I have calibrated my two Canon RF lenses, the 50mm 1.8 and the 24-240.
Only distortion and tca for now as (a) they were the simplest to do, (b) I couldnt find an appropriate diffuser to also measure vignetting.
I’m attaching here the lensfun.xml
file that you can copy to your local lensfun directory ~/.local/share/lensfun
(linux). I have also raised on issue on github to have these two lenses added to the official lensfun database.
Thanks to @asn for the how-to guide and Torsten Bronger for his screencast on how to use Hugin for the distortion data.
A couple of additional notes:
- I had to use conda to install all the packaged needed and listed https://gitlab.com/cryptomilk/lens_calibrate. Following the given procedure I wasn’t able to install
py3exiv2
. So if somebody else has the same problem and uses conda that’s what I did:
conda create --name lens_calibrate
conda activate lens_calibrate
conda install -y numpy scipy PyPDF2
pip3 install py3exiv2
- I have a bit of a mess in my darktable database now, looking at the collections by lens. The jpgs reports these two lenses as “Canon RF 24-240mm F4-6.3 IS USM” and “Canon RF 50mm F1.2 L USM” – the latter is striking as this is definitely NOT the lens that I’ve bought for about 200 euros – while the raw files report “RF 24-240mm F4-6.3 IS USM” and “RF 50mm F1.8 STM” or “RF50mm F1.8 STM”; see screenshot below:
Any idea why that is? As I’m using dt 3.6.1 all my CR3 files are converted to dng; right now I’m using the excellent dnglab, but probably a large number before dnglab were converted with the Adobe dng converter. I would like to have all the photos actually shot with the 50mm for example appear as under one and only lens and not three versions. Should I change manually the lens tag with exiv2 or exiftool? Would this result in a mess or lose the automatic association to the right lens in the lensfun db?
Examples for the most egregious and needed case (the 24mm end on the 24-240):
before
after
before
after