This has been the work around for several versions. Glad you found it.
Well it seems not everyone knew about it, X @paperdigits
Probably doesn’t apply to you, but for the Arch or Arch Linux-based distro users can always build darktable-git very easily as I do frequently in Manjaro Linux to see the new commits merged into the main branch. It’s in the AUR so you can access it from your GUI package manager or terminal via YAY or whatever AUR package manager you use.
2 Likes
R&Darktable is also available in the Arch AUR:
https://aur.archlinux.org/packages/rdarktable-git
It is mutually exclusive with the darktable-git package; if you install one it will replace the other.
4 Likes
Ah nice, you just made it. Upvoted! 
R&Darktable is also available in the Arch AUR:
https://aur.archlinux.org/packages/rdarktable-gitIt is mutually exclusive with the darktable-git package; if you install one it will replace the other.
That’s great. I assume it will happily coexist with Darktable from the regular repos, correct? Does it install to /opt or to /usr?
Looking at the PKGBUILD it appears it will replace any other instance of darktable installed via pacman (either the official package or any other AUR package) since it’s listed as “providing” and “conflicting with” darktable. Similarly, the PKGBUILD suggests it installs in /usr.
1 Like
Right, it is mutually exclusive with the official build, in line with AP’s wiki article:
Specifically he says:
R&Darktable uses the same scripts and building stack as upstream darktable, so everything works the same but you need to be aware that it will get installed in place of darktable and as darktable, so it uses the same pathes and default config/database directories. The CLI commands are all the same and R&Darktable will be registered as darktable as far as your OS is concerned.
I’m not sure at this point if it’s worth it or even desirable to have it installed somewhere else and using different config pathes and CLI commands.
So I set up the PKGBUILD accordingly. If I want to use a second instance of DT independently of the first, then I normally build it manually and install it into a separate dev directory.
The issue with replacing the official 3.8.1 build with R&Darktable is that the database will be upgraded to master, and no longer compatible with the “stable” build. If AP tags a “stable” version of R&Darktable in the future, I’ll make an AUR package for that, but I’m not sure he’ll be too keen to jump into the whole stable release/minor release cycle process anyway…
2 Likes
One other thing I noted for windows users…if you make a package and intstall…it will prompt you to uninstall DT…You can think its talking about R&Darktable but it will uninstall your version of regular Darktable…so for instance you install R&Darktable…set it to run along with Darktable. Then you build a new version and then go to install it again or over top of R&Darktable in the R&Darktable folder that you previously used and it will prompt you to uninstall first…if you do that it uninstalls your Darktable install not perhaps what you are expecting… I am sure this could be corrected but it might be something that floats around for some time…
The chromatic aberration module remove CA, if you know how to do it!
It requires a lot of work for only okay results. I think the manufacturer lens correction module also corrects for CA, so that should help if it eventually gets finished and merged.
1 Like
You do it once for each lens, you keep it as a preset. If you use the correct image which show you CA, it’s perfect. With the “manufacturer lens correction module”, you must do the same job, but with a different method.
Fringes are a mix of CA, blue & red, on the 2 lenses, the one mounted on the front of the gear & the lens on each photosite. In Dt, you can correct the 2 with 2 modules, but with CA lens, you had to duplicate the module & create a correction for red, green & blue.
I don’t agree, it’s a different technology than in RAW CA.
I use the chromatic aberrations module in addition and after RAW CA.
2 Likes
Different technologies?! No, you have a lens mounted on the gear, & a lens on each photosite. You must begin by correcting raw CA & then lens CA with a module for each primary color. With modern lenses, CA & fringes are taking into account when lenses manufacturers compute lenses.
Yes, the modules lens correction and raw ca use similar technologies but chromatic aberrations does something different. The latter also hurts healthy zones and is only to be used if nothing else works. I dont know the technical details but ask the devs.
1 Like
With some lenses, there are interactions between the lense & photosite lenses. The corrections begin with the inverse of the appearence of CA in the 2 lenses, so the first to correct is the raw CA, & then the lens CA, that you can refine. You need to prepare an image which make appear CA & fringes. I’m not a devs, i learned optical, i’m a professional photographer & even if the correction is done by software, it’s an optical problem. Why don’t you try what i explained before saying No? I’s take less one hour, to try it!
please explain your method better, I don’t understand much of your first description.
which modules do you use? how many instances of them do you use etc etc? what are the exact settings?
can you post a screenshot of your settings? maybe a video?
How can a lens covering one photosite cause visible chromatic aberration over the whole image?
(Note that I’m not saying such a lens could not have chromatic aberration)