Reverse engineering ICC colour profile
October 26, 2018 12:57 AM Subscribe
If I have a set of hex, RGB and CMYK colours, can I find out which ICC colour profile they fit into?
For example, in Photoshop, one colour profile gives the CMYK correspondence for the Hex code #bd0017 as c17/100/100/9, but a different colour profile will give it as C27/M100/Y100/K1.
It's easy to change the colour profile and get new colour settings - but how about in reverse - what if I don't know the profile used and only have a set of colours? Is there somewhere I can enter a set of RGB, CMYK and Hex codes for the same colour and have it spit out the colour profile used to generate them?
For example, in Photoshop, one colour profile gives the CMYK correspondence for the Hex code #bd0017 as c17/100/100/9, but a different colour profile will give it as C27/M100/Y100/K1.
It's easy to change the colour profile and get new colour settings - but how about in reverse - what if I don't know the profile used and only have a set of colours? Is there somewhere I can enter a set of RGB, CMYK and Hex codes for the same colour and have it spit out the colour profile used to generate them?
Colorimetry is such a deep rabbit hole that I'm not sure where to start with this.
The short answer is "not really".
There is no real direct answer to "what's the CMYK equivalent of this RGB colour". It depends on far more than a single profile. RGB is used to describe colours formed from three colours of transmitted light, CMYK to describe colours formed by inks on a reflective surface.
When you ask Photoshop for the CMYK equivalent of an RGB value in some given colour space, you're asking it "hey, if I wanted to print this out with CYMK inks on a piece of white paper under some given lighting conditions, what combination of ink values would give the closest perceptual fit to what I'm seeing on this screen for that RGB colour?". That's a complicated question, and the answers are going to depend on exactly how that particular bit of software has been written to address it, and what assumptions and configuration settings it's using to do so, including exactly what inks, what paper, what lighting, and what approximation of human perception it's using for comparison.
And Photoshop is a professional tool that handles colorimetry about as well as anything, so you can bet that the way it does it is pretty complicated and won't be exactly duplicated by any other tool without a lot of careful setup.
One thing I will say is that the two examples you've given are already probably pretty poor matches, because they both involve taking M and Y to 100%, which means they're maxed out. That suggests that the results are as close as it's able to get but that an actual match is out of reach of the output profile.
posted by automatronic at 1:46 PM on October 27, 2018
The short answer is "not really".
There is no real direct answer to "what's the CMYK equivalent of this RGB colour". It depends on far more than a single profile. RGB is used to describe colours formed from three colours of transmitted light, CMYK to describe colours formed by inks on a reflective surface.
When you ask Photoshop for the CMYK equivalent of an RGB value in some given colour space, you're asking it "hey, if I wanted to print this out with CYMK inks on a piece of white paper under some given lighting conditions, what combination of ink values would give the closest perceptual fit to what I'm seeing on this screen for that RGB colour?". That's a complicated question, and the answers are going to depend on exactly how that particular bit of software has been written to address it, and what assumptions and configuration settings it's using to do so, including exactly what inks, what paper, what lighting, and what approximation of human perception it's using for comparison.
And Photoshop is a professional tool that handles colorimetry about as well as anything, so you can bet that the way it does it is pretty complicated and won't be exactly duplicated by any other tool without a lot of careful setup.
One thing I will say is that the two examples you've given are already probably pretty poor matches, because they both involve taking M and Y to 100%, which means they're maxed out. That suggests that the results are as close as it's able to get but that an actual match is out of reach of the output profile.
posted by automatronic at 1:46 PM on October 27, 2018
« Older How do I prevent condensation on my windows?... | tfw your sounding board is the topic of your askme Newer »
This thread is closed to new comments.
posted by london explorer girl at 1:28 AM on October 26, 2018 [2 favorites]