Blur Busters Forums

Discorz wrote: ↑

07 Jan 2023, 05:48

You need much faster moving speeds to notice a difference. At 1000 px/sec you won't notice sub 1ms MPRT (persistence). For 1000 px/sec motion to look perfectly sharp you need 1ms of MPRT, for 2000 px/sec you need 0.5ms MPRT, and so on... Zowie is always sub 1ms, I think 66K at premium is about 24% duty cycle which makes it 0.66ms MPRT at 360Hz.

Had to try it Ufo Speed 2880px/s. The difference is still not huge, certainly not visible to eyes. But yeah, looks like you can squeeze a little tiny bit of extra out, with a huge penalty to brightness. But damn this monitor be fast. AMA 12 in these shots btw.

And one has to realize something moving accross 1080p screen 2880px/s is insanely fast. Too fast for eyes (atleast mine) to track consistently. 1,5 times through the whole screen in 1 second. Yet it still remains highly recognizable.

what do you think is dyac harmful to the eyes in the long run? Don't feel like having to wear glasses at some point or getting any eye disease..

planart wrote: ↑

07 Jan 2023, 07:04

Had to try it Ufo Speed 2880px/s. The difference is still not huge, certainly not visible to eyes. But yeah, looks like you can squeeze a little tiny bit of extra out, with a huge penalty to brightness. But damn this monitor be fast. AMA 12 in these shots btw.

And one has to realize something moving accross 1080p screen 2880px/s is insanely fast. Too fast for eyes (atleast mine) to track consistently. 1,5 times through the whole screen in 1 second. Yet it still remains highly recognizable.

xl2566k_2880pxs_blurbusters_persistence.jpg

Thank you for your Ufotests!

Does anyone know what's the stock Persistence value for Dyac High and Premium? I was thinking about settling for 19~18.

Discorz wrote: ↑

07 Jan 2023, 05:48

You need much faster moving speeds to notice a difference. At 1000 px/sec you won't notice sub 1ms MPRT (persistence). For 1000 px/sec motion to look perfectly sharp you need 1ms of MPRT, for 2000 px/sec you need 0.5ms MPRT, and so on... Zowie is always sub 1ms, I think 66K at premium is about 24% duty cycle which makes it 0.66ms MPRT at 360Hz.

I wonder what I should use when gaming. I find myself making movements faster than 3840px/s when aiming in games like Overwatch or Apex.

To quote TFTCentral:

The only difference between the two modes is the “on” period of the strobe. In the ‘High’ mode the on period is measured at 0.8125ms, while in the ‘Premium’ mode it’s slightly shorter at 0.625ms. In theory this could help reduce perceived motion blur a little more, but in practice there’s very little visual difference. Still, every little helps when it comes to highly competitive gaming so we’d probably recommend sticking with the ‘Premium’ DyAc+ mode.

source: https://tftcentral.co.uk/reviews/benq-zowie-xl2566k

Assuming that someone wants to stick to a value of 25, is it possible to calculate the duration of the off duty cycle based on the utility's Persistence values when we already know that the refresh window lasts for 2.778ms?

planart wrote: ↑

07 Jan 2023, 07:04

Had to try it Ufo Speed 2880px/s. The difference is still not huge, certainly not visible to eyes. But yeah, looks like you can squeeze a little tiny bit of extra out, with a huge penalty to brightness. But damn this monitor be fast. AMA 12 in these shots btw.

And one has to realize something moving accross 1080p screen 2880px/s is insanely fast. Too fast for eyes (atleast mine) to track consistently. 1,5 times through the whole screen in 1 second. Yet it still remains highly recognizable.

Very good tracking for such high speed!

axaro1 wrote: ↑

07 Jan 2023, 10:36

I wonder what I should use when gaming. I find myself making movements faster than 3840px/s when aiming in games like Overwatch or Apex.

To quote TFTCentral:

The only difference between the two modes is the “on” period of the strobe. In the ‘High’ mode the on period is measured at 0.8125ms, while in the ‘Premium’ mode it’s slightly shorter at 0.625ms. In theory this could help reduce perceived motion blur a little more, but in practice there’s very little visual difference. Still, every little helps when it comes to highly competitive gaming so we’d probably recommend sticking with the ‘Premium’ DyAc+ mode.

source: https://tftcentral.co.uk/reviews/benq-zowie-xl2566k

Persistence is already low enough that it doesn't really matter. But I guess whichever fastest speed your eyes can physically track determines the persistence u need. Now this is highly dependant on screen width for a horizontally moving object. It's easier to track on larger than smaller screens. Resolution and object size also play a role here.

On a bright backlight like this I'd keep reducing the persistence until I'm satisfied with the overall brightness because its a personal preference thing. That way you get most out of it.

axaro1 wrote: ↑

07 Jan 2023, 10:36

Assuming that someone wants to stick to a value of 25, is it possible to calculate the duration of the off duty cycle based on the utility's Persistence values when we already know that the refresh window lasts for 2.778ms?

Yes! Ideally as you know this can be done with photo diode measuring tools, but if you don't have such equipment you can use this simple pattern I recently made for Smooth Frog. It allows you to read out your display's average persistence. Lower the persistence, the faster speeds you need to detect it. And faster the speed, smaller the margin of error. Try 4000 px/sec, but it might be hard for 1920p screen.
viewtopic.php?p=87695#p87695

For my M32Q I measured between 0.75-1ms. Checking rtings chart reveals in fact persistence is about 0.8ms.

Discorz wrote: ↑

07 Jan 2023, 11:29

Hi Discorz, thank you for the pattern!! Smoothfrog being limited to 3840px/sec shouldn't be an issue when using the 4000px/sec pattern, correct?

axaro1 wrote: ↑

07 Jan 2023, 12:09

Hi Discorz, thank you for the pattern!! Smoothfrog being limited to 3840px/sec shouldn't be an issue when using the 4000px/sec pattern, correct?

For this test to work properly its necessary to use Pixels/Frame mode and speed must be closest possible to target. Avoid Pixels/Second mode as motion will look jittery, and you'll be able to achieve much faster speeds. Also check the instructions on original post for more info.

Further findings with using Blur Busters Calibration Utility - and eye health.

So, I've had some eye health issues with strobing technologies in the past. It has been very apparent with Zowie TN DYAC implementation. Every time, even after short session I started to feel a bit ill. Nothing huge, but just enough to notice it. Eye strains, slight headache, little bit of nausea, various symptoms really. This made me dissapointed, because strobing is something which I truly feel sets Zowie apart from the rest. Especially XL2456K without strobing felt kind of moot purchase.

Sadly same symptoms came back with XL2566K. Not quite as strong. But still definitely there. XL2566K even without strobing is ofcourse quite impressive and I was quite settled on the idea that no Dyac for me this time either.

I'm not 100% sure on this yet, but after full days of use and playing it seems setting "Persistence" to minimum value of 1 and I don't feel ill one bit. This really is great news for me if it proves out to be true in the long run. Even with the minimum setting the visual difference is huge if you compare it to not having it at all.

So in case Dyac has caused you eye strains or some negative effects like that, this could be worth a try for you too. Very happy that I found this. Cant understand why Zowie has not implemented a more granular setting in OSD directly. I can't be the only one that needs this.

The rtings review is such a joke. Docking points for bad viewing angles is actually braindead for a 25inch monitor for competitive gaming lmfao.

axaro1 wrote: ↑

07 Jan 2023, 12:09

Have you tried the test? I'm interested to hear about your readouts.

I forgot to mention you can do the same test for other refresh rates and get the average. It works this way because Zowie strobes at fixed percentage of refresh cycles for one setting so same duty cycle applies to all refresh rates.

If its hard to eye-track you can start with slower speeds like 1000pps to see where the lines meet first, than increase to 2000pps for slightly more detailed readout, and than 4000pps for even more accuracy. The universal pattern is good for this. You can also edit the PNG image to your liking if it's too large.

If you were to plot out the measured data it would look something like this:

Discorz wrote: ↑

10 Jan 2023, 06:46

Here's my results at 3960px/sec (360 * 11).
The brighter the screen the harder it was to capture motion pursuit results without camera artifacts (for reference, I'm using a Oneplus 8T capturing at 2160p 57~60fps, I don't have better tools), having to track at this speed didn't help either: you basically have to move your phone with no reference point and hope that you end up matching the ruler, being slightly slower or faster will end up ruining the final result.

That being said, we can definitely see the improvements of Persistence tweaking especially when looking at the lines for 0.5ms.

I'm extremely surprised by Persistence = 25, too bad that it's way too dim.
I added an extra close-up for Persistence = 20 since I'm not 100% satisfied with the first pursuit result.

- Persistence = 1



- Persistence = 10



- Persistence = 15



- Persistence = 20

Better close-up, still with Persistence = 20



- Persistence = 25