We have a brand new look! Take a tour with us and explore the latest updates on Adobe Support Community.
Hey Photoshop color management experts, sorry in advance for the lengthy post...
I have a very specific task to accomplish via Photoshop and it seems I am missing something. I am a hobbyist postage stamp collector and my aim is to scan and print my stamps so that they look exactly like the original material in all aspects (with color accuracy being of paramount importance).
I am using the latest version of Photoshop CC and Windows 10. So far, I have:
- calibrated my display with an X-Rite i1 Studio device
- profiled my Epson scanner with a brand new Monroe 2019.2 IT8 target and created a corresponding ICC v2 profile using the i1 Studio software
- scanned a sample of my material (making sure I get a raw output - as in, without any image manipulation/enhancement/profile embedding by the scanner's driver and software)
- imported this image to Photoshop and assigned it the scanner ICC profile
- selected "Photoshop Manages Colors" at the Color Handling drop-down menu of the Photoshop Print Settings window
- selected the correct Epson printer/paper combination ICC profile in the Printer Profile drop-down menu of the Printer Settings window for the paper I bought and use
- disabled color management at my printer's device driver settings window
- played with both absolute and relative colorimetric intent, both in Photoshop's Color and Printer Settings windows.
The results I get are close to the original material, but lack something that I would best describe as vividness. If I switch from relative to absolute intent in both the Photoshop Color Settings & Printer Settings, the image "brightens" and apears much closer to the original, but when I try to print with the absolute setting, I get the exact same ("dull") result as with relative intent - it's like I never changed it. So, I don't know if that is the setting that needs to change, because it never gets applied (=printed).
Where do you think I go wrong in Photoshop and/or what are my options to get an as accurate as possible printing of my original material? Answers are highly appreciated, as I have read and experimented a lot to make this process work and the results are not as I would expect them...
Thanks all and again sorry for the long post!
Sounds like you correctly during printing changed to Phothsop Manages color, then assigned your printer profile after that became avaiable.
Postage stamps are often printed with spot colors which might be difficult to get the vividness.
How many inks does your printer have?
Thanks for your answer. My printer is Epson SureColor P900 with 10 inks.
What also really puzzles me is why after changing to absolute colorimetric intent in both Photoshop color & printer settings and I see a difference on my screen, I don't see it at my printing output at all. The result on screen with absolute intent is much closer to the real-world colors I see.
Absolute colorimetric differs from relative colorimetric by not remapping the white point. So if absolute gives a better result than relative, something is off. The white point has to be remapped. The scanner light is probably not the same color as your print viewing light.
Instinctively I wouldn't trust a fully automatic process for this, just relying on profiles. There's this thing called metamerism (look it up), and one ink may not behave like another in different types of light. I'd concentrate on the output profiles, and adjust the input manually. That's what you normally do regarding input from a digital camera. The input profile gets you close enough, and then you finish visually with manual adjustments.
Take the monitor out of the equation first. Calibrate it so that it matches printed output, again by adjusting the white point. Set the monitor white point to be a visual match to paper white. Then adjust monitor black to match maximum ink (black). Then at least you know that what you see is what you get, and you can work visually on screen to match the original.
Thank you for the answer.
In my case, the monitor does not really matter, I can really take it completely out of the equation, as the devices that I need to "behave" in this sequence are the scanner and the printer. So, what I care in the end is the output, which I then have the luxury to promptly compare with the original material -- and it's there where I see the differences. As in, the monitor can show anything as long as I get prints out of my printer that are color-accurate compared to the original material.
I was under the impression that what I just mentioned would be the case if I just create precise ICC profiles for the scanner and the printer and use those. But you are suggesting that a fully automatic process based on profiles might not work in this case and I should adjust the input manually - what would you suggest as manual changes in Photoshop to achieve that perfect "photocopy" that I am looking for? Or you are referring to chamges in the monitor or scanner?
To be frank, I am not sure I understood how to (and why) change the monitor white point to match the printed output, because the latter is already not the expected one - or it does not matter? What I can easily do, if it helps at all, is switch my workload to a (mostly) color-accurate, factory-calibrated monitor that I recently purchased.
Please excuse my ignorance and stupid assumptions and questions, color management is a relatively new and difficult space for me.
You're trying to translate from one printing ink to another, and that's always inherently uncertain. There are too many variables here.
I don't have a lot of experience with profiling a scanner, but I do have a lot of experience with photographic art reproduction. There is no way I would ever expect "accurate" color (whatever that is) out of the camera, no matter how good my camera profiles.
Consistency on the output side is a lot easier, and getting a monitor to match printed output almost to perfection is fairly straightforward. It's a matter of white point and black point and a good monitor profile.
So what I'd do is adjust on screen until it visually matches the original. Print. Done.
Now, I realize a scanner is a much more controlled and consistent input than a camera and various light sources. So maybe it's possible, I don't know. I'll leave that to someone who has actually done it.
After reading your first paragraph, I was about to answer with what you wrote in your last 🙂
A scanner has its own light source, and it is not affected by any outside source of lighting since the lid is closed while scanning. This should mean that it always achieves a consistent reproduction of the material being scanned.
My understanding is that correcting its (rather small, to be honest) delta with the use of an IT8 target and the subsequent creation of an ICC profile should make it reproduce an exact copy of the original material - and the same goes for the printer, due to its corresponding printer/paper combination profiling. This is what puzzles me - the whole process is controlled end-to-end, yet the result is not very close to the original material.
I've done a lot of scanner profiling.
Even scanners, despite the inbuilt light source you mention, have issues related to metamerism.
You mention the IT8, I prefer the HCT because the IT8 was not designed for ICC profiling.
In this case you are scanning reflective material.
Consider the inks used to make the scanner characterisation target (in the case of the reflective HCT it's a photoprint). basICColor make a ["Rescan"] target which uses printing inks.
You are scanning a stamp, its not the same ink, of course, so metamerism raises its head.
What does that mean - 2 dissimilar items, which under a certain light match - to the eye, may not produce identical RGB values if scanned. They may not even mastch under different lighting - to the eye. This is related to metamerism.
Shoes and handbag match in the shop, but not in daylight. Hope that makes sense.
Without a known accurate display (set up properly to match a reference printed original such as http://www.colourmanagement.net/products/icc-profile-verification-kit ) you are shooting in the dark.
I think D.Fosse is right that its likely you'll need to make corrections in Photoshop to match the original item. You'll need an accurate screen for that.
He also mentioned that a screen to print system match can be pretty accurate. That’s true.
I hope this helps
neil barstow, colourmanagement net :: adobe forum volunteer
google me "neil barstow colourmanagement" for lots of free articles on colour management
[please only use the blue reply button at the top of the page, this maintains the original thread title and chronological order of posts]
Reading your steps/setup and your description of the results, I believe that I know part, if not all of the issue.
You bring in raw scans, apply the scanner profile and then directly print.
I would suggest that you need to convert to profile from input device/scanner ICC to a working ICC profile such as Adobe RGB. Then perform the following recommendations in the working RGB space.
There has been no tonal range compensation for the white/black points of the capture. I'm guessing that the highlights read something around 230rgb or lower (but not perfectly even/neutral) and the shadows something like 20rgb or higher (but not perfectly even/neutral). Setting the endpoints of the scan to optimise printing is a basic step. The tonal range is likely "highly" compressed and that there may be a very slight colour cast, depending on the differences between the profiling chart and the different paper/inks used in the stamps. You don't want to blow out the highlights or plug the shadows, however, you do need to expand the tonal range so that whites and blacks in the input file translate into appropriate values for print.
You should also look into sharpening, as both the capture and the output are going to be softer than the original.
All of these things were standard in days gone by.
Thanks all, seems a lot of things I need to check and probably a few manual settings that I need to apply/change, something that I was under the impression that I could avoid... Any other ideas/proposals welcome, of course!