Inconsistent in-game frame rate limiters and vsync.

Ask about motion blur reduction in gaming monitors. Includes ULMB (Ultra Low Motion Blur), NVIDIA LightBoost, ASUS ELMB, BenQ/Zowie DyAc, Turbo240, ToastyX Strobelight, etc.
Post Reply
insight2
Posts: 5
Joined: 13 Mar 2020, 11:22

Inconsistent in-game frame rate limiters and vsync.

Post by insight2 » 15 Apr 2020, 13:49

Lately I played the new FPS game Valorant. The in-game FPS limiter seemed good, it gave me a steady 235 fps value. I turned on my monitor built-in fps counter to check, and I realized the framerate was fluctuating a lot, often reaching 239 fps (vsync cap) even tho I had a 235 fps cap in game with v-sync on. Testing with riva tuner confirmed this, by lowering the fps refresh delay from 500 to 50 on RTSS, I noticed the framerate and frametimes were, in fact, really inconsistent. So I decided to disable V-sync and test again. This time without the vsync cap, I noticed the framerate was exceeding often my fps cap for some fraction of seconds. This bevhavior might trigger the vsync input lag, I suppose (correct me if i'm wrong).
Now I can't choose. My choice is to either cap my fps in game to something like to 225 hoping the framerate won't exceed my monitor's refresh rate, or disable vsync. But with vsync disabled the image isn't smooth at all cause the frametimes are inconsistent.
Using an external frame cap like riva tuner fixes everything, but I am not sure the additional input lag is worth it, considering FPS games need super low delays.
I am not sure what to do. I am a competitive player, but I also want a good image quality, tearing is DISTRACTING. I hate it.
I hope i explained myself properly, if not I apologize.
Hopefully you guys will test the game and find the perfect settings, i'm sure many people would be interested as well.
Thank you!

knypol
Posts: 68
Joined: 18 Aug 2016, 03:40

Re: Inconsistent in-game frame rate limiters and vsync.

Post by knypol » 15 Apr 2020, 14:50

Noticed this as well in many games when limiting fps to 234-237 . Not sure what is more accurate monitor fps counter (gsync monitor) or the in-game counter. Only when i limit to 229 (or less) monitor counter doesn't exceed 237.

Avean
Posts: 25
Joined: 13 Feb 2020, 07:15

Re: Inconsistent in-game frame rate limiters and vsync.

Post by Avean » 21 Apr 2020, 06:02

No way! I trusted the in-game fps limiter as well. Also playing Valorant. Checking now, in-game fps limit to 167. Monitor showing me hitting 170 several times enabling vsync. So i tried setting 160 fps limit in-game and im STILL hitting my max refreshrate 170 several times. Tried setting fps limiter through NVCPL and also there its hitting 170. I assume the monitor is telling me the correct framerate.

I had no idea this was an issue, thanks for posting this! Knowing this now i feel disabling v-sync is the way to go even with more unstable frametimes but its only jumping between 2-3 ms in my case if i let my fps fly up to 350. Also i get lower input lag. 2.2ms instead of 5.8ms.
But i wonder how many of us is having this issue cause g-sync but v-sync is pretty much a norm these days. They would have no idea they are hitting v-sync territory without checking theyre monitor.

insight2
Posts: 5
Joined: 13 Mar 2020, 11:22

Re: Inconsistent in-game frame rate limiters and vsync.

Post by insight2 » 21 Apr 2020, 10:24

Yeah, that's exactly what I wonder as well. If all these spikes trigger vsync, it's a huge problem. (maybe that's why nvidia ultra low latency caps fps at 225? to avoid this issue).
For now I locked my fps to 225 from the in-game option and they are steadily under 239, still I am considering disabling vsync for the moment. The problem is that the built-in in game fps cap is not so stable and it the game doesn't look really smooth with uncapped fps. Lower input lag is nice, but it's not worth the tradeoff in my opinion.
Maybe someone who knows more than us will be able to tell us more.

User avatar
jorimt
Posts: 1505
Joined: 04 Nov 2016, 10:44
Location: USA

Re: Inconsistent in-game frame rate limiters and vsync.

Post by jorimt » 21 Apr 2020, 10:37

I've discussed this multiple times on the forums before, but to sum-up, take the monitor's built-in refresh rate meter during G-SYNC operation with a grain of salt (especially those on FreeSync monitors running in G-SYNC Compatible mode), as it doesn't use the same averaging as system-side FPS readouts, and is prone to what you describe.

For what it's worth, during my original input lag test for the G-SYNC 101 article, I saw similar spikes on my monitor's refresh rate meter, and it did not end up directly affecting average input lag; as long as you use G-SYNC + V-SYNC + a minimum -3 FPS limit, the V-SYNC option should compensate for any of the frametime variances that occur (including the "spikes" you are seeing on the refresh rate meter).

My advice is to rely on the system-side FPS readouts for average FPS instead, and to only rely on the refresh rate meter to determine whether G-SYNC is engaged in the given game (fluctuating = VRR mode, constant static number = Fixed refresh mode).
(jorimt: /jor-uhm-tee/)
Author: Blur Busters "G-SYNC 101" Series

Displays: Acer Predator XB271HU / LG 48CX OS: Windows 10 MB: ASUS ROG Maximus X Hero CPU: i7-8700k GPU: EVGA RTX 3080 FTW3 UG RAM: 32GB G.SKILL TridentZ @3200MHz

insight2
Posts: 5
Joined: 13 Mar 2020, 11:22

Re: Inconsistent in-game frame rate limiters and vsync.

Post by insight2 » 21 Apr 2020, 11:07

jorimt wrote:
21 Apr 2020, 10:37
I've discussed this multiple times on the forums before, but to sum-up, take the monitor's built-in refresh rate meter during G-SYNC operation with a grain of salt (especially those on FreeSync monitors running in G-SYNC Compatible mode), as it doesn't use the same averaging as system-side FPS readouts, and is prone to what you describe.

For what it's worth, during my original input lag test for the G-SYNC 101 article, I saw similar spikes on my monitor's refresh rate meter, and it did not end up directly affecting average input lag; as long as you use G-SYNC + V-SYNC + a minimum -3 FPS limit, the V-SYNC option should compensate for any of the frametime variances that occur (including the "spikes" you are seeing on the refresh rate meter).

My advice is to rely on the system-side FPS readouts for average FPS instead, and to only rely on the refresh rate meter to determine whether G-SYNC is engaged in the given game (fluctuating = VRR mode, constant static number = Fixed refresh mode).
Thank you. That's all i wanted to know. I use a PG258Q G-Sync monitor, i'll lock the FPS to 235 then, just to be sure.

Post Reply