Page 61 of 65
Re: Blur Buster's G-SYNC 101 Series Discussion
Posted: 20 May 2022, 13:58
by Bubu
Ok, i tried these both settings, surprisingly, the first one (when i selected G-Sync, it automatically enabled Vertical Sync: 3D application parameter), i got 161/162 fps as expected while the recommend setting gaves me 157 fps
Did you know why there is fps difference between these settings?
Re: Blur Buster's G-SYNC 101 Series Discussion
Posted: 22 May 2022, 11:10
by jorimt
Bubu wrote: ↑20 May 2022, 13:58
Ok, i tried these both settings, surprisingly, the first one (when i selected G-Sync, it automatically enabled Vertical Sync: 3D application parameter), i got 161/162 fps as expected while the recommend setting gaves me 157 fps
Did you know why there is fps difference between these settings?
Both G-SYNC + NVCP V-SYNC + LLM "Ultra" and G-SYNC + NVCP V-SYNC + Reflex will set an automatic ~157 FPS limit at 165Hz to keep G-SYNC in range.
Re: Blur Buster's G-SYNC 101 Series Discussion
Posted: 23 May 2022, 23:20
by Chief Blur Buster
157fps more consistently keeps input lag consistent. There begins to be sudden increases between "INSIDE VRR RANGE" versus "OUTSIDE VRR RANGE", and this begins to occur near the thresholds. The Reflex analyzer is simply working to keep lag consistent by staying away from the lag danger. 157fps with less lag than 161fps, in a manner of speaking.
If this is something disliked, I always tell people to purchase a GSYNC monitor with more VRR range than they need, so that framerate ranges are always inside VRR ranges, so that they don't have to worry about framerate capping.
Besides, mathematically 1/161sec versus 1/157sec is only a ((1/157)-(1/161)) = 0.158ms difference = 158 microsecond difference in frametime (& also MPRT persistence of that said sample-and-hold frame rate).
Not worth striving for, especially if VRR-range-transition latency changes are much worse than that.
The Reflex analyzer is supposedly doing its job in keeping latency low by automatically sticking to the ideal VRR-cap-below differential. Sometimes the sweet spot is more than 3fps-below, but your mileage will vary depending on the monitor and refresh rate.
(That being said, for others who play GSYNC in esports, and haven't purchased GSYNC yet. it is useful to purchase more VRR range than you need, so you don't have the VRR-range enter/exit latency-change issue. This is why some in esports 360Hz GSYNC to play CS:GO without needing to worry about being capped far below CS:GO max frame rates, so they don't have to be frustrated at having to cap. Capping is simply a band-aid to avoid a worse latency, because VRR has its properly low latency only when staying inside VRR range)
Re: Blur Buster's G-SYNC 101 Series Discussion
Posted: 04 Jun 2022, 23:14
by smoothnobody
does setting NVCP frame limiter -3 have any effect if not able to hit the maximum frame rate?
would it be accurate to say the biggest reduction in lag comes from -1 and limiting beyond that has significant diminishing returns?
Re: Blur Buster's G-SYNC 101 Series Discussion
Posted: 05 Jun 2022, 09:49
by jorimt
smoothnobody wrote: ↑04 Jun 2022, 23:14
does setting NVCP frame limiter -3 have any effect if not able to hit the maximum frame rate?
An in-game or external framerate limiter is technically only "active" whenever the average framerate is sustained at/above the set limit.
However, whenever the framerate can't reach the set limit, it usually means the system is GPU-limited, resulting in render-queue latency instead, although there are cases in this scenario where the system is CPU-limited (due to system specs or the game, sometimes both), at which point there is no render queue latency.
smoothnobody wrote: ↑04 Jun 2022, 23:14
would it be accurate to say the biggest reduction in lag comes from -1 and limiting beyond that has significant diminishing returns?
That would be accurate if the combined frametime performance of every system, game, and framerate limiter was identical, and said frametime performance was rock solid enough to allow it, but that's certainly not the case.
As such, no, -1 is usually not enough to prevent the framerate from occasionally exceeding the G-SYNC range.
From my testing, -3 has been determined to be the safest
minimum starting point, but more may be needed depending on the system, framerate limiter, max refresh rate, and game combo in question.
Re: Blur Buster's G-SYNC 101 Series Discussion
Posted: 11 Jun 2022, 05:43
by s7gamers
Does this setting work also with FreeSync Monitors That Are Compatible With G-Sync ?
Also there is any recommendation NVCP Setting for my Setup
AOC 24G2 144 Hertz
AMD 5600X
3060Ti Rog Strix
B550 Rog Strix
Re: Blur Buster's G-SYNC 101 Series Discussion
Posted: 13 Jun 2022, 18:32
by RealNC
s7gamers wrote: ↑11 Jun 2022, 05:43
Does this setting work also with FreeSync Monitors That Are Compatible With G-Sync ?
The frame limit setting? It doesn't have anything to do with gsync or freesync. It limits the frame rate. Whether you use g-sync or not doesn't affect it.
Re: Blur Buster's G-SYNC 101 Series Discussion
Posted: 01 Jul 2022, 10:42
by drimzi
I noticed when Low Latency is set to Ultra, it is capping the frame rate to:
Code: Select all
RefreshRate - RefreshRate^2 / 60^2
Is there any significance to this approach or is it arbitrary?
This formula provides the following refresh rate and frame rate limit combinations:
- 1000 Hz = 722.2222 FPS
- 500 Hz = 430.5555 FPS
- 480 Hz = 416 FPS
- 360 Hz = 324 FPS
- 240 Hz = 224 FPS
- 200 Hz = 188.8889 FPS
- 180 Hz = 171 FPS
- 165 Hz = 157.4375 FPS
- 144 Hz = 138.24 FPS
- 120 Hz = 116 FPS
- 100 Hz = 97.2222 FPS
- 75 Hz = 73.4375 FPS
- 60 Hz = 59 FPS
Re: Blur Buster's G-SYNC 101 Series Discussion
Posted: 01 Jul 2022, 12:12
by jorimt
drimzi wrote: ↑01 Jul 2022, 10:42
I noticed when Low Latency is set to Ultra, it is capping the frame rate to
Code: Select all
RefreshRate - RefreshRate^2 / 3600
Is there any significance to this approach or is it arbitrary?
Probably not arbitrary on Nvidia's part, but they haven't revealed exactly
why they auto limit at the numbers they do with LLM Ultra/Reflex, and why said numbers are lower the higher the max physical refresh rate is; the auto limit is merely -1 (59 FPS) at 60Hz, whereas it's -16 (224 FPS) at 240Hz.
I assume it's generous breathing room to account for frametime variances as the physical refresh rate increases.
Re: Blur Buster's G-SYNC 101 Series Discussion
Posted: 13 Aug 2022, 13:31
by PWN3D_
Is limiting your frame 3 frame below your monitor refresh rate really doing everything if for example you are using a "170hz" (an XG27AQ 144hz/170hz with overclocking) monitor that VRR ranges goes from 48Hz~170Hz for DisplayPort, but only use the out of the box 144hz native frequency?
In this setup, is 144hz considered display’s maximum refresh rate and therefore a 141fps limit through NVCP is requiered for that smooth g-sync experience, or 170hz is still considered the maximum refresh rate, and a 144fps limit through NVCP(or in-game settings as it's often an option there) is the way to go?