240hz on 60FPS games

[Chief Blur Buster]

CRTs cannot bypass scanout lag (e.g. slow-scanning 60Hz video signal).

While their absolute lag is practically zero, there's a mandatory cable scanout lag.

What is the standard defination of "absolute lag"?

Generally speaking (BlurBusters definition, at least) -- "absolute lag" = lag from pixel being transmitted out of GPU framebuffer, to the corresponding pixel (at X,Y coordinates) emitting light. CRT has (in practical terms) zero absolute lag since there's no buffer or delay between pixel from GPU framebuffer, and pixel display. A perfectly synchronous cable scanout to display scanout, in a lagless 1:1 mapping. Even though there is zero absolute lag, there's still the mandatory scanout lag (overcome by higher Hz).

While absolute lag of LCD monitors are not as fully zeroed out, absolute lag of modern gaming monitors have fallen pretty low for many models (something like 2ms) -- this is from several factors like digital codec overhead (micropackets), monitor processing, LCD GtG, etc. One can minimize absolute lag on an LCD by using synchronous scanout, minimized processing (linebuffered processing), and fast pixel response (1ms TN GtG).

[Chief Blur Buster]

240Hz for 60FPS gaming is a waste. The difference in latency is minimal (2.8ms difference to 144Hz, which is virtually nothing.)

Difference is minimal, yes.
Waste, I don't dictate that word "waste" -- it's the reader's choice to decide.

The reader may play other games, or playing a ultra-high-framerate FPS where the extra visual information matters a lot more than during emulating low-lag VSYNC ON (e.g. for an arcade/console game) for example.

But yes -- in this particular case, if you're only concerned about 60fps content, I'd save money and get 144Hz GSYNC. 60fps @ 144Hz GSYNC is much cheaper and you get 90% of the lag savings (VSYNC ON eliminated, plus the 6.9ms scanout). The further improvement to 4.1ms scanout is extremely tiny. While the scanout lag is lower, the absolute lag offset of a 240Hz monitor might actually be +1-2ms above a very good well-selected cheaper 144Hz monitor, wiping out the scanout lag advantage. (To be tested -- absolute lag offsets can be a completely different lag variable than the lag gradient of scanout lag).

240Hz GSYNC is only if you've got unlimited budget and want the absolute fastest scanout money can buy.

[crossjeremiah]

I understand what you're saying. My XL2430T I can't really notice any tearing in Melee, or anything negative.
So a GSYNC monitor's input lag at 144hz + GSYNC(novsync) at 60fps has a lil bit more input lag than a 144hz(novsync) at 60fps. Just no tearing.
I know the new benq 240hz beat the response times on a crt .
I'm currently looking at this https://www.dell.com/en-us/shop/alienwa ... ccessories . If theres a cheaper monitor out there or more cost efficient let me know. because stuff about input lag on certain gsync monitors is a rare sight. the only sources i know are tomshardware and displaylag. tom usually updates more often but uses a different method than the rest. displaylag has been awol for a bit.

[lexlazootin]

If you just go to your site of choice and type in 240hz G-Sync you should see all of your options, they will perform pretty much the same since it's all the same chip so just go with the cheapest.

You will often find the Dell or the Acer 240hz G-Sync are the cheapest but it depends on when you look or the shop you look at

[KKNDT]

Summary:

• Stutter is not caused by repeat refresh cycles on sample-and-hold displays (non-strobed LCDs, no black periods)
• Stutter can be caused by repeat black periods (same frame strobed multiple times: LightBoost/PWM dimming/etc)
• Stutter can be caused by varying frame visibility times
• Stutter can also have other causes too (e.g. hard disk accesses) which influences the above (e.g. varying frame visibility times)
• For VSYNC ON, yes -- 60fps @ 120 Hz can look better than 60fps @ 144Hz
• For GSYNC, games with good framepacing, 60fps at any GSYNC look exactly like 60fps @ 60Hz VSYNC ON
• For GSYNC, games with terrible framepacing, 60fps at any GSYNC can be worse than 60fps @ 60Hz VSYNC ON
• For GSYNC, RTSS can help fix framepacing issues
• GSYNC at higher Hz is lower lag. 60fps @ 240Hz GSYNC has less lag than 60fps @ 144Hz GSYNC
• For non-strobed, 60fps@60Hz and 60fps@120Hz and 60fps@240Hz and 60fps@(any Hz GSYNC/FreeSync) look identical (no stutter) provided game has perfect framepacing. Exact same grand totalled identical-frame visibility time, if the display is sample-and-hold (non-LightBoost).
• During VSYNC OFF, many gaming monitors are now line-buffered (closer to a CRT). Cable scanout can be same as panel scanout, e.g. sub-frame latencies on a pixel-to-pixel basis. There's no full framebuffer lag when it comes to these monitors.
• Assuming monitor processing the same, the higher GSYNC Hz, the lower lag versus VSYNC ON. 60fps @ 240Hz GSYNC has much, much less lag than VSYNC ON 60fps @ 60Hz
• Different legitimate methods of input lag tests create very different results (e.g. SMTT versus Leo Bodnar versus TomsHardware approach) simply because they use different stopwatching (different stopwatch start trigger, and different stopwatch end trigger). Can't generalize based on just one lag test. Synthetic lag tests are often disconnected from actual real-world lag tests (e.g. high speed camera on VSYNC OFF CS:GO).

Very valuable summary!

[email protected] (5.67 KiB) Viewed 9461 times

I drew a very simple diagram, which is ideal 120 FPS @ 144 HZ. So, the stutter occurs in the refresh cycle 5#?

[Chief Blur Buster]

No. Divergence is 3rd and 4th, so the stutter is on the 4th object.

(Good VSYNC OFF diagram though -- it is very accurate diagram otherwise for a VSYNC OFF example -- where the black lines are refresh cycles, and the colored boundaries are VSYNC OFF tearlines).

Draw a straight line through all balls.
For a specific pixel in a refresh cycle (i.e. in the same physical row of pixels, for horizontal eye tracking), that is your corresponding horizontal eye gaze position relative to time.
Stutters are for all balls that don't go on the straight line, and jumps over the line.
So between the 3rd and 4th ball is the stutter, and you see it right at the moment of the 4th refresh cycle (4th ball).
Your ball is essentially "vibrating" centered on your moving gaze point.



I also wrote the GSYNC verison eye-track diagrams of this at:
How Does GSYNC Fix Stutters?

I also quoted Microsoft's similar eye-track diagram at:
Why Does OLED Have Motion Blur?

And others do too, Michael Abrash, Chief Scientist of Oculus, wrote eye-track diagrams at:
Down The VR Rabbit Hole: Fixing Judder

RTings has a great YouTube animation of stutters converting into motion blur. Jump to around 1 min into video:

phpBB [video]

Basically, High speed vibrations (stutters above flicker fusion threshold) = display motion blur. Which I also explain at http://www.blurbusters.com/1000hz-journey .... Yes, few people realize that stutters & display motion blur actually have the same root cause! (basically moving objects diverging from eye-track gaze point) --- low-frequency divergences show up as stutter, and high-frequency divergences blend as motion blur.

Just like a guitar string. A slow vibrating guitar string looks vibrating. A fast vibrating guitar string is blurry (vibrating too fast to see vibrations). Display motion blur is the same thing. Even at 240Hz. Even at 480Hz. We can't see the framerates (vibrations) directly but we see the side effects (blurriness of vibration) directly.

That's why we can see a 500Hz vibrating string, because the string is blurry.
That's also why 500Hz display refresh rate isn't even good enough to say goodbye to strobing (yet).
500Hz isn't "retina" refresh rate yet.

So, simply really gotta remember that stutter are just simple Geometry 101 divergences away from a constant-moving eye-track gaze point. Easy once you understand the basic concept.





Definitely look at the YouTube video, it will explain why stutters and display motion blur (sample-and-hold) is exactly the same kind of "keep-sync-with-eye-tracking" mathematics challenge. By simply visualizing stutter & blur from the eye-tracking point of view, it's much easier to figure out if you are good enough at math & geometry.

For scanout diagrams and determining stutter, you simply draw a straight line, and now you've turned a scanout diagram into an eye track diagram. (At least for determining stutter in the horizontal dimension, anyway -- but that's simply due to the vertical layout of scanout diagrams on BlurBusters).

[lexebidar]

Let me get this straight:
So if a game is internally locked to 60fps and I play on 240hz monitor with Gsync (always forced 237fps in rtss and vsync in nvcp).
-Should it be left as it is? Then gsync will just make the monitor 60hz but thanks to "vsync gsync" no lag.
-or unlock gsync to force 240hz and lock rtss to 59fps to overcome "bad" internal fps limit of a game?

Also, shouldn't there be less blur if 60fps game runs on 240hz monitor without gsync (so all the time 240hz)?

[lexlazootin]

I believe G-Sync with V-Sync off would be your best option because it has lower latency at these lower fps values.

You want to cap the game below your maximum refresh rate, if it's already capped below then you don't really have to do anything.

The blur should be identical.

[RealNC]

Let me get this straight:
So if a game is internally locked to 60fps and I play on 240hz monitor with Gsync (always forced 237fps in rtss and vsync in nvcp).
-Should it be left as it is? Then gsync will just make the monitor 60hz but thanks to "vsync gsync" no lag.
-or unlock gsync to force 240hz and lock rtss to 59fps to overcome "bad" internal fps limit of a game?

Look, we're talking in circles now. It's really simple:

• If the game has an in-game limiter, you don't need RTSS.
• In-game limiters are usually better than RTSS when using gsync.
• RTSS is better when you're using vsync instead of gsync.

Your game is using an in-game limiter, fixed to 60FPS. So you don't need RTSS, unless the in-game limiter is bad. If you see weird stutter with the in-game limiter, but when using RTSS the stutter is gone, then yeah, use RTSS. But most in-game limiters are fine.

If it turns out the in-game limiter is bad, then find a way to disable it (find a way to "unlock FPS" in your game.) If that's not possible, then you can override the in-game limiter by setting RTSS 1FPS below the in-game limiter. For a 60FPS-locked game, you can set RTSS to 59FPS and it will override the in-game limiter. You don't need to disable gsync for that. If you disable gsync for a 60FPS game, you will get more input lag and more stutter.

Also, shouldn't there be less blur if 60fps game runs on 240hz monitor without gsync (so all the time 240hz)?

No. Blur depends on how many frames per second you see with your eyes. If the game runs at 60FPS, blur will be the same regardless of the refresh rate. To get lower blur with high refresh rates, your FPS must be high too.

[Chief Blur Buster]

I believe G-Sync with V-Sync off would be your best option because it has lower latency at these lower fps values.

No, it doesn't make a difference.

When framerates are well below GSYNC maximum, any far-higher-cap has no effect (RTSS or not, ingame or not, VSYNC OFF or not, VSYNC ON or not). Doesn't matter a bleep -- if you're doing 60fps.

So don't need to disable RTSS and you don't need to change VSYNC ON versus VSYNC OFF, since none of that ever happens.
-- RTSS 235fps cap is never hit during 60fps@240Hz GSYNC
-- VSYNC OFF never occurs during 60fps@240Hz GSYNC
-- VSYNC ON never occur during 60fps@240Hz GSYNC

For games running at internally-capped 60fps during 240Hz GSYNC.

You're already getting lowest lag if:
(A) Game is permanently 60fps locked
(B) You've enabled 240Hz variable refresh rate.

Short of switching to 240Hz VSYNC OFF (non-GSYNC), doing 60fps@240Hz GSYNC is currently getting the absolute lowest lag on the market for 60fps-locked material for anything that has a VSYNC ON look-and-feel experience -- but with none of that lag.

No difference in motion blur, though. 60fps@240Hz GSYNC looks identical to 60fps@60Hz VSYNC ON -- blurwise and stutterfree -- provided the game's 60fps cap has excellent framepacing.

Look, we're talking in circles now. It's really simple:

• If the game has an in-game limiter, you don't need RTSS.
• In-game limiters are usually better than RTSS when using gsync.
• RTSS is better when you're using vsync instead of gsync.

I think he's also asking partially on a convenience-basis too:

That said, if you have RTSS enabled for other games anyway, keeping RTSS enabled will have no effect on the lag of 60fps@240Hz GSYNC.

So if you installed RTSS anyway to help other games (e.g. 237fps or 238fps cap at 240Hz GSYNC), if you accidentally leave RTSS enabled when playing 60fps@240Hz ..... there should be no deleterious effects. Because the cap is never hit, RTSS never activates, it's just ready in the background for the next time you launch a game.

Answer:
Does RTSS ~238fps cap improve 60fps@240Hz GSYNC? No, it doesn't help.
Does RTSS ~238fps cap interfere with 60fps@240Hz GSYNC? No, it's harmless.

Therefore:
If you don't have RTSS installed/active, don't need to install/activate it unless you have at least one game whose framerates hit VRR max.
If you do have RTSS active, don't _need_ to deactivate it for 60fps@240Hz VRR ...

The higher cap (RTSS) never activates if your in-game cap is lower (non-RTSS cap, or 60fps cap, etc).