Dell S2719DGF - I don't think G-SYNC CM is working...

[jorimt]

Still, there's screen tearing. The refresh rate overlay - even when staring at the sky (so a constant 141FPS) - fluctuates wildly between 120Hz~ and 144Hz. (With G-SYNC off, the overlay remains a constant 144Hz, which I guess is to be expected.)

If the monitor's overlay is fluctuating, G-SYNC mode is functioning. And yes, the overlay will only fluctuate in VRR mode; for fixed refresh rates (VRR modes disabled), it will show the currently set max refresh rate (144 in your case) at all times.

Now, with G-SYNC enabled + NVCP/in-game V-SYNC off + 141 in-game FPS limit, you should be seeing this level of tearing on a 144Hz G-SYNC monitor (containing a module) near the bottom of the screen:

phpBB [video]

Of course, you're using a G-SYNC Compatible FreeSync monitor without a hardware module (and one that isn't officially supported [by an early driver feature, mind you], to boot), so it's very possible that you will need to lower the in-game FPS limit to match that.

Try lowering the in-game FPS limit frame by frame while strafing on the training map (in the area shown in the video: take a left from the starting room up the small set of stairs, and it's the wall to the left of the last training drone in the alcove) until the tearing is first reduced to the bottom of the screen, and then is mostly or completely eliminated, then note that number.

On 144Hz G-SYNC monitor with a module, that number (with Overwatch's FPS limiter) is about 120.

Battle(non)sense recently tested a G-SYNC Compatible FreeSync monitor, and found the number to be about 110 for G-SYNC + V-SYNC "Off" (though you'll still get tearing with scenario this during frametime spikes), and 138 FPS with G-SYNC + V-SYNC "On" to fully avoid both tearing and V-SYNC latency, which isn't far off my minimum recommendation of -3 FPS (141 FPS at 144Hz).

So why is a lower limit required for both scenarios on a G-SYNC Compatible FreeSync monitor? Because G-SYNC compatible seems to currently have a higher "ping" (slower report rate) from GPU to display when directly compared to a G-SYNC module, and thus requires a slightly lower FPS limit to stay in the VRR range.

Also, on the differences between G-SYNC and standalone V-SYNC, there virtually isn't a visible/"feel-able" difference between the two with (uncapped) framerates sustained above the refresh rate 100% of the time.

And while it is true that you can prevent most of the input lag V-SYNC introduces by setting an FPS limit below the refresh rate for both G-SYNC and standalone V-SYNC, only G-SYNC can prevent delivery of repeat frames (sync-induced stutter) with framerates below the refresh rate, resulting in a further 1/2 to 1 frame input lag reduction over standalone V-SYNC in the same instance.

Whether that input lag differences at high refresh rates between the above scenarios is appreciable will depend on the person, but the reduction in the adjoining stutter of standalone V-SYNC due to repeat frame delivery that G-SYNC prevents (smooth frame pacing) is pretty much appreciable to anyone.

And to be clear here, everything I've stated to this point is assuming the system and game in question can sustain framerates above the refresh rate all times; for fluctuating framerates below the refresh rate, there really is no question to the benefits of G-SYNC over standalone V-SYNC, where you get something closer to this between the two:
http://www.testufo.com/gsync#framerate= ... &compare=1

[besweeet]

At this point, I think I'll just keep FreeSync and G-SYNC disabled. The trade-offs - and amount of testing - don't seem worth it; I don't want to further decrease my frame rate - nor deal with input lag - to eliminate tearing. Additionally, a 120FPS cap means I'm losing 22%~ of the frames the monitor can display (tearing or not). Compromises are everywhere.

[jorimt]

I'm certainly not attempting to sell you on it.

All I can do is breakdown the differences between the methods, and when properly configured, fully functioning VRR is objectively superior to standalone V-SYNC, both in input lag levels and frame pacing, especially with fluctuating framerates below the refresh rate, of which you seem you're not encountering much of in your common gaming scenarios anyway.

So that said, choice is good, and you should use what you prefer regardless.

That, and there is no guarantee that G-SYNC Compatible will work as well as an actual G-SYNC monitor w/module depending on the given FreeSync monitor model and the fact that "G-SYNC Compatible" is still a software-only driver feature, which may (or may not) yet mature.

If you were to opt for VRR, as it currently stands, I would definitely recommend an Nvidia GPU with a genuine G-SYNC monitor, or an AMD GPU with a FreeSync Monitor over G-SYNC Compatible, because I think it is important to again note that "G-SYNC Compatible" is neither fully comparable to standalone (hardware-side) G-SYNC (of which my article and testing is actually based) or standalone (cable/software-side) FreeSync.

Anyway, for lowest lag standalone V-SYNC, you'll want to use one of the three below methods (depending on the situation):
https://www.blurbusters.com/howto-low-lag-vsync-on/
viewtopic.php?f=2&t=4173&start=70#p36230
viewtopic.php?f=10&t=4916

[besweeet]

I think it may also be worth noting that, compared to your video above, my tearing was happening throughout the height of the screen and is identical regardless of whether G-SYNC was turned on or not. This was confirmed via my 240FPS slow-mo phone recordings.

Part of me is still questioning whether anything is happening with G-SYNC enabled...

[jorimt]

Which is why I suggested the G-SYNC + V-SYNC "Off" tearing test.

If you could replicate what you saw in the video I posted, you would know whether you were getting VRR behavior.

Standalone V-SYNC OFF typically does not tear in a controlled manner at the very bottom of the screen (even with an FPS limit), so it's an easy tell; if it does, VRR is indeed working (badly or otherwise), if it doesn't at any FPS limit, you're just getting plain V-SYNC OFF.

It's very possible, even with the monitor's built-in refresh meter fluctuating and everything set correctly in your monitor and in NVCP that G-SYNC Compatible isn't working properly, or at all, on that specific model, which wouldn't surprise me, since it has only been reported as "functioning" by a handful of user comments (mainly the unofficial G-SYNC Compatible monitor list on reddit) and isn't yet officially supported for this driver feature by Nvidia:



I can't test the monitor myself, as I don't own it, so these troubleshooting steps are the best I can offer you.

[Cpt.Anonymous]

@jorimt I have a cheaper Model of the one being talked about and I can confirm at least on my model that Gsync is working. I followed your steps above as well as testing others. I found The Variable refresh rate(40-144Hz) is working I do need to set RTSS at different FPS caps on a few games to get the desired out come but its way smoother than non-Gsync monitors. I do run the Gsync On - Vsync Off w/ RTSS cap's I see less input lag that way. Thanks for all the info on Cheers.

[jorimt]

Thanks @Cpt.Anonymous, good to know

[open]

try a lower in game limit. you may be too close for that monitor and variations are triggering gsync off.

[besweeet]

try a lower in game limit. you may be too close for that monitor and variations are triggering gsync off.

I opted to just deal with the limit. I paid for 155Hz and shouldn't have to decrease my frame rate.

[RealNC]

I opted to just deal with the limit. I paid for 155Hz and shouldn't have to decrease my frame rate.

I don't see how what you paid for affects what you get in any way