NVIDIA G-Sync Pulsar monitor - Asus ROG Strix XG27AQNGV

[brownvim]

The new 60 Hz ULMB2 mode is brilliant by the way — I really appreciate you nudging them towards that.
My Switch 2 is calling me to play Super Mario Bros Wonder (released today). It must look amazing on this display.

Make sure it's not DRM-locked to NVIDIA GPUs. Some features of the monitor may be locked. (AFAIK, I believe the VRR-Pulsar is, at least).

Yeah Pulsar is locked to Nvidia GPUs.

The HDMI ports are not though, so can use the 60hz mode with PS5 Pro and Switch 2.

Did you not get the Pulsar display to test off Nvidia?

[brownvim]

https://youtu.be/RAtAlYzBikA?si=-mr-OfJSXGDgKV63

[Baron of Sun]

https://youtu.be/RAtAlYzBikA?si=-mr-OfJSXGDgKV63

Thanks for sharing!

It's not the first time I hear someone saying Pulsar for OLED needs research to get it work. Can someone explain why? I don't get it. The LG C1 and CX where already able to do subrefresh rolling scan. Why would it be so difficult to get it to work in comparison to an LCD?

[kyube]

It's not the first time I hear someone saying Pulsar for OLED needs research to get it work. Can someone explain why? I don't get it. The LG C1 and CX where already able to do subrefresh rolling scan. Why would it be so difficult to get it to work in comparison to an LCD?

Just because the aforementioned TVs can do BFI at all, doesn't mean that it's worthwhile using or even comparable to LCD backlight strobing.
The appeal of the ULMB2 models is that it can do <1ms strobe 'on' period (or MPRT, if you want to use the misnomer) @ 60Hz, which cuts the total transition time by 15,66667 (16,66667 ms - 15,66667 ms = 1ms)
Strictly speaking from a fixed refresh rate strobing implementation:
To be able to get OLED to the same levels of brightness and clarity as ULMB2, you'd need:
• HW Implementation (See custom rolling scan & discussion on BFI & reminder)
• 1000Hz capable OLED and use SW BFI (CRT shader emulator)
They both need to drive the brightness high enough to achieve ~100cd/m² @ <6% duty cycle.
That's a minimum 1700cd/m² voltage amplitude target. This also assumes that the pixels don't kill themselves from the high voltage.
This simply isn't a possible brightness value achievable on any OLED right now.
I have yet to see a OLED with <1ms MPRT.

A Pulsar(VRR+PWM) implementation makes this even more impossible.

https://youtu.be/RAtAlYzBikA?si=-mr-OfJSXGDgKV63

I already knew DT was a meme, but this one...
Shills 60hz strobing, yet shits on 8K display res even though it's a minimum target for 24-32" displays

[Baron of Sun]

It's not the first time I hear someone saying Pulsar for OLED needs research to get it work. Can someone explain why? I don't get it. The LG C1 and CX where already able to do subrefresh rolling scan. Why would it be so difficult to get it to work in comparison to an LCD?

Just because the aforementioned TVs can do BFI at all, doesn't mean that it's worthwhile using or even comparable to LCD backlight strobing.
The appeal of the ULMB2 models is that it can do <1ms strobe 'on' period (or MPRT, if you want to use the misnomer) @ 60Hz, which cuts the total transition time by 15,66667 (16,66667 ms - 15,66667 ms = 1ms)
Strictly speaking from a fixed refresh rate strobing implementation:
To be able to get OLED to the same levels of brightness and clarity, you need a 1000Hz capable OLED (they don't want to invest funds into a custom rolling scan) and drive the brightness high enough to achieve ~100cd/m² @ <6% duty cycle. That's a minimum 1700cd/m² voltage amplitude target.
This simply isn't a possible brightness value achievable on any OLED right now.

A Pulsar(VRR+PWM) implementation makes this even more impossible.

https://youtu.be/RAtAlYzBikA?si=-mr-OfJSXGDgKV63

I already knew DT was a meme, but this one...
Shills 60hz strobing, yet shits on 8K display res even though it's a minimum target for 24-32" displays

1000 Hz makes sense. That sounds like such a big number, but is it really so hard to reach in terms of motion blur reduction? Naive spoken, they need to turn off every pixel 1 ms after each pixel transition to get 1 ms persistence. That shouldn't be a bandwidth issue, just a timing issue and if the electronics are able to send a "turn off" signal fast enough. OLED pixel response time shouldn't be an issue either. In my mind it doesn't sound too complicated considering LG was able to have "effective 300 Hz" pixel updates 6 years ago.

In terms of brightness: The LG G5 for example has a 10% window peak brightness of 2400 nits or so. Diving by 16.67 is still 150 nits without the need to overdrive the voltage. Full field brightness should be a problem though. Maybe that's the bottleneck?

I'd be happy if you or someone else could explain which of my thoughts from above are wrong / naive or if the only reason is that manufacturers don't want to fund raise that as you mentioned before.

[brownvim]

https://youtu.be/RAtAlYzBikA?si=-mr-OfJSXGDgKV63

Thanks for sharing!

It's not the first time I hear someone saying Pulsar for OLED needs research to get it work. Can someone explain why? I don't get it. The LG C1 and CX where already able to do subrefresh rolling scan. Why would it be so difficult to get it to work in comparison to an LCD?

It basically comes down to perceived full-screen brightness.

Pulsar uses rolling scan strobing, lighting only ~25% of the screen at any one time, but it can hit peak brightness of 2000+ nits in those lit zones. This still gives you a perceived full-screen brightness of around 500 nits.

In contrast, OLED has to light the entire panel at once (no backlight). Even the brightest OLED like the G5 only reaches ~350–380 nits full-screen. If you short-pulse an OLED the same way Pulsar does, the perceived brightness would drop to around 95 nits.

That’s simply too dim for most people, and we’re talking about the brightest OLED panel available. Other OLEDs would be even lower.

[RynoW]

Now that 60Hz ULMB2 is added, I feel like I will definitely get one (between MSI and Asus, not sure if there are any differences worth noting). The last caveat to me is that many arcade systems used not-quite-60Hz framerates and someone tested the range that works on the ULMB2 mode, and found that it only supports 59.4-60.9Hz.

Many people use the Genlock feature on the Retrotink4k to get accurate frame outputs on things like the Neo Geo MVS (57Hz) that are out of range. Then there are the sickos who use PAL systems and content at 50Hz. So with that in mind, it'd still be great to get a more flexible ULMB2 mode, even if Pulsar will never extend down to 48-60Hz.

[RealNC]

OLED doesn't need 1000Hz. It just needs rolling scan. Which can be done on the firmware side at any refresh rate, including 60Hz.

The actual problem is brightness. On the LCD, each backlight strip is boosted to who knows how many thousand nits for a very short period of time (whatever the pulse width is.) OLEDs will get damaged at too high brightness output for too long a period. This limits ULMB brightness on OLED rather severely. The engineering challenge is to find a method to boost brightness to very high levels for a very short time that doesn't result in too much wear on the OLED layer.

The rolling scan part is easy and has been done already. A Dell display from years back did it. So "ULMB2" on OLED has already existed, but didn't get any traction. I suspect because of the severe brightness loss. However, now that OLEDs can do HDR, I hope this can be overcome. If they can get it to at least usable brightness, like 150 nits at least, it would still be a win. Especially since perceived brightness on OLED is higher than on LCD. (At least according to TFTCentral. For example 300 nits on OLED is perceived the same as like 550 nits on LCD or something like that.)

Btw, this is about the 60Hz ULMB mode, not Pulsar. Pulsar on OLED does indeed require ultra high refresh rates.

[Baron of Sun]

OLED doesn't need 1000Hz. It just needs rolling scan. Which can be done on the firmware side at any refresh rate, including 60Hz.

The actual problem is brightness. On the LCD, each backlight strip is boosted to who knows how many thousand nits for a very short period of time (whatever the pulse width is.) OLEDs will get damaged at too high brightness output for too long a period. This limits Pulsar brightness on OLED rather severely. The engineering challenge is to find a method to boost brightness to very high levels for a very short time that doesn't result in too much wear on the OLED layer.

The rolling scan part is easy and has been done already. A Dell display from years back did it. So "Pulsar" on OLED has already existed, but didn't get any traction. I suspect because of the severe brightness loss.

To me it seems like it depends on what you mean with "1000 Hz". In terms of video signal processing I'd say no. In terms of how fast do the display or its pixels have to update to achieve 1 ms persistence I'd say yes.

Thanks for your explanation. That makes sense to me. I hope manufacturers will solve that brightness problem. The brightness increase is quite noticeable from year to year. LG CX 700 nits in 2020 to LG G5 in 2400 nits in 2025 sounds like a positive trend. I really hope that's the main reason they ignore BFI / rolling scan at the moment.

[RealNC]

To me it seems like it depends on what you mean with "1000 Hz". In terms of video signal processing I'd say no. In terms of how fast do the display or its pixels have to update to achieve 1 ms persistence I'd say yes.

Thanks for your explanation. That makes sense to me. I hope manufacturers will solve that brightness problem. The brightness increase is quite noticeable from year to year. LG CX 700 nits in 2020 to LG G5 in 2400 nits in 2025 sounds like a positive trend. I really hope that's the main reason they ignore BFI / rolling scan at the moment.

Sorry, I keep saying "Pulsar" instead of "ULMB" I edited my post.

I mean the 60Hz ULMB mode in the Pulsar display doesn't require any especially hard engineering. It's just the brightness that is a challenge.

The Pulsar part (meaning VRR) would indeed require ultra high refresh rates in order to do the compensation pulses. Not in the input, mind you, but internally. The OLED monitor itself could still be your usual 240/360/480Hz display, but internally the firmware would need to be able to update it much faster than that, and sync everything up correctly.