Next

G-SYNC 101: Control Panel

Published Jun 20, 2017 by in: G-SYNC

G-SYNC Module

The G-SYNC module is a small chip that replaces the display’s standard internal scaler, and contains enough onboard memory to hold and process a single frame at a time.

The module exploits the vertical blanking interval (the span between the previous and next frame scan) to manipulate the display’s internal timings; performing G2G (gray to gray) overdrive calculations to prevent ghosting, and synchronizing the display’s refresh rate to the GPU’s render rate to eliminate tearing, along with the delayed frame delivery and adjoining stutter caused by traditional syncing methods.

G-SYNC Demo

The below Blur Busters Test UFO motion test pattern uses motion interpolation techniques to simulate the seamless framerate transitions G-SYNC provides within the refresh rate, when directly compared to standalone V-SYNC.

G-SYNC Activation

“Enable for full screen mode” (exclusive fullscreen functionality only) will automatically engage when a supported display is connected to the GPU. If G-SYNC behavior is suspect or non-functioning, untick the “Enable G-SYNC, G-SYNC Compatible” box, apply, re-tick, and apply.

Blur Buster's G-SYNC 101: Control Panel

G-SYNC Windowed Mode

“Enable for windowed and full screen mode” allows G-SYNC support for windowed and borderless windowed mode. This option was introduced in a 2015 driver update, and by manipulating the DWM (Desktop Windows Manager) framebuffer, enables G-SYNC’s VRR (variable refresh rate) to synchronize to the focused window’s render rate; unfocused windows remain at the desktop’s fixed refresh rate until focused on.

G-SYNC only functions on one window at a time, and thus any unfocused window that contains moving content will appear to stutter or slow down, a reason why a variety of non-gaming applications (popular web browsers among them) include predefined Nvidia profiles that disable G-SYNC support.

Note: this setting may require a game or system restart after application; the “G-SYNC Indicator” (Nvidia Control Panel > Display > G-SYNC Indicator) can be enabled to verify it is working as intended.

G-SYNC Preferred Refresh Rate

“Highest available” automatically engages when G-SYNC is enabled, and overrides the in-game refresh rate selector (if present), defaulting to the highest supported refresh rate of the display. This is useful for games that don’t include a selector, and ensures the display’s native refresh rate is utilized.

“Application-controlled” adheres to the desktop’s current refresh rate, or defers control to games that contain a refresh rate selector.

Note: this setting only applies to games being run in exclusive fullscreen mode. For games being run in borderless or windowed mode, the desktop dictates the refresh rate.

G-SYNC & V-SYNC

G-SYNC (GPU Synchronization) works on the same principle as double buffer V-SYNC; buffer A begins to render frame A, and upon completion, scans it to the display. Meanwhile, as buffer A finishes scanning its first frame, buffer B begins to render frame B, and upon completion, scans it to the display, repeat.

The primary difference between G-SYNC and V-SYNC is the method in which rendered frames are synchronized. With V-SYNC, the GPU’s render rate is synchronized to the fixed refresh rate of the display. With G-SYNC, the display’s VRR (variable refresh rate) is synchronized to the GPU’s render rate.

Upon its release, G-SYNC’s ability to fall back on fixed refresh rate V-SYNC behavior when exceeding the maximum refresh rate of the display was built-in and non-optional. A 2015 driver update later exposed the option.

This update led to recurring confusion, creating a misconception that G-SYNC and V-SYNC are entirely separate options. However, with G-SYNC enabled, the “Vertical sync” option in the control panel no longer acts as V-SYNC, and actually dictates whether, one, the G-SYNC module compensates for frametime variances output by the system (which prevents tearing at all times. G-SYNC + V-SYNC “Off” disables this behavior; see G-SYNC 101: Range), and two, whether G-SYNC falls back on fixed refresh rate V-SYNC behavior; if V-SYNC is “On,” G-SYNC will revert to V-SYNC behavior above its range, if V-SYNC is “Off,” G-SYNC will disable above its range, and tearing will begin display wide.

Within its range, G-SYNC is the only syncing method active, no matter the V-SYNC “On” or “Off” setting.

Currently, when G-SYNC is enabled, the control panel’s “Vertical sync” entry is automatically engaged to “Use the 3D application setting,” which defers V-SYNC fallback behavior and frametime compensation control to the in-game V-SYNC option. This can be manually overridden by changing the “Vertical sync” entry in the control panel to “Off,” “On,” or “Fast.”

Next

3696 Comments For “G-SYNC 101”

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Sort by:   newest | oldest | most liked
SleepSmasher
Member
SleepSmasher

Hey Jorimt,
I’ve been revisiting this guide for over five years now—always a great read. I have a bit of an unusual question: do you know of any docking stations that support full G-SYNC (not just G-SYNC Compatible)?

I recently bought one of the best docks available, and I’m confident bandwidth isn’t the issue—my setup is a 1440p 120Hz display connected via DisplayPort 1.4 to the dock, which then connects to my laptop via USB-C 3.2 (DP Alt Mode). However, the monitor only shows G-SYNC Compatible (even though it has a native G-SYNC module), and performance is noticeably worse—more stutters—compared to using a direct DP-to-USB-C cable plugged into the laptop’s dGPU port.

I couldn’t find anyone online with the same issue, but I imagine not many people are trying to run G-SYNC through a dock. Any insights? Thanks!

Like0Vote Down 
5 months 6 days ago
jorimt
Author
jorimt

I have no experience with using a docking station with VRR, so can’t say.

You didn’t share the model of your dock, but a quick Google search shows both BenQ and Jsaux offer docks with VRR compatibility:
https://www.benq.com/en-us/knowledge-center/knowledge/gaming-vrr-freesync-gsync-docking.html
https://jsaux.com/products/rgb-docking-station-for-steam-deck?srsltid=AfmBOoqZS7bFMdPrbY7k1I4Bhy_x_-zeDusUeh5PaxuUZRKK3oH8TtYl

Whether they’re compliant with your particular configuration and how well they perform compared to non-docked would indeed be the question.

BenQ claims full HDMI 2.1 G-SYNC support, but as you say, this is an atypical VRR use case, so resources on this (both official and community-level) are probably far and few between.

Like0Vote Down 
5 months 6 days ago
mq05
Member
mq05

Hi Jorimt!

I have my setup exactly the same as your recommendation:
gsync on + vsync on

My game doesn’t have an fps limiter so I use rtss to limit it at refresh rate -3. Reflex is also not supported. The game is heavily gpu bound and the stable fps to limit at is 20 which makes the game barely playable.

I prefer the lowest latency possible so in cases like this would LLM on and ultra be better than capping my fps to 20? And if I don’t want to cap my fps, between LLM on and ultra which one is better here?

Like0Vote Down 
5 months 6 days ago
jorimt
Author
jorimt

G-SYNC + V-SYNC + framerate within the physical refresh rate already prevents sync latency, regardless of GPU usage.

If the GPU usage is still 99%+ in that scenario, however, then there will be additional render queue-level latency due to extra pre-rendered frames. This form of latency is separate from the latency G-SYNC addresses, and would occur with G-SYNC off + V-SYNC off as well.

That all said, if you don’t want to prevent 99% GPU usage with a manual 20 FPS limit in that particular game, Reflex is not available in-game, and the game in question is running DX9 or up (LLM is not currently supported in Vulkan, for instance), use LLM “Ultra.” It won’t reduce render queue latency as much as Reflex in GPU-bound situations, but it can typically shave off up to 1 1/2 frame of latency, regardless.

Finally, RTSS does allow forcing of Reflex in games that don’t natively support it, but it will probably reduce your framerate in that game at or near 20 FPS to prevent maxed GPU usage as well, so it may not be a viable option over LLM “Ultra” in your particular use case.

Like0Vote Down 
5 months 6 days ago
mq05
Member
mq05

Thank you so much for the very detailed reply!

Please correct me again if I’m wrong, this is how I understand the whole thing:

With gsync on + vsync on + fps limit at refreshrate – 3 as the baseline.

If my fps can’t reach the limit point and my gpu is 99% and there’s no nvidia reflex support ingame. For the lowest latency possible, lowering FPS limit to avoid gpu getting to 99% is always preferable. Then if the limit point is too low the next best option is using LLM Ultra, right? Is it still required to limit fps to refreshrate – 3 here or just better turn off the limit?

In the article, you recommended using LLM On instead of Ultra, why is Ultra the choice in the case above?

One more thing, I just read about forcing nvdia reflex with rtss on unsupported games, but how exactly does it compare to nvidia reflex in terms of latency? As I understand the default rtss fps limiting method is async, which does give some sort of latency. Does limiting fps with rtss nvdia reflex change anything to the recommended baseline setup and the setup in my case where gpu is 99% and my fps can’t reach refreshrate -3?

Like0Vote Down 
5 months 6 days ago
jorimt
Author
jorimt

With gsync on + vsync on + fps limit at refreshrate – 3 as the baseline.

This is why you’re still confused, I think. The above prevents sync-latency ONLY.

You can have sync latency due to standalone V-SYNC behavior, pre-rendered frame latency due to 99%+ GPU usage, and extra render queue latency due to an external framerate limiter, all at once or individually. They stack.

So when you ask about the effects of latency with maxed GPU usage and it’s interaction with settings like LLM and Reflex, this will all occur with or without VRR.

I.E. 99%+ GPU usage adds the same type and amount of latency whether you have G-SYNC on + V-SYNC on or G-SYNC off + V-SYNC off, as well as whether it occurs because the framerate is above or below the refresh rate.

Thus, the lowest possible tear-free latency is VRR on + V-SYNC option on (which acts as the strict VBLANK adherence component of VRR) + a minimum -3 FPS limit (to ensure the framerate remains within the refresh rate since VRR disengages whenever the framerate is at or above it; it’s not to prevent maxed GPU usage) + an x limit within the refresh rate instead of the -3 FPS with VRR if the average framerate cannot be sustained at that limit without the GPU usage being at 99%+.

So if you have G-SYNC on + V-SYNC on + a -3 FPS limit, and your framerate can’t reach that -3 FPS limit consistently because GPU usage is maxed, you’re now preventing sync latency, but you’ll separately be getting extra render queue latency unless you use Reflex or set a manual limit that keeps the GPU usage consistently below 99%.

Reflex does this automatically without user intervention by monitoring the GPU usage and limiting itself slightly below the achievable average framerate.

A manual limit can achieve this as well, but since it’s not dynamic, you have to find your 99% achievable average in the particular game first, and then limit.

You were doing this with your 20 FPS limit, which you found to be too low to be practically usable.

So in that case, you can achieve a higher average framerate if you remove the limit and accept the extra latency caused by maxed GPU usage. Said latency is due to an increased amount of pre-rendered frames in the render queue, which makes the higher average framerate possible in the first place.

In other words, you need to pick; achieve a higher average framerate with higher latency, or a lower average framerate with lower latency? You can’t have both in this case, unless you reduce your resolution and/or graphical settings to increase the average framerate, or upgrade your specs.

And getting back to limiters, while in-game and Reflex limiters are at the engine-level, and thus typically don’t add any extra latency when engaged, external limiters intercept and limit already rendered frames later on in the process, which typically adds anywhere from 1/2 to 1 frame of latency, but this is still better than not having a limiter when using VRR or being GPU-bound, since both add more latency without a limiter in place.

As to the RTSS Reflex setting, I haven’t formally tested it, so I can’t say for sure, but if it works exactly like in-game Reflex, then again, it will probably automatically limit your framerate to ~20 to avoid maxed GPU usage in your particular game, which I assume is also undesirable.

Like0Vote Down 
5 months 5 days ago
mq05
Member
mq05

Thank you! This was super informative and helpful!

One last thing, in the article you recommend using LLM On over Ultra. Why do you suggest otherwise in my case?

“if you don’t want to prevent 99% GPU usage with a manual 20 FPS limit in that particular game, Reflex is not available in-game, and the game in question is running DX9 or up (LLM is not currently supported in Vulkan, for instance), use LLM “Ultra.”

Also I notice in some games even when I use LLM Off with Gsync & Vsync on, the fps is still automatically capped at -3 without any kind of limiter, is this due to some hidden windows features?

Like0Vote Down 
5 months 4 days ago
jorimt
Author
jorimt

One last thing, in the article you recommend using LLM On over Ultra. Why do you suggest otherwise in my case?

I don’t, technically; my recommendations are in a descending if/else structure.

If you read it again, you’ll see the specific explanations as to why I have LLM “On” before “Ultra.”

I don’t strictly recommend either. I merely added them due to repeat requests and for completeness sake. I.E. for those that insist on combining it with VRR for whatever reason.

My core recommendations for VRR are simply G-SYNC on + NVCP V-SYNC on + framerate at least 3 FPS within the refresh rate per game. Everything else is secondary where VRR-specific configuration is concerned.

Further, while the automatic FPS limiting function with LLM “Ultra” when used with G-SYNC + NVCP V-SYNC is somewhat useful, LLM is barely effective at reducing pre-rendered frames in GPU-bound scenarios, be it “On” or “Ultra” over the default “Off” in the majority of scenarios.

It can also tend to increase frametime spikes in some cases (due to forcibly starving the render queue), when it works, that is; there are many games where it doesn’t work, even when enabled, and it’s difficult to confirm when it does.

I find Reflex or a manual limit for avoiding maxed GPU usage to be far more effective.

I personally leave LLM set to “Off.” Haven’t used it for years.

Most overrated and over-discussed driver-level setting in the history of consumer GPUs, I think.

Also I notice in some games even when I use LLM Off with Gsync & Vsync on, the fps is still automatically capped at -3 without any kind of limiter, is this due to some hidden windows features?

No. If you’re seeing that and you indeed have LLM set to “Off,” assuming the physical refresh rate of your display is 100Hz or so (Reflex sets a -4 FPS limit at 120Hz), then Reflex must be enabled in the game, even if it’s not exposed in the settings.

Reflex is typically combined with settings such as FG, for instance.

That, or you have RTSS open and/or “Max Frame Rate” set in the NVCP with a global or game-specific -3FPS limit set that you forgot to clear.

Like0Vote Down 
5 months 4 days ago
mq05
Member
mq05

Thank you! This pretty much answers every question I have. Appreciate all the effort you put into your work!

Like1Vote Down 
5 months 3 days ago
Sequinoz
Member
Sequinoz

Hi Jorimt! Really appreciate your effort for making this guide and still answering questions to this day.

Quick question: With your recommendation of -3 FPS below the monitor’s Hz and using RTSS, is this with Passive Waiting enabled or disabled?

Like0Vote Down 
5 months 8 days ago
jorimt
Author
jorimt

Since this is a G-SYNC-specific and not a general latency/performance article, my minimum -3 FPS limit recommendation has the sole purpose of ensuring VRR remains engaged by keeping the framerate within the refresh rate; it does not take into consideration further reduction of non-sync latency.

As such, it assumes RTSS is set to its out-of-box defaults if used as a limiter with VRR, at which point the user can deviate from there, as desired.

There are plenty of other things you can do to further reduce non-sync-latency over G-SYNC on + V-SYNC on + minimum -3 FPS, as well as improve frametime performance consistency, but that’s ultimately not the goal of this VRR-centric article or its recommendations.

As for RTSS passive waiting, there are detailed tooltips available for each setting:

comment image

“Enable passive waiting” is the default, and reduces CPU power consumption when on. When off, it may increase framepacing precision vs. on.

Like1Vote Down 
5 months 8 days ago
Sequinoz
Member
Sequinoz

Thanks for the response Jorimt. If I understand correctly, strict framepacing precision doesn’t really matter so long as the FPS is within the VRR range.

However, your recommended -3 FPS limit seems to take into account frametime variance. So, if I disable passive waiting, does that mean there’s less frametime variance, allowing me to get away with -1 FPS limit and still avoid engaging V-Sync?

Like0Vote Down 
5 months 7 days ago
jorimt
Author
jorimt

How little you need to limit below your physical refresh rate with VRR is entirely dependent on the frametime performance your system + game can achieve with the limiter engaged.

-3 FPS is simply a “safe” minimum starting point.

A -1 FPS limit that would 100% avoid reversion to standalone V-SYNC behavior with G-SYNC + V-SYNC probably isn’t practically achievable in 99% of scenarios, even with a -1 FPS RTSS limit and passive waiting disabled.

Like1Vote Down 
5 months 7 days ago
leovaz
Member
leovaz

Hey, might be kind of dumb question (even after reading this article multiple times). In the conclusion you guys say that “and when used with G-SYNC + FPS limit, it will never engage,” that means that with an fps limiter the NVCP v-sync wont engage because the fps wont surpass the g-sync range.

So, what is the point of having v-sync on? If fps limiter is enough to make gsync on all the time, i dont get the need to leave v-sync on.

Im surely misunderstanding something, there are parts of this article that are way beyond my tech knowledg3.

Like0Vote Down 
5 months 10 days ago
jorimt
Author
jorimt

In the conclusion you guys say that “and when used with G-SYNC + FPS limit, it will never engage,” that means that with an fps limiter the NVCP v-sync wont engage because the fps wont surpass the g-sync range.

So, what is the point of having v-sync on? If fps limiter is enough to make gsync on all the time, i dont get the need to leave v-sync on.

Short answer is standalone V-SYNC behavior won’t engage in that scenario, but the V-SYNC option will force G-SYNC to adhere to the VBLANK as to prevent all possible tearing while G-SYNC on + V-SYNC off won’t.

Did you also read my Closing FAQ, including entry #2? Because it’s the most frequently asked question I get:
https://blurbusters.com/gsync/gsync101-input-lag-tests-and-settings/15/

Like2Vote Down 
5 months 10 days ago
leovaz
Member
leovaz

Thanks for the quick response, just read the FAQ and understood.

So if im not having any tearing at all, v-sync on x off wont make a diference right?

Does leaving V-sync on have any affect on improving stutter (as g-sync does)?

Like0Vote Down 
5 months 10 days ago
jorimt
Author
jorimt

So if im not having any tearing at all, v-sync on x off wont make a diference right?

If you read the FAQ, you have your answer; G-SYNC on + V-SYNC off can tear, G-SYNC on + V-SYNC on can’t tear.

Whether the given user can perceive the tearing artifacts with the former is the question, especially the higher the physical refresh rate.

Does leaving V-sync on have any affect on improving stutter (as g-sync does)?

G-SYNC, with or without the V-SYNC option, cannot reduce non-sync-related stutter.

Further, while forcing VBLANK adherence can’t cause stutter, it can elongate frametimes spikes over letting it tear, but there’s no way around that if you want zero tearing. Pick your poison, essentially.

Like0Vote Down 
5 months 10 days ago
tmdangp
Member
tmdangp

My monitor has a 165hz refresh rate. As some custcenes, video content and some games are displayed at 30hz/30fps, would it be more beneficial if I set my refresh rate to 120hz? Would g-sync/free-sync still work as intended? And, would I need to change another setting on the control panel?

Like0Vote Down 
5 months 14 days ago
jorimt
Author
jorimt

That only applies to fixed refresh rate scenarios; if you’re using G-SYNC, 165Hz 30 FPS games won’t have the issue, since 30 FPS 165Hz VRR essentially = 30Hz (w/LFC).

Like1Vote Down 
5 months 14 days ago
tmdangp
Member
tmdangp

I see thank you.

Like1Vote Down 
5 months 13 days ago
wpDiscuz

Related Content

Featured Content

Blur Buster Forums

Blur Busters Forums

TestUFO Motion Tests

TestUFO Motion Tests