240 HZ FOR OVERWATCH, HELP ME DECIDE!

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!
User avatar
Chief Blur Buster
Site Admin
Posts: 11653
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: 240 HZ FOR OVERWATCH, HELP ME DECIDE!

Post by Chief Blur Buster » 18 Sep 2018, 13:52

Blur reduction (DyAc, ULMB) reduce brightness a lot. That's because a backlight is turned off between refresh cycles to hide LCD motion blur (hide GtG & also reduce persistence). So the backlight must be brighter to compensate during the strobe mode (DyAc/ULMB/etc).

The "DyAc" brand is brighter -- which means XL2546 has a brighter blur reduction mode than XL2540. Otherwise, the blur reduction experience is similar, it's just that you might find blur reduction to be too dim on certain monitors.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

Colt3d
Posts: 11
Joined: 13 Sep 2018, 16:53

Re: 240 HZ FOR OVERWATCH, HELP ME DECIDE!

Post by Colt3d » 18 Sep 2018, 13:59

Chief Blur Buster wrote:
The "DyAc" brand is brighter -- which means XL2546 has a brighter blur reduction mode than XL2540. Otherwise, the blur reduction experience is similar, it's just that you might find blur reduction to be too dim on certain monitors.
Ty for reply, I think brightness wont be an issue for me since Im used to playing in a strobed VG248QE at 10% lightboost, benefit from getting used to play dark is that no matter how long the game session I never have headaches or my eyes strained.

Chief, aside from brightness do you think other aspects of the strobing such as crosstalk is improved on the XL2546 model as well?

User avatar
Chief Blur Buster
Site Admin
Posts: 11653
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: 240 HZ FOR OVERWATCH, HELP ME DECIDE!

Post by Chief Blur Buster » 18 Sep 2018, 14:20

VG248QE at 10%? Wow.

That definitely clarifies my recommendation quite surgically, right there, then. That tells me motion clarity is your numero uno priority, above else.

That said, lots has happened in the last few years. Many monitors can do the equivalent of LightBoost 10% much more brightly now, with better colors. But not all of them. So I'll help you bypass the caveats.

Yes, you'll get much brighter with XL2546. But one thing you should know is that the BenQ-Zowie monitors is capable of shorter strobe lengths that are 4x clearer than LightBoost 10%. So if you want ultrashort 0.25ms strobe pulsing (0.25ms MPRT), which is 1/4th the brightness of the strobe length used by LightBoost 10% -- then you definitely want extra overkill brightness in order to get shorter persistence.

Unfortunately you will get more strobe crosstalk with 240Hz DyAc, simply due to the function of refresh cycles enroaching into LCD GtG speed, and both the VG248QE vs XL2546 is the same 1ms GtG. On the VG248QE, GtG is only 1/8th of a refresh cycle, but on the XL254X series, GtG is 1/4th of a refresh cycle.

Necessarily, this hits some laws-of-physics and amplifies strobe crosstalk quite significantly (that's why we need 0.5ms GtG pixel transition time and faster for clearer strobed 240Hz and future 480Hz/1000Hz monitors explained at http://www.blurbusters.com/1000hz-journey ...)

1ms GtG is often hyped as being unimportant, but when it comes to strobing, GtG is supremely important for strobe crosstalk.

Now, the good news is that there is a hacked 182Hz mode with major reductions in strobe crosstalk, at http://www.blurbusters.com/crosstalk ... The great thing is that both XL2540 and XL2546 can strobe at any refresh rate in 1Hz increments from 75Hz through 240Hz. Mathematically as a rule of thumb, the more Hertzroom, the less crosstalk. So 144Hz strobed on the 240Hz monitors will have much less strobe crosstalk than either 240Hz strobing on a 240Hz-max monitor, or 144Hz strobing on a 144Hz-max monitor.

If you are very leery of strobe crosstalk, and you love LightBoost 10% and wish for even clearer than LightBoost 10% (without being dimmer than LightBoost 10%), then you must do both:
1. You'll want to use a hacked 182Hz or custom large-Vertical-Total mode (see Appendix A) on the BenQ 240Hz monitor
2. You'll want to choose the XL2546 instead of XL2540 to get the extra brightness headroom to have extra range below "LightBoost 10%" that both XL2540 and XL2546 are capable of.
3. You might want to slightly reduce your dynamic range (slightly brighten blacks and slightly dim whites) to get the extra overdrive-overshoot headroom necessary to eliminate strobe crosstalk on brighter/darker scenes. LightBoost does this automatically (it's poorer colors) but you do have the option to do this manually if you want LightBoost-league ultralow strobe crosstalk.

As we know, MPRT is not the same as GtG
LightBoost 100% is 2ms MPRT
LightBoost 10% is 1ms MPRT
BenQ is cpable of 0.25ms MPRT at minimum "Persistence" setting (minimum Pulse Width setting).
MPRT = 1ms of MPRT translates to 1 pixel of motion blurring per 1000 pixels/sec motion.

So minimum-Persistence-setting BenQ is roughly 1/8th the motion blur of LightBoost 100% (excluding crosstalk) but eliminates 8x more lumens than LightBoost 100%. So you need super excess brightness headroom in order to gain sufficient brightness that can out-compete LightBoost 10%.

It is not as user-friendly as a ToastyX LightBoost hack, and you will get worse crosstalk at default strobing, so expect to do some tweaking trial-and-error with a bunch of HOWTOs, to create a simultaneously "low-crosstalk" AND "ultra-low-persistence" custom DyAc mode.

You can get better than LightBoost 10% but it just takes a lot more work / tweaking. The larger the Vertical Total you get, the less strobe crosstalk there is, so if you are unfamiliar with custom resolution tweaking, that is something you will need to learn how to do. It can be aggravating for some doing so much tweaking, since you have pros/cons. Most people prefer brighter & colorful, and so strobe crosstalk was a little less important with DyAc. But you can at least tweak-tweak-tweak-tweak (custom Hz + use Hz lower than max Hz + use custom large VT + dynamic range reduction + custom strobe phase) to get quite close to low-crosstalk LightBoost -- at least you have the tweakability though it can be a learning curve.

If you don't care about crosstalk, then DyAc is easy to turn on/off -- much easier to turn on/off than LightBoost, no software needed -- and many love its defaults. It's much brighter than LightBoost. Imagine your VG248QE with LightBoost turned off -- the XL2546 in strobed mode is as bright as your VG248QE in non-strobed mode!!!!!! (Some say the new XL2546 was brighter, at least compared to a worn-out VG248QE that has been run for several years). So all that extra delicious lumens overhead can be fully directed into "better MPRT than LightBoost 10%" if you wished. However, it has way more crosstalk, being optimized for brightness & colors, and ability to strobe at full 240Hz.

Tricks to reduce strobe crosstalk on post-LightBoost monitors
Generally, there are several tricks:
-- Intentionally add extra hertzroom (lower Hz than max Hz, gives more GtG headroom to avoid strobe crosstalk)
-- Intentionally slightly decrease dynamic range (brighten the blacks & slightly dim greys), 90-95% dynamic range is sufficient
-- Use largest blanking intervals if monitor supports large vertical totals (e.g. many BenQ Zowie monitors) since that hides GtG pixel transitions better between refresh cycles. Some monitors do this automatically (scan-velocity-acceleration to create large VBIs), LightBoost already does it, but sometimes you have to do it manually on the GPU side to get further improvements over monitor default strobe crosstalk.

In both cases, reduce your dynamic range slightly to fix any amplified strobe crosstalk caused by lack of overdrive overshoot headroom below-black and above-white. It helps roughly ~75% of center-screen strobe crosstalk disappear if I do that. It'll still be better colors than LightBoost, but worse colors than non-strobed, so it's a pill you have to swallow trading contrast-ratio for strobe crosstalk -- but at least you have a choice with newer monitors (manually adjusting contrast) rather than permanent LightBoost unadjustable low-contrast.

Intentional dynamic-range reduction to improve crosstalk downwards to similar to LightBoost 10%. You simply play with Contrast/Brightness while watching http://www.testufo.com/crosstalk .... It's normal to have more crosstalk at top/bottom edges with optimally adjusted strobe phase, but your center-screen crosstalk should mostly be gone.

If tweaking is too scary, I recommend a 240Hz GSYNC monitor and using ULMB instead which is fairly low strobe crosstalk out of box (a bit worse than LightBoost, but is much brighter/colorful). Fortunately you can remove most of ULMB crosstalk via intentional dynamic-range reduction (slightly reduced contrast), at least you have the choice of that unlike old LightBoost. Yet ULMB also has an optional Pulse Width adjustment capable of going to as low as ~0.25ms too (at minimum setting) -- one-quarter the MPRT of LightBoost 10%.

Personally MPRT 0.5ms versus MPRT 1.0ms is still visible to my human eyes during TestUFO Panning Map Test 3000 Pixels/sec. The street map labels are slightly blurry at 1ms MPRT (LightBoost 10%) but clear at 0.5ms MPRT. They're perfectly pixel-sharp on ULMB/DyAc minimum pulse setting on the newer adjustable-persistence monitors.

240Hz monitor recommendations for "LightBoost 10%" lovers

If your god is LightBoost-10%, and want to beat LightBoost 10% (even less motion blur) -- then my personal recommendation for you is either:

1. Ease of use -- Go for any 240Hz GSYNC monitor and use 144Hz ULMB at low "ULMB Pulse Width" settings.
2. Flexible & Tweakable -- Go for BenQ XL2546 240Hz DyAc monitor and use either 144Hz or 182Hz DyAc with large-VT tweak + lower persistence settings. On a budget, the XL2540 is an option if you don't need to go better than 10% LightBoost.

To support Blur Busters research, please feel welcome to purchase monitors through the Blur Busters monitors list -- if you want. Much appreciated :)
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

Colt3d
Posts: 11
Joined: 13 Sep 2018, 16:53

Re: 240 HZ FOR OVERWATCH, HELP ME DECIDE!

Post by Colt3d » 18 Sep 2018, 15:32

Chief Blur Buster wrote: Intentional dynamic-range reduction to improve crosstalk downwards to similar to LightBoost 10%. You simply play with Contrast/Brightness while watching http://www.testufo.com/crosstalk .... It's normal to have more crosstalk at top/bottom edges with optimally adjusted strobe phase, but your center-screen crosstalk should mostly be gone.
ohh I see some crosstalk in the top/bottom edges of my current monitor indeed, center crosstalk is almost null and martians are sharp and clear as it gets :D one single ship with 3 little separate white squares on every section in the bottom.

Chief Blur Buster wrote:If tweaking is too scary
Not at all, time to get hands dirty.. and in distress who are we gonna call? BLUR BUSTERS!! xD

Chief Blur Buster wrote: To support Blur Busters research, please feel welcome to purchase monitors through the Blur Busters monitors list -- if you want. Much appreciated :)
This was going to be my next question, how can I support you guys? without this website I would have never gotten my current 120hz VG248QE monitor nor know that strobing even existed and would have missed all the benefits of blur reduction. This website guided me in the past to do the satisfiying jump from 60 hz to 120 hz and Im looking forward for the 240hz jump and further.

I appreciate all the research and tips you have given me, Ill def make sure to use that link, thank you Chief

User avatar
k2viper
Posts: 293
Joined: 23 Jan 2018, 06:30

Re: 240 HZ FOR OVERWATCH, HELP ME DECIDE!

Post by k2viper » 19 Sep 2018, 02:39

Btw there is at least one "newer" model around, ASUS XG248Q, it uses kinda new 24" panel, previous models were mostly using older 24,5" panel like one in XL2540 and XL2546.
XG248Q also supports strobing (ASUS ELMB) at full 240hz. https://www.youtube.com/watch?v=5vj1HcsHNX8

User avatar
fenderjaguar
Posts: 168
Joined: 19 Apr 2014, 11:57

Re: 240 HZ FOR OVERWATCH, HELP ME DECIDE!

Post by fenderjaguar » 19 Sep 2018, 08:07

Chief Blur Buster wrote:VG248QE at 10%? Wow.
I used to use the VG248QE at 0% brightness in lightboost mode on the desktop at all times. That's the only reason I used to use it, because it offered a lower brightness than without lightboost (never used it in games)

Colt3d
Posts: 11
Joined: 13 Sep 2018, 16:53

Re: 240 HZ FOR OVERWATCH, HELP ME DECIDE!

Post by Colt3d » 19 Sep 2018, 14:24

k2viper wrote:Btw there is at least one "newer" model around, ASUS XG248Q, it uses kinda new 24" panel, previous models were mostly using older 24,5" panel like one in XL2540 and XL2546.
XG248Q also supports strobing (ASUS ELMB) at full 240hz. https://www.youtube.com/watch?v=5vj1HcsHNX8
Wow didnt know that monitor existed and at a very affordable price too , gotta research into what it has to offer vs the XL2546 :roll:

Colt3d
Posts: 11
Joined: 13 Sep 2018, 16:53

Re: 240 HZ FOR OVERWATCH, HELP ME DECIDE!

Post by Colt3d » 25 Sep 2018, 15:16

Hey Chief, I pulled the trigger on purchasing the XL2546 and after like 8 hours of tweaking this is what I could achieve:


Note: I took this pic with my cellphone at 1/250 exposure (monitor refresh rate is 230hz), please let me know if theres a better way to capture strobed pics with a phone

Image

I used this custom resolution created with the Custom Resolution Utility by ToastyX:
Image

While motion blur was non existant right from the box and default settings, I was having a lot of issues in reducing the crosstalk which at 240 hz was much much worse than current pics, I was experiencing this sharp and ugly negative crosstalk/corona. One funny thing I noted is that using the lower hz refresh settings from windows defaults , ie going from 240 hz to 120 or 60 made the crosstalk EVEN WORSE which was weird considering the info I have been reading here.

So I proceeded to install the CRU Utility from ToastyX and followed the steps for increasing the vertical totals at different refresh rates (240, 220, 200, 120) but realized simply increasing the vertical totals and keeping max pixel clock (594 Mhz) wasnt doing anything to my crosstalk issue, the tests I did looked something like this:

Image


so I looked further for other things to tweak and saw the 182 hz numbers posted in this website which are these:

Image

One thing I quickly noticed about the 182 hz numbers posted on this website is that they not only messed with the VT, they also changed the front porch, sync width, back porch, blanking and sync polarity settings (vertical polarity is set to negative), all these result in a higher pixel clock that I was using too 599.64mhz instead of the (594 Mhz) I was using-


Triying out all these numbers at 182 hz worked better to reduce the crosstalk, (what is this undocumented witchcraft?) and after entering the factory menu in my XL2546 (turning monitor off, holding button 4 and 5 at the same time and turn it on, then pressing button 5) to tweak the "overdrive gain" to 38 which seems to me its just the AMA setting but tweakable in number increases instead of just high/premium, AMA/Overdrive gain seemed to have a big effect in the crosstalk/corona I was experiencing, so I ended up setting it into a somewhat medium/low value which seemed to be the 38.


Then I tweaked all the color modes and settings to this:

Dyac: Premium
Black Equalizer: 2
Color Vibrance: 10
Low Bluelight: 1
Instant mode: ON
Picture mode: Gamer 2
Brightness: 35
Contrast: 20
Sharpness: 9
Gamma: 1
Color Temperature: Normal
AMA: Premium
OD Gain: 38 *factory hidden setting (turn monitor off, hold button 4 and 5 at the same time and turn it on, then press button 5)*

Then I used Blur busters strobe utility to configure these settings:
Persistence/Intensity: 15
Crosstalk Area: 13

With ALL this the results finally started to look smooth and acceptable, and Finally but not least I decided to try using the 182 hz dark magic numbers to try and achieve a 230 hz refresh rate, I tried this out mostly because I felt 182 hz is not the refresh rate I paid for with this monitor and in principle compromising any input lagg and responsiveness kind of defeated the purpose on ever getting this monitor in the first place, so I tried out these numbers:

Image


And it seemed to do the job, somehow the crosstalk is the exact same as with 182 hz and the pixel clock is still higher than what I was using in previous tests 599.64Mhz instead of 594Mhz.

All in all my current XL2546 crosstalk with 230hz and 182hz is still slightly worse than my 120 hz VG248QE with lightboost 10%, so Id be interested to know what else I could improve in my current config, these are just my best results Ive gotten so far, I also still need to find a way to fix text bleeding or ghosting (specially in a browser) when I scroll up and down slowly (if I spin the wheel fast its not noticeable), this bleed is also noticeable in the BLUR busters vertical scrolling text test btw.

User avatar
Chief Blur Buster
Site Admin
Posts: 11653
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: 240 HZ FOR OVERWATCH, HELP ME DECIDE!

Post by Chief Blur Buster » 25 Sep 2018, 18:47

Colt3d, I appreciate your follow up!
You can't say I didn't warn you it would be VERY time consuming at first. ;)
Brightness: 35
Persistence/Intensity: 15
Why don't you increase brightness in order to reduce Persistence?
Same brightness, clearer motion -- they cancel each other out.

Importantly, when you adjust persistence/intensity up/down -- do you see differences in the 3000 pixels/sec TestUFO Panning Map street-name-label readability test?
ie going from 240 hz to 120 or 60 made the crosstalk EVEN WORSE which was weird considering the info I have been reading here.
Did you increase VT when going to 120Hz?
You need much bigger VTs to fix th strobe crosstalk.
VT2000 has less strobe crosstalk than VT1500
VT1500 has less strobe crosstalk than VT1250

Some 240Hz monitors can accept up to ~VT4500(ish) at 60Hz using the Appendix A trick at http://www.blurbusters.com/crosstalk
However, your report is another report that the XL2546 seems to behave a little counter-intuitively with crosstalk behaviour. The crosstalk increase-decrease effect is extremely clear on a XL2720Z with large vertical totals.
Colt3d wrote:I also still need to find a way to fix text bleeding or ghosting (specially in a browser) when I scroll up and down slowly (if I spin the wheel fast its not noticeable), this bleed is also noticeable in the BLUR busters vertical scrolling text test btw.
You can fix most of it by tightening your dynamic range. Dim your whites, brighten your blacks.
(Not via backlight, but via LCD pixel-driving GtG voltage aka RGB color ranges)

Reduce your colorimeter contrast ratio by about 10-20% if you have a Spyder or i1 device. Essentially, on the cable, you want to use digital RGB colorrange 10-245 or 5-250 rather than 0-255 because the intensities at the end of the greyscale range are the ones that has the nastiest overdrive artifacts. Another trick is the HDTV colorrange setting in some drivers; "16-235" but that's overkill, you don't really need to reduce dynamic range THAT much.

If your monitor controls are not letting you do this (e.g. it adjusts backlight) then you can use either NVIDIA Control Panel, AMD Catalyst Control Center to tighten your greyscale. Brightness controls the backlight, it's okay to go 100% brightness. What you want is to adjust stuff that doesn't touch the backlight and instead touches the voltage range sent to the panel (aka RGB color values) -- so adjust "Contrast" instead and other adjustments.

Failing that, just calibrate monitor adjustments to brighest, and instread use NVIDIA/AMD adjustments to reduce your dynamic ranges to avoid the overdrive from slamming against fullblacks/fullwhites. Avoid doing that, and more than 50% of your ghosting/corona artifacts are gone (if you're using the middle AMA setting), and strobe crosstalk goes down quite significantly. Avoid the endrange greyscale and you've got much, much less strobe crosstalk and ghosting.

This is why LightBoost had poor quality colors -- it also intentionally tightened the dynamic range in order to avoid the problematic overdrive color values (the fullblacks/fullwhites) because there's no overdrive overshoot headroom below black or above white with full dynamic range; so artifacts are amplified with the min/max color values.

At least you have the continuum of choice of keeping full dynamic range (but more strobe crosstalk) or reduced dynamic range (but reduced strobe crosstalk) with a continuum in between -- but it does take a learning curve to understand the relationship between strobe crosstalk and contrast ratio.

If you already did this, then great. But if you haven't tested this (successfully), this is another thing to try!
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

Colt3d
Posts: 11
Joined: 13 Sep 2018, 16:53

Re: 240 HZ FOR OVERWATCH, HELP ME DECIDE!

Post by Colt3d » 25 Sep 2018, 22:05

Why don't you increase brightness in order to reduce Persistence?
Same brightness, clearer motion -- they cancel each other out.
Ohh thx good catch, I now increased brightness from 35 to 90 and increased Intensity/persistance from 15 to 22 and overall picture seems about the same , I didnt quite understand well the relationship between brightness in the monitor menu and persistence in the blur buster strobe utility (they both seemed to do the exact same thing image wise) so figuring what values to use on each was not that intuitive.
Importantly, when you adjust persistence/intensity up/down -- do you see differences in the 3000 pixels/sec TestUFO Panning Map street-name-label readability test?
Adjusting persistence/Intensity up and down on this test unfortunately doesnt make any difference in the readability of the text it just makes the image brighter/darker and the readibility at the Intensity value Im using now (22) is still very poor since there seems to be some sort of crosstalk or 2 layers of the same texts phased with each other which makes it akward to read. Is this important?

Did you increase VT when going to 120Hz?
You need much bigger VTs to fix th strobe crosstalk.
VT2000 has less strobe crosstalk than VT1500
VT1500 has less strobe crosstalk than VT1250
When first downscaling the Hz refresh rate from out of the box 240 hz to 120 hz I first tried it with the windows default settings for 1920x1080 @120 hz which resulted in a much more ugly crosstalk than out of the box 240 hz crosstalk, then I tried a Custom resolution for 121 hz with 2230 Vertical totals (Increased from 1125) but the crosstalk didnt seem to improve that much versus the windows 120hz default, whenever I increased the hz rate in the windows default settings the crosstalk seemed to improve for some reason, I focused more on the custom resolution utility on 182 hz after that point.

Any reason for this happening? maybe the way I was increasing vertical totals was not entirely correct since in the suggested 182 hz configuration you guys used in Toasty's custom resolution utility you changed the front porch, sync width, back porch, blanking and sync polarity settings and vertical polarity that is set to negative instead of the default positive, I wasnt touching any of that except vertical totals during my first tests and I wasnt making any progress until I found this 182hz config.
You can fix most of it by tightening your dynamic range. Dim your whites, brighten your blacks.
(Not via backlight, but via LCD pixel-driving GtG voltage aka RGB color ranges)

Reduce your colorimeter contrast ratio by about 10-20% if you have a Spyder or i1 device. Essentially, on the cable, you want to use digital RGB colorrange 10-245 or 5-250 rather than 0-255 because the intensities at the end of the greyscale range are the ones that has the nastiest overdrive artifacts.
Any more specific tip on how to find that setting on the XL2546 monitor? other than color temperature in picture settings (which shows as options: normal, blueish, redish and user define with RGB values from 0 to 100) I havent seen anywhere how to adjust the settings youre talking about, when you said dynamic range and dimming whites/brightening blacks I thought you just meant tweaking brightness/contrast and strobe intensity values but it seems this setting youre talking about is much more color based.

Im not familiar on how to do this via the NVIDIA panel either.
Last edited by Colt3d on 26 Sep 2018, 21:21, edited 1 time in total.

Post Reply