- This topic has 16 replies, 4 voices, and was last updated 1 year, 3 months ago by
.
-
Posts
-
Hey guys,
I have an Eizo CG2700x which is hooked up to a decklink 4k mini monitor, connected to my iMac GUI. I am using an i3d display probe, CN7 and DisplayCal to the callibration to REC709 Gamma 2.4. I have been following this guide on youtube: https://www.youtube.com/watch?v=KQyuJ-EL3DA
Brief overview of the process:
In CN7, I select the CAL SLOT and set the following target settings:
- Brightness: 100 cd/m^2,
- Black level: minimum
- White point: D65
- Gamma 2.4
- Priority: Standard
- Gamut: Native
Compensation method is set to color managed and I have checked the box ‘Do not apply an ICC profile’.
Then I go into DisplayCAL, and set the following settings for Video 3D LUT for Resolve (D65, Rec. 709 / Rec. 1886): See images, screenshots are from the video
Afterwards, I load the 3D LUT into the monitor via CN7, and do the verification. I am getting some weird / off results in my measurement reports, I added a screenshot of the results. I have done the exact same thing a month ago, and results seemed great. Feels like I am missing something?
I did notice that when I got the start measurement pop-up before callibration with displaycal, my RGB values were not within target. Am I supposed to do a manual adjustment of RGB and brightness in CN7 before proceeding with callibration in displaycal?
No. Color Navigator 7 in its preferences sets the white point. It is in the training video. You use it to check the monitor is set and working from screens profile white point.
Hey Ben, thank you for the reply. I did a new test: I did the callibration with CN7, continued with displaycal and when the RGB measurement pops up when profiling and calibrating, it says that the RGB values are not within target. I opened up CN7 again and did a manual adjustment for the RGB values, went back into displaycal and only then it shows me that the RGB values are within target. After that, I proceeded the profiling with displaycal, created the 3D LUT and made a new measurement report. In this report, everything seems solid, except some light variance of delta e 1.5 for blue. I will post the new report once I’m home, but basically I shouldn’t be doing the manual adjustment even though it gives me a good report?
Compensation method is set to color managed and I have checked the box ‘Do not apply an ICC profile’.
I did notice that when I got the start measurement pop-up before callibration with displaycal, my RGB values were not within target. Am I supposed to do a manual adjustment of RGB and brightness in CN7 before proceeding with callibration in displaycal?
Compensation method:
color managed :
on CG-X using built-in colorimeter, use the factory matrix correction made for them
on i1d3 use the generic matrix correction for a randomi1d3 on eizo lab and a random unit of each modelNo compensation:
on an i1d3 use RG_phosphor EDR (GB-LED) which is not a suitable EDR for all these new WLED PFS phosphor displays like newer CG or CS, but t is all what Eizo is giving you.Hence… it will be a missmatch between WP measured by CN ****using that configuration*** and DIsplayCAL using RG_phosphor EDR.
You can switch CN7 to “no compensation” so DisplayCAL and CN7 are using the same “wrong” correction
Or you can switch CN7 to “no compensation” replace CN7 C:\ProgramData\EIZO\Sensor\i1DisplayProEDR\RG_Phosphor_Family_25Jul12.edr (or equivalent macOS folder, IDNK right now, use “SuspiciousPackage” app for macOS on Eizo PKG instaler to locate actual path) for Stuart Pointon pack of EDR for these WLED PFS displays(CG319X and CS2731 as family representatives) or use some equivalent EDR from community. Then on DisplayCAL use CCSS translation of these EDR.Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.Afterwards, I load the 3D LUT into the monitor via CN7, and do the verification. I am getting some weird / off results in my measurement reports, I added a screenshot of the results. I have done the exact same thing a month ago, and results seemed great. Feels like I am missing something?
Regarding errors as a general rule for any display and manufacturer when you spot something weird:
1-Check if HW cal and DisplayCAL use same set of corrections if you want to preserve HW cal white or corret both to a better match. Do this just after CN7 native gamut D65 calibration.
2-Check (without LUT uploading) if custom ICC “3d mesh” profile (XYZLUT) matches display behavior, so you did not change display bahevior after profiling.
3-Check (without LUT uploading) if computed “ICC equivalent” to LUT3D, a “device link icc” (the bigger ICC file generated at the same time as the LUT3D file) is able to transform from source to destination.
4-LUT uploading and verification.IDNK where is your issue, we know that step 1 is wrong on your setup and errors are cumulative.
Also do not use screenshots with report. Attach full HTML. Some HDMI level mismatch can be spotted by under/over saturation of latest CIE a*b* plot. Gamma ~2.2 may point to this, that at some point on this pipeline something did not match but I cannot know where.
I’d redo all steps using EDR replacement with an actual WLED PFS phosphor EDR, verifying on each step.
Hey Vincent, thank you so much for the advise. I am very new at this, so bare with me as I try to understand everything correctly;
So using i1d3 probe is a bit pointless because RG_phosphor EDR is not suitable for any newer Eizo model with a WLED PFS panel. I need to create a 3D LUT with displaycal for the work I will have to do though. So my best bet is to set compensation to none, replace the RG_Phosphor_Family_25Jul12.edr file with Stuart’s EDR made for the CG319X (I found the file RG_Phosphor_Family_25Jul12_CG319X.edr on google drive).
The only part I don’t understand is “Then on DisplayCAL use CCSS translation of these EDR”, can you describe this process in simple what do I have to click/select terms? Again, apologies for the newbie questions Vincent, I appreciate your help tremendously! Thank you for the tips about the report too!
Hey Vincent, thank you so much for the advise. I am very new at this, so bare with me as I try to understand everything correctly;
So using i1d3 probe is a bit pointless because RG_phosphor EDR is not suitable for any newer Eizo model with a WLED PFS panel.
No. That’s false. You can make i1d3 measure properly an it has been explanied how to do so.
I need to create a 3D LUT with displaycal for the work I will have to do though. So my best bet is to set compensation to none, replace the RG_Phosphor_Family_25Jul12.edr file with Stuart’s EDR made for the CG319X (I found the file RG_Phosphor_Family_25Jul12_CG319X.edr on google drive).
Yes
The only part I don’t understand is “Then on DisplayCAL use CCSS translation of these EDR”, can you describe this process in simple what do I have to click/select terms? Again, apologies for the newbie questions Vincent, I appreciate your help tremendously! Thank you for the tips about the report too!
Next to that EDR should be a CCSS file with the same data, but as text. EDR = binary spectral distribution of a sample display . CCSS = text data with spectral power distribution of a sample display.
EDR -> CCSS use “oeminst” app from ArgyllCMS
CCSS -> EDR a python tool called ccss2 edr.Okay Vincent, thank you once again.
I have replaced the RG_Phosphor_Family_25Jul12.edr with Stuart’s EDR for the CG319x, callibrated in CN7 with compensation off. I am having issues converting the EDR to CCSS though. As an alternative solution, I used HP_DreamColor_Z24x_NewPanel.ccss to callibrate in displaycal. Results look a lot better now, will post measurement report when I’m home tonight. My python knowledge is unfortunately very limited, I managed to convert EDR to CCSS with a custom script ChatGPT built for me, but DisplayCal won’t let me import the CCSS and says it’s faulty, thus I will try with again with oeminst tonight.
Oeminst just keeps giving me the same message, no matter what option I give in. Paths are verified and correct, just keeps giving me this:
Install OEM data files, Version 3.1.0
Author: Graeme W. Gill, licensed under the GPL Version 2 or later
usage: oeminst [-options] [infile(s)]
-v [level] Verbose
-n Don’t install, show where files would be installed
-c Don’t install, save files to current directory
-S d Specify the install scope u = user (def.), l = local system]
infile Manufacturers setup.exe install file(s) or .dll(s) containing install files
infile.[edr|ccss|ccmx] EDR file(s) to translate and install or CCSS or CCMX files to install
If no file is provided, oeminst will look for the install CD.
Oeminst just keeps giving me the same message, no matter what option I give in. Paths are verified and correct, just keeps giving me this:
Install OEM data files, Version 3.1.0
Author: Graeme W. Gill, licensed under the GPL Version 2 or later
usage: oeminst [-options] [infile(s)]
-v [level] Verbose
-n Don’t install, show where files would be installed
-c Don’t install, save files to current directory
-S d Specify the install scope u = user (def.), l = local system]
infile Manufacturers setup.exe install file(s) or .dll(s) containing install files
infile.[edr|ccss|ccmx] EDR file(s) to translate and install or CCSS or CCMX files to install
If no file is provided, oeminst will look for the install CD.
Because you provide no data as parameter.
Example: oeminst.exe C:\share\edr\RG_Phosphor_Family_25Jul12_CG319X.edr
Instead of using this path, put the file path where you stored CG319X EDR
The measurement report is from the following workflow:
Replaced RG_Phosphor_Family_25Jul12.edr file in EIZO probe path with RG_Phosphor_Family_25Jul12_CG319X.edr. In CN7, I select the CAL SLOT and set the following target settings:
Brightness: 100 cd/m^2,
Black level: minimum
White point: D65
Gamma 2.4
Priority: Standard
Gamut: Native
Compensation method is set to NO COMPENSATION and I have checked the box ‘Do not apply an ICC profile’.Then in DisplayCAL, I used Stuart’s settings here https://www.youtube.com/watch?v=KQyuJ-EL3DA but for correction I selected HP_DreamColor_Z24x_NewPanel.ccss, because I have not been able to convert RG_Phosphor_Family_25Jul12_CG319X.edr to ccss yet using oeminst..
Oeminst just keeps giving me the same message, no matter what option I give in. Paths are verified and correct, just keeps giving me this:
Install OEM data files, Version 3.1.0
Author: Graeme W. Gill, licensed under the GPL Version 2 or later
usage: oeminst [-options] [infile(s)]
-v [level] Verbose
-n Don’t install, show where files would be installed
-c Don’t install, save files to current directory
-S d Specify the install scope u = user (def.), l = local system]
infile Manufacturers setup.exe install file(s) or .dll(s) containing install files
infile.[edr|ccss|ccmx] EDR file(s) to translate and install or CCSS or CCMX files to install
If no file is provided, oeminst will look for the install CD.
Because you provide no data as parameter.
Example: oeminst.exe C:\share\edr\RG_Phosphor_Family_25Jul12_CG319X.edr
Instead of using this path, put the file path where you stored CG319X EDR
I put the following command (paths are correctly copied, cd is where oeminst is located and I stored the EDR file in ArgyllCMS folder) 🙁
cd “/Users/Lizzie/Library/Application Support/DisplayCAL/dl/Argyll_V3.1.0/bin/”
oeminst -n -i “/Library/ArgyllCMS/RG_Phosphor_Family_25Jul12_CG319X.edr”
It gives me the exact same response
There is no “-i” param. I tested on macOS and it denies execution on 1st run because unknown developer. Enable it on security, run again and macOS will ask you if want to run.
If will output something like:
./oeminst -n /Users/XXX/Desktop/compartido/share/edr/RG_Phosphor_Family_25Jul12_CG319X.edr
Would install ‘/Users/XXXX/Library/Application Support/ArgyllCMS/RG_Phosphor_Family_25Jul12_CG319X.ccss’
Would install ‘/Users/XXXX/Library/Application Support/ArgyllCMS/CRT.ccss’Since you choose “-n”, which means “do not install”, just show where CCSS would be placed. Run without params, just the EDR file to install it in /Users/XXXX/Library/Application Support/ArgyllCMS/
