Blur Buster's G-SYNC 101 Series Discussion

Talk about NVIDIA G-SYNC, a variable refresh rate (VRR) technology. G-SYNC eliminates stutters, tearing, and reduces input lag. List of G-SYNC Monitors.
User avatar
jorimt
Posts: 2481
Joined: 04 Nov 2016, 10:44
Location: USA

Re: Blur Buster's G-SYNC 101 Series Discussion

Post by jorimt » 16 Feb 2022, 20:58

gameinn wrote:
16 Feb 2022, 20:46
Thanks. The thing concerning me was the right SIM value being much higher than the other screenshot.

I don't know how long the values take to change so I expect if I kept it open more than 10 seconds it would stabilise lower but then it doesn't explain why the non reflex enabled screenshot didn't show this behavior.
The answer is frametime variance.

That 7.5ms means at least one frame rendered at an average of 134 FPS during that session, so since that SIM column shows the maximum frametime that has been reached in the current game session, it will only change if an even higher maximum is reached in the same session. It would eventually happen with the non-Reflex scenario as well.

Again, the SIM readout doesn't measure input latency, it measures frametime, which will vary by achievable framerate scene-to-scene, session-to-session.
(jorimt: /jor-uhm-tee/)
Author: Blur Busters "G-SYNC 101" Series

Displays: ASUS PG27AQN, LG 48CX VR: Beyond, Quest 3, Reverb G2, Index OS: Windows 11 Pro Case: Fractal Design Torrent PSU: Seasonic PRIME TX-1000 MB: ASUS Z790 Hero CPU: Intel i9-13900k w/Noctua NH-U12A GPU: GIGABYTE RTX 4090 GAMING OC RAM: 32GB G.SKILL Trident Z5 DDR5 6400MHz CL32 SSDs: 2TB WD_BLACK SN850 (OS), 4TB WD_BLACK SN850X (Games) Keyboards: Wooting 60HE, Logitech G915 TKL Mice: Razer Viper Mini SE, Razer Viper 8kHz Sound: Creative Sound Blaster Katana V2 (speakers/amp/DAC), AFUL Performer 8 (IEMs)

smoothnobody
Posts: 10
Joined: 13 Feb 2022, 17:04

Re: Blur Buster's G-SYNC 101 Series Discussion

Post by smoothnobody » 16 Feb 2022, 21:54

jorimt wrote:
16 Feb 2022, 09:58
With G-SYNC, "Ultra" sets an auto fps limit slightly below the refresh rate to keep G-SYNC in range, whereas "On" doesn't.
this sounds no different than LLM on + frame rate limiter. only difference you manually set it below refresh rate instead of letting ultra do it.
jorimt wrote:
16 Feb 2022, 09:58
"Ultra" uses a different method to reduce the pre-rendered frames in the render queue with a "just in time" delivery component, whereas "On" simply sets the max pre-rendered frames to "1."
this clearly does sound different though. i guess this is where ultras secret sauce comes in.
jorimt wrote:
16 Feb 2022, 09:58
there's no harm in having it enabled.
good to know. thank you sir.

User avatar
jorimt
Posts: 2481
Joined: 04 Nov 2016, 10:44
Location: USA

Re: Blur Buster's G-SYNC 101 Series Discussion

Post by jorimt » 16 Feb 2022, 22:37

smoothnobody wrote:
16 Feb 2022, 21:54
this sounds no different than LLM on + frame rate limiter. only difference you manually set it below refresh rate instead of letting ultra do it.
But for that, yes, where G-SYNC and the Ultra auto limiter is specifically concerned.
smoothnobody wrote:
16 Feb 2022, 21:54
this clearly does sound different though. i guess this is where ultras secret sauce comes in.
The upside of the "secret sauce" is potentially less pre-rendered frames than with "On," but the downside is potentially more stutter. Again, depends on the system/game. Varies.
smoothnobody wrote:
16 Feb 2022, 21:54
jorimt wrote:
16 Feb 2022, 09:58
there's no harm in having it enabled.
good to know. thank you sir.
To be clear, there's no harm in having LLM enabled if you're playing in an unsupported game/engine, since LLM won't apply, even if globally active. If LLM is supported, depending on the scenario, it can increase the frequency and/or severity of frametime spikes by starving the render queue in GPU-bound scenarios.

Unlike LLM (both "On" and "Ultra"), Reflex isn't at risk of causing this, so it should be used in LLM's place when and where available.
(jorimt: /jor-uhm-tee/)
Author: Blur Busters "G-SYNC 101" Series

Displays: ASUS PG27AQN, LG 48CX VR: Beyond, Quest 3, Reverb G2, Index OS: Windows 11 Pro Case: Fractal Design Torrent PSU: Seasonic PRIME TX-1000 MB: ASUS Z790 Hero CPU: Intel i9-13900k w/Noctua NH-U12A GPU: GIGABYTE RTX 4090 GAMING OC RAM: 32GB G.SKILL Trident Z5 DDR5 6400MHz CL32 SSDs: 2TB WD_BLACK SN850 (OS), 4TB WD_BLACK SN850X (Games) Keyboards: Wooting 60HE, Logitech G915 TKL Mice: Razer Viper Mini SE, Razer Viper 8kHz Sound: Creative Sound Blaster Katana V2 (speakers/amp/DAC), AFUL Performer 8 (IEMs)

gameinn
Posts: 43
Joined: 16 Nov 2020, 10:11

Re: Blur Buster's G-SYNC 101 Series Discussion

Post by gameinn » 17 Feb 2022, 02:02

Forgive me if it feels like I'm tying to undermine the research done/suggestions given in the 101 article (definitely not my intention) but I always like to try and get info from various sources before I make an informed decision on what should and shouldn't be done.

I've been looking around and this video says most of the same but instead says that any in game vsync option should actually be turned on, for Overwatch anyway.

https://www.youtube.com/watch?v=Gub1bI12ODY

Both in the essentials and smooth scenarios.

smoothnobody
Posts: 10
Joined: 13 Feb 2022, 17:04

Re: Blur Buster's G-SYNC 101 Series Discussion

Post by smoothnobody » 17 Feb 2022, 02:16

i thought of another question. ultra has it's own built in FPS limiter, but ultra doesn't work in every game. so i'm assuming i still want my NVCP FPS limiter turned on for games that don't play nice with ultra. but what happens when a game does work with ultra? who gets to control the FPS, ultra, or the NVCP limiter?

User avatar
RealNC
Site Admin
Posts: 3741
Joined: 24 Dec 2013, 18:32
Contact:

Re: Blur Buster's G-SYNC 101 Series Discussion

Post by RealNC » 17 Feb 2022, 03:16

smoothnobody wrote:
17 Feb 2022, 02:16
i thought of another question. ultra has it's own built in FPS limiter, but ultra doesn't work in every game. so i'm assuming i still want my NVCP FPS limiter turned on for games that don't play nice with ultra. but what happens when a game does work with ultra? who gets to control the FPS, ultra, or the NVCP limiter?
Whatever is the lower cap. For example at 144Hz, ultra caps to 138FPS. If you set a limit of 120FPS, then 120 wins.
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
jorimt
Posts: 2481
Joined: 04 Nov 2016, 10:44
Location: USA

Re: Blur Buster's G-SYNC 101 Series Discussion

Post by jorimt » 17 Feb 2022, 09:32

smoothnobody wrote:
17 Feb 2022, 02:16
i thought of another question. ultra has it's own built in FPS limiter, but ultra doesn't work in every game. so i'm assuming i still want my NVCP FPS limiter turned on for games that don't play nice with ultra. but what happens when a game does work with ultra? who gets to control the FPS, ultra, or the NVCP limiter?
What @RealNC said; the limiter with the lowest cap applies.
gameinn wrote:
17 Feb 2022, 02:02
Forgive me if it feels like I'm tying to undermine the research done/suggestions given in the 101 article (definitely not my intention) but I always like to try and get info from various sources before I make an informed decision on what should and shouldn't be done.

I've been looking around and this video says most of the same but instead says that any in game vsync option should actually be turned on, for Overwatch anyway.

https://www.youtube.com/watch?v=Gub1bI12ODY

Both in the essentials and smooth scenarios.
The "jorimt vs. Battle(non)sense NVCP V-SYNC or in-game V-SYNC w/G-SYNC" subject has been covered in these forums and elsewhere multiple times.

The fact is, neither of us is "wrong," per say, and I only recommend NVCP V-SYNC in my optimal settings because it's virtually always guaranteed to engage, whereas not all in-game solutions do, and/or some may introduce non-V-SYNC behaviors that may or may not be beneficial, neither of which would directly relate to G-SYNC operation.

And ultimately, I say to try both in my original article if any undesired behavior is seen, because it can sometimes be game-dependent:
https://blurbusters.com/gsync/gsync101- ... ttings/14/
Nvidia Control Panel V-SYNC vs. In-game V-SYNC

While NVCP V-SYNC has no input lag reduction over in-game V-SYNC, and when used with G-SYNC + FPS limit, it will never engage, some in-game V-SYNC solutions may introduce their own frame buffer or frame pacing behaviors, enable triple buffer V-SYNC automatically (not optimal for the native double buffer of G-SYNC), or simply not function at all, and, thus, NVCP V-SYNC is the safest bet.

There are rare occasions, however, where V-SYNC will only function with the in-game option enabled, so if tearing or other anomalous behavior is observed with NVCP V-SYNC (or visa-versa), each solution should be tried until said behavior is resolved.
Again, when enabled with G-SYNC, the only thing the V-SYNC "option" does (I say "option" because G-SYNC + V-SYNC is technically G-SYNC + G-SYNC) is strictly adhere to the VBLANK as to prevent tearing inside the refresh rate whether the given frame is in-time, early, or late.

Finally, if you use G-SYNC + in-game V-SYNC instead of G-SYNC + NVCP V-SYNC, the LLM "Ultra" auto limit typically won't engage, same goes for Reflex in most cases, so that's something to consider as well.
(jorimt: /jor-uhm-tee/)
Author: Blur Busters "G-SYNC 101" Series

Displays: ASUS PG27AQN, LG 48CX VR: Beyond, Quest 3, Reverb G2, Index OS: Windows 11 Pro Case: Fractal Design Torrent PSU: Seasonic PRIME TX-1000 MB: ASUS Z790 Hero CPU: Intel i9-13900k w/Noctua NH-U12A GPU: GIGABYTE RTX 4090 GAMING OC RAM: 32GB G.SKILL Trident Z5 DDR5 6400MHz CL32 SSDs: 2TB WD_BLACK SN850 (OS), 4TB WD_BLACK SN850X (Games) Keyboards: Wooting 60HE, Logitech G915 TKL Mice: Razer Viper Mini SE, Razer Viper 8kHz Sound: Creative Sound Blaster Katana V2 (speakers/amp/DAC), AFUL Performer 8 (IEMs)

gameinn
Posts: 43
Joined: 16 Nov 2020, 10:11

Re: Blur Buster's G-SYNC 101 Series Discussion

Post by gameinn » 17 Feb 2022, 17:52

jorimt wrote:
16 Feb 2022, 20:58
That 7.5ms means at least one frame rendered at an average of 134 FPS during that session, so since that SIM column shows the maximum frametime that has been reached in the current game session, it will only change if an even higher maximum is reached in the same session. It would eventually happen with the non-Reflex scenario as well.
I tried to see if this was the case and unless I need more time I stared at reflex vs non reflex for 5 minutes each. The same behavior. Non reflex would be rock solid but reflex on would always have the highest value fluctuate up to 7-8.

I don't understand why because unless I mis understood nvidia reflex, it only activates or has an effect if I am gpu bound. In no way is a 3070 being GPU bound at ~170 fps on Overwatch.

User avatar
jorimt
Posts: 2481
Joined: 04 Nov 2016, 10:44
Location: USA

Re: Blur Buster's G-SYNC 101 Series Discussion

Post by jorimt » 17 Feb 2022, 18:36

gameinn wrote:
17 Feb 2022, 17:52
I tried to see if this was the case and unless I need more time I stared at reflex vs non reflex for 5 minutes each. The same behavior. Non reflex would be rock solid but reflex on would always have the highest value fluctuate up to 7-8.

I don't understand why because unless I mis understood nvidia reflex, it only activates or has an effect if I am gpu bound. In no way is a 3070 being GPU bound at ~170 fps on Overwatch.
If your GPU usage is always below 99% in that game (or any game), you don't need to have Reflex enabled. Disable Reflex and set a manual 162 or under in-game limit instead.

As for Reflex not having an effect in non-GPU-bound scenarios, that's not necessarily the case. Reflex sets a dynamic FPS limit slightly below the currently achievable FPS, so if the system can only reach, say, 134 FPS in a given moment, Reflex will set a limit slightly below that as to avoid the GPU maxing.

My guess is you're seeing higher maximum frametimes in that scenario (at least in the SIM readout, and who knows how accurate it actually is) because whenever there is a significant enough frametime variance, Reflex kicks in and overcompensates with too low a limit in the split second of that variance or spike, causing a higher maximum frametime than with it off.

There can (but not always will there) be detriments to having settings like LLM (either "On" or "Ultra"), Reflex, and even HAGS on when they're not needed. As I keep saying, any setting involving the render queue has highly variable and situational effects, and is primarily only beneficial in further reducing latency in GPU-bound scenarios.
(jorimt: /jor-uhm-tee/)
Author: Blur Busters "G-SYNC 101" Series

Displays: ASUS PG27AQN, LG 48CX VR: Beyond, Quest 3, Reverb G2, Index OS: Windows 11 Pro Case: Fractal Design Torrent PSU: Seasonic PRIME TX-1000 MB: ASUS Z790 Hero CPU: Intel i9-13900k w/Noctua NH-U12A GPU: GIGABYTE RTX 4090 GAMING OC RAM: 32GB G.SKILL Trident Z5 DDR5 6400MHz CL32 SSDs: 2TB WD_BLACK SN850 (OS), 4TB WD_BLACK SN850X (Games) Keyboards: Wooting 60HE, Logitech G915 TKL Mice: Razer Viper Mini SE, Razer Viper 8kHz Sound: Creative Sound Blaster Katana V2 (speakers/amp/DAC), AFUL Performer 8 (IEMs)

gameinn
Posts: 43
Joined: 16 Nov 2020, 10:11

Re: Blur Buster's G-SYNC 101 Series Discussion

Post by gameinn » 17 Feb 2022, 21:53

I think Reflex is fighting against itself or something. My monitor has an OC mode to 180 Hz so if I use that reflex seems to cap fps at 171.

If I then set a frame limit below 171 then the game behaves fine. As soon as I set 172 the issues start to happen.

Post Reply