I have a Canon RF 24-70mm f2.8 lens for a EOS R6 II, and when I process raw files using those kit in Darktable, I noticed a strange bright ring around where the vignetting starts. This is making the correction for vignetting very difficult. It is most noticeable in the smaller preview window on the top left in darkroom, as well as in the previews in lightable.
Is this a defective lens or is this normal? If it is normal, can you guys give some idea of correcting the vignetting?
I have attached two example raw files here, please have a look.
Which version of darktable ? OS ? Which modules applied ? Is module “Lens correction” active ? If so, what are the settings ? Can you share the xmp sidcar files ?
Loading your images with master 5.1.0+483~gab5cdbd3c6 they look “normal” for me, no bright ring.
I compiled the source from github, it is the latest version (5.1.0+516~g869316abfc). I have the lens correction turned on, the profiles for the lens and the camera are there and it’s set to “all”.
You are right, it is not very obvious when only turning on the lens correction. But After the edit it is quite visible, especially in the smaller previous. From the histogram, the bright spot can be seem as a “compression” in the profile. Please see the screenshots I have uploaded.
I really don’t think it is the software, I also went back to some older photos taken with this kit, most of them look fine, but a very small set did display this artifact as well. Not sure if it is the weather (<-20C) or something to do with the image stabilization.
Does the problem go away if you disable lens corrections, or at least set it not to do vignetting correction?
Have you tried updating the lensfun database? (Search the forum or the web to find how to do that.)
Without knowledge of your complete development history (xmp sidecars) it’s not possible to help. For me, with basic processing, no bright ring visible, neither in main view nor in navigation.
One thing to think about when taking the sample files is to not focus beyond infinity. For infinity I try to focus on some clouds far away so that it doesn’t over correct later.
I noticed the lens was wide open at f2.8 and I wonder if this added to the problem or minimised it?
Maybe do some tests. Vignetting is plainly obvious when you photograph the sky or any object with low detail and a constant luminosity. To detect it, you can defocus the image so that all texture and details on your subject (sky, white wall, …) are blurred, leaving vignetting as the only visible information on the photograph.
I just updated Canon’s DPP to their latest version and it seems the lens in not in their database based on the message given when I hover on the thumbnail or open the image. I do see the vignetting problem in DPP but it is not super obvious, possibly due to the lack of any editing steps by me.
This shows the history module but not the settings of all applied modules / steps. To reproduce your issue the xmp sidecar files are needed.
edit :
you have three instances of tone equalizer applied. This module internally works with guided masks. If the masks of these three instances overlap in an unfavorable way, the artifacts shown could result. But without the xmps : all just a vague guess…
