21
Replies
Thanks, @Bob_Hallam, but maybe we have a misunderstanding here.
I don't plan to be a photographer and to work with RAW files. My question is not related to that.
I simply have sRGB images that I have to work with. I'm making books. I'm asking are my color settings correct (currently they are sRGB, which I find logical). Some people say it should be Adobe RGB, which I find not so logical, at least not for my case.
Depending on the CMYK color gamut of the printers (and we can examine that), you ARE clipping colors using sRGB that you would not if you used Adobe RGB (1998). It is that simple.
Book printing is an important clue. What is the printer's profile for these books? That will help me determine which RGB profile will be most efficient.
The settings do not determine any image's appearance unless you have the settings ignoring embedded profiles (incorrect for a color-managed workflow). So long as you have "Preserve embedded profiles" checked, and all of your files include embedded profiles, the system will use what's in the files and display them properly. What those settings will do is provide a standard way to change between color spaces, and are what's used as default standard profiles when new documents are created or converted.
By @Bob_Hallam
Thank you, @Bob_Hallam , but what do you mean by "determining a color profile"?
Maybe you meant converting from sRGB to another (wider gamut) RGB? Or just assigning another RGB profile to the sRGB content?
If you meant assigning a wrong profile, then this trick is a bit "dirty" and @TheDigitalDog seems to disapprove it. Opinions seem to be divided.
You also adviced me to honor the embedded profiles in Indesign. Its already set that way, the profiles are not ignored. But if I set Indesign to Adobe RGB, the embedded (sRGB) profiles will override that. AFAIK, they'll take precedence over the app's color settings (cause the embedded profiles are not ignored). So what would be the benefit of setting Indesign to Adobe RGB then?
The printer (book or otherwise) should have an ICC output profile that defines the output conditions. This is also used to convert from (Working Space) RGB to CMYK. If you can find out what is used, then it can be plotted to show the color gamut of the ouput device compared to the RGB Working Space. There are litterly thousands of possible CMYK color spaces! The profile deciphers this for us in many ways. For example, here is a book publishing company using an Indigo digital press and a comparison to sRGB in 3d (so only part of the entire story):
The overall larger red color gamut is sRGB. But you can see that outside that gamut (greens/teels as expected, sRGB shortcomings) and yellows, the press has a WIDER color gamut! So if you used sRGB to this press, and you had saturated greens and yellows, they would clip to that red shape of sRGB.
Now I plot Adobe RGB (1998) over the same press: you can see that a small portion of greens/blues fall outside even Adobe RGB (1998) but far less so. I've altered the 'drawing' of Adobe RGB (1998) as a wireframe so it is hopefully easier to see this small area of blues/teels that indeed fall outside its gamut:
IF I had a profile from your printer, and if you wished, actual images you print, I could plot similar gamut maps.
What you see above is clear: Even Adobe RGB (1998) isn't wide enough gamut for this kind of printing press!
Of course, rotating this 3D map in a video shows more, but hopefully you get the point about the size of RGB Working Space and that of output devices.
The printer (book or otherwise) should have an ICC output profile that defines the output conditions. This is also used to convert from (Working Space) RGB to CMYK. If you can find out what is used, then it can be plotted to show the color gamut of the ouput device compared to the RGB Working Space. There are litterly thousands of possible CMYK color spaces! The profile deciphers this for us in many ways. For example, here is a book publishing company using an Indigo digital press and a comparison to sRGB in 3d (so only part of the entire story):
The overall larger red color gamut is sRGB. But you can see that outside that gamut (greens/teels as expected, sRGB shortcomings) and yellows, the press has a WIDER color gamut! So if you used sRGB to this press, and you had saturated greens and yellows, they would clip to that red shape of sRGB.
Now I plot Adobe RGB (1998) over the same press: you can see that a small portion of greens/blues fall outside even Adobe RGB (1998) but far less so. I've altered the 'drawing' of Adobe RGB (1998) as a wireframe so it is hopefully easier to see this small area of blues/teels that indeed fall outside its gamut:
IF I had a profile from your printer, and if you wished, actual images you print, I could plot similar gamut maps.
What you see above is clear: Even Adobe RGB (1998) isn't wide enough gamut for this kind of printing press!
Of course, rotating this 3D map in a video shows more, but hopefully you get the point about the size of RGB Working Space and that of output devices.
By @TheDigitalDog
@TheDigitalDog, I just replied to one of your posts and I saw this one a bit later. Thank you for these 'sci-fi' diagrams (or how should I call them). Its very interesting to hear about the wide gamut of the press and the shortcomings of sRGB.
But what options do I have to get a more vivid print? I only have sRGB content. I'm afraid that the solution could be assigning of a different (wide gamut) RGB profile to my sRGB content (basically "lying"), which you might find controversial.
My workflow is RGB and the result will be converted to CMYK in the end, so I still don't know the printer's profile.
Don't be surprised to hear that some cheaper printeries in my area are not profiled at all (such still exist). But if they are profiled, it will probably be something like PSO Uncoated (FOGRA52) and PSO Coated, which is common in this area.
PSO Coated vs sRGB. Similar story.
Uncoated fits fully within sRGB (thankfully) 😋
PSO Coated vs sRGB. Similar story.
Uncoated fits fully within sRGB (thankfully) 😋
By @TheDigitalDog
Very interesting, thanks, @TheDigitalDog. Btw PSO stands for Process Standard Offset, so this will go to an offset press.
And thank you very much for all your detailed explanations and advices. To be honest, I don't know which answer to mark (click) as 'correct'.
Wider gamuts are for wide gamut printing processes.
By @Bob_Hallam
And the hundreds of millions of wide gamut displays 🤔
Thank you for the reply and the video, I appreciatte it. I noticed these things in the video:
- At 18:07, the video says: "Assign profile command - using it incorrectly shows incorrect preview of the numbers".
That's exactly what I'm talking about from the very beginning - one should not view the content through a wrong "prism". And these "prisms" obviously exist.
Correct! There is a prism because you are using color management. You've got it (the answer to the first question).
As the video confirms, asigning a wrong profile is basically lying that X is Y. Its like putting a label "banana" to an apple.
Correct again! Assigning the wrong profile to the numbers is silly and pointless. Recommending this 'workflow' to someone initially new to the concept is ridiculous and not helpful. Now you fully understand what's happening under the hood. But then I did try to warn you earlier about such advice <g>.
As the video confirms, asigning a wrong profile is basically lying that X is Y. Its like putting a label "banana" to an apple.
Yes, it is a lie. That serves no purpose. You can label an apple a banana, it doesn't alter what the fruit actually is.
Then why should my content (originally encoded as sRGB) be viewed through the Adobe RGB's "prism" - as some people suggest? It makes no sense to me.
That isn't what they are suggesting per se. In a color-managed system, sRGB is being previewed and handled in a sRGB prism just as Adobe RGB (1998) is dealt with and previewed as Adobe RGB (1998) as are all other color spaces in their own prism along with your display profile for previews. If you start with sRGB (from a camera, from a new document in Photoshop), it is what it is. As is Adobe RGB (1998) or any other RGB Working Space. So the question becomes, why even deal with sRGB from the get-go? The answer today is, no real reason! There are hundreds and hundreds of millions of wide gamut displays with color-managed web browsers out in the world today. Every iPhone since version 6, iPads of the same time frame, and others all using wide gamut and color management. You can provide them data in sRGB or Adobe RGB (1998) and all will preview in the prism correctly; they ARE color managed. Now, what do you gain and lose by posting sRGB? The gamut volume of sRGB is 35% (of Lab), and the gamut volume of Adobe RGB (1998) is 50.6%. That gives you a numeric idea of how much more colorful an image may (repeat may) appear in the wider gamut color space and display. If you insist on using sRGB, you clip those colors you can see and reproduce. In a fully color-managed app, it doesn't matter which you use. Both will work, one will appear less saturated. At this point, use whichever you prefer. I prefer to show all the colors I can capture and reproduce. And on the smaller gamut display, it will be 'fine' just less saturated.
Then why should I view my images in a wider gamut (Adobe RGB), when they were encoded in a narrower gamut (sRGB)? Its like forcing the image to show something that is not there. Correct me if I'm wrong.
See above. Again, your choice. You can stick with sRGB with color-managed wide gamut systems. Or Adobe RGB (1998) or anything else.
As to the video and the images, yes; they all came from raw to ProPhoto RGB but could have been encoded into sRGB or Adobe RGB (1998). For my work, Adobe RGB (1998) still clips colors I can capture and output so it's too small. All the nonphotographic elements in my Gamut Test File document you can download are created in Photoshop. You can use the Gamut Test File to test output even to your display. IF you have a wide gamut display, duplicate the ProPhoto RGB version, convert to sRGB and look at the two. One will appear (IMHO) superior than the other. The point is this:
1. You can move directly to sRGB. Fine with me. You'll clip colors in many images. If you work with a color-managed application, all is fine.
2. You can move directly into Adobe RGB (1998) and that too is fine with me. You'll clip less colors by a significant percentage depending on image content. AND you can output those colors correctly with color management.
3. The idea of assigning the wrong profile to the right set of numbers is rather stupid and counter-productive and fails in the fact of sound color management practices. But again, if you do so, fine with me.
4. Without color management, sRGB and ALL color spaces are moot and an unknown attribute. There is no prism. It would be as if I typed all of the above text in a language you didn't understand nor could translate. Without color management sRGB is not a prism. It is a set of RGB values that amount to RGB mystery meat.
AdChoices