Is ELMB-Sync Really That Bad? [Reviewer Consistency? / Model Differences?]

Ask about motion blur reduction in gaming monitors. Includes ULMB (Ultra Low Motion Blur), NVIDIA LightBoost, ASUS ELMB, BenQ/Zowie DyAc, Turbo240, ToastyX Strobelight, etc.
User avatar
Chief Blur Buster
Site Admin
Posts: 8832
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Is ELMB-Sync Really That Bad? [Reviewer Consistency? / Model Differences?]

Post by Chief Blur Buster » 15 Mar 2020, 04:59

Easy!

Both sides are getting a bit hot on this. Both of you, both sides, please re-read the forum rules on the "Be Nice To New Users" rule, even to users you disagree with or to users who are wrong.

I'm not going to close this otherwise useful thread. But I'm going to change the subject. This thread a "Consistency of Reviewer Tests" type of thread, and the thread stays open. I understand some of the reasons of the inconsistencies between reviewers.

From what I've seen of ELMB-SYNC on 1ms panels (TN or IPS), the use of ELMB-SYNC already looks reasonably good at max-Hz because it doesn't start multi-strobing. The strobe crosstalk in Hardware Canucks is roughly similiar to what I saw with my eyes in person. There's a bit of error margin going on with the tests, but it's reasonably small enough that it's representative of what I see go on in ELMB-SYNC at fps-Hz situation. So stop arguing about whether those YouTubes are authentic; they are. They're not exactly following the standard pursuit camera instructions (I'll have to ask them for more info), but as long as they're using 360-degree-shutter (no closed-shutter moments) during video pursuit camera tests, it's passable. I will contact them to make sure they're doing that, or to switch to long-exposure multi-refresh pursuit camera tests -- so that the ladder looks more accurate. But yes, I will talk to them so they're more experienced with pursuit camera -- I didn't even know they were using my free pursuit camera invention until recently --

The thing is that ELMB-SYNC (at least some implementations) go multi-strobe at framerates below Hz. But the fps=Hz strobe of ELMB-SYNC for screen middle is tolerable at least during 1ms panels.

What I am REALLY interested in, is how ELMB-SYNC behaves at fps below Hz on some of the newer ASUS models. Now, you launch CS:GO with ELMB-SYNC, it actually looks reasonably decent because it's always capping out at fps=Hz. But as soon as framerate fluctates, the sudden multistrobing effects starts to happen... More tests are needed for framerates below Hz.

As for the photos of ELMB-SYNC in this thread, the fake/not-fake should not be argued on (because they're real photos, I looked at them closely -- but that doesn't answer if they're perfectly identical to what was seen in real life), but "What are the camera parameters? And is he camera parameters accurate?".

The topic of reviewer consistency is a fascinating subject to me. But please, dial down the heat in this thread a bit, and follow the "Be Nice To New Users" rule (even those you disagree with / even those who are wrong). Even if this means this topic needs to get redirected to the reviewer-consistency discussion. I've changed the topic title to reflect this.
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

Dmoney405
Posts: 47
Joined: 01 Mar 2020, 01:14

Re: Is ELMB-Sync Really That Bad? [Reviewer Consistency? / Model Differences?]

Post by Dmoney405 » 15 Mar 2020, 12:37

Chief Blur Buster wrote:
15 Mar 2020, 04:59
Easy!

Both sides are getting a bit hot on this. Both of you, both sides, please re-read the forum rules on the "Be Nice To New Users" rule, even to users you disagree with or to users who are wrong.

I'm not going to close this otherwise useful thread. But I'm going to change the subject. This thread a "Consistency of Reviewer Tests" type of thread, and the thread stays open. I understand some of the reasons of the inconsistencies between reviewers.

From what I've seen of ELMB-SYNC on 1ms panels (TN or IPS), the use of ELMB-SYNC already looks reasonably good at max-Hz because it doesn't start multi-strobing. The strobe crosstalk in Hardware Canucks is roughly similiar to what I saw with my eyes in person. There's a bit of error margin going on with the tests, but it's reasonably small enough that it's representative of what I see go on in ELMB-SYNC at fps-Hz situation. So stop arguing about whether those YouTubes are authentic; they are. They're not exactly following the standard pursuit camera instructions (I'll have to ask them for more info), but as long as they're using 360-degree-shutter (no closed-shutter moments) during video pursuit camera tests, it's passable. I will contact them to make sure they're doing that, or to switch to long-exposure multi-refresh pursuit camera tests -- so that the ladder looks more accurate. But yes, I will talk to them so they're more experienced with pursuit camera -- I didn't even know they were using my free pursuit camera invention until recently --

The thing is that ELMB-SYNC (at least some implementations) go multi-strobe at framerates below Hz. But the fps=Hz strobe of ELMB-SYNC for screen middle is tolerable at least during 1ms panels.

What I am REALLY interested in, is how ELMB-SYNC behaves at fps below Hz on some of the newer ASUS models. Now, you launch CS:GO with ELMB-SYNC, it actually looks reasonably decent because it's always capping out at fps=Hz. But as soon as framerate fluctates, the sudden multistrobing effects starts to happen... More tests are needed for framerates below Hz.

As for the photos of ELMB-SYNC in this thread, the fake/not-fake should not be argued on (because they're real photos, I looked at them closely -- but that doesn't answer if they're perfectly identical to what was seen in real life), but "What are the camera parameters? And is he camera parameters accurate?".

The topic of reviewer consistency is a fascinating subject to me. But please, dial down the heat in this thread a bit, and follow the "Be Nice To New Users" rule (even those you disagree with / even those who are wrong). Even if this means this topic needs to get redirected to the reviewer-consistency discussion. I've changed the topic title to reflect this.
Yeah fake was the wrong word, I was using faked as perhaps intentionally (or unintentionally due to lack of experience) showcasing a comparison that doesn't represent both sides accurately. Similar to how they take normal photos and dial the digital vibrance up 10x and say this is the difference you will see with HDR. I didn't mean faked as they were created in different software or the like. I came to this forums to escape things like that, this should be a place about hard data points and unbiased information sharing so I got a bit carried away with this guy....when people intentionally ignore important data to support an agenda is just a big pet peeve of mine, I apologize.

I support the theory of requiring max FPS, I even stated that in my original infodump of this monitor. I believe I said anything under 165hz is unusable, but even 165-170 still doesn't look great. The middle 3rd is okay, still had crosstalk but it's mild enough to improve the clarity over using normal overdrive. The top and bottom areas lose to it though. Overdrive has quite the advantage in those areas.

On a side note I'm not sure information from Asus can be entirely trusted when they say the new 280hz will be better because I found some fascinating conflictions when looking up the panel for my monitor, the xg279q. It uses the M270DAN06.6 panel which is rated for 165hz native, yet Asus claims its a 144hz and allows you to 'overclock' it to 170.... So since it's 165hz panel their OC feature only gets you 5hz. Why would they claim this? Not only that, they claim that the screen has 95% DCI-P3 coverage but the box it comes in only states 90% DCI-P3 coverage. Something is going on at Asus and this is kinda making me question any other claims they may have especially if they are claiming the new 280hz are going to be better.

Dmoney405
Posts: 47
Joined: 01 Mar 2020, 01:14

Re: Is ELMB-Sync Really That Bad? [Reviewer Consistency? / Model Differences?]

Post by Dmoney405 » 15 Mar 2020, 13:24

Also I thought of a potential good idea to help streamline these tests. First off we can already instantly tell they were testing crosstalk with the ghosting lanes instead of the actual crosstalk test which is incredibly biased by only showcasing the best case scenario (middle of screen). We know this because their tests have the pursuit pixels, while the crosstalk test doesn't. So while that is useful to know it still doesn't show us what part of the screen is shown. If each lane had a number associated with them in the bottom corner of each block it would help to validate their claims.

Going further would be to only show these block numbers should only show up if the app is in full screen because in windowed mode they can simply scroll down the page and put the lane of their choice in the best part of the screen. I do not know if it's possible to make it full screen only but it would help a lot in the verification process.

User avatar
Chief Blur Buster
Site Admin
Posts: 8832
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Is ELMB-Sync Really That Bad? [Reviewer Consistency? / Model Differences?]

Post by Chief Blur Buster » 15 Mar 2020, 15:00

One thing I really need to create is a Version 2 of the really old Crosstalk FAQ at www.blurbusters.com/crosstalk to at least educate people on the behaviour of strobe crosstalk (GtG leaking over 2 refresh cycles) versus multi-strobing (strobe backlight intentionally flashing more than once contiguously per Hz).

We're the ones that are the most experienced with strobe backlights, so we've got a public-service responsibility to create more articles about this in this year -- and I shall create quite a few in the coming months. (On that, keep tuned!)
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

permaximum
Posts: 4
Joined: 13 Mar 2020, 19:58

Re: Is ELMB-Sync Really That Bad? [Reviewer Consistency? / Model Differences?]

Post by permaximum » 16 Mar 2020, 02:37

According to "that guy":

Hardware Canucks: Inexperienced paid testers that post fake (well not fake but...) photos.
Techspot: Reviewers who make critical errors that completely breaks the whole point of reviewing. They mislabeled the comparison photo.
Rtings: No comment.
The Site Admin who came up with these monitor test tools and says the crosstalk in the reviews are similar to what he saw with his own eyes: No comment.
Me: Have an agenda.

No problem. Techspot and Rtings may not sound reputable enough for some. Let me come up with another reputable reviewer than. This time it's Prad.de. I hope it's reputable enough. And this time it's a different model that has ELMB-Sync. XG279Q. That guy's own monitor.

PRAD.DE:
With ELMB-Sync it has now been possible to adapt the strobing of the backlight to the variable refresh rate. The result is really worth seeing.
ELMB-Sync Off
Image

ELMB-Sync On
Image
The picture certainly darkens somewhat, but is still bright enough with a luminance of around 185 cd / m² . Overdrive is grayed out when ELMB sync is activated and the strength cannot be adjusted. Although there is a slight afterglow on moving objects, we did not find it particularly disturbing. On the following two pictures you can see very well how the picture changes in terms of motion blur in a positive way.

However, the effect is still achieved by the flickering backlight, which leads to faster fatigue and headaches for many people. Apart from this problem, those who love fast shooter games will be particularly pleased, as the reduced motion blur makes aiming easier.
BTW it's sad that this one has such amount of input lag compared to VG27AQ according to Prad. It having a worse implementation of ELMB-Sync on top of the bad input lag while being more expensive would make it a bad choice to buy.

Last but not least, this is the bonus from Hardware Canucks' VG27AQ review. I don't think those trees or transmission tower are on the middle of the screen.

Image

There's certainly a reviewer inconsistency between Prad, Techspot, Rtings, Hardware Canucks, Chief Blue Buster and that guy called Dmoney405 along with a few other forum members, a noname reviewer or two. At least the two sides are consistent between themselves. Prad, Techspot, Rtings, Hardware Canucks, Chief Blue Buster pretty much agrees on what they see with the tech. On the otherside Dmoney405 and some forum member 708 and another one also agrees on their view on the tech.

I was just trying replace my old monitor with a fast-paced gaming monitor and I didn't want to make a bad purchase by going with something like XG279Q which has too much input lag because I'm too sensitive to input lag. I also hate motion blur and at some point I got particularly interested in the possibility of BFI and sync at the same time. It just happens to be Asus is the only brand that releases monitors which allow users to open a BFI mode and a sync mode at the same time, as of this moment. That's why I'm talking about Asus. The brand is not important and no one here tries to promote the brand.

Dmoney405
Posts: 47
Joined: 01 Mar 2020, 01:14

Re: Is ELMB-Sync Really That Bad? [Reviewer Consistency? / Model Differences?]

Post by Dmoney405 » 16 Mar 2020, 12:49

permaximum wrote:
16 Mar 2020, 02:37
Permaximum continuing to ramble on and using screenshots with no data behind them
Listen dude we were already warned about making this personal, continuing to do so is childish.

Again stop putting words into Mark's mouth. He said that the ELMB representation that is shown by them seems to be what he has personally witnessed. I AGREE WITH THAT, IN THE MIDDLE OF THE SCREEN IT IS A GOOD TECHNOLOGY AND LOOKS GREAT. I even stated that in my infodump of this tech on my post weeks before you escaped reddit and poisoned these forums with subjective and untested proofs. Then I again stated in it my first response to you. Then I stated it multiple more times in this same topic. I never denied the "ELMB-ON" photos. They look very similar to what I see as well, and I have stated that multiple times. My argument was that >>>>>>>>>>>They are specifically ONLY showing the best part of the screen, INCLUDING the new picture you showed with the cars, which can mislead people to an expectation that is unrealistic<<<<<<<<<<<<< The camera was inches from the screen with that picture taken (you can literally see the PIXELS in the photo proving this) and you see the top of the buildings but that is probably only inches away from the crosshair location. These are 27" monitors that have a vertical viewing area of 13.25 inches. If the butterzone is 1/3 of the screen then you have a 4.41 inch area that will look vastly better than the others. Since the camera is so close you can see the pixes on a 109 PPI screen we know it is not an accurate representation of the entire scene. If we do not see the bezel of the monitor in question then we have nothing to relate the screen capture to. Since in racing games the car you are in control of is essentially a pivoting static image anchored to the very bottom of the screen then "ELMB on" and "ELMB off" would make no difference. (as seen in your supplied picture, both cars have the same level of clarity). This means that the buildings above the car would land smack dab in the middle of the screen! So, once again, you are basing judgement on the best case scenario.

My entire argument has been;

1. We have no data behind the "ELMB-OFF" photos. -What trace free setting if any? -What recording hardware and techniques used? -How fast was the in-game camera panned? -How bright was the screen because lowering the brightness can introduce PWM artifacting on some monitors.

2. We can already tell by the tests that all lanes used for testing were on the GHOSTING test and not the CROSSTALK test. This is a fact, and is easily proved. This is important because the ghosting test only has lanes in the middle of the screen, which just so happens to be the best case scenario with this technology. Yes their pictures match mine in the middle of the screen. If the entire screen looked like the middle we wouldn't be having this conversation in the first place since every single gamer in the world would be forced to use it because it would be an amazing and unbeatable gaming experience. This is NOT the case and there are very real tradeoffs that you need to personally be okay with. Tricking people, intentionally or unintentionally, by only showing the best part is what I am fighting against.

You have openly admitted to your refusal to look at actual data and base your entire argument on a couple snapshots. Then you try to renew your theory with more photos from the center of the screen with no data behind them! This is your quote;
I don't care about how the tech happens and what it does behind the scenes. I don't even bother checking those graphs
So honestly just kick rocks dude.
Last edited by Dmoney405 on 16 Mar 2020, 13:58, edited 2 times in total.

Dmoney405
Posts: 47
Joined: 01 Mar 2020, 01:14

Re: Is ELMB-Sync Really That Bad? [Reviewer Consistency? / Model Differences?]

Post by Dmoney405 » 16 Mar 2020, 13:38

permaximum wrote:
16 Mar 2020, 02:37
BTW it's sad that this one has such amount of input lag compared to VG27AQ according to Prad. It having a worse implementation of ELMB-Sync on top of the bad input lag while being more expensive would make it a bad choice to buy.
If you are trying to somehow take a shot at me by saying that the xg279q is a bad buy you are not doing anything that I haven't already done myself. I specifically stated in my review that the monitor was a rip off, but not for the reasons you state it. I have no personal connections with this screen and the only reason I haven't sent it back yet is because there are still no gl850s in stock.

But I will, however, defend poor information, which you seem to be full of.

1. Prad.de never once mentions the XG279Q having a worse implemintation of ELMB-Sync than others in the ELMB section. If it is sourced elsewhere please link it because that would actually be valuable data on diagnosing ELMB issues.

2. Prad.de does not even have a review for the VG27AQ. So best case scenario you are comparing Prad.de's input data on the XG279q to some other reviewers input data for the VG27AQ. You should never do this, people have different testing methods that can possibly result in different scores for the same test.

3. Prad.de input delay data was based off of the XG279Q in 144hz mode. The panel is a native 165hz screen that is OSD overclockable to 170. There is a lot of good data on this site that you can read that shows input lag is typically worse on a high hz screen when set to a lower hz, than a screen designed to do that lower hz natively. (there may not be pictures for this though, maybe have a friend read it to you).

So as you can see this is why I am so critical to reviewers. Within 10 seconds on that page I was able to find a massive discrepancy in their testing, which renders it void. Reviewers are people. People make mistakes. Stop drooling over the pictures and READ THE DAMN DATA!!

User avatar
Chief Blur Buster
Site Admin
Posts: 8832
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Is ELMB-Sync Really That Bad? [Reviewer Consistency? / Model Differences?]

Post by Chief Blur Buster » 16 Mar 2020, 16:14

Thread closed for now.

Appreciate certain discourse, like analysis of reviewer inconsistency, but the frustration levels is a smidge high.
Frustration at reviewers should not be directed at each other.

In the future, frustration at reviewers should be tempered into a productive critique whereupon I can play a role in properly educating reviewers. I will contact HardwareCanucks and correct them on their pursuit camera technique to use the multi-exposure method (solid ladder per frame, typically 1/30sec exposure).

There has been useful information this thread, which I will definitely be acting upon (correcting HardwareCanucks' pursuit camera procedure).

Steam Achievement Unlocked: First thread closed on Blur Busters Forums in more than 1 year
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
Chief Blur Buster
Site Admin
Posts: 8832
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Is ELMB-Sync Really That Bad? [Reviewer Consistency? / Model Differences?]

Post by Chief Blur Buster » 16 Mar 2020, 17:04

Reviewers Reading This Thread
To readers, please feel free to email my tweet to reviewers: twitter.com/BlurBusters/status/1239669830300717060

Two reviewers are making mistakes with some aberrations away from test guidelines -- that cause me to be unable to trust the integrity of the pursuit camera photography, because the tickmarks are the evidence of scientific accuracy.

When photographing www.testufo.com/ghosting or other pursuit-camera compatible test pattern,
Please Follow Guidelines
Please Confirm Scientific Integrity

This tweet tagged 10 reviewers, most of whom is doing it correctly, but some simply needs process refinement:
Image

Image

Image

Image

Image

[!] CRITICALLY IMPORTANT FOR TRUST [!]: Just because photo (or individual video frame) looks correct despite bad tickmarks, doesn't mean the photo can be trusted. The tickmarks is the trust. The tickmarks is hardcoded evidence of pursuit camera accuracy. Please help us trust the photograph by showing accurate tickmarks, please. Camera exposure must be at least 2x refresh cycle length minimum (and preferably about 4x refresh cycle), to represent aggregate human vision integration time accurately. Example: For 120Hz refresh rate -- that means 1/30sec camera exposure for photographs -- or 1/30sec exposure per video frame if doing video. For end users doing the easy rail-less method via smartphone hand-wave, try to get less than 1 pixelwidth error margin. For professional reviewers, try to get less than 0.2 pixelwidth error margin, if you can!

Pursuit Camera Instructions
For those wondering how to use a pursuit camera for free:
Please Follow Guidelines
Please Confirm Scientific Integrity

To Reviewers Reading This Thread, Thank You For Using Blur Busters Tests!
To readers, please feel free to email my tweet to reviewers: twitter.com/BlurBusters/status/1239669830300717060
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

Locked