LG 27GK750F-B - 240Hz, 27" 1ms Blur reduction, Freesync

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!
daniloberserk
Posts: 25
Joined: 22 Aug 2017, 16:50

Re: LG 27GK750F-B - 240Hz, 27" 1ms Blur reduction, Freesync

Post by daniloberserk » 03 Dec 2018, 10:00

k2viper wrote:daniloberserk, thanks for your efforts. As you may already got, this 60fps youtube video is barely useful for measuring such high refresh rate displays like 144/240hz. I am not sure, that even being set at 2x timespeed, youtube (flash/html5) player plays it with actual 120 frames per second and not dropping frames, even if I set your videos to 0,25 and watch these slow-mo sections there are some dropped/skipped 60fps video frames seen, I think it's a youtube player flaw, not your technique.

But still, we can have an idea which modes are faster. Your Benq is in Instant mode, right? I dont think DVI-D is superior to Displayport in terms of input lag, but I suppose that 144hz mode is not somewhat "native" for many of 240hz displays and their scalers/panel electrionics drivers.

1:1 seems to be clearly the slowest mode. 144hz Benq is often a frame ahead of 1:1 LG.

Original is fastest, in Original you more often can catch a 16ms higher frame number on the left (LG) screen, there are even some double frames ahead stopwatches I found.

I dont know other then SMTT 2.0 tools that would not be browser based, though.
Yep. They're not precise enough to know exactly values, but we can be sure it's faster at 240Hz at least!

Still no luck with the vtricks, I'm stuck with 1182, it seems to work fine and without stutter so I'm sticking with this. Every other trick I tried gave me stuttering...

Edit: Yes, Benq is on Instant mode!

User avatar
XAKE
Posts: 8
Joined: 12 Aug 2018, 02:10

Re: LG 27GK750F-B - 240Hz, 27" 1ms Blur reduction, Freesync

Post by XAKE » 03 Dec 2018, 10:35

daniloberserk wrote:
k2viper wrote:
daniloberserk wrote:k guys, I don't know how "precise" this tests I made is but I hope it can help for something.
I have a Sony Xperia XZ premium, which can record 960fps for about 1/4sec. I made some comparisons with the LG 27GK750F and my Benq XL2420z

144 Hz on both displays, aspect: full wide on LG:
https://www.youtube.com/watch?v=zrCryMk ... e=youtu.be

144Hz on both displays, aspect: original on LG:
https://www.youtube.com/watch?v=MG_QGNM ... e=youtu.be

240Hz vs 144Hz, aspect: original on LG:
https://www.youtube.com/watch?v=vVgINau ... e=youtu.be

The Benq is connected via DVI-DL and the LG on d-port on my R9 390x.

I'm note sure how "precise" this test is but it seems the LG is slight slower on 144Hz, but faster on 240Hz. Fullwide x Original seems about the same for me...

Edit:

Two more tests...

Full Wide - 240Hz: https://www.youtube.com/watch?v=Dav8tR5 ... e=youtu.be
Original - 240Hz: https://www.youtube.com/watch?v=rC2oojz ... e=youtu.be

Again... I don't know if this can help or something... It seems strange, sometimes the LG is faster and other times, the Benq is faster?
Well, some notes on measuring technique:
Display cloning dont work well for this purpose, a clone source display should have about a frame less delay then cloned one. Use display panning.
Any browser based stopwatch timers are innacurate due to Vsync is forced in Windows and broswer. That's why browser-based measured results are always multiple to given frame time. It is innacurate but still can give an idea which mode is faster and which is slower.
You show in OSD that freesync is active. Disable it for display lag measuring purpose.
Nice, interesting notes there! Do u know any software who make this easier to test? Also, maybe isn't better for the chief to make a standalone software with the tests of testufo.com ? Since the browsers always have issues here and there...

Edit: So, after reading some posts from the Chief, the article gsync 101, etc... I came to a conclusion there's no way I can made tests precise enough. But I'll do some more tests anyway since they can provide at least some margin of error about the displays.
Seems that D-Port have more lag then DVI... I'm quite shocked honestly. This is a incredible hot topic with tons of info, it seems the Chief is working for a substitute to the SMTT 2.0, which is great... As soon as I can I'll post some more videos. This time with freesync off and panning videos (still browser tests though since I don't have a clue of how can I test effectivelly). So I just got this video: https://www.youtube.com/watch?v=e8ZtPSIfWPc and set to 2x timespeed. Panned the two screens, put the video in the middle and recorded.

Here are the tests:
144Hz
https://youtu.be/ETVDT7O5sH0
https://youtu.be/ZzzFpkRmhBY
https://youtu.be/UzIqwKSuIFs

240Hz
https://youtu.be/3v7kZWMdJU4
https://youtu.be/o1UQiYQsji0
https://youtu.be/wnS1KIXiwKU
https://youtu.be/0jLEE6olv-8

So, my conclusion (at least from this tests):

On 144Hz, the Benq is ALWAYS faster.
On 240Hz, the LG is faster EXCEPT on 1:1 Aspect Ratio mode.

Fullwide and Original seems to be about the same.

Not sure if this lag happens because of the display, or the cable, or whatever. Benq is on DVI-DL, LG on D-Port.

So, if you're planning to use freesync on high competitive gaming with low FPS, it seems not a great deal (still, much faster then the average display).

240Hz seems always great comparing to 144Hz. I'm not sure what people say about the "lag feeling", played as Ashe only on OW on one of my accounts since I got this display and the results are impressive: https://www.overbuff.com/players/pc/dou ... ompetitive got almost 20k of heroscore being a Diamond pleb, which scores me at the top 4% on Overbuff, played a bit with freesync and the majority of the time with MBR off and 240Hz (fps uncapped). Even on this conditions (at least for me), I can feel a difference between 240Hz and 144Hz, to bad my setup can't hold 200FPS + all the time, so I'm not sure where is my threshold, but feels incredible snappy and smooth to track targets, so that's my anecdotal evidence regarding my feeling about the display.

Thanks for the test !!! Say the firmware version of panel, make a screen service menu
1. Power off, then power on
2. Push left 3 times, then right 1 time, and then press the joystick button
3. Then press left or right and you should get in service menu.
Attachments
my_panel.jpg
my_panel.jpg (92.5 KiB) Viewed 894 times

daniloberserk
Posts: 25
Joined: 22 Aug 2017, 16:50

Re: LG 27GK750F-B - 240Hz, 27" 1ms Blur reduction, Freesync

Post by daniloberserk » 03 Dec 2018, 15:25

XAKE wrote:
daniloberserk wrote:
k2viper wrote:
daniloberserk wrote:k guys, I don't know how "precise" this tests I made is but I hope it can help for something.
I have a Sony Xperia XZ premium, which can record 960fps for about 1/4sec. I made some comparisons with the LG 27GK750F and my Benq XL2420z

144 Hz on both displays, aspect: full wide on LG:
https://www.youtube.com/watch?v=zrCryMk ... e=youtu.be

144Hz on both displays, aspect: original on LG:
https://www.youtube.com/watch?v=MG_QGNM ... e=youtu.be

240Hz vs 144Hz, aspect: original on LG:
https://www.youtube.com/watch?v=vVgINau ... e=youtu.be

The Benq is connected via DVI-DL and the LG on d-port on my R9 390x.

I'm note sure how "precise" this test is but it seems the LG is slight slower on 144Hz, but faster on 240Hz. Fullwide x Original seems about the same for me...

Edit:

Two more tests...

Full Wide - 240Hz: https://www.youtube.com/watch?v=Dav8tR5 ... e=youtu.be
Original - 240Hz: https://www.youtube.com/watch?v=rC2oojz ... e=youtu.be

Again... I don't know if this can help or something... It seems strange, sometimes the LG is faster and other times, the Benq is faster?
Well, some notes on measuring technique:
Display cloning dont work well for this purpose, a clone source display should have about a frame less delay then cloned one. Use display panning.
Any browser based stopwatch timers are innacurate due to Vsync is forced in Windows and broswer. That's why browser-based measured results are always multiple to given frame time. It is innacurate but still can give an idea which mode is faster and which is slower.
You show in OSD that freesync is active. Disable it for display lag measuring purpose.
Nice, interesting notes there! Do u know any software who make this easier to test? Also, maybe isn't better for the chief to make a standalone software with the tests of testufo.com ? Since the browsers always have issues here and there...

Edit: So, after reading some posts from the Chief, the article gsync 101, etc... I came to a conclusion there's no way I can made tests precise enough. But I'll do some more tests anyway since they can provide at least some margin of error about the displays.
Seems that D-Port have more lag then DVI... I'm quite shocked honestly. This is a incredible hot topic with tons of info, it seems the Chief is working for a substitute to the SMTT 2.0, which is great... As soon as I can I'll post some more videos. This time with freesync off and panning videos (still browser tests though since I don't have a clue of how can I test effectivelly). So I just got this video: https://www.youtube.com/watch?v=e8ZtPSIfWPc and set to 2x timespeed. Panned the two screens, put the video in the middle and recorded.

Here are the tests:
144Hz
https://youtu.be/ETVDT7O5sH0
https://youtu.be/ZzzFpkRmhBY
https://youtu.be/UzIqwKSuIFs

240Hz
https://youtu.be/3v7kZWMdJU4
https://youtu.be/o1UQiYQsji0
https://youtu.be/wnS1KIXiwKU
https://youtu.be/0jLEE6olv-8

So, my conclusion (at least from this tests):

On 144Hz, the Benq is ALWAYS faster.
On 240Hz, the LG is faster EXCEPT on 1:1 Aspect Ratio mode.

Fullwide and Original seems to be about the same.

Not sure if this lag happens because of the display, or the cable, or whatever. Benq is on DVI-DL, LG on D-Port.

So, if you're planning to use freesync on high competitive gaming with low FPS, it seems not a great deal (still, much faster then the average display).

240Hz seems always great comparing to 144Hz. I'm not sure what people say about the "lag feeling", played as Ashe only on OW on one of my accounts since I got this display and the results are impressive: https://www.overbuff.com/players/pc/dou ... ompetitive got almost 20k of heroscore being a Diamond pleb, which scores me at the top 4% on Overbuff, played a bit with freesync and the majority of the time with MBR off and 240Hz (fps uncapped). Even on this conditions (at least for me), I can feel a difference between 240Hz and 144Hz, to bad my setup can't hold 200FPS + all the time, so I'm not sure where is my threshold, but feels incredible snappy and smooth to track targets, so that's my anecdotal evidence regarding my feeling about the display.

Thanks for the test !!! Say the firmware version of panel, make a screen service menu
1. Power off, then power on
2. Push left 3 times, then right 1 time, and then press the joystick button
3. Then press left or right and you should get in service menu.
Oh, I did that already, I'll post a pic as soon as I can, but I remember being the 3.07 ver!

Kheri
Posts: 59
Joined: 29 Sep 2018, 13:39

Re: LG 27GK750F-B - 240Hz, 27" 1ms Blur reduction, Freesync

Post by Kheri » 03 Dec 2018, 15:59

k2viper wrote:
Kheri wrote:k2viper, wondering if you'd be willing to help me out, I can see in your strobe crosstalk pics you have one maybe two duplicated images at the center line, right? Is that reflective of reality or how your eyes see it? On mine, I can see 2 of the crosstalk images clearly, and a very faint third one. It actually makes the street names in the panning test quite difficult to read.

I think it might be a good idea to put the crosstalk pics in the first post, sort of as a way to show the results.
I can read street names without much effort, I may just be used to strobing since I was on full-time Lightboost from 2012.
These are the most accurate photos I have to reproduce what an eye is actually seeing in testufo.com/crosstalk test - photos made with hand pursuit :)

ImageImage

At center, crosstalk is very faint.
I'm fairly sure it's the duplicated images that make the text hard to read. Although it's still more readable than with MBR off (I leave MBR on always). I can read Bay street without MBR for example, but the text isn't anywhere as clear.

From your pics (thank you) i see one crosstalk image that is pretty noticeable but kinda faint(?) and second one that's very faint. Is that reflective what you actually see with your own eyes? (I really want clarification here, maybe I'm just being too picky with my crosstalk.) I'm still trying to investigate what's going on.

I have two very noticeable crosstalk images. And one (the third one) is super faint, 8th line/center of the screen. my SECOND crosstalk image looks like your first crosstalk image. But I'm not sure about that because I'm not viewing both our screens IRL. It's kinda bad to be quite honest, although I probably still prefer MBR on.. I'm using your settings, 1304 vtotals.


Also, I've checked this about 5 times. There doesn't seem to be any difference between 1180 vtotals and 1304 vtotals (and I think i'm switching between them right, because if I set CRU to 1150 vtotals, and restart64.exe, and switch monitor refresh rate, restart chrome, the test gets stuttery). There might be some difference with the very faint third crosstalk image (I get 3 crosstalk images in addition to the normal image.) My monitor is Version 3.07 and panel M270HTN02.3 (from the service menu). I'm curious what the Output Type(s) under ToolOption are all about: VByOne (current), LVDS 8Bit, LVDS 10Bit, eDP.

User avatar
k2viper
Posts: 290
Joined: 23 Jan 2018, 06:30

Re: LG 27GK750F-B - 240Hz, 27" 1ms Blur reduction, Freesync

Post by k2viper » 04 Dec 2018, 03:12

Well, probably yes, there is one faint crosstalk double-image like behind 1mm after the UFO body. Cant say that there is third one, at least not at center.
"Normally" crosstalk is more visible and thick at top and bottom, but at center zone it is very acceptable. Also testufo is an synthetic scenario and you barely will see that much crosstalk in real games.

I cant really say, is there a problem with your crosstalk or you're just too picky. Maybe I can say if you post pics like mine, with hand pursuit.

About VT1304. You should understand that it only increases Vertical blanking interval from 100 (default) to 224 (VT1304) lines. While frame is still being 1080 lines, there is not that much of improvement when vblank becomes 124 lines thicker.

We probably need VTs at 2500 or more, so Vblank could be huge enough to eliminate crosstalk completely. But I dont know 240hz monitors that are able to work with such big VTs.

Kheri
Posts: 59
Joined: 29 Sep 2018, 13:39

Re: LG 27GK750F-B - 240Hz, 27" 1ms Blur reduction, Freesync

Post by Kheri » 04 Dec 2018, 07:19

k2viper wrote:Well, probably yes, there is one faint crosstalk double-image like behind 1mm after the UFO body. Cant say that there is third one, at least not at center.
"Normally" crosstalk is more visible and thick at top and bottom, but at center zone it is very acceptable. Also testufo is an synthetic scenario and you barely will see that much crosstalk in real games.

I cant really say, is there a problem with your crosstalk or you're just too picky. Maybe I can say if you post pics like mine, with hand pursuit.

About VT1304. You should understand that it only increases Vertical blanking interval from 100 (default) to 224 (VT1304) lines. While frame is still being 1080 lines, there is not that much of improvement when vblank becomes 124 lines thicker.

We probably need VTs at 2500 or more, so Vblank could be huge enough to eliminate crosstalk completely. But I dont know 240hz monitors that are able to work with such big VTs.
Yes, I'm talking about the center. crosstalk at the top and bottom look pretty similar between your pics and what I see on my screen. But there's clearly 3 duplicated images (meaning I see 4 images) at center. The first two are quite obvious. The third one is pretty faint, although surprisingly it's noticeable enough on the street map (it's still quite faint, to be clear). And text in some 240fps games still hard to read sometimes (honestly not different over non-MBR). Although the MBR is still nice, just the crosstalk seems to be a bit problematic even in games. I don't think I'll be able to share pics with you, although I wish I could. (One of my bigger problems is not being able to use black frame insertion (not LG's MBR) for 60hz games that "support" 240hz but don't actually go above 60hz, and I can't do it with ReShade either.)

"there is one faint crosstalk double-image" I might unfamiliar with the term "double-image" or if it has a special meaning. From your pics I could see you had ONE crosstalk image, and sometimes it looked like you had a second one (So 2-3 images, including the one you're supposed to see). This is at center of the screen (roughly the 8th row from the top).

I looked pretty closely and several times.. there is zero difference when I add 124 or 126 VBI lines. It really makes me wonder if I'm messing up somehow, or if I'm just being too picky. I noticed no difference in the size of the crosstalk images, their position, opacity or how many there were. I know we're using different versions of the same monitor. And I know it'd be much better if I could take a pic an share it with you; I don't think I can, but I'll look into it later to make sure. I need to check what FPS on the camera is required, hand-pursuit vs stationary, etc. So I'll try this next. I know I tend to be a bit picky, and the difference 124 VBI lines adds isn't much, but there's no difference as far as I can see, and you're claiming to get 1 double-image at center, and I'm getting 3 (with 2 of them being rather bold). I'm running the crosstalk and moving photo/street map at 960 pixels/second, 240hz.

As a side note, Bay Street (and this is the lower part of the screen) is considerably or more readable on 120hz strobing than 240hz strobing, even though I can see the crosstalk images spread about.. I think they get pushed out of the way of the text, perhaps. I think that's normal though, right?

User avatar
k2viper
Posts: 290
Joined: 23 Jan 2018, 06:30

Re: LG 27GK750F-B - 240Hz, 27" 1ms Blur reduction, Freesync

Post by k2viper » 04 Dec 2018, 09:17

Sometimes I'm getting a bug, since I have MBR full-time on, after boot and Windows start, I see strobing toggles on with much higher brightness then was intended and then I'm used to. Sometimes it even dont strobe, yet displaying MBR on in OSD. All strobing bug variants are fixed by power cycling. It may be your case.

Easy check if strobing works - fast move your hand between screen and eyes, with normal strobing you will clearly see a stroboscopic effect.

Also, what brightness/contrast do you use. The higher brightness/contrast are, the more crosstalk you get.

And finally, provide pics of what you're complaining about. Dont mess up with video, do just hand pursuit and normal photos with fast shutter speed.

Kheri
Posts: 59
Joined: 29 Sep 2018, 13:39

Re: LG 27GK750F-B - 240Hz, 27" 1ms Blur reduction, Freesync

Post by Kheri » 04 Dec 2018, 18:31

All I could give you is slow shutter picture or video. I can't get it anywhere near as clear as your pics, either.
I was using 44 brightness, 60 contrast, gamma 2. But I switched it to gamme 4, 45 brightness, 50 contrast, not much of a difference.

Did you notice a difference when you went from 1180 vtotals to 1304?

dogroll
Posts: 36
Joined: 24 Aug 2017, 16:20

Re: LG 27GK750F-B - 240Hz, 27" 1ms Blur reduction, Freesync

Post by dogroll » 04 Dec 2018, 19:29

Anyone know the difference between 3.06 and 3.07 firmware? I got 3.06 since I bought my 27GK750F at release.

Kheri
Posts: 59
Joined: 29 Sep 2018, 13:39

Re: LG 27GK750F-B - 240Hz, 27" 1ms Blur reduction, Freesync

Post by Kheri » 04 Dec 2018, 21:54

I don't have fast shutter speed. My pictures came out blurry, unfocused and faint, even the good ones, so it doesn't really show the crosstalk. (I used hand pursuit)

I can't really measure this, but I compared against a screenshot and it looks like the crosstalk image trails behind by about 1.5mm (same as yours by the looks of it). And on 120hz it's closer to 3mm trail. IRL, that seems like it's probably accurate. So it looks similar but I get, at center: [dupe3] [dupe2] [dupe1-image] [normal-image] and the first two dupes are pretty distinct. I'm using version 3.07.

I compared 1180 vtotals and 1562 vtotals on 120hz (multiple times, like I did with 240hz). I didn't see any difference, and it's clear it's actually switching resolutions when I change them in Nvidia; if I go too high with the vtotals or horizontal refresh rate the LG just goes blank. I've looked very hard for a difference. Looks like the monitor blanks out due to the horizontal refresh rate being too high?

I'm probably being too picky, but it's striking that there's no difference when I switch between vtotals; k2viper saying he gets 1 (or 2) double-images and I get 2-3; my crosstalk images look like they have more color/opacity than his (maybe they don't, though, because I see his through pictures); the street map is difficult to read. I can't read the top portion due to crosstalk, can't make out "the natrel pond/rink," so maybe not that bad, but still, the text is difficult to read. I wonder if a hand-pursued picture of the street map would look any different than what I see.

Why is there no difference when I change the vtotal lines, even when I add nearly 300 lines (still not much)? It's too bad this monitor can't handle 2600 vtotal lines even at 120hz. Think I'd really be able to tell if there is a difference between vtotals then.

Post Reply