I thought that when a DCP profile for a certain model of camera is produced (using for example X-Rite ColorChecker) there is a guarantee that the colors from raw data will be very close to what I can see for anyone who uses that profile. Right?
However, when I compare the output of Capture NX-D and Rawtherapee on my photos of art, I can see a difference. Where does this difference come from? Can I somehow compensate so I can use only one software (Rawtherapee) for postprocessing?
So for me this means that I will always have to export icc profile from the first photo of a new collection using Capture NX-D and import it into rawtherapee. ((different collection = different lightning))
You only need the profile once … it will not change with each photo.
In additiom NX-D applies some stuff zto the image and if you did not change anything in RT a curve is applied which uses some info from the embedded jpeg
what is your intent? do you want to reproduce the output of Capture NX-D or generate accurate colors ? In my experience, Capture NX-D was not good at reproducing colors, but can output enticing photos (vivid colors…) .
according to http://rawpedia.rawtherapee.com/How_to_create_DCP_color_profiles,
One of the steps of this development involves translating the numbers (in the raw file) into accurate colors, and for that you need to profile the camera, to map the numbers to specific known colors.
(and for that don’t use the embedded tone curve).
The embedded Ratherapee DCP are good at reproducing colors. You can also use the Adobe standard DCP.
Some DCP are made also to reproduce the in-camera settings to the detriment of color accuracy.
beyond the DCP,there are a lot of things that can change the perceived color of the output photo. Among the simplest, processing tools and settings used (they are not the same in different SW), viewing software( color managed or not), profiled display and wether the used softwares (processing and viewing) uses this profile for display…
for instance, in your sceencast above, I don’t know if in CaptureNX-D and rawtherapee you set up the same display profile if any.
And if you want to print, it’s still an other story
So welcome in the unmerciful world of color management
One thing I immediately notice in your screenshots is that the colors are more vibrant in the upper image, more dull in the lower one, in a way which looks like one program is using a wide-gamut monitor profile and the other is not. So possibly completely unrelated to DCP. Who knows.
Don’t crop screenshots, as in doing so you’re discarding important information that could give us clues.
Supply a sample raw file + relevant files (sidecar, etc.).
Describe your operating system and hardware (monitor(s)).
Describe your color management settings, both in your operating system and in relevant programs (raw processor, image viewer, etc).
And then trash NX-D
I was working with NX-D from 2018-2019, converted 12.000 photos, before switching to RawThwerapee. If you would like to have real accurate colors, forget NX-D! It uses some hidden internal adjustments, which you can’t reproduce or retrace with other software.
Please set up a proper color management, with correct and fitting profiles (no chance with NX-D), and keep going with RawTherapee.
Regards
Morgan_Hardwood here it is:
2. Boris B / color-management · GitLab
3. Ubuntu, Dell Latitude E5570 monitor
4. I have not touched any setting, so default
gaaned92 My goal is to have accurate colors. However, with only white balance applied, NX-D provided more realistic colors than RT. Therefore this topic.
For this moment, I solved it by applying the icc profile from NX-D. Next step I would like to understand more about the topic from Color management in raw processing.
I experimented with IT 8.7 Scanner Calibration Target C1 (Shooting the color target in specific lighting situations for yourself from How to Create DCP Color Profiles) and after applied as input profile the colors were more precise then before.
However, I was surprised that I no longer needed to apply white balance using white card. I mean, after applied white balance using white card, the result was a less precise. How’s that possible?
Let’s return to the basics: did you make a DCP profile for this shot yourself? I’m not a 100% sure it is done correctly… Simply applying it on the NEF you provided changes the colors and the tone quite a bit more than I’d expect.
Keep in mind that this doesn’t change the RAW, only the in-camera JPEG. So again this boils down to the question: do you want an accurate profiled starting point for further editing your images, or do you want to obtain the in-camera look?
Hi guys, thank you very much for your interest and help.
However, I realized I made a very dumb newbie mistakes (not calibrating my monitors, shooting color target with a lot of noise, not understanding the impact of input, working, output color profiles on what I see on monitor) and my questions here became all irrelevant.
I’ll clean my workflow first, read a bit more and if I have more questions I’ll probably open a new topic for them. Maybe just to peer review my refined workflow for producing accurate photos.
And by accurate I mean make the photo of the paiting as close as possible to what a person sees in the gallery/exhibition.
