I have to admit I am completely lost.
I have searched this forum and the rawpedia (e.g. Editor - RawPedia) but I can’t figure out the correct workflow for soft proofing. So maybe you can help me. Thanks very much in advance.
I have an icc profile from the company I want to have my picture printed with. For my monitor I have an icm profile from the manufacturer (and an inf driver but that is probably irrelevant). I do not have a hardware monitor calibration device. They are quite expensive …
In the preferences tab under colour management there are two sections. monitor and printer. Under monitor I can set the default color profile. In the dropdown box there are no monitors listed but profiles that are hardware independent. There is also the option to set “none” or using the operating systems main monitor colour profile. I am lost what is the best/correct option here.
Under printer I can select the icc profile I downloaded from the company I print my work with. I can also set options like rendering intent and black point compensation. The website of the company provides further settings like preserve cmyk numbers and simulate paper color. I guess there is no way to set them in RT.
Am I using these preferences correctly? What is the correct way to set the monitor profile?
In the editor, tab colour, section colour management. I can set an input profile, working profile and output profile. Input profile has none, embedded and custom as options. With the custom button I can use the icc from the company. Does not sound right to me. between “none” and the one from the printing company there is no difference in appearance, the embedded one gives a huge difference. Is it correct to use the embedded one? I am using a dng converted from the camera raw.
Working profile: I can select from hardware independent ones. They have no visible effect. Which one is reasonable? I remember adobe RBG has a greater range than sRGB. But most jpgs are transferred in sRGB. so … sRGB??
Output profile: There is a wide range to choose from. There are some of the printing companys profiles there, but not all. Any ideas why? Is this an RGB vs. CMYK thing? I have no idea what to set here. Chosing any does not have a visible effect.
Under the preview window there is a dropdown box (most left control) with “None” preselected. There I can select from a number of profiles, most of them I have never heard. There is also the ICM from my monitor and a system default. Changing from none has a visible effect. Is it correct here to chose the one from the monitor?
Next to the right there is a dropdown for rendering intent. No idea what this is for. selecting one does not have a visible effect.
Next to the right is the soft-proofing button. The tooltip distinguished between “when printed” and “when viewed on a display”. I don’t understand this. I am always viewing on the display am I not?`Pressing it has a significant effect, but with the tooltip I am completely lost which effect I am inflicting on the image. Also I would have no idea how to export the correct jpg which I then hand in to the printing company.
Sorry for the long post, but I wanted to make my painpoints clear so that maybe somebody else can relate and benefit from your answers. Keep up the great work. RT is amazing.
Thank you very much for your kind words. Both of those resources I consulted before posting this thread.
What I learned in the meantime is that the working profile does have a visual effect and I should not change it. I can only guess but my assumption is that the input profile should be the one from my camera and the output profile in the color management tool is the one used for the softproof button. the profile left to the softproof button is the one that is used for displaying the preview, so should be the one from my monitor. I further assume that I have to set the output profile when toasting the jpg (something that is explicitly discouraged in LR or PS).
Even if all my assumptions are correct I have no idea what I use the printer profile in the preferences for.
I will try to answer your questions as I understand them. If a more knowledgeable forum member sees anything incorrect, please correct me.
Let’s start with an analogy: even if you don’t realize, when you do something, you do a lot of things one after another, and RT mostly works the same way. Let’s say you are looking at an object, your eyes are capturing light, that your retina transforms into electrical signals, and then your brain interprets all of it into something that you understand. Finally, you tell someone else what you have just seen. There are quite a few transformations there, one from retina to neurons, another in the brain (interpreting the neurons signals), and another from brain to mouth. Any raw processing software works the same way.
you have an image and you need RT to know how to work with it: you have to tell RT how that image was encoded (how it was transformed from real life colors to pixels colors), and you do that by setting the appropriate input profile in the Color Management tool
then RT works with that image internally using the working profile (it’s better if it’s a wide enough color space, like ProPhoto, although I personally use REC2020)
finally you wish an image after all the processing done in RT, and you have to tell RT which profile it has to use to re-encode the resulting image: you do that with the output profile. To show the image in a website, it’s recommended sRGB. If it has to be sent to print it, I guess it should be a color space wider than that of the printer profile, or you won’t be using the full capabilities of the printer (so you would usually set the same output profile as the working profile)
in the middle of this, to process the image, you need to see onscreen what RT is doing, so the program keeps converting the image in its internal working profile to an image encoded using your display profile, that then is sent to your monitor (keep in mind that this process is just for you to see the current state of the image onscreen, but the image itself is not modified internally)
(hover them and read the tooltip to get some useful info)
And just to give an answer in advance: you set the printer profile in preferences to see how the image would look after printing, but only if you click the appropriate button (the middle one in the previous screenshot), not to export the image with that profile (unless the print service tells you to do so).
Thank you very much @XavAL that is helping me confirm my workflow:
Import with the camera profile, display with the monitor profile, recommended working profile, softproofing is done with the preferences printer profile and output shall be the working or sRGB/AdobeRGB.
I wonder why this is not shown/said more intuitively in the wiki or tooltips.
Again, thx very much!
Hi XavAL,
Actually, I know the maximum gamut capacity for a monitor is AdobeRGB ( I don’t know if there are more capable ) …but working profiles are more bigger… how do you behave about this difference in your workflow when you use Prophoto or REC2020 or let’s say ACES ?
Well, I think that it’s quite a lot of work to even have RawPedia online, so it’s even harder to keep it updated with the latest enhancements, and richly explained. If anybody thinks that he/she is able to help documenting RawTherapee, just ping @Morgan_Hardwood.
There are a few monitors that actually exceed AdobeRGB (at least is what manufacturers say…)
My own way of facing it is working with the biggest color space appropriate for the task: I used to set ProPhoto as my working profile, until I was persuaded that it’s better to use a bigger working profile than the output profile, but not much bigger. So I use now REC2020.
My goal is to make sure that no matter how crazy my editings are, I won’t get clipped colors in the final image: while working with an image, the hues around the boundaries of a small profile (color space), will be thrown in and out of limits, so the idea is that with a bigger color space, you will always be inside boundaries.
The problem when you send images for printing is that displays and printers are very different devices (one works in RGB, the other one in CMYK; one is additive, the other subtractive), and with good printers is not unusual that it’s gamut exceeds even AdobeRGB gamut (well, at least in some colors).
Even though I consider this image as just a sketch, what I say is something like this:
In this example, you can see that the printer profile (cmyk gamut) exceeds even the wide gamut profile in the blue-green colors, so if I were to print an image with that colors present, I would choose an even wider output gamut, so I wouldn’t get any clipped color in the print.
Ok, but how do you behave with colors you can’t view because they are out of gamut of the monitor (adobeRGB is smaller than REC2020) ? Do you control they with OOG preview??
Oh! I think I didn’t understood your question, sorry.
Somewhere I read that most of the colors captured in photography have a strong tendency to fall around the sRGB color space, with a low amount of colors being out of it. I can’t remember if those fell inside AdobeRGB or where spread out on a wider area.
But, either because of my way of taking pictures, my lack of skills, or either because of my somehow old camera, I usually don’t end up with strong saturated colors in my images, so setting the rendering intent as Relative colorimetric, and being not so crazy when processing images seems to be enough. Those areas that get saturated when converting to sRGB (for display viewing) seem not to be bothersome. At least until now.
I think that if I ever face a problem with saturated colors (out of Output Profile gamut), I would play first with the Vibrance tool, tuning down a bit the saturated tones (uncoupling it from the pastel tones).
But whatever I face, I’m pretty sure I will be more comfortable with fully saturated colors than with clipped areas, just because with Relative colorimetric at least I would have a smooth gradation of tones up to fully saturated colors. I bet that in the same situation and clipped colors, I would get posterization. And I wouldn’t like that at all.
I hope that I’ve properly answered your question, now.
P.S.: and if I have big areas of out of gamut colors, then yes, I control them with the OOG preview
I’m always in doubt what is the better working profile to use… I saw others use Prohoto, REC2020 ( as you), others ACESap1… I think if you want to capture all visible colors you need to use the bigger color space as ACESap0
but with it during the post processing you could go very simple out of visible colors and (you can’t view which colors are OOG with RT) … hence what is the better solution??
Perhaps the better solution is to have a working profile which encompass little more your camera gamut?
