- This topic has 9 replies, 2 voices, and was last updated 6 years, 10 months ago by
.
-
Posts
-
Hi,
First of all, thank you for this great software, it’s a big step over the bundled software included with my Color Munki Smile.
I own an Asus Pb278q, an 8bit panel (no fcr), with 1000:1 contrast ratio. Setting my monitor to its default settings, and using the “report on uncalibrated display device”, I get the following results.
The panels osd default settings in “user mode”
Contrast : 80
Color temp : 6500k
Gamma 2.2
Uncalibrated report :
00:10:44,043 uncalibrated response:
00:10:44,045 Black level = 0.2883 cd/m^2
00:10:44,045 50% level = 58.92 cd/m^2
00:10:44,045 White level = 240.83 cd/m^2
00:10:44,045 Aprox. gamma = 2.03
00:10:44,046 Contrast ratio = 835:1
00:10:44,046 White chromaticity coordinates 0.2999, 0.2924
00:10:44,046 White Correlated Color Temperature = 7938K, DE 2K to locus = 11.3
00:10:44,046 White Correlated Daylight Temperature = 7966K, DE 2K to locus = 13.9
00:10:44,046 White Visual Color Temperature = 9349K, DE 2K to locus = 10.5
00:10:44,046 White Visual Daylight Temperature = 9999K, DE 2K to locus = 13.0
00:10:44,046 Effective Video LUT entry depth seems to be 8 bits
00:10:44,048 Black drift was 0.000000 DE
00:10:44,048 White drift was 0.000000 DE
00:10:44,048 The instrument can be removed from the screen.
00:10:44,164 DisplayCAL: Reached EOF (OK)I’m aiming for a target white of 6500k.
Even though the display is set to 6500k via osd, DisplayCal is reporting a much higher value. However, when I set the monitor, via osd, to 5500k, the reported Correlated Color Temperature is much much closer to 6500k.
I’m guessing the 6500k setting on the monitor is bugged somewhat if the 5500k setting is closer to 6500k.
When calibrating, the osd 5500k setting requires less osd rgb value changes than osd 6500k, and also produces less banding than 6500k.
Would you recommend i stick with osd 5500k and a target of 6500k in DisplayCal, as they are closely matched.
Thanks for your time.
I apologize for so many edits, I’m trying to word it the best way so it makes sense.
Something else which is bugging me.. With my monitor set to its default gamma of 2.2 via osd, uncalibrated report shows the gamma as 2.03, which is way off. When i calibrate to 2.2 gamma I’m seeing quite a bit of gradient banding in videos / video games. Would it be best to aim for 2.03 gamma instead of 2.2 to minimize banding? Would it affect color quality much?
Hi,
I’m guessing the 6500k setting on the monitor is bugged somewhat if the 5500k setting is closer to 6500k.
It depends what you trust more: The monitor preset or the measurement instrument. Unless you aim to do some form of softproofing or want color accuracy on a professional level for other reasons, you don’t need to concern yourself with the whitepoint too much though.
Would it be best to aim for 2.03 gamma instead of 2.2 to minimize banding?
You can do that and see if it helps. Outside of color managed applications, you don’t get accurate color anyway.
Would it affect color quality much?
Depends what you define as “color quality”. If you define it as accuracy, then no, the calibration tone curve doesn’t affect it.
Hi again,
Using the interactive display, i can reduce the delta to .02 using osd, and all looks good.
Looking at the below greyscale image (downloaded and opened inside windows, not the browser) the greyscale seems to be quite accurate and the colors are all blending from white to black, vice versa, an ideal grey scale, but with a tiny bit of red, noticeable when i zoom in on the image. I’m also seeing quite a lot of banding.
Finishing the calibration process, (using default settings), I can achieve average srgb coverage of around 95 – 98%. I apply the profile.
I fire up the image above ^, after applying the calibration, I’m seeing much more emphasized pink/green vertical hues at different intervals across the gradient. I tried playing with rgb on my monitor and all it does it shift the hue colors left or right horizontally, so they move place, but I can’t seem to remove them.
Using “reset video card gamma table”, the pink and green hues almost disappear, not visible unless zoomed in.
I’m puzzled now as to whether I should be using the calibration or not. My aim is to have the best possible color reproduction/image quality when gaming/watching videos and when I can see these pink/green hues in the grey skies whilst gaming it bugs me.
As i can already see a very subtle pink hue using the interactive display to calibrate alone, I’m assuming the monitor is mostly to blame, but then adding a profile on top and making it look 10x worst, isn’t desirable
I have tried using the report on uncalibrated option, and set the reported gamma into the target, which doesn’t seem to help at all. Ideally I was to use the calibration file made in DisplayCal, without green, pink hues.
I’m using the monitors default contrast of 80 and default gamma of 2.2, I’m only adjusting osd brightness and rgb settings during the interactive display adjustment, not after.
Any advice would be appreciated. Thanks again!
I picked up a ColorMunki Display, to replace a ColorMunki Smile, I’ve found that the contrast ratio reported is way higher (100+), and I trust the Display more.
I found out whilst reading about the panel my monitor uses, on several forums, the panel (Samsung S2716DG) is known to have severe banding issues, especially after calibrating.
I’ve also found that the “user mode” preset on my monitor introduces banding, where all the other profiles are much more smoother, and only changing the Contrast setting helps smooth out the gradients, but I can not match the smoothness of the other modes. I blame the monitor.
Also adjusting green on the Rgb scale (on screen or via software) introduces all sorts of dark banding in the Grayscale, wheres altering Red and Blue do not introduce banding at all.
Also i tried using the ColorMunki’s own software to calibrate and i’m seeing discoloration after calibrating, so of course DisplayCal wasn’t responsible.
Time to buy another monitor.
Just one question..
The correction file “Spectral: LCD White IPS (WLED AC LG Samsung), is it suitable for an Asus Ips display? Does Ac stands for Ancor communications?
Thanks for your time, btw I donated the other day 😉
Calibrite Display SL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.oops, a typo :
The correction file “Spectral: LCD White IPS (WLED AC LG Samsung), is it suitable for an Asus Ips “white led” display? Does Ac stands for Ancor communications?
I’ve also found that the “user mode” preset on my monitor introduces banding, where all the other profiles are much more smoother, and only changing the Contrast setting helps smooth out the gradients, but I can not match the smoothness of the other modes. I blame the monitor.
Did you test this with linear videoLUT to make sure it’s the monitor, not the graphics card (AMD cards should not be affected due to reliable dithering, while nVidia will often be undithered)?
The correction file “Spectral: LCD White IPS (WLED AC LG Samsung), is it suitable for an Asus Ips “white led” display?
Yes.
Does Ac stands for Ancor communications?
The description comes directly from the original X-Rite EDR file. Only X-Rite will know specifics about abbreviations.
Thank you Florian for replying.
It’s definitely the monitor which is at fault.. I know that Nvidia does not using dithering, so I expect banding. But the issue i’m facing is the banding is very uneven in tone.
I’ve found that adjusting the Green setting, via osd or whilst DisplayCal creates a profile, is causing all sorts of issues with my monitor, adjusting red and blue are fine.
Here are two uncalibrated reports. One is the monitors Standard preset, and the other is the User preset. Set to defaults. Notice their readings are pretty close to each other.
Standard mode preset
12:20:37,881 DisplayCAL: Starting interaction with subprocess
12:20:37,882 Setting up the instrument
12:20:38,845 Product Name: Colormunki Display
12:20:38,845 Serial Number: CM-16.B-02.191458.11
12:20:38,845 Firmware Version: v2.28
12:20:38,846 Firmware Date: 29Jan14
12:20:39,161 Place instrument on test window.
12:20:39,161 DisplayCAL: Waiting for send buffer
12:20:40,671 DisplayCAL: Skipping place instrument on screen message…
12:20:40,720 DisplayCAL: Sending buffer: ‘ ‘
12:20:40,721 Hit Esc or Q to give up, any other key to continue:
12:20:49,345 Patch 3 of 3
12:21:11,275 Patch 17 of 17
12:21:11,276 Uncalibrated response:
12:21:11,276 Black level = 0.1716 cd/m^2
12:21:11,276 50% level = 36.86 cd/m^2
12:21:11,276 White level = 164.82 cd/m^2
12:21:11,276 Aprox. gamma = 2.16
12:21:11,276 Contrast ratio = 961:1
12:21:11,276 White chromaticity coordinates 0.3029, 0.2927
12:21:11,276 White Correlated Color Temperature = 7680K, DE 2K to locus = 12.6
12:21:11,278 White Correlated Daylight Temperature = 7710K, DE 2K to locus = 15.0
12:21:11,278 White Visual Color Temperature = 9280K, DE 2K to locus = 11.9
12:21:11,278 White Visual Daylight Temperature = 9940K, DE 2K to locus = 14.1
12:21:11,278 Effective Video LUT entry depth seems to be 8 bits
12:21:11,279 The instrument can be removed from the screen.
12:21:11,496 DisplayCAL: Reached EOF (OK)User mode preset
12:18:33,614 DisplayCAL: Starting interaction with subprocess
12:18:33,615 Setting up the instrument
12:18:34,575 Product Name: Colormunki Display
12:18:34,575 Serial Number: CM-16.B-02.191458.11
12:18:34,575 Firmware Version: v2.28
12:18:34,576 Firmware Date: 29Jan14
12:18:34,904 Place instrument on test window.
12:18:34,905 DisplayCAL: Waiting for send buffer
12:18:36,404 DisplayCAL: Skipping place instrument on screen message…
12:18:36,411 DisplayCAL: Sending buffer: ‘ ‘
12:18:36,413 Hit Esc or Q to give up, any other key to continue:
12:18:45,039 Patch 3 of 3
12:19:06,970 Patch 17 of 17
12:19:06,970 Uncalibrated response:
12:19:06,970 Black level = 0.1719 cd/m^2
12:19:06,970 50% level = 36.29 cd/m^2
12:19:06,970 White level = 164.82 cd/m^2
12:19:06,970 Aprox. gamma = 2.18
12:19:06,970 Contrast ratio = 959:1
12:19:06,973 White chromaticity coordinates 0.3028, 0.2925
12:19:06,973 White Correlated Color Temperature = 7688K, DE 2K to locus = 12.7
12:19:06,973 White Correlated Daylight Temperature = 7718K, DE 2K to locus = 15.1
12:19:06,974 White Visual Color Temperature = 9300K, DE 2K to locus = 11.9
12:19:06,974 White Visual Daylight Temperature = 9964K, DE 2K to locus = 14.2
12:19:06,974 Effective Video LUT entry depth seems to be 8 bits
12:19:06,974 The instrument can be removed from the screen.
12:19:07,142 DisplayCAL: Reached EOF (OK)
12:19:07,177Switching between each preset on the monitor, there is no immediate visible tone difference, and that is reflected in the report ^ (other than gamma).
When I open up a grayscale gradient picture (using windows photo viewer) :
In standard mode, the grayscale is intact, each band is the same distance from the other, the gradient is smooth and looks perfect, no color cast at all. Just as it should look.
In user mode, the greyscale seems to be quantized, and each gradient is twice as big, less bands, less steps of grey in fact it looks like there are half as many. Like it has lost half as many colors.
I found that adjusting the monitors green offset by a value of 1 from default. 80 -> 79 .. or 80 -> 81.. minus or plus removes the quantization.
It seems that the user mode, has had its green offset adjusted somewhat (via error), that has caused the quantization to happen using default “user mode” settings.Again, adjusting the red/blue on screen does not alter the banding in the grayscale. Only adjusting green does.
When DisplayCal creates a profile, I’m assuming when Green value is adjusted it is causing the grayscale to look blotchy, with green/red cast to some of the bands. So even if I leave the green gain alone on the screen, it gets adjusted via software and is causing a very uneven gradient with color cast.
I’ve just seen a very bright green line (artifact) appear on my screen, when adjusting the green offset to -1 +1 from the default of 80.
It seems to happen at the same place each time.I don’t expect you to diagnose what’s wrong with my monitor 😉 Time for an upgrade i think.
Thanks again for your help.
