please help me understand some things about 240Hz and G-Sync

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!
Post Reply
chocobo
Posts: 4
Joined: 19 May 2018, 21:11

please help me understand some things about 240Hz and G-Sync

Post by chocobo » 20 May 2018, 00:03

I'm relatively new to the world of using a quality monitor for gaming, I discovered 144Hz in January but got a subpar HP Omen monitor and was still looking for a better option... recently I caught the Alienware AW2518HF on sale and I'm pretty happy with it. But I'm still trying to better understand the advantages of certain monitor features and could use some help. BTW I play games like CSGO and Fortnite on low-mid settings on a GTX 1060 in order to get high framerates, especially now that I have a monitor that can display more than 144 per second.

1) If I set my monitor to 144Hz and in-game framerate limit to 144 FPS, and keep my settings low enough to maintain a constant 144, will I have no screen tearing at all and will it function the same as if I had G-Sync?

2) When I have the monitor on 144Hz and the game is in the 200-240 FPS range, I see a little bit of screen tearing. At 60Hz, there's plenty of screen tearing because 60Hz is terrible. But at 240Hz, the screen tearing is virtually undetectable unless I'm looking closely and trying hard to find some.

Am I correct that higher refresh rates make G-Sync less important? I was wondering if maybe I made the wrong choice buying this monitor instead of a 144Hz G-Sync monitor that was on sale. But at 240Hz it seems like the tears are smaller, and of course they're visible for a much shorter time, so the result is that it seems (to me) like adaptive sync would be only a very tiny benefit. Am I understanding this right?

3) If that's right, then is there an argument to be made for ~210 fps 240Hz over 144Hz G-Sync? Extra smoothness at the cost of imperceptible screen tears sounds like a decent tradeoff, right? (of course my settings have to be low to do this, and the 144Hz G-Sync option could use higher graphics settings... but tbh I don't care about graphics in this games, I just want smooth and easy to see gameplay)

4) I've seen people say things like "100 FPS G-Sync looks smoother than 144 FPS without it". How can 100 FPS look smoother than 144? They just mean they prefer the G-Sync option over a higher framerate + screen tearing, right? G-Sync and Freesync don't accomplish anything outside of eliminating screen tearing, right?

Any info would be appreciated, please let me know if I have badly understood how some of these things work. I'm hoping that I made the right monitor choice and won't end up regretting passing up the 144Hz G-Sync one.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: please help me understand some things about 240Hz and G-

Post by Chief Blur Buster » 20 May 2018, 14:31

chocobo wrote:1) If I set my monitor to 144Hz and in-game framerate limit to 144 FPS, and keep my settings low enough to maintain a constant 144, will I have no screen tearing at all and will it function the same as if I had G-Sync?
No, you will still have tearlines. Tearlines will become a beat-frequency relative to frame rate.

141fps at 144Hz = rolling tearlines 3 cycles downwards a second (faster rolling)
142fps at 144Hz = rolling tearlines 2 cycles downwards a second (fast rolling)
143fps at 144Hz = rolling tearlines 1 cycles downwards a second (slow rolling)
144fps at 144Hz = stationary tearline
145fps at 144Hz = rolling tearlines 1 cycles upwards a second (slow rolling)
146fps at 144Hz = rolling tearlines 2 cycles upwards a second (fast rolling)
147fps at 144Hz = rolling tearlines 3 cycles upwards a second (faster rolling)

The more precise the capping software (e.g. RTSS instead of in-game), the less the tearline vibrates and becomes more smoothly rolling. So random tearline positions make it more invisible.
chocobo wrote:2) When I have the monitor on 144Hz and the game is in the 200-240 FPS range, I see a little bit of screen tearing. At 60Hz, there's plenty of screen tearing because 60Hz is terrible. But at 240Hz, the screen tearing is virtually undetectable unless I'm looking closely and trying hard to find some.
Tearline visibility length is controlled by two things:
(1) Refresh cycle length. A tearline appears/disappears in 1/240sec at 240Hz.
(2) Position of next tearline versus previous tearline.
If they reappear in the same location (e.g. like the above scenario), 144fps cap at 144Hz, that amplifies visibility of tearline, but does kind of contain the tearline visibility.

Even without tearing -- microstutters will still be more visible.
chocobo wrote:Am I correct that higher refresh rates make G-Sync less important? I was wondering if maybe I made the wrong choice buying this monitor instead of a 144Hz G-Sync monitor that was on sale. But at 240Hz it seems like the tears are smaller, and of course they're visible for a much shorter time, so the result is that it seems (to me) like adaptive sync would be only a very tiny benefit. Am I understanding this right?
To a certain extent for some people, but it depends on your stutter-sensitivity and your tearing-sensitivity.

Personally, during my 480Hz monitor tests I could still barely see tearing & microstutters at 480Hz. So the diminishing points of returns have not yet ended -- though it requires a continual doubling of Hz to keep milking the diminishing point of returns (120Hz -> 240Hz -> 480Hz -> 960Hz).

A good educational journey of the diminishing returns curve, see Blur Busters Law And The Amazing Journey To Future 1000Hz Monitors -- it doesn't disappear fully till the quintuple digits Hertz in the most extreme use case scenario. It does get increasingly difficult to tell differences without increasingly massive jumps in Hz however, but the vanishing point is still currently very far away.
chocobo wrote:3) If that's right, then is there an argument to be made for ~210 fps 240Hz over 144Hz G-Sync? Extra smoothness at the cost of imperceptible screen tears sounds like a decent tradeoff, right? (of course my settings have to be low to do this, and the 144Hz G-Sync option could use higher graphics settings... but tbh I don't care about graphics in this games, I just want smooth and easy to see gameplay)
Yes, 210fps at 240Hz looks smoother with VRR than without. Microstutter has harmonic-frequency / beat-frequency effects with Hz -- such that when framerates modulate near harmonic frequencies (e.g. 121fps at 240Hz), you can witness 1 microstutter per second. VRR eliminates those kinds of microstutters from fluctuating frame rates.

Go to http://www.testufo.com/gsync and watch how VRR eliminates microstuttering. The real thing (real GSYNC) looks even better than that simulated animation.
chocobo wrote:4) I've seen people say things like "100 FPS G-Sync looks smoother than 144 FPS without it". How can 100 FPS look smoother than 144? They just mean they prefer the G-Sync option over a higher framerate + screen tearing, right? G-Sync and Freesync don't accomplish anything outside of eliminating screen tearing, right?
Again, go to http://www.testufo.com/gsync to understand. 100fps means the display is currently 100Hz.
The refresh rate of a GSYNC display can change hundreds of times per second seamlessly. Frame rate is the refresh rate. Refresh rate is the framerate. 47.159fps means the monitor is 47.159Hz. The monitor waits for the software before beginning to refresh. The refresh cycle is directly controlled by the software itself. When a frame is presented by the game, the monitor immediately refreshes at that instant.

Image

Versus:

Image

From an eye-tracking perspective, object positions in sync with frame positions:

Image

As you can image, "perfect 51.943fps at 51.943Hz" looks smoother and better than "51.943fps at 60Hz". GSYNC makes variable framerate look like permanent perfect VSYNC ON. Meaning random fluctuating 50-70fps looks near identical to perfect 60fps@60Hz VSYNC ON but without the input lag of VSYNC ON. Random framerates have no stutter if it's in a tight varying range.

So generally, the stutter-removal feature gives you roughly a 30-50% framerate-prettiness premium. In other words, many people prefer perfect-smooth 40fps over stuttery 55fps@60Hz fixed-Hz. VRR becomes 40Hz because fps permanently equals Hz, and Hz permanently equals fps, whenever in VRR range.

For a wonderful demonstration of how framerate randomization are successfully destuttered, see these simulated animations: So as you can see, VRR eliminates stutter caused by changes in framerates (as long as the frametimes continually stay within the VRR range -- e.g. 1/50sec frametime versus 1/100sec frametime -- it can't do anything to major frametime spikes like hard disk accesses like 1/10sec frametimes caused by disk accesses and such which is well below typical VRR range). Fortunately, most frametime variations are caused by GPU rendering variations, and thusly, VRR is able to fix stutters caused by framerate variance.

Make sure you run these animations in a 100% stutterfree web browser (continuously green "READY") so that browser stutters don't interfere with the framerate-change-caused stutters.

<MATH POINT OF VIEW>
Technical Numbers View:

Also, if you're a math person, this may help put sense in how much "milliseconds matters" in some of the stutter mathematics:
Imagine horizontal panning motion, 4000 pixels per second horizontally panning (1 screen width per second at 4K). A 1 millisecond error means a 4 pixel misalignment (4 pixel stutter). So milliseconds matter during stutter. Now the milliseconds difference between 1/100sec = 10ms and 1/144sec = 6.9ms. That's a 3.1 millisecond difference. Now, with worst-case refresh-cycle misalignment, e.g. missing a VSYNC interval -- perfect framepacing at 10ms (VRR) or rounded-to-nearest-refresh-cycle (non-VRR) ... half of 6.9ms is approximately 3.4
milliseconds. 3.4 milliseconds out of 1000 is still 0.34%. Now, 0.34% of 4000 pixels/second movement (1 screenwidth panning) is equal to a 13.6 pixel jump (stutter) from (4000 * 0.34) = 13.6 ....

Those situations can mean a several pixel-jump of microstutter during near-harmonics situations (e.g. 239fps at 240Hz, 118fps or 121fps at 240Hz, 181fps at 240Hz, and whatnots) -- as the framerate floats up and down, the microstutter mechanics appears and disappear. The higher the framerate, the fainter the microstutter does become. But not all your games can run at 500 frames per second, so, VRR saves the day by completely eliminating this.

Obviously, the amount of pixel jump totally depends on how much the frametime is misaligned with refreshtime when the game is slaved to a fixed monitor Hz. But with GSYNC, the monitor is slave to the game's gametime instead, and the game software directly controls the refresh cycles instead -- framebuffer present timing equals timing of photons hitting eyeballs.
</MATH POINT OF VIEW>
chocobo wrote:Any info would be appreciated, please let me know if I have badly understood how some of these things work. I'm hoping that I made the right monitor choice and won't end up regretting passing up the 144Hz G-Sync one.
Maybe. Maybe not.
- Some of us are very stutter sensitive. Others may not be.
- Some of us are very tearing sensitive. Others may not be.

Just like other vision behaviours:
- Some of us are color sensitive. Others may not be (e.g. slight color blindness -- 8% of population is colorblind)
- Some of us are brightness sensitive. Others may not be. (e.g. eyestrain from brightness)
- Some of us are very flicker sensitive. Others may not be. (PWM-free monitors versus blur-reducing strobe backlights)
- Some of us are very blur sensitive. Others may not be. (Blur? Hah! We are BLUR BUSTERS ;) )
- Etc.

TL;DR: Everybody's vision is different.
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!

User avatar
RealNC
Site Admin
Posts: 3740
Joined: 24 Dec 2013, 18:32
Contact:

Re: please help me understand some things about 240Hz and G-

Post by RealNC » 20 May 2018, 14:48

I recommend reading through the "G-Sync 101" series of articles:

https://www.blurbusters.com/gsync/gsync ... d-settings

The "TL;DR" of g-sync is basically: It looks as smooth as when using vsync, but without the input lag of vsync and regardless of the frame rate you get in games. Knowing how to properly configure g-sync however (especially the frame limiter) is highly recommended, so you should read the above articles.
SteamGitHubStack Overflow
The views and opinions expressed in my posts are my own and do not necessarily reflect the official policy or position of Blur Busters.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: please help me understand some things about 240Hz and G-

Post by Chief Blur Buster » 20 May 2018, 15:06

Yes, Jorim's 14-part GSYNC 101 on Blur Busters is the best on the Internet for GSYNC tweaking information.

Don't let it overwhelm you -- most of it is easy tweaking info such as simply adding a frame rate cap as simply an additional lag-improving optimization to already beautiful GSYNC motion.
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!

chocobo
Posts: 4
Joined: 19 May 2018, 21:11

Re: please help me understand some things about 240Hz and G-

Post by chocobo » 20 May 2018, 20:18

Chief Blur Buster wrote:No, you will still have tearlines. Tearlines will become a beat-frequency relative to frame rate.
Thanks for your extremely detailed answers, especially the links to other articles. I've got some reading to do.
Again, go to http://www.testufo.com/gsync to understand. 100fps means the display is currently 100Hz.
Ahh I understand now, people were talking about G-Sync vs Vsync, and how Vsync produces a choppy/stuttery experience as an alternative to screen tears. I don't use Vsync, I think the side effects it causes are worse than the screen tearing it fixes.

chocobo wrote:3) If that's right, then is there an argument to be made for ~210 fps 240Hz over 144Hz G-Sync? Extra smoothness at the cost of imperceptible screen tears sounds like a decent tradeoff, right? (of course my settings have to be low to do this, and the 144Hz G-Sync option could use higher graphics settings... but tbh I don't care about graphics in this games, I just want smooth and easy to see gameplay)
Yes, 210fps at 240Hz looks smoother with VRR than without. Microstutter has harmonic-frequency / beat-frequency effects with Hz -- such that when framerates modulate near harmonic frequencies (e.g. 121fps at 240Hz), you can witness 1 microstutter per second. VRR eliminates those kinds of microstutters from fluctuating frame rates.
I meant 210 fps at 240Hz without G-Sync or Freesync or Vsync or anything, vs 144Hz with G-Sync. AFAIK when compared to the 144Hz G-Sync monitor, on my 240Hz I'm getting smoother visuals (as in more frames per second) at the cost of lacking G-Sync (some screen tearing), but the tearing is so small and hard to notice I think the extra FPS is worth it. The tears are so small and quick compared to 60 fps (where the top half of the screen can be severely misaligned with the bottom and that frame stays visible for four times as long... I've got 1/4 of the misalignment for 1/4 of the duration, so I can't even see it unless I try.)
Maybe. Maybe not.
- Some of us are very stutter sensitive. Others may not be.
- Some of us are very tearing sensitive. Others may not be.
I think my level of sensitivity to those things is lower than some people's. At 144Hz I see a tolerably small but still slightly annoying bit of tearing, at 240Hz I don't normally notice it (though I can find the tearing if I try hard to look for it). It seems like 240Hz minimizes tearing/stuttering enough that it's almost as good as G-Sync, and since I got a good price on the Alienware I think I'm satisfied with what I've got.

I'll have to find a way to try out a G-Sync monitor sometime though. There's a chance I just think what I have is good but I'll be blown away by G-Sync. I used to think 60Hz was just fine, I tried out 144 and was like "what's the big deal, it's basically the same" but then when I went back to 60 I could easily see the difference and learned about all that I've been missing out on.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: please help me understand some things about 240Hz and G-

Post by Chief Blur Buster » 22 May 2018, 13:03

chocobo wrote:I meant 210 fps at 240Hz without G-Sync or Freesync or Vsync or anything, vs 144Hz with G-Sync. AFAIK when compared to the 144Hz G-Sync monitor, on my 240Hz I'm getting smoother visuals (as in more frames per second) at the cost of lacking G-Sync (some screen tearing), but the tearing is so small and hard to notice I think the extra FPS is worth it. The tears are so small and quick compared to 60 fps (where the top half of the screen can be severely misaligned with the bottom and that frame stays visible for four times as long... I've got 1/4 of the misalignment for 1/4 of the duration, so I can't even see it unless I try.)
There is a percentage factor.

Overkill helps a lot. 500fps at 240Hz VSYNC OFF non-GSYNC will often still look better than 40fps GSYNC anyway.

But 210fps GSYNC -- to my eyes -- still looks better than 250fps non-GSYNC. There's 10 microstutters per second (250 MOD 240) and I still see them when the microstutter frequency gets sufficiently low enough (e.g. when framerates fluctuates through harmonic multiples/divisors of fixed refresh rates).

The approximate percentage (penalty) is about 30-50%. Meaning a GSYNC-smoothed framerate about 30%-50% lower begins to be the threshold of preference for many people. In other words, GSYNC 45fps looks better than VSYNC OFF 55fps. And GSYNC 210fps looks better than VSYNC OFF 250fps. And so on.

Sometimes it's very marginal, much like most people cannot tell the difference between 275fps and 325fps VSYNC OFF -- you need progressively bigger jumps. So 45fps-vs-60fps (33% difference) is more noticeable than 145fps-vs-160fps (10% difference). So you must compare framerate-feel differences by percentage jumps instead, not by framerate count jumps.

VSYNC OFF 200fps versus 500fps definitely is noticeable, even at 144Hz -- the microstutter amplitude is smaller because the frametimes of the visible frameslices are more well-aligned to pixel visibility times. In programming terms, "gametimes" are more in sync with photons hitting human eyeballs = less microstutter for 500fps@144Hz VSYNC OFF than 200fps@144Hz.

The VSYNC OFF frameslices that are streamed onto the display scanout (scanout interruptions = tearlines) simply means next frameslice has that much less gametime-vs-photons aliasing effect, ala less microstutter at progressively bigger excess framerate far beyond refreshrate. It's yet another Advantage Of Framerates Above Refreshrates, which also has a diagram showing the progressively-smaller tearline-offsets of higher framerates too.

In other words, the framerate-vs-refreshrate temporal aliasing effect is smaller (fps vs Hz aliasing = stutter!)

Just like for framerates -- milking the diminishing returns curve also applies to refresh rates too -- 120Hz -> 240Hz -> 480Hz -> 960Hz is definitely noticeable if you jump up bigger -- see Blur Busters Law: And The Amazing Journey To Future 1000Hz Displays.
chocobo wrote:I think my level of sensitivity to those things is lower than some people's. At 144Hz I see a tolerably small but still slightly annoying bit of tearing, at 240Hz I don't normally notice it (though I can find the tearing if I try hard to look for it). It seems like 240Hz minimizes tearing/stuttering enough that it's almost as good as G-Sync, and since I got a good price on the Alienware I think I'm satisfied with what I've got.

I'll have to find a way to try out a G-Sync monitor sometime though. There's a chance I just think what I have is good but I'll be blown away by G-Sync. I used to think 60Hz was just fine, I tried out 144 and was like "what's the big deal, it's basically the same" but then when I went back to 60 I could easily see the difference and learned about all that I've been missing out on.
That is possible.

If you are insensitive to 60fps versus 144fps, the GSYNC-smoothing effect might be less noticeable.

However, sometimes it's a "I don't know what it's like until I see it" factor. Just like you noticed when you downgraded 144Hz to 60Hz.

To help judge better, one can try out the multiple different TestUFO GSYNC stutter dems I've created, to gauge an impression of the approximate GSYNC-smoothing effect (simulated, as unavoidably may be, it's one of the best animations available before buying a GSYNC monitor).

The good news is that if you live near a major city -- the resale values of 240Hz monitors are currently fairly good at the moment (while avoiding expensive shipping which can cost >$100 which is why I recommend in-person pickup instead, unless the purchaser pays the expensive shipping).

So if you decide to upgrade soon, selling your 240Hz will usually return most of your original purchase price if you got a 240Hz at a great discount (sometimes you'll even make a profit, especially if you got a 240Hz GSYNC during Black Friday last year). Since local purchase of 240Hz is usually higher than the price you paid. Essentially you've equivalently "rented" a 240Hz monitor by reselling it and getting most of the money back (if you're lucky enough to live in a dense area that makes it easy to sell on eBay without shipping).

At least, you should owe it to yourself to demo GSYNC for real, but remember -- GSYNC improvement varies a lot from game to game. GSYNC improvement is not very big in some games, but night-and-day in others.

I do indeed recommend getting VRR. It works! But of course, it's not a solve-all for 100% of human population.
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!

User avatar
RealNC
Site Admin
Posts: 3740
Joined: 24 Dec 2013, 18:32
Contact:

Re: please help me understand some things about 240Hz and G-

Post by RealNC » 22 May 2018, 13:18

As Chief mentioned, some games benefit a huge amount by g-sync. It's a freakin' godsent for many games. If you play a game where just looking in one direction gets you 120FPS and then looking at another direction gets you 80FPS, using g-sync results in you not caring about these FPS differences. It always looks perfect.

For games where you're always getting really high frame rates, there's not much of a problem to begin with. Even plain old vsync is "good enough" if you're running 144Hz and the game always reaches 144FPS. You can instead opt for motion blur reduction in these cases. But for the games where you get 144FPS here, you move a bit and look somewhere else and suddenly you're at 110FPS, then you go to another area and you're getting 80FPS, then 100, then 90... Well, g-sync is such a huge improvement, it's not even funny.
SteamGitHubStack Overflow
The views and opinions expressed in my posts are my own and do not necessarily reflect the official policy or position of Blur Busters.

Post Reply