I have an issue with my new fuji X-T5 and my workflow.
My workflow =
- convert RAF into DNG with Pureraw 3 (or PL6) with “denoise and optical correction only” option
- import DNG into Lightroom Classic for adjustement
With the X-T5, burned highlight turn to purple with recovering highlight in LR on the DNG (not on RAF)
See this photo
I had an X-H2S without this issue with the same workflow !! (see this photo)
Do you have an idea ?
Does somebody have the same issue ?
Looks a bit like CA shift.
Purple shift is often also visible in the sky around branches.
Try to set WB (if that’s possible) and then export as DNG. (optical module changes CA corrections when you changes WB in PL.)
Never forget, burned highlights are never truly recoverable. Whatever the software promises, it has to invent something, which, in most cases is going to be shear guesswork.
Hello OXiDant : same probleme after changing WB
Yes I know, but why NO issue with X-H2S (and also with my old RAF file from X-T4, X-T3 and X-T30).
I’ve never see that before the X-T5 !
Hi, if you own only PureRaw you could upload your troubled rawfile and then one of us runs it through Photolab. upload the DNG for you and see if it’s different.
I tried on PureRaw 3.3.1 and PhotoLab 6.8.0 with same issue
Like I said, the problem arises because the image is over-exposed.
All PureRAW or PhotoLab is doing is converting the over-exposed RAW file into an over-exposed DNG file.
From the on, if you are using Llightroom to try an “fix” unfixable over-exposure, the problem rests fairly and squarely with Adobe and how their software handles irrecoverable over-exposure
ok and if you fully develop the rawfile inside PL?
do you get also purple overcast?
(please if you could upload the rawfile so we can investigate the scale of overexposure and the colorshift by lowering exposure. What @Joanna wrote could be the answere of your problem.)
Next step would be develop inside PL to recover the overexposure and such then export as lineair DNG ( with corrections) or 16b Tiff
and see if it’s again shift to purple inside Adobe’s enviroment.
PhotoLab doesn’t use purple, it just turns grey instead
There was a similar problem with Affinity Photo some time back that I think Serif fixed. From memory my work around was to open the dng in Capture One Express (its free for Sony and Fujifilm) then send a tiff to Photo. I haven’t done it for a while as I was evaluating the free trial of PureRaw and because of that problem didn’t buy it.
do you mind to shot a fully over-exposed ( all raw channels - point to some bright object and give enough exposure time to fully clip everything in the frame) RAF from X-T5 and post it ? then we can use Rawdigger to check the difference between what DxO PL6 generates in its linear DNG and what other software like Adobe and Iridient will generate in linear DNG …
w/o looking into linear DNG generated by DxO PL6 it is too early to say where the error lies - why do you automatically assume that DxO PL6 does not make some error the way it generates linear DNG in this specific case ? did you see it ? w/o seeing the data it is pure (but expected from you) fangirl statement
I think the problem is when only one channel is clipped.
I’m not being a “fangirl”, just a photographer and printing specialist with around 60 years experience.
I don’t care what software you use and how you scan or convert, if a negative or RAW file contains totally over-exposed areas, there is nothing that will “recover” what isn’t there.
Sure, Adobe claim to be able to but only by using a magic sauce or AI to invent something.
The way to determine whether there is something there is to pass the mouse over the area, watching the pixel values under the histogram and if they show anything less than 255, you stand a chance. If not, there is simply no detail. PL colours this grey as you reduce the exposure, Adobe uses purple.
Here is un completely burned image from my X-T5.
In PL with exposure at -3 : RAF goes gray
In LR, with exposure at -3 :
- RAF goes gray
- RAF enanced by LR (DNG file) goes gray
- DNG from DXO with “denoise & optic corr” goes purple
- DNG from DXO with “all correction” goes gray
But I don’t want to use “all correction” because the result is not exactly the same and also I lose Fuji Film Simulation in LR.
If I still use “denoise and optic corr” I still have Fujim Film Simulation
DSCF5498.RAF (6,7 Mo)
I’m out shopping. I’ll take a look when I get home
This one is clipped beyond any repair.
it 's white in FRV and DPLv6
Please shoot a Tree or something meaningless against a bright sky same like your personal image like without the person.
I just tested in CONE and Darktable, and in both of them, DXO DNG goes gray with exposure -3.
But I put a reservation because on a true photo color are not very good in CONE with DNG.
The issue seems to come from LR which see pure white on the DXO DNG as a R=99.6% G=99.6% B=100%
But I repeat my main question, why there is no issue with X-H2S and other fuji bodies ? With the same workflow, a pure white from X-H2S in a DXO DNG is read by LR R=100% G=100% B=100%
DSCF5499.RAF (34,4 Mo)
And here the result in LR after DXO processing :