money for the monitor.

The Zowie XL2566K has much worse colour accuracy according to two reviews I've seen (Brandon Taylor and Monitors Unboxed) which is no surprise because when I tried a Zowie monitor from a few years ago, the colours and contrast were horrible. I haven't tried the XL2566K but I guess all you would be paying for is motion clarity and nothing else.

I'd say if you're worried about frame rate then lower your graphical detail and you should be fine. Check the Monitors Unboxed review here to see the stats between the different refresh rates. You probably won't even notice too much difference between 240hz and 360hz.

I didn't even really notice much difference between the different levels of overdrive on the monitor. It looked ok without any overdrive. Maybe it was because I kept it in the default profile and not set it to esports? .. not sure.

For OLED, I haven't tried any because they're all dealbreakers with not looking correct with Windows font/text due to sub-pixel layout.

If you want a monitor that is accurate with colours, brightness, contrast and is decent at motion clairty I'd say go for the PG27AQN. I returned mine since I felt it was too big for me and the panel uniformity (on the left and right hand side of the screen) was slightly dimmer than what I'd like which is only noticable on the desktop (Monitors unboxed also faced the same issue). ULMB2 wasn't out at the time of owning the monitor so I can't comment on it. But what I will say is that I'm not paying 1.1K to have an issue like that when the previous monitor I have (standard 60hz) doesn't have any panel uniformity issues.

[Chief Blur Buster]

ULMB2 sucks, edit, well, it doesn't "suck" that's a bit strong, but there's certainly a lot of room for improvement. At 360hz the strobe phase is off, has no adjustment. The crosstalk is worse than it should be, the bottom 20% of the display is still showing the previous frame more clearly than the current frame

It's a tradeoff, pick-poison.

You could adjust it so you have 10% badness at top edge, and 10% badness on bottom edge.

Max Hz, such as 360Hz, will typically have no large vertical total possible (without Pixel Clock overclocking capabilities), so you're definitely stuck with 20% at max Hz, so adjusting is a tradeoff between top and bottom edge.

However, making the bottom 20% edge bad has lower input lag than making the top 20% edge bad. The input latency is a latency gradient along the vertical axis, starting from a zero-basis at the crosstalk band. In other words, input lag is lowest right above crosstalk band, and input lag is highest right below crosstalk band, due to the dissonance between LCD scanout (high speed video) and global strobe flash.

What I am super impressed by NVIDIA is that they managed to make it reasonably good for 80% of the pixels of the screen. That's really good strobe tuning for 360Hz. It's "not good enough" for many, and there are many who prefers to center the good zone (80% good middle, 10% bad bands equally balanced top and bottom).

Cramming LCD GtG into the VBI between refresh cycles, is like pushing an elephant through a drinking straw!

But I understand word-of-mouth is very powerful, and strobe backlight modes are often used by advanced Blur Busters users. And bad word-of-mouth can reduce ULMB2 sales.

It would be in NVIDIA's interests to contact Blur Busters to get ULMB2 Blur Busters Approved, by adding:
(A) Any-Hz strobe tuning (59Hz-360Hz). We've begun using algebra regression curves on strobe presets to curve-fit between-presets strobe tuning, so if an indie like me can do it, NVIDIA almost certainly could.
(B) OD fine-tune offset for existing advanced overdrive (good for panel temperature, panel aging, and panel lottery)
(C) Strobe Utility compatibility (including strobe phase, pulse width, and fine-tune OD fine-tune offset)

If NVIDIA adds those, it would be pretty much Blur Busters Approved!

[bumbeen]

If NVIDIA adds those, it would be pretty much Blur Busters Approved!

All my complaints with the thing would be solved with strobe phase adjustment, that's all it needs

The zowie 360 has slower response time(excepting blacks) than pg27aqn, yet it doesn't have this issue of showing more of the prior frame than the current frame at the bottom of the display and still looks just as good at the top and far better in middle of the screen. Perhaps they're doing some trickery with the scanout that isn't possible at 1440p.

Do you think the delayed frame in the bottom 20% of the screen causes games like valorant to be unplayable?

It's not unplayable, just annoying to pay so much for a monitor and have it not function properly. ULMB2 should be defacto best for a game like valorant, but I often just leave it on gsync.

It's just dumb that they added this pulse width setting and then it does absolutely nothing when it could be improving the image significantly

[Nebulusedge]

Thanks for the feedback although I had my problems understanding everything you guys talked about

In the end I ordered the ASUS 360hz. After further thinking about it, even thinking about a 1080p monitor feels so wrong at this point. OLEDs seem to have a lot of problems with G-Sync, flickering and burn-ins.

I have to say I feel a little bit betrayed paying 1300€ for a monitor. That’s a lot. In the end I’m still paying for it but if I would consider price/performance in my buying decision then this monitor never would have been bought. Nonetheless I’m very excited to see what this monitor is capable off and I will try to leave my own subjective feedback as a rather “noob“ customer for future reference and if I can really notice the problems you guys are talking about.

[Chief Blur Buster]

It's just dumb that they added this pulse width setting and then it does absolutely nothing when it could be improving the image significantly

Are you sure it does nothing? www.testufo.com/map#pps=3000 and look at the changes in clarity to the street labels.

To see human visible MPRT differences, you need horizontal motionspeeds faster than what the pulsewidth suggests. 1ms pulsewidths need to be benchmarks at minimum ~1500 pixels/sec and 0.5ms pulsewidths need to be benchmarked at ~3000 pixels/sec

Basically, enough time for your eyes to track over 1ms on 1500 pixels/sec (1.5 pixels of motion blur) and 0.5ms on 3000 pixels/sec (1.5 pixels of motion blur).

It's very subtle. You must have perfect VSYNC ON framerate=Hz *and* you must have very fast motionspeeds, to benchmark ultra short pulse widths. My favorite test for pulse widths is 3000 pixels/sec on 1080p and 4000 pixels/sec on 1440p as that is the edge of my eye tracking speed for a desktop display.

Note: Add any jitter (even VSYNC OFF jitter) and it will totally wreck your ability to tell apart ULMB pulsewidths. So if you want more lagless VSYNC, try using RTSS Scanline Sync or SpecialK Latent Sync. Marriage made in heaven with strobing to bring all that TestUFO-smoooooooth delicious pans/turns/strafes into your games! But if your game runs at a lower frame rate, intentionally lower your refresh rate to make it syncable jitter-free. And, to prevent mouse jitters with strobing, use high-pollrate, high-DPI, low-ingamesens, and try to use ~2KHz pollrate (if game handles it without lagging down). Games like Rocket League (capable of >600fps) usually behave relatively well with Scanline Sync even at 240Hz+, making it easier to track the flying ball jitter-free and blur-free. Try testing 1600-3200dpi whenever using strobing. Higher DPI benefits really become visible during strobe situations (improved slowmoves/tracking/eyetracking while panning for scouting scene), if you also do slowmoves in addition to your flick moves. Experiment though!

________________________

Interested related subject -- someday, we will say goodbye to flicker-based blur reduction and use brute framerate-based blur reduction instead.

For strobless blur reduction (brute framerate method, via 10:1 lagless frame generation of the future) -- incidentally, fastest eye-tracked motionspeed in pixels/sec is also a good estimate of the very approximate required "retina refresh rate" (~3000Hz for 24" 1080p, and ~4000Hz for ~27" 1440p, with an error margin of 50%-200%). This assumes the best case of a perfect GtG=0 and flicker-0 (flickerfree blurless sample and hold, with motion blur solved by brute frame rate instead for strobeless sub-millisecond MPRTs).

Retina refresh rate of no further humankind benefit, is a function of your fastest eye tracking speed on perfect jitter-free motion (VSYNC ON style) over a finite time span long enough, over still-human-visible pixels, to identify the existence of motion blur that makes it blurrier than a stationary image. So the moving object can't disappear offscreen too fast, or you can't tell it's blurrier than a stationary image. So wider-FOV higher-rez displays have a higher "retina refresh rate". It goes to 20,000Hz (ish) for 180-degree FOV 16K display (VR).

The approximate 20,000 Hz number (roughly the max display refresh rate of no further humankind benefit) is due to the stroboscopic effect (PWM style), ala mouse-dropping effect of The Stroboscopic Effect of Finite Frame Rate. Even though in theory stroboscopic effect (e.g. 50,000 pixels/sec mouse flicks on a theoretical 8K esports display of a future decade) could still reappear, this is solved by adding an imperceptible 1/20,000sec of GPU blur effect, and then call it a day.

The number is mentioned in 3 locations:

(A) Lighting industry paper (Page 6), where they found they needed 20,000Hz for electronic ballasts for fluorescent lights because a tiny % of human population saw stroboscopic effects at 5000Hz-10000Hz;
(B) The Vicious Cycle section of my article, Blur Busters Law: The Amazing Journey To 1000 Hz Displays
(C) This new YouTube video,

phpBB [video]

Bet it will not be till the late parts of this century where we can have 16K 20KHz displays, where it's spatially AND temporally maxed out. Ha.

But I definitely will be pushing to incubate the 1000fps 1000Hz ecosystem. It can be done (eventually) much more cheaply than many think. Expensive at first, though.

FSR/XeSS/DLSS needs 10:1 framegen ratios, and done laglessly (yes, lagless esports framegen is possible). 1000Hz displays are arriving by mid-decadeish in LCD format, followed by OLED versions before end of the decade (2027ish earliest ETA, I think).

[bumbeen]

It's just dumb that they added this pulse width setting and then it does absolutely nothing when it could be improving the image significantly

Are you sure it does nothing? www.testufo.com/map#pps=3000 and look at the changes in clarity to the street labels.

Oh well for sure you're absolutely correct it DOES improve the MPRT blur absolutely since it is further reducing the static image time. I guess I should have said it could have improved the crosstalk more than it does. You do in fact get a tiny bit of improvement in crosstalk at the top of the display by lowering pulse width from 100% to 80%. The very top edge goes from having slight crosstalk to zero crosstalk.

Around 70% vertical distance of the screen is the point (at 100% pulse width) where the previous frame crosstalk becomes more visible than the current frame. So sadly lowering pulse width actually shifts this point further up the screen. It's effectively a one way strobe phase adjustment in the wrong direction since it doesn't maintain the mid-point of the flash.

I have some interesting news however! By a miracle of God I received a substantial response from Asus on the ticket I had open with them on this. It's a little hard to understand what they're saying, but it sounds like Nvidia are completely aware of the issues I find with the pulse width control and another firmware update will be coming in the future:

Nvidia Reply: Based on the user description you provided, we know that the "Best location" for ULMB 2 is the top of the monitor, not the middle of the monitor, which is correct.
A firmware update is required if modification is desired, we will refer to this user's suggestion and add future updates, but we cannot provide an exact update date.
The above is the reply from NV. Our PM hopes to complete the update by the end of October this year.
However, the technology of ULMB2 comes from NV. We cannot control it to avoid disappointing users’ expectations. After discussing with PM, we have suggestions and users reply to his suggestions. Pass it to Nvidia, Nvidia will take it into consideration, if the user does not accept it, please discuss the Local policy with CSM.