Experience & Opinion: 240hz displays are blurry

Everything about displays and monitors. 120Hz, 144Hz, 240Hz, 4K, 1440p, input lag, display shopping, monitor purchase decisions, compare, versus, debate, and more. Questions? Just ask!
Locked
User avatar
Chief Blur Buster
Site Admin
Posts: 7868
Joined: 05 Dec 2013, 15:44
Location: Toronto, Ontario, Canada
Contact:

Re: 240hz displays are blurry

Post by Chief Blur Buster » 27 Dec 2017, 22:46

Easy, easy, please.
Be nice to each other. Forum rules. Thanks.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

       To support Blur Busters:
       • Official List of Best Gaming Monitors
       • List of G-SYNC Monitors
       • List of FreeSync Monitors
       • List of Ultrawide Monitors

darzo
Posts: 211
Joined: 12 Aug 2017, 12:26

Re: 240hz displays are blurry

Post by darzo » 27 Dec 2017, 23:22

The photos do indeed speak for themselves, doubly so with the TFT review that spells out their meaning for you. That's the extent of attention you'll get from me.

yehaw
Posts: 71
Joined: 21 Dec 2017, 21:41

Re: 240hz displays are blurry

Post by yehaw » 28 Dec 2017, 01:40

There are multiple sources across the web complaining about this mysterious blur that I think I've tracked down:

Acer 240hz:
Acer Predator XB252Q is really blurry and sh***y
BenQ 240hz:
The BenQ Zowie XL2540 was met with decent reviews but worries about motion trails (overshooting pixels that create a blurry visual) made us look at different monitors.
BenQ 240hz:
Also compared to the xl2450, everything is a lot more visible to my eyes when moving, it doesn't seem to have the extreme fuzziness/blurriness I was experiencing on the zowie.
Asus 240hz:
During fast movements with PG258Q, it feels more like "smooth motion blur". Objects in the game lose their contrast around their edges and sharpness during a fast "swipe" and everything becomes more "blurry". Therefore I recommend ULMB on PG258Q for games where the fps is constantly very high (like csgo). It is possible that the PG258Q feels more blurry due to the higher refresh rate and limitations with your eye? Perhaps this is because the higher refresh rate requires a better overdrive / AMA to have the same contract / sharpen as shields with lower refresh rate?

I do not know why it feels better on the XL2411t with overdrive, but it is noticeable that it affects the flicks in play. An easy way for me to see the difference is when I run both screens in mirroring, and just spin the mouse pointer in Windows. The PG258Q mouse pointer can actually be "dropped" due to the higher motion blur, while the XL2411T makes it easier to keep track of the mouse pointer, as it is clear even during motion.
Acer 240hz:
Comparing 1080p 240hz gsync to dell 1440p 24 inch gsync i choose the sharpness of 1440p 165hz over 1080p 240hz. Also the acer seemed to have ghosting or blur at 240hz but the dell being lower hz still looked clearer. I will miss the lower input lag though but thats about it.


The last Acer link and quote above are exactly my experience to a tee if you go back and read my original post. And the person has identical panels. Dell 1440p is same panel in the AOC 1440p I have, and Acer 240hz is identical panel as Alienware 240hz I have. And even downscaled 1440p to 1080p with the added interpolation, the clarity of overdrive is still superior on the AOC compared to the native 1080p 240hz Alienware panel.

Probably more complaints, but I'm too lazy to look. But I think it's pretty clear there is a pattern here with 240hz displays. I'd love if someone more knowledgeable could do an in-depth comparison between all 240hz and 144/165hz panels and confirm or deny my findings. Personally, I think it's a limitation of the AU Optronics panel, which seems like premature tech. Why else would we have nearly pixel perfect overdrive on nearly all current gen 144hz/165hz/180hz, then all of a sudden we have this mess at 240hz?
Last edited by yehaw on 01 Jan 2018, 09:53, edited 1 time in total.

open
Posts: 223
Joined: 02 Jul 2017, 20:46

Re: 240hz displays are blurry

Post by open » 28 Dec 2017, 04:05

When you increase refresh rate, it can make still eye observations see the effect of pixel response times more.

When you increse the sample and hold time, (like you would when your fps dips on a gsync monitor) it can make moving eye observations see more motion blur.

So even though the 240hz gsync monitors have the best yet pixel response and lowest sample and hold times:
-There are times when the fast refresh rate can empasize the pixel response bluring. You might not be able to notice it at a lower refresh rate.

-There are also times when gsync fps dips can make eye tracking motion blur varry. Because the fps changes the sample and hold times and the sample and hold times determine eye tracking motion blur.

This is normal and you will still have the best pixel response and best potential sample and hold time on the 240hz.

If you are not noticeing what I am talking about then there is something wrong with your monitor and that is not normal for a 240hz. Best to speak clearly and not get the two things confused with eachother. If there is a problem then there is a problem. If we are talking about the subtle variations in motion blur depending on certain conditions then that is something else and is just part of the tech of modern lcd's that is getting better with each iteration.

yehaw
Posts: 71
Joined: 21 Dec 2017, 21:41

Re: 240hz displays are blurry

Post by yehaw » 28 Dec 2017, 04:22

open wrote:When you increase refresh rate, it can make still eye observations see the effect of pixel response times more.

When you increse the sample and hold time, (like you would when your fps dips on a gsync monitor) it can make moving eye observations see more motion blur.

So even though the 240hz gsync monitors have the best yet pixel response and lowest sample and hold times:
-There are times when the fast refresh rate can empasize the pixel response bluring. You might not be able to notice it at a lower refresh rate.

-There are also times when gsync fps dips can make eye tracking motion blur varry. Because the fps changes the sample and hold times and the sample and hold times determine eye tracking motion blur.

This is normal and you will still have the best pixel response and best potential sample and hold time on the 240hz.

If you are not noticeing what I am talking about then there is something wrong with your monitor and that is not normal for a 240hz. Best to speak clearly and not get the two things confused with eachother. If there is a problem then there is a problem. If we are talking about the subtle variations in motion blur depending on certain conditions then that is something else and is just part of the tech of modern lcd's that is getting better with each iteration.
If you read my original post, I stated I did an A/B test side by side with a 144hz display and the 240hz display and set both to 144hz with the tesfuo site. I don't think the test drops many frames once it hits the green valid, but both displays had G-Sync on. I originally stated I observed the Alienware looked much blurrier:
I do a comparison, I set my AOC to 144hz vs Alienware at 144hz on the testufo site using the alien invasion test. I stretch the browser window across both screen, and sure enough, the Alienware looks incredibly blurry or fuzzy in comparison. Even if I downscale the AOC 1440p resolution down to 1080p, the Alienware still looks incredibly blurry in comparison. I don't know if it's the pixel overdrive, the coating, or what - but it's terrible when you see it side by side.
So I think that test helps rule the higher refresh rate, fps, or g-sync being the culprit. I also played with the Alienware with G-Sync off and didn't notice a difference.

User avatar
Chief Blur Buster
Site Admin
Posts: 7868
Joined: 05 Dec 2013, 15:44
Location: Toronto, Ontario, Canada
Contact:

Re: 240hz displays are blurry

Post by Chief Blur Buster » 28 Dec 2017, 11:22

open wrote:When you increase refresh rate, it can make still eye observations see the effect of pixel response times more.
In a way, this is correct:

The closer that GtG gets to a frame duration cycle, the more visible GtG issues become again. Even 1ms GtG problems can indeed still be human-visible during ultra-high refresh rates.

Picture it this way, technologiclly, inaccuracies in GtG such as the last 10% of certain colors taking 4ms longer in a ghosting trail would still be 4 pixels worth of discoloration (such as a faint ghost or inverse ghost) per 1000 pixels/second under Blur Buster's Law of persistence. This is also easily seen when turning strobing on: ghosting turns into sharp strobe crosstalk (clear ghost afterimages that is often tinted very differently). The better a display can do with less ghosting or after-artifacts, the less strobe crosstalk during ULMB too.

Whatever GtG is doing, it superimposes (stacks) on top of display motion blur. A GtG curve smears across your vision as you track your eyes while the pixel is undergoing its GtG curve. At a given point, GtG+time, it's following Blur Busters Law in smearing that distorted instantaneous color across your vision. That includes GtG+0.00ms, GtG+0.01ms, GtG+0.02ms.......(all the way thru GtG completion time).....every single point on the GtG curve! Blur Busters Law (1ms persistence = 1 pixel motion blur per 1000 pixels/sec) is very simple and predictable when it comes to squarewave stuff (e.g. GtG 0ms or full-flash strobe backlights).

Throw a non-instant pixel transition curve (GtG) into it, what is simple Blur Busters Law might as well instead become complex calculus, needing to be applied to all of the infinite intermediate values along the GtG curve... That's why Blur Busters Law wasn't easy to witness on monitors until GtG became a tiny fraction of a refresh cycle (e.g. 1ms GtG on a 120Hz or 60Hz display) and people started noticing that fast GtG was no longer improving motion blur... Old 33ms LCDs of the 1990s which became 16ms. Then 10ms. 8ms 5ms 4ms 3ms 2ms 1ms... The motion blur stopped really improving once GtG became noticeably less than a refresh cycle. That's when MPRT begins to dominate (persistence of sample-and-hold) as the main motion blur limiting factor.

If you want to apply persistence (of Blur Busters Law) to a GtG curve, you have to blur the whole GtG curve -- every instantaneous color in the GtG curve is smeared, time-basis, tracked along the axis of your eye tracking motion (or camera tracking motion, in the case of pursuit camera). While you're traking your eyes, the pixel is executing its own GtG curve. Different photoreceptors on your retina is capturing a different part of the GtG curve as a result. Flaws in GtG shows up as ghosts and coronas that only show up on one side of a motion blur. That's why we can tell apart 2ms bad overdrive apart from 2ms good overdrive -- it's very clear in motion tests -- GtG curve changes are noticed as asymmetries in the motion blur. It's really complex how GtG inaccuracies can manifest itself. Mathematically it's simple to my brain and I can wrap my head around it easily, but most people can't... (I try to explain things in Plain English to the best of my abilities though! But sometimes it ends up becoming long posts like these.)

So eye-tracking is essentially revealing those GtG imperfections. If GtG is mostly perfect (even just 90%+ perfect) then it looks mostly pretty good to most eyes, most remnants (1%) that takes more than 10ms often is not noticed. Minor blur asymmetries even doesn't really bother me (as much as the general AMOUNT of blurriness is relatively low). But some people do notice motion blur asymmetries (aka ghosting/coronas) much moreso than others. Ghosting and coronas are visible to human eye, as seen in LCD Motion Artifact and LCD Overdrive Artifacts. So you see curve distortions as those asymmetric blur artifacts such as more ghosting/coronas on left edge or right edge, instead of symmetric motion blur (same blur at left/right edges of horizontally moving objects).

Image
This image demonstrates how amazingly human-visible the millisecond-timescale GtG artifacts are!

This is an animated-PNG (view in Chrome, Safari, FireFox) pursuit image of XB252Q at 240Hz, at different overdrive settings. The clearest looks much better than 144fps@144Hz to my eyes, but, there's still slight minor differences between left/right edge. Apparently, this bothers some people -- asymmetry in blur being a big bother. Who's me to judge such people -- some people are flicker sensitive, blur sensitive, stutter sensitive, tearing sensitive, or all the above. So I presume, there's people still extremely sensitive to slight asymmetries. I've seen way worse 144Hz monitors though...but I wouldn't disagree that a few good performing 144Hz monitors have less asymmetries in its blur at 144Hz, than the 144fps on 240Hz monitors. I haven't noticed. Many haven't. But, it's there. (Historically, many 120Hz/144Hz monitors didn't do *that* good a job at 60Hz -- e.g. more ghosting at. 60Hz than native-60Hz monitors -- so pretty much a similar thing is happening, all over again)

It seems, sometimes, apparently, blur asymmetry bothers people more than the blur itself. Ghosting/coronas versus motion blur is very hard to separate because they're relatively interwined with each other, but some people really pick the two out apart really clear as bell.

Ghosting/coronas is asymmetry away from simple perfect linear motion blur asymmetry
Ghosting/coronas/etc (discolorations, asymmetries and colorshifts in motion blur) also follow Blur Buster's Law of 1ms = 1 pixel of motion blurring per 1000 pixels/sec. A perfect motion blur is linear motion blur, like the one you can do in PhotoShop. But display motion blur is rarely a perfect linear motion blur -- that only happens with perfect 0ms GtG with instantaneous sample-and-hold frame transitions. Any deviation away from linear motion blur is an artifact like ghosting, coronas, etc.
open wrote:When you increse the sample and hold time, (like you would when your fps dips on a gsync monitor) it can make moving eye observations see more motion blur.
Or more ghosting. GtG response time with variable frame lengths means that ghosting can become more visible at specific frame durations (e.g. specific refresh rates). E.g. The overdrive may not be perfectly calibrated at a specific refresh rate like 173Hz or 159Hz or whatever you will -- for it to be perfect, the VRR's variable overdrive (Which must dynamically vary properly) to tune GtG in a custom way to minimize ghosting/corona artifacts during all frame duration times (all Hz in the VRR range).

In a very bad overdrive not tuned for variable refresh rate (e.g. early FreeSync monitors) ghosting can suddenly appears or disappears (after-images appear or disappears) at certain frame rates. This means that at certain frame rates, ghosting or corona after-images appear and disappear (or intensifies/fades) depending on what frame frame rate you are running at.

Far more common on older FreeSync monitors (even 144Hz FreeSync monitors) before FreeSync 2, it's present to a certian degree on many G-SYNC monitors (even 144Hz). The effect of ghosting changes is so subtle (often like a 10% intensity change in the duplicate after-image trail behind an image) but sometimes the ghosting changes are so dramatic and some people are truly sensitive to ghosting asymmetries.

Both yehaw and darzo are right to varying extents, so there's no reason for them to be putting each other down.

It's possible to have clearer (primary-images) but worse after-image-consistency. A human may see that type of motion as "wow, that's much better" and a different human may see that as "ugh that has worsened". That includes pursuit camera photographs. Like artwork, one person admires the UFO eyes, but a different person might be analyizing the symmetry of motion blur on the left edge and right edge. A third person says "I see it, but meh" and just admires how much less overall average blurring there is compared to their monitor. Or yadda, yadda. You know.

Some people are more motion-artifact sensitive than others -- uncannily 100x moreso than the next person. That "Faint 5% asymmetry" in linear motion blur, then, truly bothers the person. I'm not going to argue against such human specific sensitivities. I know some are flicker sensitive but don't care about tearing. Others are super stutter sensitive but don't care about motion blur. Yet others see discolorations pretty quickly (e.g. nonuniformity effects, ghosting, etc). In that sense, I'm seeing some correctness in both yehaw and darzo. It's not 100% on target, but I'm strongly suspecting that differing sensitivites is also at play here.
open wrote:So even though the 240hz gsync monitors have the best yet pixel response and lowest sample and hold times:
Yes, but that doesn't necessarily mean they do 144fps (during 240Hz G-SYNC) as clearly as a native true 144fps @ 144Hz non-GSYNC monitor. The degradation may be tiny, like 1% or 5% worse ghosting. Many of us do not notice, but it's often there -- and some humans are just remarkably sensitive to asymmetries in the linear motion blur (aka ghosting, coronas, etc).

This is unfortunately true for almost all variable refresh rate monitors -- this was true back in the 144 Hz G-SYNC and FreeSync days where 100fps during 144Hz VRR didn't look quite as good as 100fps @ native 100Hz (with well-adjusted overdrive).

There's been almost 10 years for 120Hz monitors to get really good overdrive, and some of them now have nearly perfect linear-looking motion blur during full framerate motion. And some people have gotten used to that. Overdrive for 240Hz monitors are still in its infancy and calibrations for 145fps-240fps frame rates cannot recycle 144Hz overdrive lookup tables (but many probably did it) without making certain kinds of ghosting artifacts reappear.

Variable-refresh-rate overdrive is extremely complicated, and is one of the reasons the first G-SYNC monitors used an FPGA. There's also an element of predictiveness required in VRR overdrive too, since it also now has to predict the future frametime, to attempt to come up with the 'perfect overdrive algorithm without ghosting and coronas' for a specific refresh cycle! Every. Single. Refresh. Cycle. (It also happens to be part of the G-SYNC premium too -- better-calibrated VRR overdrive than in most FreeSync monitors).

I think I now have an idea of how to test for VRR overdrive quality that no other sites has ever done before.

However, I have to say that I've seen 240Hz monitors be able to do 144Hz pretty well, but that's fixed-Hz 144Hz, and not 144fps@240Hz (which should in theory look exactly the same, but does not always). Another way to test how 144fps@240Hz looks like on a 240Hz monitor using fixed-Hz is to use large vertical totals (240Hz timings / horizontal scanrate but with 144Hz refresh rate) and such modes will have different ghosting effects than its default sync settings. This allows TestUFO capture of imperfect overdrive in these situations.

This thread has given me ideas.
open wrote:-There are times when the fast refresh rate can empasize the pixel response bluring. You might not be able to notice it at a lower refresh rate.

-There are also times when gsync fps dips can make eye tracking motion blur varry. Because the fps changes the sample and hold times and the sample and hold times determine eye tracking motion blur.

This is normal and you will still have the best pixel response and best potential sample and hold time on the 240hz.
Agreed, that's normal, too. In an ideal monitor, 144Hz or 144fps on a 240Hz monitor should look as good as 144fps@144Hz -- assuming darn near perfect overdrive creating the best possible symmetry in motion blur (and no discolorations).

But there are deviations from normal -- and that "normalish" to an extent even in 120Hz, 144Hz and 240Hz monitors. 144Hz VRR doesn't always do 75fps as ghost-free as a 75fps@75Hz fixed-Hz. But it's so subtle most people don't notice the difference. However, in 240Hz where we're starting to hit LCD GtG limitations sooner, there are some humans that may notice ghosting variability issues more, especially with GtG being a bigger percentage of frametimes. (See: human sensitivity differences)

What I do know is that ghosting/coronas do vary at different framerates on a VRR monitor. Much worse on early FreeSync monitors. This is not the only factor (fixed-Hz 240Hz can have different amounts of motion-blur-asymmetry artifacts (ghosting/coronas/etc) too on different models, most subtle, some worse) -- but variable refresh rate tends to amplify ghosting/corona consistency depending on the frame rate.
open wrote: If you are not noticeing what I am talking about then there is something wrong with your monitor and that is not normal for a 240hz.
Whoa, hold on. What one human notices, another doesn't. Not everyone is flicker sensitive. Not everyone is stutter sensitive. Not everyone is tearing sensitive. Not everone notices 3:2 pulldown. Same thing. Not everyone instantly sees motion blur asymmetries (ghosting/coronas).
open wrote:Best to speak clearly and not get the two things confused with eachother. If there is a problem then there is a problem.
Correct, it's important to speak clearly. Alas, when it comes to such subtleties like this, it gets horrendously difficult to speak clearly about it.

As with other "monitor things" (other than VRR, blur and ghosting), one person's nitpick is another person's eyestrain. This adage may be true in this thread.

If fighting resumes in this thread, I may have to close this thread, but to a certain extent, it seems darzo and yehaw may both simultaneously be right (in many "line items") even though this is all further muddied by inconsistencies between monitors.

For example, I also like UlMB, yet I acknowedge ULMB is eye-searing pain to some super-flicker-sensitive people. Eveyone is different until the perfect real-life display. My opinion remains unchanged: There are good 240Hz monitors out, people like them. But I also acknowledge there appears to be issues that bothers some (but not all) people.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

       To support Blur Busters:
       • Official List of Best Gaming Monitors
       • List of G-SYNC Monitors
       • List of FreeSync Monitors
       • List of Ultrawide Monitors

User avatar
BTRY B 529th FA BN
Posts: 155
Joined: 18 Dec 2013, 13:28

Re: 240hz displays are blurry

Post by BTRY B 529th FA BN » 28 Dec 2017, 20:15

TL:DR

So what was the verdict? Was something not configured right? Was it hardware? Considering one myself,

Thanks!

yehaw
Posts: 71
Joined: 21 Dec 2017, 21:41

Re: 240hz displays are blurry

Post by yehaw » 28 Dec 2017, 21:02

BTRY B 529th FA BN wrote:TL:DR

So what was the verdict? Was something not configured right? Was it hardware? Considering one myself,

Thanks!
There is no verdict, just opinions, unless someone who has proper testing methodology can give comparison results. Some say no issue, some say issues, some say bad panels and have to be swapped, etc. But for me, I think the ghosting test ufo images from review sources speak for themselves and matchup with my experience.

To me, it's clear these 240hz panels are flawed. You kind of have confirmation of it from Chief Blur Buster:
Overdrive for 240Hz monitors are still in its infancy and calibrations for 145fps-240fps frame rates cannot recycle 144Hz overdrive lookup tables (but many probably did it) without making certain kinds of ghosting artifacts reappear.
If you're thinking about spending $500 on a monitor, the least you can do is spend 15-20 minutes reading through this and making your own decision. In that post, I link to 2 other posts, I use sourced images from reviews that test ghosting to draw my conclusions. If you can follow along and see the flaws I point out, chances are you won't enjoy 240hz. If you still want to buy one, I suggest you order from a place that has a good return policy, so you won't have to pay $40 to send it back.

I say if you have current 144/165/180hz panel, just wait a while longer for more mature 240hz panels. It's just not worth $500 at the moment in my eyes. Once they improve the overdrive to be more sharp, precise, no artifacts, I'll gladly hand over the $500.

User avatar
BTRY B 529th FA BN
Posts: 155
Joined: 18 Dec 2013, 13:28

Re: 240hz displays are blurry

Post by BTRY B 529th FA BN » 28 Dec 2017, 22:09

Thanks for the reply. I do have a 144Hz first gen Asus panel, and was just thinking about what kind of difference would be gained.

Thanks for the reply, i'll look through the posts and see what's what,

Cheers!

User avatar
lexlazootin
Posts: 1251
Joined: 16 Dec 2014, 02:57

Re: 240hz displays are blurry

Post by lexlazootin » 28 Dec 2017, 23:32

240hz displays are fine. If you get one with G-Sync your over-drive is also going to be fine.

Although this is not at all the best way to check OD i recorded some 240hz samples on my Iphone with the SlowMo. It's pretty damn good and as good as my XL2420G (Benq's 144hz G-Sync)

https://streamable.com/6kwa6

Since the Iphone records at 239.~ you can see the overdrive shifting slightly as time passes, but that's just a artifact of 240hz on a 239.~hz camera. As you can see it's not nearing as bad as yehaw is making it out.

https://streamable.com/ce99g

I captured some In-game footage so people can see that it looks fine, i did play at 200fps because HL likes to play at 1000/n fps values (100/125/200/250) so it will look like it's shimmering but that's just a camera artifact.

Most of the confusion is just the way you can capture the screen, it's very easy to get what ever result you want.

Here's a image captured on my XL2420G

Image

And here's a image i took on my Acer 240hz with my iphone #WorstMonitorEver Kappa

Image

Most of what yehaw says is pretty flawed.
If you're thinking about spending $500 on a monitor, the least you can do is spend 15-20 minutes reading through this and making your own decision.
"your own decision" after recommending someone to read some cherry picked bad things about the product. :lol:

Locked