Blur Busters Forums

Hello,

Is it worth to use G-sync Compatible with ELMB + v-sync at ON?
Im asking about that with the VG259QM /or VG279QM (https://www.tftcentral.co.uk/reviews/as ... g279qm.htm)

I mean if someone have more than 200 fps I believe there is no need to run G-sync, but what about if we have less than 200 fps, is it worth to use ELMB which block OD to 80, so I might see some overshoots but actually 240hz with 80 OD provides lowest inpust lag with only a little bit overshoots, so that is why I thought to use as a counter an ELMB, so my screen will be perfect without any blur.

But then I asked myself, why wouldn't use also G-sync with that, and then... why don't use also g-sync + V-sync (enabled both only in NVCP) + ELMB.

it is quite confusing for me and I am not sure if for example G-sync is worth only for 200 fps less or so.

I don't see any difference using g-sync or not, my game is actually more responsive without that, but maybe ELMB works better with g-sync? Or... It is better than using g-sync, so I could use ELMB instead of g-sync

ELMB-SYNC works best near max-Hz. You can use framerate capped ELMB-SYNC with G-SYNC, as a strobed-lag-reduction technique. It's lower lag than VSYNC ON anyway. Just make sure you stay near max Hz, to prevent low-framerate ELMB-SYNC artifacts.

Chief Blur Buster wrote: ↑

10 Apr 2020, 23:55

ELMB-SYNC works best near max-Hz. You can use framerate capped ELMB-SYNC with G-SYNC, as a strobed-lag-reduction technique. It's lower lag than VSYNC ON anyway. Just make sure you stay near max Hz, to prevent low-framerate ELMB-SYNC artifacts.

Wait, so actually turning G-sync ON with ELMB reduce input lag compare to ELMB without g-sync/vsync off?

So its actually a must have with ELMB? what if ELMB works only with 240hz and 144hz? Does it mean when g-sync change the 240hz to 200hz=200fps it goes down to 144hz???

Last question, what about if I use g-sync only (No ELMB) / and no v-sync to avoid input lag) - does it gives any input lag? Im confused actually, because all pros use g-sync/v-sync off, because they say g-sync gives input lag, but it doesnt - thats what I heard. Dunno what is true and what is false..

Slow down on the questions --

forii wrote: ↑

14 Apr 2020, 15:56

Wait, so actually turning G-sync ON with ELMB reduce input lag compare to ELMB without g-sync/vsync off?

VSYNC OFF at framerates beyond Hz will have lower average lag. That is already shown in G-SYNC 101. But high refresh rates reduce that differential greatly (240Hz VRR).

forii wrote: ↑

14 Apr 2020, 15:56

So its actually a must have with ELMB?

For other brands of motion blur reduction, there are some brand-specific easter eggs and techniques to reduce input lag. There are other tricks such as Quick Frame Transport with PureXP+ (custom 120Hz timings) combined with RTSS Scanline Sync, that can reduce strobe latency massively. But that's an advanced trick.

forii wrote: ↑

14 Apr 2020, 15:56

what if ELMB works only with 240hz and 144hz? Does it mean when g-sync change the 240hz to 200hz=200fps it goes down to 144hz???

It won't work unless you have a monitor that supports "ELMB-SYNC". Older ELMB monitors do not have the ability to operate concurrently with G-SYNC. However, be warned there are some disadvantages too -- you have to choose if the trade-offs are worth it or not.

forii wrote: ↑

14 Apr 2020, 15:56

Last question, what about if I use g-sync only (no v-sync to avoid input lag) - does it gives any input lag?

For framerates below refresh rate, there's no difference in lag between "GSYNC + VSYNC OFF" versus "GSYNC + VSYNC ON". (To be clear, I'm talking about the configuration of a fallback sync technology, which only activates for framerates above Hz whenever GSYNC is enabled).

Most people simply enable GSYNC and forget to configure a frame rate cap, creating some penalities whenever frame rates go beyond Hz. This happens more often in some games (CS:GO) than other games (PUBG). So some people use VSYNC OFF with CS:GO while they use G-SYNC with PUBG.

Each gamer is different and have different playing preferences.

forii wrote: ↑

14 Apr 2020, 15:56

Im confused actually, because all pros use g-sync/v-sync off, because they say g-sync gives input lag, but it doesnt - thats what I heard. Dunno what is true and what is false..

Both answers are true.

If you want to understand why, please read All 14 Pages of GSYNC-101

Semi-complex one-paragraph answer:
A. GSYNC is much lower lag than most sync technologies (example: VSYNC ON)
B. GSYNC is only a tiny bit more lag than ultra-high-framerate VSYNC OFF (e.g. 500 frame per second VSYNC OFF).
C. Also, low-refresh rates enforce a high scanout latency (high speed videos of refresh cycles). By choosing 240Hz, 280Hz or 360Hz with VRR, you can dramatically reduce scanout latency, improving the advantage that can be obtained with G-SYNC -- with fewer situations that G-SYNC is worse.

To read more, check out GSYNC 101 as well as Battle(non)sense tests, etc.

If you play ONLY CS:GO, framerates can be almost permanently above refresh rates. For that situation, GSYNC is not as much of a help as for other games, unless you hate tearing. You might just go with the most popular monitor found on prosettings.net and others BenQ ZOWIE XL2546 VSYNC OFF 500fps

If you play a variety of games, then you might strategically enabe G-SYNC for games that have a framerate range that falls within the VRR range, and also strategically use a frame rate cap. For example, PUBG benefits more from G-SYNC than CS:GO -- because PUBG is often more stuttery than CS:GO -- the smoothing of stutters can improve your aiming in those less-optimized games.

I know my reply creates more questions. First, doublecheck the 14 pages to make sure your question is already answered there --

Chief Blur Buster wrote: ↑

14 Apr 2020, 16:08

VSYNC OFF at framerates beyond Hz will have lower average lag. That is already shown in G-SYNC 101. But high refresh rates reduce that differential greatly (240Hz VRR).

I still don't know correct answer in my case.
I know that v-sync off = g-sync(?) at framerates above 240Hz with have lower avarage lag (thats obvious) but... I almost never have more than 240+ fps anyway (I play CoD MW - its quite bad optimized game = gpu bound game)

I wouldn't use g-sync in games when I can have always stable 240+ fps. We talking here about fps much bellow 240

So with fps much bellow 240 and actually with avg fps of 175 (in my case) with g-sync ON -> do I have less input lag than with g-sync OFF and v-sync OFF?

+My gpu usage is actually max 96-97%, I use NULL at ultra to lower the latency, still also don't know if NULL is good with g-sync or its only good only without g-sync.

I cap fps with RTSS at 218 (only because I never reach more than 220+ fps anyway, so I leave the space for fps to go max to the 218 + have less framerate) - i can't cap fps in game, I can do that only at 166, 200, 250, or 300 (sadly)

Whats best option for me for most competive play in a shooter game? when the input lag actually matters, because whoever see the enemy faster and pre-fire = he wins.

forii wrote: ↑

14 Apr 2020, 16:41

Chief Blur Buster wrote: ↑

14 Apr 2020, 16:08

VSYNC OFF at framerates beyond Hz will have lower average lag. That is already shown in G-SYNC 101. But high refresh rates reduce that differential greatly (240Hz VRR).

I still don't know correct answer in my case.
I know that v-sync off = g-sync(?) at framerates above 240Hz with have lower avarage lag (thats obvious) but... I almost never have more than 240+ fps anyway (I play CoD MW - its quite bad optimized game = gpu bound game)

Sometimes it is not possible to give a correct answer without putting lots of analysis equipment such as photodiode oscilloscopes and other stuff to your game -- and not all games are analyzed. One of the best people who do this stuff, is Battle(non)sense, and you've already linked to his stuff. He does a great job at this already and has already replicated a lot of Blur Busters findings in many cases.

forii wrote: ↑

14 Apr 2020, 16:41

So with fps much bellow 240 and actually with avg fps of 175 (in my case) with g-sync ON -> do I have less input lag than with g-sync OFF and v-sync OFF?

Absolute lag? Only for screen top edge region (since VRR has a scanout latency at bottom edge).
Latency consistency? Yes. Better latency consistency than VSYNC OFF.

TL;DR: Lag is not a single number

Most arguments, testers, and reviewers contradict with each other because it's difficult to standardize lag measurements because of so many lag variables.

forii wrote: ↑

14 Apr 2020, 16:41

Whats best option for me for most competive play in a shooter game? when the input lag actually matters, because whoever see the enemy faster and pre-fire = he wins.

Correct. This is true for CS:GO

But not all games are identical.

For example, in Battlefield 3, there's that high-speed low-altitude flyby. That is very blurry, unless you see blur reduction. For motionblur-camoflaged games like that, slightly laggier-but-clearer gives you a competitive advantage. Or when you're always scrolling/panning/turning in games that don't have a crosshairs -- a game such as Rocket League. Seeing a very clear ball a millisecond second later is often superior to seeing a blurry ball a millisecond sooner. It is possible to aim to the goalposts better.

Also don't forget, for different games:
Absolute lag -- the average lag
Latency gradient -- bottom edge of screen is sometimes laggier than top edge when not using VSYNC OFF
Latency volatility -- latency varying continually

In CS:GO, absolute lag is often king. But are all games CS:GO? Nope....

For some heavily stuttery games, some players have discovered that fixing latency volatility is more important than absolute latency. That's where RTSS can be a match made in heaven. It sometimes feels lower lag (despite higher average lag) because of reduced latency volatility.

For some games, playing at a consistent 5ms lag is often better than playing at a volatile 2ms-7ms lag. You'll undershoot/overshoot more, because 1000 pixels per second movement can translate to a 2 pixel overshoot/undershoot or 7 pixel overshoot/undershoot with 2ms to 7ms lag. Remember 1ms differences translates to 1 pixel for every 1000 pixels per second. Imagine turning fast and shooting immediately after a turn. If lag is predictable, it's easier to stop on a frag without overshoot/undershoot. Even if it's a smidge higher average lag.

But CS:GO doesn't have a latency volatility problem, and VSYNC OFF fixes latency gradients (for non-strobed modes). The great news is that strobed ELMB-SYNC (near max Hz) doesn't have a noticeable latency gradient issue (as long as you're cappign just below Hz).

I am sorry, but Blur Busters never gives simple answers -- latency is a complex problem when not all pixels have the same lag (scanout latency), and also the lag volatility versus lag average tradeoffs. Even switching configurations can create changes to the latency advantages, e.g. increased latency volatility but lower absolute lag. Competitive advantages will vary from game to game.

Well, alright, after that here where I ended with my 240hz ips panel and fps much often around 180~ in one game: Call of Duty Modern Warfare (2019), and sometimes even lower = at 155 avg - based on the game mode.

-Due to limited in-game fps limiter (only possible to cap 166/200/250 fps) and like most pro players I decided to do not cap fps (even with RTSS) to avoid any input lag. I mean - I cap fps at 250 with in game limiter, but I dont know if that can be called a "cap" because I never reach 230+ fps anyway. I could cap the fps at 200 with in game limiter, but I sometimes hit more than 200 fps, so I better have these few fps sometimes which might help.

-I decided to be better like all pro players and use the g-sync because even if it does add some input lag It should be much less than RTSS and thanks to that my game is much smoother (no stutter). I do not use v-sync because I do not have any tearing anyway at 240hz.
It is also trade off with using no rtss (i might get some input lag from g-sync but I don't get any from rtss cap - cuz I dont use it)

-I also use NULL at Ultra due to unlimited fps + gpu bound game (even with unlimited fps = aka = caped at 250) i still have gpu usage at max 97% - I know its not perfect with input lag but NULL helping in that case. (as I said in previous post it doesnt matter if I cap fps at 200, 180, 160 - it is still max at 97% usage anyway - not always but maximum, most often actually), only at 144 fps cap I have max 90% gpu usage, but that's not worth to have I guess.

I would not use g-sync for games where I can reach stable fps 240+ ( I believe its good only if the fps drops bellow)
Same with NULL - i wouldn't use it at ultra in no gpu bound game (97+% usag) - only at ON in that case.

I will also consider using in same time ELMB but not full time. (thats why I also prefer to cap with in game limiter fps at 250 to keep as much fps as I can for using elmb-sync)

For the VG259QM/VG279QM - i use Override 80 and with g-sync it helps to avoid seeing overshoots though.

I run 1000hz mouse (wireless g305)

I alsu run MSI Mode Utility on my GPU (Message Signaled-Based Interrupts) which gives me even less latency compare to Line Based which has almost all gpu cards as default.

I also have HPET on in bios but... turned off in windows (it has to be like that) + TimerResolution to have 0.5ms. ( I run app called Intelligent standby list cleaner which helps with cleaning ram + setting lowest timerresolution)

To have as much fps I can I OC my GPU with Msi afterburner, Overclocked my CPU + OC the ram (3600mhz /cl15) - make sure guys it has to be stable thought, it takes a lot of time to stabilize this.

+ Due to windows 10 I actually run a lot of tweaks just to have higher fps as possible, like turning off some services, etc...