How many frames of lag are there with the following

Everything about latency. Tips, testing methods, mouse lag, display lag, game engine lag, network lag, whole input lag chain, VSYNC OFF vs VSYNC ON, and more! Input Lag Articles on Blur Busters.
drmcninja
Posts: 112
Joined: 09 May 2017, 10:26

How many frames of lag are there with the following

Post by drmcninja » 03 Sep 2017, 12:15

So, let's take Overwatch for this.

Let's say the baseline is 300 FPS (maximum cap), V-Sync OFF, G-Sync OFF, Fast-Sync OFF. Lightboost OFF.

Let's say this gives 0 frames of input lag. A baseline.

How many frames of lag do you get with the following options:

1. V-Sync On, FPS cap set to just below refresh rate (i.e, 120.012Hz refresh rate, 120fps cap)

2. V-Sync On, FPS cap set to 1 or 2 frames per second below refresh rate (i.e, 120Hz refresh rate, 118 or 119fps cap)

3. V-Sync On, G-Sync On, FPS cap set to 118 (and then 119) with 120Hz refresh rate

4. Fast-Sync On, FPS uncapped at 300, Automatically limited by Fast-Sync to 240-260 fps (with no FPS drops) at a 120Hz refresh rate.

5. V-Sync On, FPS cap set to max 300 (so you get the full V-Sync lag effect)

6. Lightboost ON.

I figure from the frames, everyone can calculate the milliseconds of lag for their refresh rates (1 frame at 120Hz is 8.33ms, 1 frame at 240Hz is 4.17ms, etc)

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

Re: How many frames of lag are there with the following

Post by jorimt » 03 Sep 2017, 13:55

Looks like you're asking how many frames of delay syncing methods add. I already answered this in my G-SYNC 101 article/charts.

1 & 2: 1/2 to 1 frame of added delay.

3: Virtually no sync-induced delay at 2 frames below the refresh rate, but at 1 frame below, there could be the occasional latency spike when hitting the G-SYNC ceiling, thus the recommended 2-3 frame below.

4. 1/2 to 1 frame of added delay.

5. 3 to 3 1/2 frames of added delay (with "Reduced buffering" enabled in the game options).

6. I think it's around 1-2ms (don't quote me), but the Chief knows for sure.

Also, if you're using RTSS to limit frames, add up to 1 frame of delay to each of the above scenarios, and if you're using Nvidia Inspector, 2 to 2 1/2. If you're using the in-game limiter, it doesn't add frame delay (actually decreases it as long as it is the framerate's limiting factor).
(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)

Sparky
Posts: 682
Joined: 15 Jan 2014, 02:29

Re: How many frames of lag are there with the following

Post by Sparky » 03 Sep 2017, 16:11

Jormint's post is accurate, though there are some differences depending on what you use to limit framerate(use an in game cap if available), and on how you're counting latency(first change anywhere on monitor, first change at single point of monitor, top vs middle of screen, etc.)

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

Re: How many frames of lag are there with the following

Post by RealNC » 04 Sep 2017, 04:59

jorimt wrote:Also, if you're using RTSS to limit frames, add up to 1 frame of delay to each of the above scenarios, and if you're using Nvidia Inspector, 2 to 2 1/2. If you're using the in-game limiter, it doesn't add frame delay (actually decreases it as long as it is the framerate's limiting factor).
We talked about it in the past. I thought the consensus was that RTSS reduces frame delay, it doesn't add to it. In-game limiters reduce it more, RTSS still reduces it, just not as much as a good in-game limiter.
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: How many frames of lag are there with the following

Post by jorimt » 04 Sep 2017, 09:06

Yes, but that wasn't in context for his baseline/scenarios:
Let's say the baseline is 300 FPS (maximum cap), V-Sync OFF, G-Sync OFF, Fast-Sync OFF. Lightboost OFF.

Let's say this gives 0 frames of input lag. A baseline.
With that hypothetical baseline and scenario 1-4, the assumption is, the framerate limiter is the limiting factor at all times, and my answers (which isolate the approximate frame delay by syncing method) would be based on the in-game limiter for that game, thus RTSS would appear to "add" up to a frame if used instead of the in-game limiter in that specific context.

I'd love to, but I can't included the "in-game/RTSS limiter reduces latency at the same framerates as framerates not being limited by a framerate limiter below the refresh rate, but by the way, the in-game limiter reduces latency more than RTSS because it is part of the engine" explanation in every passing post, especially if we're only covering the basics.
(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)

drmcninja
Posts: 112
Joined: 09 May 2017, 10:26

Re: How many frames of lag are there with the following

Post by drmcninja » 04 Sep 2017, 10:57

I had to ask because I can actually notice a difference between V-Sync ON, 240.012Hz refresh rate and 240fps in-game cap versus V-Sync OFF, 300fps in-game cap.

There's stuttering, but I can actually aim better. Though I don't do bad at all with V-Sync On. It's very playable and I've done almost as well. It's just that I can hit much sharper twitch shots with it off and get a little precision not there before.

1 frame at 240Hz is 4.17ms. I am surprised I can notice that.

This was not the case when I was running 120Hz/144Hz. There the stuttering/blurriness made things worse. I couldn't aim. Especially since Lightboost 120Hz was so dim/dark. I moved to 144Hz without Lightboost and 154fps cap and tearing/blurriness and fluctuating mouse lag was alleviated. Plus the brightness helped.

Then I moved to 120Hz with Fast-Sync ON at 240-260fps (along with some computer upgrades so FPS wouldn't fluctuate), and that was amazing all around (on my laptop I use 120Hz, 118fps G-Sync On, V-Sync On, it also feels good, but not as good as Fast-Sync). 120Hz with Fast-Sync felt almost as good as uncapped/300fps, or just as good. The mouse felt like it had noticeably less input lag than 120Hz/118fps.

So the stuttering/tearing/blurriness can still mess me up now, but I can just force myself to focus and then hit the shots again. I guess it must be that the BenQ DyAc feature doesn't reduce brightness. And I've upgraded my computer so my FPS doesn't drop below 290-300.

Side question: My laptop's refresh is actually 120.12 (not 120.012, but 120.12). Does this mean I can run G-Sync On, V-Sync On and FPS cap at 120 on the dot? Or should I stay with 118? G-Sync/V-Sync @ 118fps/120.12Hz is better than V-Sync @ 120fps/120.12Hz? Or is there a difference at all?

drmcninja
Posts: 112
Joined: 09 May 2017, 10:26

Re: How many frames of lag are there with the following

Post by drmcninja » 04 Sep 2017, 10:58

By the way, has anyone seen this reddit thread:

https://www.reddit.com/r/Competitiveove ... _the_next/

A Blizzard programmer actually responded (with a pretty lame response, but still).

Can "subframe input polling" eliminate input lag linked to FPS? I.e, give 1000fps mouse feel at all FPS?

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

Re: How many frames of lag are there with the following

Post by jorimt » 04 Sep 2017, 11:42

@drmcninja,

Have you read my article? Because it already answers the majority of your questions:
http://www.blurbusters.com/gsync/gsync101-input-lag/

Unless G-SYNC on your laptop isn't setup properly, I'm sorry to say that the perceived improvement over G-SYNC with Fast Sync is almost certainly a placebo at both 120Hz and 240Hz (and when I say "almost," I'm being kind):

Image

Image

While they're close, as long as G-SYNC is properly setup, it always has less latency than Fast Sync with those specific refresh rate/sustained framerate ratios. Fast Sync has to have at least 5x times the sustained framerate of the current refresh rate to match G-SYNC latency levels (see 60Hz scenario in my article).

As for the Overwatch programmer's post, you are aware that syncing methods aren't the only source of input latency, correct? Your mouse, keyboard, display, system, and game engine all add varying levels of input latency as well. There's no game or setup that offers instant response to user input.

That, and even if all the above had no latency, let's say you're playing at 120Hz/120 FPS sustained; it would take the graphics card a minimum of 1 frame (current frametime would dictate latency induced) to render, and, if we're talking synced, the scanout 8.3ms to completely display the frame on-screen.
(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)

drmcninja
Posts: 112
Joined: 09 May 2017, 10:26

Re: How many frames of lag are there with the following

Post by drmcninja » 04 Sep 2017, 12:22

Yeah, I've read it.

That's a strong placebo effect, considering my accuracy % is different between the two consistently. Could it just be down to better mouse feel due to more input from the mouse at the higher framerate? Even though they're both the same refresh rate, the game engine is taking mouse input twice as often with the other one (240fps vs. 120fps). I would have thought Fast Sync's buffer would make the mouse feel worse (since it's throwing out half the frames, right?), but it feels better.

In other words, the mouse at 240fps is different from 120fps. So with Fast-Sync, the game engine is doing the former (mouse input 240 frames a second), but the GPU is cutting it down to appear like the latter (120 fps hits the monitor screen). But the 120fps delivered to the display with Fast-Sync seem qualitatively different from the 120fps delivered to the display when the game engine is limited to no more than 120fps.

Just as how there's a difference between 240fps/120Hz and 120fps/120Hz (with no refresh sync). The mouse under Fast-Sync feels more like the former than the latter. Like somewhere inbetween the two.
As for the Overwatch programmer's post, you are aware that syncing methods aren't the only source of input latency, correct? Your mouse, keyboard, display, system, and game engine all add varying levels of input latency as well. There's no game or setup that offers instant response to user input.
The thread was about framerate-associated input lag specifically. Just as how there's less input lag at 300fps than there is at 60fps. The user claimed that using this other technique ("subframe input polling") mostly eliminated that difference. So you can have the mouse feel almost as responsive as it does at 300fps while only at 60fps.

Sparky
Posts: 682
Joined: 15 Jan 2014, 02:29

Re: How many frames of lag are there with the following

Post by Sparky » 04 Sep 2017, 12:25

drmcninja wrote:By the way, has anyone seen this reddit thread:

https://www.reddit.com/r/Competitiveove ... _the_next/

A Blizzard programmer actually responded (with a pretty lame response, but still).

Can "subframe input polling" eliminate input lag linked to FPS? I.e, give 1000fps mouse feel at all FPS?
Nope. What matters is where the render pipeline is bottlenecked. The farther down the render pipeline the bottleneck is, the more input lag. Doesn't matter if you gather input every 1ms if the frame sits in the context queue for 50ms.

Post Reply