- This topic has 5 replies, 4 voices, and was last updated 6 years ago by
.
-
Posts
-
I have a problem with difference in colour of images between PS CC 2018 and image viewers. I switched to using Spyder 5 and DisplaCal earlier this year and am not sure whether that or Win10/PSCC updates triggered the problem. I wondered if anyone had seen anything similar or could suggest what to try next.
Using Win 10 with Nvidia GT 730 and Dell 2407WFPHC monitor. Software PS CC 2018, Gimp, Firestone Viewer, Breeze Browser and XNViewMP. All including the GT 730 are up to date, CM is enabled as necessary. PS workspace is sRGB.
I usually use PS CC and BreezeBrowser and added the other applications for comparison.
sRGB .jpg images straight from the camera and opened in PS and Gimp look the same. The same image in the viewers is significantly different and all viewers show the same difference.
I have the ICC CM test images, all applications show a pass.
The profile in use is a 1Curve+MTX, correction for the Spyder 5 is “auto(none)” and the monitor identified as “Wide Gamut LCD(RGB LED). Monitor colour temp native is approx. 4950K, set to 5200K with monitor controls for the profile.
The following compares PSCC and Faststone, saved as an untagged snip. Gimp to XnViewMP looks the same.
I would appreciate any suggestions
2407WFPHC seems to be a wide gamut CCFL, not LED. Maybe you want to try that correction if avaliable in Spyder5 corrections.
Since it’s a widegamut (and unless you use some sRGB emulation preset), images will look very different in editors with color management than in editors without color management.
PS and GIMP use color management so they render your images as intented. Other editors won’t color manage your image or maybe they need to be configured to do so.***
PS workspace default RGB profile is used to tell PS what color profile should apply when user opens an image without embeded profile.
A lot of JPG images from the internet do not have an embeded profile so sRGB as default colorspace seems a good choice.The following compares PSCC and Faststone, saved as an untagged snip. Gimp to XnViewMP looks the same.
FastStone isn’t using the monitor profile. This can be determined with the “color cast” test profiles from the Color Management Test page.
Shouldn’t PS be configured to use the display profile as well, instead of sRGB?
Photoshop will automatically use the display profile, and it can’t be configured.
Thanks for the suggestions.
As I understand, setting the working space allows PS to trap missing or different image profiles. By having a defined image profile, the monitor profile “knows” the starting point when applying adjustments to suit the monitor.
You are correct Vincent, it is wide gamut and I had selected the wrong monitor type. Thanks, changed to Wide Gamut LCD(CCFL). I did load Spyder 5 corrections when installing DisplayCal, there are two mentions for Dell. I have left it on auto(none). I also suspect that they will have a small effect on final monitor accuracy when what I am seeing is substantial difference between editors (PS and Gimp) and viewers (Faststone and XnViewMP).
Florien, I had tested PS, Gimp, Faststone and XnViewMP with the .jpg versions of the ICC Test files. All showed the correct display. I rarely open anything other than .jpg camera images which are tagged sRGB. Should I try a different Profile type?
The only other thing I notice is that if I use the assign profile tool in PS and view the result by selecting Adobe RGB, the PS image looks as it does in the viewers. I believe that the viewer image is correct.
I have run DISM and SFC, no issues were found. PS CC has been removed using Adobe Clean CC Removal tool and reinstalled.
Any suggestions of where else to look will be appreciated.
Florien, I had tested PS, Gimp, Faststone and XnViewMP with the .jpg versions of the ICC Test files. All showed the correct display.
Did you install (and set as default) any of the color cast test profiles to your monitor? The test is only correct if it shows the right rendering and doesn’t have a color cast when one of the test profiles is assigned to the monitor (which contains calibration that will introduce a strong color cast to the whole monitor that is only cancelled in color managed applications). As I said, FastStone fails this (color cast on the gray gradient, so not using the monitor profile).
