Page 15 of 33
Re: G-Sync 101 w/Chart (WIP)
Posted: 28 Mar 2017, 12:14
by MaxiJazz

Here is screen from forum. It's really have visual effect, but he's talking about "decreasing input\camera lag " with this settings.
Re: G-Sync 101 w/Chart (WIP)
Posted: 28 Mar 2017, 12:34
by jorimt
@MaxiJazz, I am aware that he is referring to "TableFogEnable" decreasing input/camera lag, but what I'm saying is, it's offtopic for this thread.
I aim to document the effect syncing methods and framerate limiters have on input latency, while he aims to document the effect CPU/GPU-based graphical settings have on input latency; these are two entirely separate sections of the rendering chain.
And again, as I stated in my previous post, even if "TableFogEnable" did directly have to do with what I'm testing, it would be extremely difficult to come to a definitive conclusion, as the aforementioned setting is heavily dependent on the given hardware, system, game, and perhaps even driver version being tested.
Re: G-Sync 101 w/Chart (WIP)
Posted: 28 Mar 2017, 14:23
by jorimt
RealNC wrote:jorimt wrote:RTSS does indeed appear to introduce up to 1 additional frame of latency, even with G-Sync
So next candidate is dxtory? On the search for the lowest input-lag frame limiter...
Sorry RealNC, I missed this post until now.
As far as I know, RTSS is still the best external solution. And at higher refresh rates, as long as you aren't a competitive gamer, even 1 additional frame of input latency is still better than the additional 2 frames standalone double buffer v-sync or G-Sync + v-sync on would introduce without a limit.
And again, fps limits are only necessary if your system can exceed your display's refresh rate in the given game. As for my system, at 1440p & 144Hz, even with a factory overclocked 1080, I average 60-100 fps in the latest games when maxed out.
As for less performance-intensive competitive online shooters like CS:GO, Overwatch, etc, if input latency is that much of a concern, you might as well uncap the framerate and play un-synced anyway.
Though, it's worth mention, even this will only net you an average of 1-3ms less latency, max, when compared to G-Sync + v-sync on + in-game fps limiter, and only if you count bottom-of-screen updates (frames rendered with either v-sync off or G-Sync + v-sync on + in-game fps limiter applied reach the middle of the screen at the same time; I will have more info about this in my upcoming article).
Re: G-Sync 101 w/Chart (WIP)
Posted: 28 Mar 2017, 18:27
by RealNC
On heavy games I cap to ~80FPS. The overall experience is much better if it only fluctuates between 70 and 80 compared having it run between 70 to 120.
If it drops near 70 though, then 1 frame starts becoming something you can feel. At 100+ it's no issue though. But if you play at 1440p with everything on "ultra", 100+ is difficult to maintain in most modern games
Anyway, the most popular limiters out there are RTSS, nvidia's built-in limiter (through Inspector), Radeon Pro (should work with NVidia too), and dxtory. I'll test them out. It's actually possible to do this with a low framerate. Capping to 30 should make 1 frame of additional lag stand out visually without the need for a mouse-camera test setup. In theory...
Re: G-Sync 101 w/Chart (WIP)
Posted: 28 Mar 2017, 21:19
by jorimt
Ah, I see, you're capping for consistency.
I myself am not too sensitive to large fluctuation (or, at least not bothered), say, between, 60 and 144, so I leave it uncapped. That's not to say I don't notice; anything below 85 fps is notably less fluid/responsive, but I'd rather that, then have that extra one frame of latency with RTSS.
But, yes, for that reason, I can see why you'd be in need of a latency-free external framerate limiter. I don't know of any, but if you find any promising limiters, I'd be happy to test them and post the results.
Re: G-Sync 101 w/Chart (WIP)
Posted: 29 Mar 2017, 10:55
by MT_
Yeah I'd probably cap for consistency too If i had a 144hz~ screen, and take the minimum fps you will encounter with as a guideline. At least for competitive games I suppose.
My new 7600k @ 5GHz and DDR4-3600mhz PC also further reduced game stutters and engine stalls so G-sync's bottom tear line is also much more stable, sometimes even absent compared to before!
Wheeeeeee.
Anyone know when a new G-sync version will be about? I thought I'd read in a BB article that at some point Nvidia will address this polling rate stuff?
Re: G-Sync 101 w/Chart (WIP)
Posted: 29 Mar 2017, 11:06
by jorimt
I believe that was the Chief's original part #2 article published in 2013:
http://www.blurbusters.com/gsync/preview2/
We currently suspect that fps_max 143 is frequently colliding near the G-SYNC frame rate cap, possibly having something to do with NVIDIA’s technique in polling the monitor whether the monitor is ready for the next refresh. I did hear they are working on eliminating polling behavior, so that eventually G-SYNC frames can begin delivering immediately upon monitor readiness, even if it means simply waiting a fraction of a millisecond in situations where the monitor is nearly finished with its previous refresh.
No official word about that from Nvidia since, at least to my knowledge.
Anyway, I pretty much debunked the polling rate's affect on G-Sync input latency in the OP of this thread, if it does indeed still exist.
Currently, with G-Sync + v-sync off, you're basically getting fancy adaptive sync (tearline pushed to the very bottom of the screen), which reverts to full-on v-sync off during frametime spikes (complete middle-screen tear).
The only tear-free G-Sync experience is G-Sync + v-sync on (original intent w/frametime compensation), and I don't see that changing any time soon.
Re: G-Sync 101 w/Chart (WIP)
Posted: 29 Mar 2017, 14:09
by akirru
So would you say for the most part unless you are some sort of pro gamer ... it might be best to go with G-sync + Vsync on and no cap?
Andy
Re: G-Sync 101 w/Chart (WIP)
Posted: 29 Mar 2017, 14:32
by jorimt
That entirely depends on if you're exceeding the maximum refresh rate of your monitor or not.
If you are, then with G-Sync + v-sync on and no fps limiter in place, you're getting double-buffer v-sync level input latency (up to two additional frames). Two additional frames at 60Hz is up to an extra 33.2ms input latency, @100Hz an extra 20ms, @120 an extra 16.6ms, @144Hz an extra 13.8ms, @165hz an extra 12.2ms, and so on.
If an in-game limiter isn't available, then RTSS is the best alternative. It, however, adds up to 1 frame of latency. This goes for when you're using it with v-sync off, standalone v-sync, G-Sync + v-sync off, or G-Sync + v-sync on.
If you're averaging well below your display's max refresh rate in a given game (say fps in the 80's on a 144Hz display), then no, with G-Sync + v-sync on, you don't need a framerate limit to avoid additional sync-induced input latency.
Re: G-Sync 101 w/Chart (WIP)
Posted: 29 Mar 2017, 23:01
by RealNC
akirru wrote:So would you say for the most part unless you are some sort of pro gamer ... it might be best to go with G-sync + Vsync on and no cap?
If you are a pro gamer, you don't use g-sync.
It does not make sense to disable vsync when using g-sync. I don't even know why people wanted this in the first place.
IMO, it's because people are generally completely clueless and don't know what they actually want
