I am planning to send some files to the print lab. Every time I did it in the past, that was TIFF files exported from C1, with the AdobeRGB profile. I am now using exclusively ART and RT and I am wondering how to deal with the output profile.
I always thought that there was one sRGB ācolor spaceā (or should I say profile?) , and the information embedded in the output image file was just a tag referring to an enumeration listing all the existing profiles, maybe located somewhere on color.orgā¦ But now I see that there are lots of profiles, and for example RT/ART use āRTv4_sRGBā by default, and the documentation lists āRTv4_Mediumā as an AdobeRGB equivalent.
=> Can I assume that the profile for the image is actually fully encoded in the file?
That is what I think I understand from the ICC specification (https://www.color.org/specification/ICC.2-2019.pdf).
If my assumption is correct, can I also assume that the lab will be able to convert my image files correctly for the printer and paper, whatever exotic profile I may use, because all the information needed is encoded in the file?
Additional questions:
Am I correct that the rendering intent specified under the output profile menu is the intent which is used by RT/ART when encoding the image from the working profile to the specified output profile?
When I open the image file in digikam, under ICC profile, there is a field called āRendering intentā. Thie field is always set to āPerceptualā. What does that mean, is that a tag which says āas the owner of this file, I would prefer that any color space conversion would be done using this rendering intentā?
Thatās the way itā supposed to work, but you need to confirm with the lab that theyāll handle it this way.
Ideally, they should give you a copy of their printerās profile, and you should export your image with it, which should convert your image from the internal colorspace to the printer colorspace AND embed the profile in the image. That way, either 1) they recognize the embedded profile and donāt do their normal convert, or 2) they just blat the image to the printer and it comes out okay because you did your due diligence in export.
You need to ask how they handle color management. I recently went to our local camera shop, which has in-house printing, asked the fundamental questions and they had no idea what I was talking about. I thanked them for their time and leftā¦
Thatās only half the story. They should give you the profile that is used for that printer in combination with the paper thatās been used/chosen.
Printing paper (all paper for that matter) isnāt white and comes in many shades/gradations of white. Varying from colder white to warmer white, each having their use. Printing is an additive process and the colour of the base layer (the paper) does play a roll in the end-result.
Printing paper manufacturers have many profiles for their paper types and specific printers combination. They donāt have profiles for all combinations, but do for most of the high(er) end printers and their paper types. Iām assuming that the more serious print shops have the higher end ones.
The Lightroom bit isnāt all that important. A colour profile is a colour profile. As long as you can load it and have soft proofing functionality it is basically the same.
They do provide such profiles but they are for soft proofing indeed, their goal is only to simulate the final rendering in the software. They should not be embedded in the image file.
I guess the final conversion from the fileās color space to the printerās one is done by the printerās driver, depending on several parameters including the paper type and probably a drop-down where they select the rendering intent to use.
which is indeed what I think (see above), but my concern is that I am not sure the conversion can read a profile such as āRTv4_Mediumā for example.
That is what I really need to do now, I just hope they handle color management better than your local shop! I know they do actually
The Printer Driver - in most cases a thing treated with so much disregard by the general photography consumer - BIG mistake.
It converts the colour numbers of an archival colour space such as ProPhotoRGB to those of the ink/paper/print head combo (commonly called the printer profile/paper profile) on the fly.
It also upsamples dpi from ppi on the fly.
And if that wasnāt enough, it also controls the incremental line feed of the paper/print media, and controls the vital platten gap. Platten gap controls the distance between the print head and print surface therefore has a major effect on dot width which in turn effects colour mixing, whilst at the same time preventing head strikes.
And most people think it just makes the printer talk to the computer and never give it a second thought.
Online print comes in two flavours: standard on line print (e. g. CEWE) or professinnal print shop.
In the first case use SRGB
In the second case, they must be able to help to choose the print method (C-print or fine art), the paper, to specify numerical file ( dimensions, margin, embedded profile, resolutionā¦), profile for soft proofing corresponding to your choice, eventually sending you some print examples to help you.
This is a professional one, which asks for AdobeRGB, and provide profiles for soft proofing. They donāt provide, however, the actual printer profile for each paper, but Iāll let this conversion to them (and to their printerās driver, as aforementioned). I usually print on Hahnemuhle Photo Rag 308 for those who know about papers, and soft proofing has proved to be very useful for such a matte paper.
I found this short explanation: Embedded color space information
which gives some details about embedded colors profiles and give the exiftool commands to show and extract them.
I have played a bit with exiftool and gimp: extracting the darktable AdobeRGB compatible profile, and assigned it to a file actually encoded with RTv4_Medium, and as expected the display changes a bit, as the profiles arenāt a 100% match even if they are close. darktableās AdobeRGB compatible profile seems to be interchangeable with the real AdobeRGB one though, at least I canāt see any difference on my (not so good) screen when swapping the embbeded profile in the file. (I really assign the profile with Gimp without changing the actual encoding, i.e. I donāt convert).
I downloaded the original Adobe RGB 1998 icc file and for now, my workflow will be to export to this color space directly when sharing the file with any third-party, so that I wonāt have any bad surprise.
Hello, I donāt understand this. How can you soft proof without the actual paper profile that you choose to print on? Youāre soft proofing with what?
Another set of complications is, if the soft-proofing profile is in CMYK (as my local print shop is doingā¦), then things get interesting. The only somewhat reliable results I ever got was with recent versions of GIMP for soft proofingā¦
The softproofing profiles are not the same as the printer profiles. They are created in order to simulate the render. They should be loaded in your software so that you can preview the final print. They should certainly not be used as output profiles for your picture.
If soft proofing profiles are not the same as printer profiles, what are they?
I make my own paper/ink/printer profiles using Argyll CMS and a photospectrometer from X-Rite, using this very informative guide written by Anders Torger.
When I want to print a photo, I assign the paper/ink/printer profile as the Output profile in my photo software. Then I activate soft proofing to see on my calibrated screen (also done with Argyll) how the image is rendered for that specific type of paper.
Well Iām not sure about that so take my words with care:
I think that the softproofing profile is made by analyzing a print for a couple (printer, paper). This profile defines a color space which is reduced because the paper and the printer have limitations. And by comparing with the original file, the profile can also describe the transformation made from the original color space to the final output.
When you apply this profile in the raw software, it shows you a kind of preview.
Now thatās the part Iām not so sureā¦
If you convert your picture to this color space with this profile, you lose information at a step where itās not required. If the printer drivers are very well implemented, I think that theoretically the while print process could be a ānoopā to your image, but I canāt believe this will really be the case in practice. Thus you could lose information a second time.
Yes itās better to kept a large gamut color space for archival purpose, however for printing we want a high quality conversion to the printer color space, nowdays open source tools use floating point precision so probably itās better to do the conversion before sending to lab or printer