Blur Busters Forums

_zebracrossing666 wrote: ↑

21 Feb 2025, 04:44

How does it prevent VRR from working if it's removing screen tearing from Cyberpunk? There was noticeable tearing on neon lights when moving the camera, but now it is smooth. So it's not subjective. Unless that's due to me also enabling Vsync in Nvidia app along with Gsync, but the latency was good and my fps around 80-90, far below 240hz(if the method is having zero effect, and I thought Vsync only prevented tearing at fps above max refresh rate?)

Shifroval wrote: ↑

21 Feb 2025, 05:30

_zebracrossing666 wrote: ↑

20 Feb 2025, 10:42

Why are you colour temps so low in general though? Lowest I have seen was putting G at 96 and B at 95

Low? I can't say, last time I touched them was long time ago, I guess I liked these values? Bigger numbers look more 'bright' to my eyes, and I don't like bright displays, apart from hdr mode in games, it's one story when you're gaming (dynamic picture with bright and dark scenes) and another when sitting at desktop and reading text. Just for the sake of it I upped them to 95/90/97. That completely destroyed white balance, so white became more yellow. As I said I like more bluish whites, and these numbers provide the white that I like. Iirc last time I was fiddling with rgb I saw some sort of dependecy, like when one of the colors exceeds a specific value (with other two upped accordingly), the display suddenly changes overall color temperature and becomes more red/green/blue. For example when I set 78 it's 'ok', but when I set 79 it becomes more of that color. But when I lower to 76 it's not that apparent. I'm no color expert and just assume all of this completely from a layman's point of view.

_zebracrossing666 wrote: ↑

20 Feb 2025, 10:42

edit: Also I tried using an ICC profile for the sRGB clamp, but it adds this horrible cloudy filter over everything. For some reason enabling EDID primaries instead clamps the colours to sRGB without the cloudy filter. Idk how it doesn't work for you, when I enable the EDID primaries, I can notice colours are no longer in Wide Gamut and look less saturated

Edid primaries change all other colors, but do not fix the black crush. Just checked it again. Opened a colorful picture and switched to primaries. Some colors indeed became more tame and less saturated, but lagom black test still shows only the lowest row.
And btw it's not a filter, nor it's cloudy. Novideo can't produce such things, it only applied the values set in the profile on the driver level.
If you looked at your crushed deep blacks long enough you start taking them at face value, like that's how it looks. But when black levels suddenly become more distinguishable, of course your brain resists that and you see a duller picture. Purely a psychological thing.
The only correct answer to this is to put another correctly calibrated display (without black crush problems) near this one and see the difference. Other colors still look bright and vivid enough for me.

RealNC wrote: ↑

21 Feb 2025, 05:50

_zebracrossing666 wrote: ↑

21 Feb 2025, 04:44

How does it prevent VRR from working if it's removing screen tearing from Cyberpunk? There was noticeable tearing on neon lights when moving the camera, but now it is smooth. So it's not subjective. Unless that's due to me also enabling Vsync in Nvidia app along with Gsync, but the latency was good and my fps around 80-90, far below 240hz(if the method is having zero effect, and I thought Vsync only prevented tearing at fps above max refresh rate?)

It prevents VRR because there's judder when FPS does not match Hz. Also, it has duplicate frames. See first post in this thread about the borked 180Hz mode. If you apply the VT/pixel clock tweak to any other mode so that VRR flicker "goes away," you end up with the same problem the 180Hz mode has.

_zebracrossing666 wrote: ↑

21 Feb 2025, 05:59

Okay, well if there is indeed judder, for me it's certainly smoother than playing with VRR off and no sync at all. The screen tearing on Cyberpunk made moving the camera around feel terrible, as it was small screen tears but very noticeable when moving the camera slowly and around objects with a lot of light. Vsync by itself adds a ton of latency and judder and should be noticeable, which according to you should be the only sync in play here, yet latency feels the exact same, and on top of that screen tearing is gone.

So idk which sync is improving all this, I'm so confused as I was playing RDR2 a week ago with just Vsync and no VRR and it felt like crap. Juddery like hell when moving the camera. But I played it yesterday with apparently just Vsync????(with the non-working VRR enabled) and it felt smooth af.

_zebracrossing666 wrote: ↑

21 Feb 2025, 05:50

And about the EDID primaries, there is almost no black crush for me on Lagom's black level test. I can see square 3 clearly and square 2 if I concentrate on the overall picture. This is with 100 brightness though, which I think it should be to make it fair. Like obv there will be more black crush if you have it at 50 brightness or something. Yes, I can see even square 1 very clearly with the ICC profile, but to me the black levels look raised as square 1 should be barely distinguishable. I'm not an expert either tho, just square 1 blends in more with the black background with EDID.

It's weird your black crush is so much more apparent, maybe this is panel dependent stuff and is slightly better or worse on each monitor?

RealNC wrote: ↑

21 Feb 2025, 06:27

_zebracrossing666 wrote: ↑

21 Feb 2025, 05:59

Okay, well if there is indeed judder, for me it's certainly smoother than playing with VRR off and no sync at all. The screen tearing on Cyberpunk made moving the camera around feel terrible, as it was small screen tears but very noticeable when moving the camera slowly and around objects with a lot of light. Vsync by itself adds a ton of latency and judder and should be noticeable, which according to you should be the only sync in play here, yet latency feels the exact same, and on top of that screen tearing is gone.

So idk which sync is improving all this, I'm so confused as I was playing RDR2 a week ago with just Vsync and no VRR and it felt like crap. Juddery like hell when moving the camera. But I played it yesterday with apparently just Vsync????(with the non-working VRR enabled) and it felt smooth af.

I don't know what exactly you're seeing, but here I can tell there is a difference in motion with working VRR vs broken VRR. Especially visible when there's side scrolling, where the difference becomes quite big. The broken VRR modes are jittery.

Jittery motion is also easier to see when something bright and sharp is moving in front of a dark background. And the faster the motion, the more apparent the jitter becomes.

Another thing to keep in mind is that when there's no VRR flicker, that just means there's no VRR. Gamma response differs between different refresh rates. This is the source of VRR flicker. When the gamma response doesn't change, that means the refresh rate doesn't change. A dynamically changing refresh rate is what VRR is supposed to be doing.

Shifroval wrote: ↑

21 Feb 2025, 06:33

_zebracrossing666 wrote: ↑

21 Feb 2025, 05:50

And about the EDID primaries, there is almost no black crush for me on Lagom's black level test. I can see square 3 clearly and square 2 if I concentrate on the overall picture. This is with 100 brightness though, which I think it should be to make it fair. Like obv there will be more black crush if you have it at 50 brightness or something. Yes, I can see even square 1 very clearly with the ICC profile, but to me the black levels look raised as square 1 should be barely distinguishable. I'm not an expert either tho, just square 1 blends in more with the black background with EDID.

It's weird your black crush is so much more apparent, maybe this is panel dependent stuff and is slightly better or worse on each monitor?

That's odd. I tried with brightness 100 and uniform brightness off/on and primaries, the most I can see is the second lowest row, square 10. How does this method work for everyone, but not for me? Rtings profile works better if I have higher brightness, but I don't want that.
Maybe you use another profile for the monitor, like racing mode? I use user mode and brightness 25, because my eyes will melt with higher values if I spend half of the day using the monitor. Contrast is 80.
Something is amiss here. Ideally all units should be more or less the same and definitely not exibit such different behavior when using the same settings.

_zebracrossing666 wrote: ↑

21 Feb 2025, 07:33

Shifroval wrote: ↑

21 Feb 2025, 06:33

_zebracrossing666 wrote: ↑

21 Feb 2025, 05:50

And about the EDID primaries, there is almost no black crush for me on Lagom's black level test. I can see square 3 clearly and square 2 if I concentrate on the overall picture. This is with 100 brightness though, which I think it should be to make it fair. Like obv there will be more black crush if you have it at 50 brightness or something. Yes, I can see even square 1 very clearly with the ICC profile, but to me the black levels look raised as square 1 should be barely distinguishable. I'm not an expert either tho, just square 1 blends in more with the black background with EDID.

It's weird your black crush is so much more apparent, maybe this is panel dependent stuff and is slightly better or worse on each monitor?

That's odd. I tried with brightness 100 and uniform brightness off/on and primaries, the most I can see is the second lowest row, square 10. How does this method work for everyone, but not for me? Rtings profile works better if I have higher brightness, but I don't want that.
Maybe you use another profile for the monitor, like racing mode? I use user mode and brightness 25, because my eyes will melt with higher values if I spend half of the day using the monitor. Contrast is 80.
Something is amiss here. Ideally all units should be more or less the same and definitely not exibit such different behavior when using the same settings.

Here is a photo of my Lagom black level test, I know the camera won't pick it up exactly like eyes see it, but it's roughly the same and square 2 is noticeably visible but also not raised compared to the background. This is on User mode + 100 brightness, even as low as 60 brightness is good enough to see it though.

https://imgur.com/a/VLSg9p1

edit: This is it with ICC profile https://imgur.com/a/Z2jkQn5 square 1 looks like it's raised too much, I'm not 100% sure but I think square 1 should be BARELY noticeable and almost fully blend in with the black background

Shifroval wrote: ↑

21 Feb 2025, 08:08

_zebracrossing666 wrote: ↑

21 Feb 2025, 07:33

Shifroval wrote: ↑

21 Feb 2025, 06:33

_zebracrossing666 wrote: ↑

21 Feb 2025, 05:50

And about the EDID primaries, there is almost no black crush for me on Lagom's black level test. I can see square 3 clearly and square 2 if I concentrate on the overall picture. This is with 100 brightness though, which I think it should be to make it fair. Like obv there will be more black crush if you have it at 50 brightness or something. Yes, I can see even square 1 very clearly with the ICC profile, but to me the black levels look raised as square 1 should be barely distinguishable. I'm not an expert either tho, just square 1 blends in more with the black background with EDID.

It's weird your black crush is so much more apparent, maybe this is panel dependent stuff and is slightly better or worse on each monitor?

That's odd. I tried with brightness 100 and uniform brightness off/on and primaries, the most I can see is the second lowest row, square 10. How does this method work for everyone, but not for me? Rtings profile works better if I have higher brightness, but I don't want that.
Maybe you use another profile for the monitor, like racing mode? I use user mode and brightness 25, because my eyes will melt with higher values if I spend half of the day using the monitor. Contrast is 80.
Something is amiss here. Ideally all units should be more or less the same and definitely not exibit such different behavior when using the same settings.

Here is a photo of my Lagom black level test, I know the camera won't pick it up exactly like eyes see it, but it's roughly the same and square 2 is noticeably visible but also not raised compared to the background. This is on User mode + 100 brightness, even as low as 60 brightness is good enough to see it though.

https://imgur.com/a/VLSg9p1

edit: This is it with ICC profile https://imgur.com/a/Z2jkQn5 square 1 looks like it's raised too much, I'm not 100% sure but I think square 1 should be BARELY noticeable and almost fully blend in with the black background

Ok, that's even more strange. I can't reproduce the same thing no matter what I do or did in the past. What's left is compare everything I know that can alter the end result.

1. You use display port connection?
2. Do you have a loaded icc profile that shows in windows color management? If you installed a driver from asus, you probably have one. Open it with colorcpl command (win + r then paste the command).
3. Any custom nvidia control panel options? Especially that color reference options, or whatever they're called.
4. 8 bit or 10 bit? I use 12 bit tweak, but I doubt it changes anything.
5. Uh, you have Win 11, right?
6. Have you ran windows hdr calibration app?
7. Do you have dldsr/dsr modes enabled?
8. I also used cru to rise vrr floor value to 70, but that shouldn't affect colors in any way.
9. Just to rule out this as well, hwinfo reports that my unit was manufactured on week 25, year 2024. What does it show for you?
10. You have vrr enabed? In nvidia panel and monitor settings? I assume it based on your discussuon with RealNC, but still better to know for sure.
11. And the question that deserves an award for stupidity, you use sdr mode in windows, right?

Please tell me all your monitor settings, just in case. Like from every tab, if you have something enabled or changed.

For me all that changed when I enable edid primaries are some colors, not the depth of the blacks. I was genuenely surprised when I first tried it to fix black crush like it was suggested in the first post, because it changed almost nothing and I thought I was doing something wrong.

RealNC wrote: ↑

21 Feb 2025, 05:50

It prevents VRR because there's judder when FPS does not match Hz. Also, it has duplicate frames. See "Update 2" in the first post of this thread about the borked 180Hz mode. If you apply the VT/pixel clock tweak to any other mode so that VRR flicker "goes away," you end up with the same problem the 180Hz mode has.