NVIDIA G-Sync Pulsar monitor - Asus ROG Strix XG27AQNGV

[kyube]

The fundamental limitation remains the IPS panel itself. The advertised “1 ms” response time is a best-case gray-to-gray marketing figure measured under extreme overdrive conditions, often with overshoot artifacts. In real use, response times are significantly higher and more variable depending on transition, refresh rate, and overdrive mode. IPS panels rely heavily on aggressive overdrive to approach those numbers, which introduces inverse ghosting, overshoot, and transition inconsistency—especially above 300+ FPS.
Pulsar also disables variable overdrive, which means overdrive tuning no longer adapts dynamically to refresh rate. That results in either underdrive at lower Hz or overshoot and crosstalk at higher Hz. In practice, effective response behavior becomes less consistent, not more. I also observed refresh rate compliance inconsistencies, added processing overhead, and signs of short buffer-induced latency when combining certain sync modes.

The “limitation” of Pulsar isn't it's G2G RT's in the way you're trying to frame it.
Yes, G2G_RT > refreshtime are always a limitation, even in backlight strobing mode, but they're not physically making the panel blurrier like in sample & hold
In my opinion, the actual “limitation” is the use of a too high duty cycle for the entire refresh rate range, as it'a a variable pulse width implementation. This is very likely what you're experiencing, not the former.
Not to mention that Pulsar adds ~3ms of additional processing latency to the total system latency, as per Battle(non)sense.
TFTCentral should be having a review soon enough for more data!

With V-Sync ON + Reflex Ultra, the system hard caps at ~320 FPS and latency increases from roughly 4 ms to ~12 ms. 1% lows worsen and frametime consistency degrades.

It's called “Nvidia Reflex”, not “Reflex Ultra”.
The supposed total system latency increase you've mentioned is completely absurd and very likely false.
What tools have you used to come to such findings?
“1% lows worsen” because you're displaying the ACTUAL freshest frame your system can output as possible.
Frame rate is a misleading metric to evaluate performance.

The behavior feels similar to buffered rendering, which defeats the purpose of a 360 Hz esports panel. In contrast, running G-Sync with a manual 357 FPS cap and V-Sync OFF or Fast results in better responsiveness, tighter frametimes, and no meaningful clarity loss.

All you're doing in that scenario is jumping in & out of the GSYNC range, as internal/external frame rate limiters aren't precise enough.

Marketing frames Pulsar as a competitive advantage, but it mainly reduces perceived blur.
It does not change panel speed, pixel accuracy, or transition consistency.

? What are you talking about?
What are these ambiguous word choices explaining the same thing? :DDD

Motion clarity on IPS remains limited by response behavior, not refresh rate alone.
Crosstalk and ghosting become more noticeable once you push beyond ~320–350 FPS.

Again, you're using a panel in a impulsed mode (fixed OR variable refresh rate backlight strobing), the claim of “limited by response behavior” is just completely nonsensical in the way you're trying to purport it.
The latter claim (crosstalk part) is plausible, as this is a physical RT limitation of +300Hz refresh rate strobing.

In direct comparison, OLED remains fundamentally superior for motion. Near-instant pixel response, absence of overdrive artifacts, stable transitions across refresh rates, and higher contrast all improve motion definition and target visibility. Even at lower FPS, OLED appears cleaner and more responsive because it does not rely on aggressive tuning trade-offs.

If you're genuinely belief is that 120Hz OLED (sample & hold) is cleaner than a Pulsar panel running at 120FPS (in GSYNC+VSYNC+Reflex mode), then either:
• Something is completely wrong with Pulsar (a very likely culprit, given the constraints it has)
• Something is completely wrong with your setup
• You're not sensitive to sample & hold blur of <1000Hz refresh rates as you think you are (though, considering the terrible duty cycle target Pulsar has & the bug mentions, I somewhat doubt this is the case....)
Or something else entirely.

After a week of testing every relevant configuration, I returned the monitor and switched to a 540 Hz WOLED panel. The difference in motion stability and responsiveness is immediately noticeable. IPS still depends on refresh rate scaling, overdrive compromises, and artifact balancing. OLED avoids most of those structural limitations.
For competitive gaming, I see no meaningful advantage from Pulsar over a fast OLED panel.

540Hz OLED is fundamentally still blurrier than good backlight strobing (<1ms MPRT)...
I do agree that Pulsar shouldn't be considered for any use-case though.

Its not just the stutters due to 1% lows. Its also the full screen blackouts, which you cannot avoid in some older games when there just is no borderless windowed option (spaghetti kart and other emulation games which look incredible, but black out for some seconds ingame if you play them with 360 Hz)... Yes, with 240 Hz they do not (yet) and look identical, but just the fact that I have to switch between Windows refresh rates for some games is annoying and... it does not "just work".

And this is not even a pulsar thing. Seems like many Asus oled screens with DP1.4a connection and 360 Hz suffer from the same behavior... since years (!). So to me it sounds like there is no quick solution for that.

Mayyyyybe... if they manage one day to bring this to a mini LED panel (which I think makes a lot more sense and is possible way earlier than with an oled panel), with DP2.x support, no more blackouts, maybe full 360 or more Hz without using DSC... I will jump right on it.

For now, its just waiting for new drivers, firmwares etc. to get most of this sorted out...

The “blackout” is likely a combination of FSE & DSC.
The latter cannot be fixed with the DP2.1 support route, as they're sticking to the Mediatek MT9810 scaler IC, which is only capable of DP1.4 HBR3 & HDMI 2.1 FRL6.
Of course, we didn't get the latter because of planned obsolescence...
Ideally, we'd have gotten a MiniLED backlight & HDMI 2.1 FRL6 on these QHD 360Hz models, but we can't have nice things

And these facts do NOT bother me at all and I will keep mine because I will play my retro games, emulators, side scrollers etc with 60-120 fps with this pure clarity which is possible although 60 Hz needs to be proven still with the new firmware but 90 fps.. omg this can look so damn fine with pulsar, OLED just cannot deliver this without bfi - which then brings other downsides with it. Zelda Breathe of the wild with fps++ at around 110-160 fps looks mindblowing. I could go on forever. There are so many titles.

^ ~1500 px/s SmoothFrog pursuit photograph (general info: ~1500 Hz would be ~0,6667 ms refreshtime)
I personally think that you'd have a far better experience with a <1ms MPRT fixed refresh rate backlight strobing model than with the Pulsar models for this use-case. :shrug:

[mawi]

The “blackout” is likely a combination of FSE & DSC.
The latter cannot be fixed with the DP2.1 support route, as they're sticking to the Mediatek MT9810 scaler IC, which is only capable of DP1.4 HBR3 & HDMI 2.1 FRL6.

I also do not think a DP2.1 card on one end should make a difference if the display on the other is just capable of DP1.4.
On our pulsar display, there are not many reports yet. But on the OLED side, it seems like this only happens with RTX 30 and 40 cards. AMD cards seem to work much better on 360 Hz DSC and also RTX 50 cards do so. I think AMD has DP2.1 since 7900 XT... Nvidia since RTX 50, so it couuuuuuuuld be related to that. But I am just repeating many reddit posts without knowing exactly what is going on. Someone claimed the RTX50 cards are using different "heads", like one guy mentions here:

https://www.reddit.com/r/OLED_Gaming/co ... _dsc_your/

"Apparently from reading the CRU patch notes: "5000-series GPUs have a higher single-head limit" which I believe is what drives the output of the display (there's 4 on Nvidia GPUs, generally one for each port). On the 4000 series and below, with DSC, a lot of times the GPU used 2 heads to drive that display, which caused crazy issues when switching anything."

I do not have a lot of hope this will change with new nvidia drivers or ASus firmwares. This problem is out there for too long to be easily adressed. But if DP2.1 on the GPU side is a known solution to this... I am at least happy to know i couuuuld do sth about it. Which I probably wouldnt

[kyube]

I also do not think a DP2.1 card on one end should make a difference if the display on the other is just capable of DP1.4.

Yes, you're limited by what the display can support, regardless of GPU architecture you get.
AMD doesn't support full BW (DP80) of DP2.1 on any of their dGPU consumer models.
Your only option is sticking to QHD@240Hz@8bit to avoid DSC :/

[mawi]

Your only option is sticking to QHD@240Hz@8bit to avoid DSC :/

Which I do and it is fine for me (for now). But is there any chance for me to check if DSC is really switched off then? OSD is not showing anything on the ASUS model. I read sth about "if nvidia offers DSR / DLDSR scaling, then DSC is surely turned off". But with the Asus pulsar model, there is no DSR / DLDSR option anyway, no matter what resolution or Hz I use, even with 120 Hz and pulsar turned off.

I have read the Gsync module (which I know is within the mediatek scaler now) is different to most other displays and does not allow this. Also its not possible to add any other resolutions or frequency rates than the pre-defined ones (60, 120, 240 and 360).

So yes, I am using 240 Hz most of the time and I keep believing its a non-compressed picture... But I have no proof.

[bobbie424242]

I never understood why no tool (software) that I can think of can tell you if DSC is enabled or not. No DSR / DLDSR scaling option can be an indication but I'm not even sure that when this option is not available it 100% implies that DSC is used.
It's been confusing on my 4K 160Hz monitor that has HDMI and DP inputs. For HDMI there are 120Hz and 160Hz modes configurable in the OSD. For DP, there is only a 160Hz mode. The only mode for which the DSR / DLDSR scaling option is available is HDMI 120Hz. The 160Hz modes of HDMI and DP never have have this option, even if the refresh rate in Windows is configured to a value not requiring DSC (<= 144Hz on my RTX 4080S). Now imagine if a tool (or the monitor OSD) told your if DSC is enabled or not. But nope, it does not seem to exist.

[brownvim]

You can do DSR / DLDSR on the Acer Pulsar monitor along with 10bit colour and HDR at the same time. Not sure if this totally rules out it using DSC but I can’t find much about it.

Im enjoying playing Indiana Jones and Resident Evil Requiem on it at the moment with Path Tracing. Had to role back Nvidia drivers though as the latest ones seemed to add stutter and blur to both games.

[hamza_tm]

What driver version you using? liquidshadowfox was suggesting a specific version somewhere earlier

[brownvim]

I rolled back to 591.86

[hash]

A lot of GSYNC Pulsar benefits disappear with 400/800dpi mice settings, common in CS2 for handshake-filter, but hurts Pulsar blur busting.

Could you elaborate on this handshake filter thing? I tried google but not much showed up.
Is it some kind of overhead when starting and stopping usb data transmission?

(i made an account just to ask this question lol)

[ashrr]

Anyone have any ideas why this monitor would have higher input lag than another 144hz monitor he also tested? The pulsar monitor gets just under 6ms while the 144hz ips gets just under 4ms and that's without pulsar. He also tested a 500hz oled monitor which is about 1ms.

phpBB [video]

phpBB [video]

phpBB [video]