Monitor Tests Forum

(04-22-2017 05:35 AM)LinCeX Wrote: [ -> ]Hey , im having a problem wich could be related to CRU .
http://imgur.com/a/W5u7d (Sorry that it is on spanish)
As you can see it gets locks and i cant change the screen resolution , but also clicking on Display adapter properties and nothing happens , it doesnt open . Also cant launch games at full screen . sorry for my bad english but its hard to explain .

What video card? If this is with AMD, it might be a driver bug. Are you using the latest driver? http://support.amd.com/en-us/kb-articles...Notes.aspx

(04-22-2017 08:56 PM)ToastyX Wrote: [ -> ]

(04-22-2017 05:35 AM)LinCeX Wrote: [ -> ]Hey , im having a problem wich could be related to CRU .
http://imgur.com/a/W5u7d (Sorry that it is on spanish)
As you can see it gets locks and i cant change the screen resolution , but also clicking on Display adapter properties and nothing happens , it doesnt open . Also cant launch games at full screen . sorry for my bad english but its hard to explain .

What video card? If this is with AMD, it might be a driver bug. Are you using the latest driver? http://support.amd.com/en-us/kb-articles...Notes.aspx

R9 270x . Tested with latest drivers .

(04-22-2017 08:55 PM)ToastyX Wrote: [ -> ]

(04-22-2017 06:56 AM)falkentyne Wrote: [ -> ]Is the laptop's mini displayport bandwidth capped at 330 MHz or something? Why wouldn't the custom 120hz resolution show up?

DisplayPort 1.0/1.1 is limited to 360 MHz with 8-bit color and 480 MHz with 6-bit color. The monitor must support DisplayPort 1.2 and HBR2 (High Bit Rate 2) mode to go higher. You might be able to squeeze in a 1500 vertical total by reducing the horizontal total to 2000 (with the front porch or sync width reduced to accommodate) if the monitor can handle that.

Yeah I know. I was at 359.94 MHz by reducing the HT to 2001, but the resolution still didn't show up. I didn't even have an option to switch to 6 bit color, because the refresh rate didn't show up. At 359.94 MHz and lower on my R9 290x, I can choose 6 or 8 bit. At 360 MHz+, it's 6 bit only. I had no problem with this on the R9 290X.

The problem is, this isn't working on the Geforce...only on the 290X.

Also, the "advanced adapter settings" didn't open at all when I did the 120hz custom resolution. I'll remove it later and only do the 100hz one and then see if I can get that window open. I hope I didn't run into a windows 10 bug ...(I'm extremely new to 10).

Anyway, to refresh: with the mini displayport on the Geforce 1070 on the laptop, the refresh rate doesn't even show up at all. It's like, not there. And the limits are in range.

2080/1500 should be 374.40 MHz and 6 bit color only, but it didn't show up in the NVCP.
2001/1500 is 359.94 MHz, and should work at 8 bit color, but again it didn't show up in the NVCP.

So I was wondering if somehow there was a limit in place that doesn't happen on desktop cards. Or maybe I'm still too new to Nvidia. I am in windows 10 on the laptop and 7 on the desktop. I was just wondering if I'm doing something wrong.

I'll mess around with it later when I hook up my laptop again, but that's really bizarre.

As always, thank you for all your help. Sorry if I'm nagging.

Hi ToastyX
There seems to be a bug with the driver here.
I am using the "Creator's update hotfix" Nvidia driver. (Forgot the version but it was from a week ago).

1) With the XL2720z plugged into mini displayport on the mobile GTX 1070, I lose the ability to choose between 6 and 8 bit. It's 8 bit only, even at 60hz or 144hz.
2) attempting to use the Nvidia control panel (NVCP) to set a custom refresh rate that has a pixel clock higher than 359.85 MHz causes the Benq to black screen and the monitor is "Frozen" until the test resolution fails and times out (or I just press "No" blindly). The OSD on the Benq does not respond.
359.94 MHz (HT: 2001, VT: 1499, at 120hz) black screens even though the pixel clock is less than 360 mhz.

3) My favorite 125hz VT tweak I use on my R9 290x also does not appear if installed with your CRU (2080 HT, 1497 VT, pixel clock 389 MHz)
If tested in NVCP, monitor black screens (OSD unresponsive) until test times out.

However 359.70 MHz and 359.46 MHz (VT 1498 and VT 1497) both TEST successfully. So it can be added in the NVCP itself and it works.
However if I use your CRU to add them, the refresh rate does not appear at all. So the driver is blocking those pixel clocks if added through CRU, but using the NVCP works. (I would prefer NOT using the NVCP because I don't trust Nvidia).

However VT 1485 (356.58 MHz) DOES appear if I add this in your CRU, but this VT is not accepted properly by the monitor (I see the windows screen but an "Out of range" error pops up anyway). I have to use 1497-1502 which clearly isn't working

This seems to definitely be a driver issue with the laptop GPUs.

I guess this explains why "Advanced adapter properties" does not even load when I have it selected to use the Benq screen.

I have NOT tried an older driver like 378.xx. I suspect it won't work and it took me a long time to test this for you.

How do I report this to Nvidia? Will they even care? I'm 100% certain a monitor capable of HBR2 without reverting to 6 BPC at >359-480 MHz would work perfectly (since no one would care if 6 bpc is missing due to a bug anyway), so this is a pretty huge bug. I have no idea if this happens on desktop 10xx series gpus as I don't have one

(04-23-2017 12:05 AM)falkentyne Wrote: [ -> ]Hi ToastyX
There seems to be a bug with the driver here.
I am using the "Creator's update hotfix" Nvidia driver. (Forgot the version but it was from a week ago).

1) With the XL2720z plugged into mini displayport on the mobile GTX 1070, I lose the ability to choose between 6 and 8 bit. It's 8 bit only, even at 60hz or 144hz.
2) attempting to use the Nvidia control panel (NVCP) to set a custom refresh rate that has a pixel clock higher than 359.85 MHz causes the Benq to black screen and the monitor is "Frozen" until the test resolution fails and times out (or I just press "No" blindly). The OSD on the Benq does not respond.
359.94 MHz (HT: 2001, VT: 1499, at 120hz) black screens even though the pixel clock is less than 360 mhz.

3) My favorite 125hz VT tweak I use on my R9 290x also does not appear if installed with your CRU (2080 HT, 1497 VT, pixel clock 389 MHz)
If tested in NVCP, monitor black screens (OSD unresponsive) until test times out.

However 359.70 MHz and 359.46 MHz (VT 1498 and VT 1497) both TEST successfully. So it can be added in the NVCP itself and it works.
However if I use your CRU to add them, the refresh rate does not appear at all. So the driver is blocking those pixel clocks if added through CRU, but using the NVCP works. (I would prefer NOT using the NVCP because I don't trust Nvidia).

However VT 1485 (356.58 MHz) DOES appear if I add this in your CRU, but this VT is not accepted properly by the monitor (I see the windows screen but an "Out of range" error pops up anyway). I have to use 1497-1502 which clearly isn't working

This seems to definitely be a driver issue with the laptop GPUs.

I guess this explains why "Advanced adapter properties" does not even load when I have it selected to use the Benq screen.

I have NOT tried an older driver like 378.xx. I suspect it won't work and it took me a long time to test this for you.

How do I report this to Nvidia? Will they even care? I'm 100% certain a monitor capable of HBR2 without reverting to 6 BPC at >359-480 MHz would work perfectly (since no one would care if 6 bpc is missing due to a bug anyway), so this is a pretty huge bug. I have no idea if this happens on desktop 10xx series gpus as I don't have one

1. NVIDIA might not support 6-bit color. I don't have a new enough card with DisplayPort to check. Even if it did, NVIDIA does not dither like AMD, so gradients would have banding.

2. Is 2001 the lowest you can go? What might be happening is 2001 is rounding up to 2002 and causing it to pass the 360 MHz limit. I recommend multiples of 8 for horizontal values. You should be able to get 2000 or lower by reducing the horizontal front porch and/or sync width if the monitor can handle that.

3. Maybe there's a 400 vertical back porch limit with EDID overrides. Try raising the vertical front porch and/or sync width to reach 1500.

(04-22-2017 09:54 PM)LinCeX Wrote: [ -> ]R9 270x . Tested with latest drivers .

I've seen people report this with the 17.4.x drivers even without CRU. Try running this: reload.zip

If the resolutions appear after running reload.exe, then it's definitely a driver bug. The last driver without resolution issues is 17.1.2: http://support.amd.com/en-us/kb-articles...Notes.aspx

(04-23-2017 10:42 PM)ToastyX Wrote: [ -> ]. 1. NVIDIA might not support 6-bit color. I don't have a new enough card with DisplayPort to check. Even if it did, NVIDIA does not dither like AMD, so gradients would have banding.

2. Is 2001 the lowest you can go? What might be happening is 2001 is rounding up to 2002 and causing it to pass the 360 MHz limit. I recommend multiples of 8 for horizontal values. You should be able to get 2000 or lower by reducing the horizontal front porch and/or sync width if the monitor can handle that.

3. Maybe there's a 400 vertical back porch limit with EDID overrides. Try raising the vertical front porch and/or sync width to reach 1500.

Thank you for the reply, ToastyX,

The hardware itself supports 6 and 8 bit color as I've seen many other people with Nvidia cards and Benq monitors talk about the 6 and 8 bit options in the past, with screenshots. I've just never seen them talk about it on *laptops.*; very few people care about using desktop monitors on laptops and even fewer care about custom resolutions, with issues like Optimus not being able to be disabled in the past.


My laptop's monitor has a 6 and 8 bit selector on it. Just the Benq gets locked to 8 bit. This absolutely does not happen to Desktop Nvidia GPU's.

Optimus is disabled. The MSI GT72VR has optimus and the intel iGPU disabled so that Gsync will always be available. Optimus can not be enabled on this series.

I know there's a bug happening (I talked to Chief blur buster about it before contacting you), because when the Benq is plugged into mini displayport on the laptop, the "Advanced adapter settings" page will NOT load at all.

Your CRU editor will not allow me to go below 2001 on the horizontal total. It turns red. It's not that big of a deal because I can use the NVCP to create a custom 120hz resolution with 2001 HT, 1498 VT but other Nvidia users told me that there are some issues they had when doing VT tweaks on the NVCP that went away when using your excellent editor.

*EDIT*
Just hooked the Benq through HDMI.
Same problem.
With BOTH displayport AND HDMI:
the "windows display properties resolution box" was GREYED OUT And unselectable. And you could sometimes cause some sort of exception error by toggling monitors.

And advanced adapter settings (or advanced display settings, whatever it was called) didn't open. Only way to change resolution was with the NVCP.

On the old laptop (with a GeForce 570m), with the Benq hooked up, everything worked perfectly. 6/8 bit depth, resolution switching, advanced properties, etc.. But that was also in windows 7.

so I don't know if this is a windows 10 issue or a Nvidia driver issue, because this should definitely not be happening.

(04-24-2017 05:05 AM)falkentyne Wrote: [ -> ]Your CRU editor will not allow me to go below 2001 on the horizontal total. It turns red.

Like I said, you need to reduce the horizontal front porch and/or sync width. The total includes the active pixels + front porch + sync width + back porch.

Any clue why this resolution isn't showing up for my Radeon 380x? It's interlaced, but total pixel count on horizontal axis is under 4,000, and vertical under 2,000, so I thought it should work

http://imgur.com/a/O3EfN

Just in case, I'd like to report that 17.4.1 had fixed the EDID overrides on startup bug for me and my R9 290. I was happy again.

17.4.2 (Creators Update / now WDDM2.2 driver) broke it again... and it's still broken on 17.4.3 WHQL.


Dammit AMD. Back to 17.4.1 for the time being, and reported it again through their problem reporting link.