I would not convert to dng and expect it to be easy to sub in the cr3.
You can build master now, import, and start editing, then just stick to stable when that is released.
And while you are at it, @aadm, create a noise profile for the RP. If you need help, I can offer some newbie guidance to the process as I did it myself the last days for the first time 
There’s a bit more to it than simply building master to get CR3 support working:
- you need to pull down rawspeed PR #271 before building darktable
- you need to update your exiv2 to 0.27.4-RC2 and build it with EXIV2_ENABLE_BMFF=ON
- you may need to put a standard matrix for your camera into adobe_coeffs.c if it is not already there.
then if you build master, you should be able to open CR3 files. Some features like lens corrections, denoise profiles and WB presets may not be there yet, but that’s being worked on too.
So, CR3 support is not quite yet ready for production use, but it’s getting very close.
4 Likes
oh definitely! if you can give me some guidance I’ll do that (noise profile I mean).
So I can finally feel like I’m contributing something!
Thanks Matt. All these things sound maybe a bit too complicated… I’ll probably just wait a little and maybe help with noise profile for the RP for now.
You have to be able to read cr3 files but I can run the script for you, no problem.
We need raws for every iso value with a bright overexposed area and a nice gradient falloff. I tried following this guide https://pixls.us/articles/how-to-create-camera-noise-profiles-for-darktable/ but this didn’t work out for me. My monitor was too dark and the gradient not so nice.
So I followed the advice in the darktable IRC to put a lamp to a (white) wall and use this as the gradient. I produced the following kind of pictures:
They need to have perfect close to perfect black around the gradient and close to perfect white in the overexposed area. Took a few attempts. Finally I used a 135 mm lens. Maybe try to reproduce those kind of pics and upload the raws somewhere. I’ll run the script and it will tell if the pictures are fine. When I’m done I’ll ask rawfiner on IRC and if he’s happy with it we can submit it. Do you know how to do pull requests on github? I can do this for you as well if you want me to.
Benjamin can you check and tell me if these are ok? I have set the camera to f4, short enough shutter speed (1/30s), I think the bulb is overexposed not sure about the gradient and the pitch black requirement… I have shot two test photos, first one with focus to infinity as recommended in the original blog post the second all the way to the opposite way (shortest focal distance possible on my rf 50mm lens). Not sure which one is the best?
focus to infinity
shortest focus distance
if you need to double check with dt, here’s the two cr3:
Also, at this moment I don’t think I would be able to run the noise profiling myself, esp. if this involves compiling darktable with the additional things outlined by @Matt_Maguire above… (I can run a simple make, but all those other things are a bit too much for me)… so if you can do the actual processing I’ll be happy to provide the files.
thanks!
1 Like
@aadm the script likes both variants so if you could do those for every iso the same way I think it will make a nice noise profile 
ok so you think this kind of images are ok for the noise profile? excellent, I’ll do that. Will do one set with focus to infinity, the other with focus to nearest distance then. Same shot, I will only let ISO vary and adjust shutter speed accordingly… let me know if I’ve missed something.
I will do them tomorrow evening.
1 Like
Yep, try to keep the exposure the same for all iso values. I’ll create two profiles for each focus setting and then rawfiner can decide which is better
Hi Benjamin I’ve prepared the raw files and will send you the link via private msg to download the zip file containing the two sets (one taken at minimum focus distance, the other at infinity).
For the shoot I have kept fixed aperture at f8, variable shutter speed and then modify the ISO from L setting (the one below the base ISO of 100) up to H1 and H2. For these last two however I had to decrease the aperture to f13 because I couldnt go below 1/4000s. I hope it’s all ok, otherwise let me know.