I am wanting to check in if I am following the correct procedure wrt the color calibration module. Apologies if this seems a bit simple, I read the module docs a few times but I wasn’t totally sure if this is the intended way.
I only have a gray 18 reference square, so it appears that I need to use the “area color mapping” part of the module on my gray 18 reference photo and use an Lhc of 18.94%, 0, 0 (determined from converting the hex value for gray 18 #2e2e2e on this site), select just the gray part of the reference image, and then selective copy the color calibration module to all of my other images. Do I have this correct? The colors seemed natural and correct, but I can never really be sure sometimes and I’m not an expert in all of the different color spaces. Thanks. I am using Darktable 4.8.1.
area color mapping is used to match color calibration between multiple images taken under the same illumination situation. So if the shot of your reference grey patch was taken under different conditions or you’re entering arbitrary values, you’ll just get a consistent deviation.
You have to account for gamma so that is why the values by default in CC are set to 50 0 and 0 and not 18. For white balance you should leave them there…
Thank you. Yes, the conditions were pretty similar. I was doing a 3D photoscan using Meshroom and I put the gray square near my object, took test photo, then took it away and did the scan within 30 minutes. Also, for the white balance I set my D7500’s white balance to a random preset K value to avoid the camera changing the white balance on each shot. All of this was shot on a cloudy day for more even ambient lighting with minimal shadows. Is this the correct setup for using the gray reference?
In that sort of static setup you could use the custom wb setting of your camera and just set it from the grey card rather than use a random number… then shoot everything at that wb…
Color spaces have nothing to do with things here. At least not in a way you have to worry about.
When shooting raw, the camera WB is only a hint to the raw processor. The raw data itself is not affected by it. In the case of darktable, it will be used to set initial parameters in color calibration, but that’s it.
The area color mapping is not intended to be used that way. This video shows how to use it:
If you have changed anything in area color mapping, be aware that it will affect all future uses of color calibration, so make sure to reset it by double-clicking the sliders. It should look like this:
The solution is to either do as @priort suggests and set a custom WB in-camera, based on the gray patch, or use CC normally - without area color mapping - and then transfer the values to other images. Doing it in-camera is by far the simplest approach.
This video explains how to use CC for white balancing:
then you simply can “measure” the gray patch on the reference patch and then copy the color calibration module setting by ctrl+shift+c and ctrl+shift+v.
no need to deal with that area color mapping in this case
