NVIDIA G-Sync Pulsar monitor - Asus ROG Strix XG27AQNGV

[liquidshadowfox]

https://youtu.be/h27x8H6n6p4?t=434

At 7:14 he says the oled is better than the pulsar.

Is this true? 720hertz using 270hz with ULMB 1 is more clear than Pulsar?

As far as I understand, he means using the 540 Hz OLED at 270 Hz with ELMB... so black frame insertion which loses a lot of brightness. however this ASUS flaship oled seems to be bright enough so its still fine in ELMB mode.

What confuses me is minute 8:05... Where 120 Hz pulsar easily beats 120 Hz OLED with ELMB. I would have expected 120 Hz OLED with ELMB strobing to be clear as hell, like the 60 Hz mode on pulsar, just with 120 Hz because of the very good oled pixel response times.
But is looks really bad compared to pulsar 120 Hz... very weird.

Oh that's easy to explain, OLED BFI currently only allows a 50% duty cycle, so 120 hz BFI on OLED is actually running at 240 hz (half frames real, half full black). IF they implemented GOOD BFI that's configurable, that could do something like 480 hz with 1 real frame and 3 black full frames and the motion clarity for 120 fps would "look" as clear as 480 hz (for a 540 hz panel I think you'd have to do like 135 fps real frames and the rest black frames to get the same motion clarity as ELMB 270 hz on OLED).

In theory Gsync pulsar "should" look better at 270 simply because it's targetting a 25% duty cycle but pixel responses times can't keep up and Nvidia refuses to reduce the duty cycle further for gsync pulsar (I think 15% duty cycle with a max nit target of 225 would be amazing given they are doing 25% with over 500 nits right now )

[liquidshadowfox]

https://youtu.be/h27x8H6n6p4?t=434

At 7:14 he says the oled is better than the pulsar.

Is this true? 720hertz using 270hz with ULMB 1 is more clear than Pulsar?

As far as I understand, he means using the 540 Hz OLED at 270 Hz with ELMB... so black frame insertion which loses a lot of brightness. however this ASUS flaship oled seems to be bright enough so its still fine in ELMB mode.

What confuses me is minute 8:05... Where 120 Hz pulsar easily beats 120 Hz OLED with ELMB. I would have expected 120 Hz OLED with ELMB strobing to be clear as hell, like the 60 Hz mode on pulsar, just with 120 Hz because of the very good oled pixel response times.
But is looks really bad compared to pulsar 120 Hz... very weird.

I have a 500hz oled, and tried comparing the ULMB 1 with changing my monitor to 240hz and compare it to a pulsar and i thought it didnt much up. But he is saying by doing 270hz (so a couple more than what I can do now) beats out the puslar completely? I'm so confused because my tests with 240hz ULMB on my oled is not even close. I easily can't see the words in the ghosting test.

Like I mentioned, Gsync pulsar "should" be able to beat oled in motion clarity due to the 25% duty cycle (vs 50% duty cycle) but there's a lot of things that have to go right in order for the clarity to be better in games.
Currently if I use Vsync + Gsync + Ultra low latency mode or reflex, the motion clarity of gsync pulsar isn't there. Some games it works perfectly without having to do anything and just having reflex on, most games I play on the other hand... reflex on = bad frame pacing = no good gsync pulsar strobing so I end up having to cap my FPS using riva tuner and front edge sync in order to get the frame pacing to play nice with gsync pulsar and then the motion clarity is absolutely pristine.

I've never thought I'd actually play my games at a locked 90 FPS on purpose XD I am really trying to be patient with them giving us the asus firmware update so I can finally enjoy my games at native 120 fps with gsync pulsar with no double images. 100 fps - 180 fps range is ironically the most important VRR range for me given that's the fps I get in most of my games with all the settings maxed out. Above 200 fps the clarity isn't as good and I REALLY hope nvidia is looking into fixing 200 fps + gsync pulsar because it works great in smooth frog but if you play an actual video game it just feels like it fires off too many compensation pulses unless the frame times are perfect, I haven't been able to get the frame times to be perfect above 200 fps even with riva tuner which is a pity. I think they need to look into a reflex mode that's aware of gsync pulsar to help frame timing match with the gsync module (don't see why they wouldn't jump on this considering they own the entire stack here: game engine integration -> Nvidia only GPUs -> Nvidia only display scaler for gsync pulsar to work).

[John Blur]

1.1.6 is now available for the ASUS XG27AQNGV on the gsync update nvidia website. I can't post the URL in here but there are 2 Nvidia websites to do the Pulsar update for some reason, and only 1 of them has the 1.1.6 update for the Asus mode. It's the one that starts with gsync-update.

Anyways, after updating, despite Nvidia telling me on the update website I'm on 1.1.6, it still says firmware MCM102 in the monitor's OSD. Wasn't that the firmware version of the 1.1.4 update?

Also this is what's coming in 1.1.7 :

new in upcoming FW 1.1.7
ULMB2 Pulse Width slider for MSI Pulsar model
fix for issue where G-SYNC (and hence Pulsar) are no longer available after monitor sleep
reduced input lag for 240 Hz Pulsar/ULMB2
improved 120 Hz ULMB2 quality (Global Flash instead of Rolling Scan)
misc bug fixes

[CriticalHit_NL]

The new 1.1.6 G-sync Pulsar firmware is now also released for the Asus XG27AQNGV, refer to this update page:
https://gsync-update.nvidia.com

As you may see, these pages are not yet updated at the time of writing:
https://www.nvidia.com/en-us/geforce/pr ... rs/update/
https://nvidia.custhelp.com/app/answers ... 1.6-update

Edit:
Updated the monitor successfully.

Did some short testing in:

• Planet Coaster 2 (~120fps & ~335 with FG)

• Life is Strange Reunion (80 fps cap in Rivatuner & ~120-130s)

• Call of Duty 2 (333 fps)

The double image issue seems to be a thing of the past.
The previous firmware around ~100-120 fps it was especially badly noticeable in the above games so it's great to see that being resolved.

Smoothfrog is still clear as ever using 15x speed in 2D test with either V-sync 327FPS lock or 335FPS rivatuner cap on Reflex mode.
The lower FPS ranges no longer show these jarring double images as seen in previous firmware in the rotating test.

My G-sync Pulsar is set to 75Hz lower limit so ~80 is where it is starting to be active, though you might see a short deactivation at this FPS limit if something fluctuates so I would probably pick a slightly higher FPS to test with so you're not too close on the activation/deactivation edge.

Anyway I believe this update is definitely a must-install to see the full potential these monitors have to offer in motion clarity.
Firmware in the OSD still reports MCM102 but maybe that is a separate thing to Pulsar internally but I don't know for sure.

[mawi]

yep, can confirm RE9 now also works perfectly. Either with DLSS 2x Frame Gen at around 170 fps or without FG at around 110 fps. Both look almost identical, no double image, perfectly clear to me.

[brownvim]

Also this is what's coming in 1.1.7 :

new in upcoming FW 1.1.7
ULMB2 Pulse Width slider for MSI Pulsar model
fix for issue where G-SYNC (and hence Pulsar) are no longer available after monitor sleep
reduced input lag for 240 Hz Pulsar/ULMB2
improved 120 Hz ULMB2 quality (Global Flash instead of Rolling Scan)
misc bug fixes

Thanks! Where did you get this information?

Looks very promising.

Edit: found it: https://www.nvidia.com/en-us/geforce/fo ... opicPage=3

Posting below what Nvidia staff wrote:

new in upcoming FW 1.1.7
ULMB2 Pulse Width slider for MSI Pulsar model
fix for issue where G-SYNC (and hence Pulsar) are no longer available after monitor sleep
reduced input lag for 240 Hz Pulsar/ULMB2
improved 120 Hz ULMB2 quality (Global Flash instead of Rolling Scan)
misc bug fixes

under study for future:
reduced input lag for 60 Hz ULMB2 on DP
reduced input lag for 60 Hz ULMB2 on HDMI
ULMB2 Pulse Width slider for AOC Pulsar
two second black screen with RTX30/40 cards at 360 Hz. Issue is very low probability, making the investigation very hard. If you have a reproducable case with latest driver, please contact us and provide detailed system configuration and repro steps

features requested but not currently considered:
adjustable pulse width for Pulsar. The Pulsar algorithms to create a flicker free variable refresh solution are complex, adding adjustable pulse width without causing flicker is not possible
Pulsar below 75 FPS (the original plan to allow Pulsar down to 48 FPS). Flicker when pulsing below 75 Hz is very visible, especially at full Pulsar brightness (this is why ULMB2 60 Hz defaults to lower brightness). Supporting dynamically variable refresh rate pulsing adds 12 dB to this already bad flicker. We implemented it and tried it. The resulting flicker was not acceptable to any viewer .

[mawi]

Thanks! Where did you get this information?

Looks very promising.

Edit: found it: https://www.nvidia.com/en-us/geforce/fo ... opicPage=3

Posting below what Nvidia staff wrote:

new in upcoming FW 1.1.7
ULMB2 Pulse Width slider for MSI Pulsar model
fix for issue where G-SYNC (and hence Pulsar) are no longer available after monitor sleep
reduced input lag for 240 Hz Pulsar/ULMB2
improved 120 Hz ULMB2 quality (Global Flash instead of Rolling Scan)
misc bug fixes

...

two second black screen with RTX30/40 cards at 360 Hz. Issue is very low probability, making the investigation very hard. If you have a reproducable case with latest driver, please contact us and provide detailed system configuration and repro steps

This is a banger news!

But I cannot understand they are not able to reproduce the 2s black outs. Have they ever really used an RTX40 card? with 360 Hz this thing blacks out with each and every alt-tab operation... for multiple seconds. And yes, also with the current driver. Just tested it last night.

When I start playing Notblood (amazing source port from Blood, same as eduke32 with Duke Nukem) and starting the game, it completely freaks out. I hear the sound and picture remains black for 30 seconds...

With 240 Hz and enabletiledisplay=0... no black screen at all, just jumps right into it.

[kyube]

new in upcoming FW 1.1.7
ULMB2 Pulse Width slider for MSI Pulsar model
fix for issue where G-SYNC (and hence Pulsar) are no longer available after monitor sleep
reduced input lag for 240 Hz Pulsar/ULMB2
improved 120 Hz ULMB2 quality (Global Flash instead of Rolling Scan)
misc bug fixes

Well, that's.... funny
Why only the 120Hz ULMB2 mode and why resort to global flash instead of (discrete) rolling scan behavior, utilising all the 10 LED strips inside the display?
Global flash is still a larger eye comfort concern than rolling scan... I fail to see how they're unable to repeat what Zowie & other vendors have done on their models, with their respective (discrete) rolling scan solutions (DyAc2, MBR+,...)
Missed opportunity to not include this change to the 240Hz & 360Hz refresh rate modes.

It's funny that the AOC & MSI models are the ones receiving a slider only now...

[liquidshadowfox]

I think if they aren't going to give us pulse width adjustment for pulsar, they should at least make another preset that favors motion clarity over flicker by allowing us to drop down to 12.5% duty cycle (x8 motion clarity and around 250 nits max brightness) or at least 16.66% duty cycle (x6 motion clarity and around 300 nits max brightness). I'm excited to see if and when they add the nvidia control panel/nvidia app controls to have per game ULMB 2/pulsar config with max hz set per application profile cause this would make it more bearable to use ULMB 2 without going through the annoying process of enabling/disabling Gsync constantly when switching games.

Also they had global flash with vertical independent overdrive on the Asus PG27AQN and ULMB 2 looks great on it at 240 hz with little to no crosstalk, if this panel is indeed a new gen version that's better (and has no red phophor trails) why wouldn't they use global flash at 240 hz ULMB 2 mode as an option? Given that rolling flash reduces eye strain is great but it makes the higher refresh rates suffer due to the light diffuse response time (I think global flash could mitigate this but it would introduce crosstalk and possibly some harsher flicker in the mix) I think it's worth considering Gsync pulsar with global flash at this point.

Even if they didn't make the above changes, I think Gsync pulsar is too strict with the frame times OR reflex needs to be made "pulsar aware" because Vsync + Gsync + ULL = on or reflex ruins the gsync pulsar clarity and I think this is my biggest concern that needs to be addressed (Along with a gsync pulsar watermark that shows if gsync pulsar module is happy or mad with current frame pacing).

[brownvim]

He added your request to the list so good sign!

“per game NVAPP profiles to set monitor mode (G-SYNC, Pulsar, ULMB2, none), refresh rate“