Can you make available one of the images having the issue?
I use Photolab as my main RAW converter and I’m happy with it.
But recently I’ve started with astrophotography and noticed that Photolab isn’t working properly. The raw conversion is causing weird concentric circles in the astro photo’s. Post processing astro photo’s is difficult because of the lack of light, the peak of the histogram is mostly around 1/4.
Knowing the difficulties in post processing astro photo’s I’ve searched the internet for possible reasons. I found a post on Cloudy Nights that using lossy compression with RAW for Nikon can cause this effect: Nikon Coloured Concentric Rings - DSLR, Mirrorless & General-Purpose Digital Camera DSO Imaging - Cloudy Nights . But I’m using a Nikon D7200 with 14 bit losless compression, so this is not the reason for these weird concentric circles in my images. And using other RAW converters (Nikon Capture NX-D, Rawtherapee and Affinty Photo) showed that this effect is almost absent.
But I think I’ve found the reason for these circles, it’s the vignette correction in Photolab. Without the auto vignette correction these circles don’t appear. To show the effect I’ve made a photo with the histogram peak around 1/4 from the left.
Converted the RAW file with Photolab using the preset ‘no correction’ and then only applied the camera and lens corrections (and vignette correction on) with the prime noise reduction. Saved this as a 16 bit TIF file and the processed this TIF file in Affinity Photo using the same techniques for astro, i.e. stretching the histogram with levels and/or curves. Doing this will show the artifacts introduced by the vignette correction of Photolab:
Not using the vignette correction of Photolab, the concentric circles don’t show up. And converting the RAW file with Nikon Capture NX-D, Affinity Photo or Rawtherapee with vignette correction on, there are no concentric circels. As a new user I’m not allowed to upload these photo’s.
Now I’ve found this topic and see that this problem is already reported in december 2019 for Photolab 3. Unfortunatly this problem hasn’t been solved and still exists in Photolab 4.
Any feedback from DXO would be appreciated.
Many thanks.
Good day @LexB,
Could you, please, provide us with this RAW file (and a sidecar if available) for the investigation? Please, upload it on upload.dxo.com with your forum name in ‘support ticket number’ field and let me know when ready?
Thank you
Regards,
Svetlana G.
Hello Svetlana,
I’ve uploaded the RAW file and the sidecar. Also the TIF file generated by Photolab and the processing of this TIF file in Affinity Photo using levels and curves for stretching the histogram.
Kind regards,
Lex
1 Like
Good morning Lex,
I’ve got them, thank you. We will investigate the issue.
Regards,
Svetlana G.
Hello Svetlana,
Could you please give a status update on this issue?
Is the cause already known and when will it be solved (update june 2021?)?
Kind regards,
Lex
Good morning @LexB ,
Well, the issue (id = DFRMWK-17944) is in the backlog but I can’t say anything about the timeframe as it’s not an easy one to fix. I will let you know as soon as it’s fixed.
Regards,
Svetlana G.
1 Like
Can you tell me if this issue is fixed/solved in PL6?
1 Like
Am new to DxO, about to purchase it if this problem is resolved.
Please update the status.
See, also:
Banding section of
and
Hi and welcome to the USER forum,
who are you talking to? If you like to contact DxO or the authors of the linked articles, you have to adress them.
You should just try photolab whith the free one month trial version and see if it fits your needs, your gear and the kind of photography you do.
I have no banding problem with vignetting filter (if this is your question since this is this topic subject) with my nikon gear :
D850 - nikon 300mm f4 VR - nikon 105 mm f1.4 - nikon 105 mm macro f2.8 - 85 mm f1.4 sigma art - 35 mm f1.4 sigma art (which I don’t like that much with D850 - maybe a bad model) - and sometime nikon 600 mm (which I don’t own myself).
Even with 300 mm f4 at full aperture (which has I think about a huge 1 stop vignetting at this aperture). Only a very slight undercompensation of vignetting for this lens, probably due to the variations between models that today’s industrial production and quality controls allow (nothing to do with my 80s Angenieux, which came with a test curve done for each individual lens).
PS : Anyway photography life review is very interresting on some points.
Transcient Eye review seems more subjective on some important points to my eyes.
PS2 : I have no banding problem of any kind with my gear. Maybe I never pushed the curve tool in extreme!!! limits (but I never had to (with the kind of photography I do ?) ).
Am talking to Barbara-S, as I hit “reply” icon on Barbara’s append.
The linked articles address banding problems of DxO as well, so, I wanted to let Barbara know that other people are concerned with this problem, too.
JoPoV,
Thanks for the suggestion on the trial version. Will do it.
May I suggest to adress her explicitely to make sure she gets your message?
And please, explain a little.
DxO support is not always following here. 
Probably the problem was with GFX100 firmware (at least in the first case). It was fixed later by Fuji and maybe they provided workaround for LR and C1. See PL7 Shadow Banding - DxO PhotoLab - DxO Forums for some analysis by @noname and GFX 100 PDAF banding is fixed - the last word (kasson.com) for some background.
EDIT: Oops, it seems that in this case GFX100 PDAF banding was a different issue. Nevertheless, I had no such problems with several Nikons (D4/D780/Z8) in PL7 and the analysis by @noname has shown that the problem was with GFX100 RAWs.
Note also that some Premium LUTs recently provided by PL7 may cause posterization in certain cases, but the reason is in LUTs themselves, being ‘locally discontinuous’.
Yes, the GFX issue is probably due to the RAW file data being bit-wise truncated sometime after demosaicing.
If you use PL7 to process the problematic GFX files to a linear DNG and then raise exposure/shadows in Lightroom or Capture One, there is no banding, whereas if you do the same exposure/shadow correction in PL7 the banding is seen.
I have not seen a similar issue with any other cameras that I have looked at in PL7, and it only affects GFX files if you need to apply an unusually large exposure adjustment (in my case deliberately underexposing high-contrast night shots to maximise highlight headroom to retain colour). So it is either something specific to the processing of GFX files, or simply that the GFX is unusual in having enough useful dynamic range to permit shadow lifts large enough to expose a problem.
There is a support case open for this, although it is not progressing much…
The Photography Life issue with the curve tool and LUTs may be related. ie what we are seeing is an insufficient bit-depth in the processing pipeline after demosaicing. If so, a solution may be a while away.
I never saw this banding issue before in my editing, I just tested it, and I could reproduce it on my first try:
What’s going on here?
Could be interresting to see what similar image processing produces with other softwares if you use some.
With the same image of course.
Okay, I tested it with Lightroom, and I believe the problem is not actually any wrong bit depth or something like that, but the way Photolab interpolates the curve based on the points that you set in the interface.
All other Software that I know does a spline interpolation, which will create very smooth gradients for your tone curve, e.g. in Lightroom:
If I set it steeper, it will just crush the blacks and whites:
Compare this with Photolab, which will just somehow connect all the points almost linearly and thus will create very steep gradients:
I had already criticised this weird interface some time ago in another post, I find it much harder to use, and now you can see, how it will also lead to banding much more easier.
I can reproduce banding in Lightroom by trying to create a similar tone curve, but it is much harder, as I really need to set many points in order to achieve such a curve, so you would hardly ever do that:
1 Like
ok. thanx for the test.
Nice to see this does not seem to be an “engine” photolab limitation.
So with fiew more points in photolab curve tool you could reproduce lightroom behavior.
But i agree and talked about this too, this curve tool interface should be rewritten.
And at least :
it should be resizable or with several sizes choice for more precision, we should be able to see points values, and a pipette should be present. This are the 3 more important missing points to my eyes.
Other points in order of importance to me :
Arrow keys should allow very small adjustments and/or CTRL key “limiter” (go up/down only or left/right only depending of the first mouse direction), user presets, histogram in background.
Let push further (a little too much perhaps) :
curve interpolations : bezier, hermite, linear, tcb …