Did I unfairly judge strobing?

[GFresha]

As an avid eSports and online gamer who doesn't play single-player games, I've been in the market for a monitor upgrade from my TN 1080p 240Hz. I was torn between opting for an LCD with Dyac+/ULMB2 or going for an OLED if I wasn't planning to use strobing.

I recently snagged a great deal on a 1440p 360Hz OLED locally and put it to the test with some Valorant games. To compare the OLED to ULMB, I switched back to my old 144Hz LCD (XB271HU). The XB271HU only had ULMB at 120Hz, so I enabled that feature and was underwhelmed by the results. In terms of motion clarity, the difference was negligible - almost imperceptible - compared to the 360Hz OLED. Moreover, the experience was even worse by the need to disable VRR, which made the gameplay feel jarring despite matching strobe rate to the frame rate to the refresh rate. The upcoming G-Sync Pulsar technology may alleviate this issue, allowing for a smoother VRR experience with strobing.

However, putting VRR and smoothness aside, I was disappointed by the lackluster motion clarity of ULMB on the XB271HU. Given that 120Hz strobing should theoretically improved clarity, I was expecting more.

This brings me to my question: is ULMB 2 a significant improvement over ULMB in terms of motion clarity? I'm not concerned with brightness, flicker, or refresh rates, but solely with motion clarity. Nvidia promised substantial improvements, claiming motion clarity equivalent to 1000Hz or 1440Hz. Does ULMB 2 genuinely offer a massive clarity boost over ULMB 1? Would a comparison between ULMB 2 at 120Hz and ULMB 1 at 120Hz reveal a night-and-day difference in motion clarity or is the difference in ULMB 2 vs 1 mainly felt in brightness/being able to use it at higher Hz/less flicker etc...

Another question: was my comparison between 120Hz ULMB and 360Hz OLED unfair in terms of pure motion clarity? Should I have compared 360Hz/540Hz ULMB2 to 360Hz OLED for a more accurate assessment?

After scratching my ULMB/strobing itch with the XB271HU, I'll be returning to the OLED 360Hz. However, I'm open to revisiting this comparison in the future, possibly with a 360Hz+ ULMB 2 pulsar setup. For now, I prioritize smoothness over clarity, and the OLED's VRR and smoothness win out over strobing.

If anyone else has faced a similar dilemma and tested both OLED and strobing, I'd love to hear about your experiences!

[RealNC]

Another question: was my comparison between 120Hz ULMB and 360Hz OLED unfair in terms of pure motion clarity? Should I have compared 360Hz/540Hz ULMB2 to 360Hz OLED for a more accurate assessment?

One thing to keep in mind here is that on OLED, motion clarity is inherently better because of near zero pixel response time. Motion clarity at 360FPS/360Hz on OLED is more comparable to 480FPS/480Hz on LCD (without strobing.) For 480Hz to look its best on an LCD, pixel response time needs to be lower than 2ms, and that's cutting it very close even with a fast TN panel.

So if you can reach that kind of motion clarity at 120FPS/Hz, that's not underwhelming at all. It's actually quite good. Which is exactly what ULMB is trying to do to begin with - achieve motion clarity that matches higher FPS/Hz without actually needing to be able to reach that higher FPS.

So yes, it makes more sense to compare 120Hz ULMB against 120Hz on the OLED. ULMB should win that battle, but with the drawback of having some crosstalk. Without ULMB, 120Hz on the OLED will look clearer though simply because of how fast OLED pixels are.

[GFresha]

One thing to keep in mind here is that on OLED, motion clarity is inherently better because of near zero pixel response time. Motion clarity at 360FPS/360Hz on OLED is more comparable to 480FPS/480Hz on LCD (without strobing.) For 480Hz to look its best on an LCD, pixel response time needs to be lower than 2ms, and that's cutting it very close even with a fast TN panel.

So if you can reach that kind of motion clarity at 120FPS/Hz, that's not underwhelming at all. It's actually quite good. Which is exactly what ULMB is trying to do to begin with - achieve motion clarity that matches higher FPS/Hz without actually needing to be able to reach that higher FPS.

So yes, it makes more sense to compare 120Hz ULMB against 120Hz on the OLED. ULMB should win that battle, but with the drawback of having some crosstalk. Without ULMB, 120Hz on the OLED will look clearer though simply because of how fast OLED pixels are.

You gave me a really good comparison with 360Hz OLED is more compariable to 480Hz LCD (without strobing)

So Its fair to assume a 480Hz OLED is going to compare to a 600-640Hz LCD (unstrobed)?

How about strobing wise, what does 120 Hz, 240Hz, 360Hz, and even 540Hz strobing compare to? Assuming we have the formula of FPS=Hz=Stroberate, and assuming pulsewidth is calibrated correctly.

For a 360Hz monitor with ULMB 2, NVIDIA claims an effective clarity of 1440Hz. Using this as a reference point, would a 540Hz strobing rate translate to an effective clarity of 2160Hz? Or am I off the mark?

Also is that the same for dyac+/dyac2? So a 360Hz ULMB 2 = 1440Hz clarity and dyac+/dyac2 360Hz is also 1440Hz clarity or we can't compare it straight like that? As they are different technologies?

[Discorz]

For a 360Hz monitor with ULMB 2, NVIDIA claims an effective clarity of 1440Hz. Using this as a reference point, would a 540Hz strobing rate translate to an effective clarity of 2160Hz? Or am I off the mark?

Also is that the same for dyac+/dyac2? So a 360Hz ULMB 2 = 1440Hz clarity and dyac+/dyac2 360Hz is also 1440Hz clarity or we can't compare it straight like that? As they are different technologies?

Mathematically speaking u're correct. ULMB 1 and 2 are 2.5%-25% duty cycle (setting of 10-100). 360Hz display at 25% duty cycle has similar motion clarity to 1440Hz display. 1000/360=2.77ms, 2.77*0.25=0.69ms, 1000/0.69=1440Hz. And 540Hz would be similar to 2160Hz. You can push that all the way up to 21600Hz with 2.5%, but at a much lower brightness...

But as you have mentioned due to GtG limitations the mathematical MPRT or Hz numbers are not always correct - the Y-axis (intensity) is also a factor, not just MPRT X-axis (time). While the two can be measured they essentially turn into some form of Z through our visual perception. Its even harder to say for strobe modes because smearing from GtG become double images.

On top of all, flickering in general messes with the perception. So its very complicated. Flicker based blur reduction only helps eye-tracking which is 1 out of 3 eye-movement vs content-movement cases. There can be:

1A. stationary eyes on stationary images
1B. stationary eyes on moving images
2A. moving eyes on stationary images
2B. moving eyes on moving images

Using term "motion clarity" automatically implies eye-tracking. If your eyes are not tracking moving objects (1B - stationary eyes on moving images) u're pretty much immune to strobing/bfi. And if something is static and eyes are moving (2A - moving eyes on stationary images) flickering introduces stroboscopic effect (see this ilustrition). Brute sample rate on the other hand helps all 3 cases. This is why impulsed MPRT and sample-and-hold MPRT are not the same and why some people see no difference with strobing on/off. I'm hoping people are aware of this.

Same applies to new DyAc 2. Only difference they're doing is rolling the backlight to reduce top-bottom crosstalk instead of globally flashing it like with DyAc 1 and other LCD strobe modes. At 540Hz I believe they're doing triangle-like ~0.5ms pulse instead of standard square-like pulse, however its not a percentage of refresh cycle so MPRT doesn't scale linearly across the range.

[GFresha]

Thanks!

[Falkentyne]

For a 360Hz monitor with ULMB 2, NVIDIA claims an effective clarity of 1440Hz. Using this as a reference point, would a 540Hz strobing rate translate to an effective clarity of 2160Hz? Or am I off the mark?

Also is that the same for dyac+/dyac2? So a 360Hz ULMB 2 = 1440Hz clarity and dyac+/dyac2 360Hz is also 1440Hz clarity or we can't compare it straight like that? As they are different technologies?

Mathematically speaking u're correct. ULMB 1 and 2 are 2.5%-25% duty cycle (setting of 10-100). 360Hz display at 25% duty cycle has similar motion clarity to 1440Hz display. 1000/360=2.77ms, 2.77*0.25=0.69ms, 1000/0.69=1440Hz. And 540Hz would be similar to 2160Hz. You can push that all the way up to 21600Hz with 2.5%, but at a much lower brightness...

But as you have mentioned due to GtG limitations the mathematical MPRT or Hz numbers are not always correct - the Y-axis (intensity) is also a factor, not just MPRT X-axis (time). While the two can be measured they essentially turn into some form of Z through our visual perception. Its even harder to say for strobe modes because smearing from GtG become double images.

On top of all, flickering in general messes with the perception. So its very complicated. Flicker based blur reduction only helps eye-tracking which is 1 out of 3 eye-movement vs content-movement cases. There can be:

1A. stationary eyes on stationary images
1B. stationary eyes on moving images
2A. moving eyes on stationary images
2B. moving eyes on moving images

Using term "motion clarity" automatically implies eye-tracking. If your eyes are not tracking moving objects (1B - stationary eyes on moving images) u're pretty much immune to strobing/bfi. And if something is static and eyes are moving (2A - moving eyes on stationary images) flickering introduces stroboscopic effect (see this ilustrition). Brute sample rate on the other hand helps all 3 cases. This is why impulsed MPRT and sample-and-hold MPRT are not the same and why some people see no difference with strobing on/off. I'm hoping people are aware of this.

Same applies to new DyAc 2. Only difference they're doing is rolling the backlight to reduce top-bottom crosstalk instead of globally flashing it like with DyAc 1 and other LCD strobe modes. At 540Hz I believe they're doing triangle-like ~0.5ms pulse instead of standard square-like pulse, however its not a percentage of refresh cycle so MPRT doesn't scale linearly across the range.

They're doing rolling scan now with Dyac2, yet still won't strobe below 100hz?
Man, some companies just don't like money, I guess....