While working in photolab I came across an overexposed image in which the light reflected on a child’s face looks clipped.
Upon reducing exposure this happens:
But in Adobe’s camera raw I can start from this:
To this:
NEF file in all examples above taken with a Nikon Zf.
Has anyone got any idea what’s going on here?
Is that file RAW or JPG?
@john4pap , please share the original RAW and associated .dop files for us to see if heights are clipped or if the developer(s) interpretation “does the trick”.
From first view, I’d say that DPL or the person operating it … have room for improvement in the field of “limiting the use of sliders to what makes sense and to avoid what looks like an over-processed image” …
as people noted please share the raw file - DxO code does clip unclipped data indeed rendering it into oblivion, but it is good at hiding this sad fact if you stay within the app (PhotoLab) and have no idea what is going on in raw channels ( so if you see this it indeed might be both unrecoverable clipping and your use of some sliders - but it might be just your use of sliders )… so do post that raw file and people here will tell you the truth, the whole truth and nothing but the truth
Thanks everyone. Please find attached.
_PAP2059.NEF (30.6 MB)
@rrblint It’s a NEF file, raw.
@platypus & @noname Perhaps I wasn’t clear enough in my first post. The first image is before any editing. Hence no dop file available.
if you go to menu and export DOP it will tell people what DxO uses by default in your case
magenta tint ( .NEF → DxO PL 7.4 → “no correctons” preset → export to DNG w/ only NR and optics correction applied / and they are not applied of course / → ACR ) - a quick tell sign of DxO usual scr$wup with the raw data … rawdigger shows that data is NOT that massively clipped at all ( there might be some specular highlight points somewhere, clipped in green ) …
rawdigger = no clipping on skin
FRV shows that even if you add 2/3 stops of “exposure” - no clipping shall be on a skin
PS: this NEF is a beautiful example of how bad DxO code is … I shall save it
one more time, obviously (c), shall I repeat - do not practice any semblance of ETTR if you intend to use DxO tools for raw conversion …
and you illustrate what exactly ?
That Affinity (like camera Raw) has no problem - I thought that was self-explanatory ;-).
Captain Obvious has to ask, you know … if is of course obvious (c) that only DxO code has the bug, DxO invented it, owns it and proud to have it … a theory - forcing people to ETTL they can then pile their AI/ML NR on top and proudly say - patience, young grasshopper - we will save your / underexposed or else we clip you / shot !
PS: a new hint for DxO Marketing, campaign under a slogan : “clipped highlights, only in Photolab !” 
Agree with your FRV observations and that DxO has probably made another mess here. However, were you able to download the proper camera / lens optics module? Probably doesn’t account for this though.
it has nothing to do w/ optics correction… this bug was illustrated previously many-many times for raws where optics module exists … what is beautiful in this specific raw ( thank you OP for sharing … no - THANK YOU / all CAPs, obviously / ! ) , that it is so spectacular in how it illustrates the bug
rawdigger is illustrating it better, FRV is not aimed for high-precision - but of course it gives an idea that even a mild ETTR will lead to troubles in DxO applications
I am agreeing with you, just think covering the usual bases might convince the more skeptical.
I am not sure what needs to be done more for DxO to see the light ? a 24x7 picket in front the their offices ? somebody, give me a yellow [ obviously ] vest …
@noname Thanks for this. Yes, I now see the point of a dop file. I had forgotten that Photolab offers default corrections. I have none of them on as far as I know. I’ve created and am attaching a dop for this image.
_PAP2059.NEF.dop (9.4 KB)
@roj Thanks. It’s the same result I get from Adobe’s camera raw. I haven’t tried it in Capture One yet as I’m experiencing a bug with its library at the moment. I expect it should be the same.
@eriepa Yes, I wouldn’t be able to open the image without the camera module (if that’s how it’s called in the case of camera support). No lens module as I’m using an old adapted Nikkor AIS lens here.
Has anyone tried to reproduce it in Photolab?
do not bother - file a bug report with DxO and TELL THEM that they need to finally fix the bug for ALL camera models - not just for your model…
