Sorry, but some people just have to try to be no it alls.
my point was solely limited to (A) finding a bug (B) explaining the bug and lastly just in case (C) offering a simple solution for topic starter’s exact workflow while additionally illustrating the bug
as you do not like to have an egg on a face about DxO bug you try to switch readers attention to (C) as if it was my main point… it was not… I agree that IF topic starter can change his workflow there be probably better solutions where he can invest more time in corrections with DxO …
This seems really pathological.
I never gave any opinion about Adobe or DxO for this “bug”. Just about how to fix it.
It is very dangerous to see things we want to see when they don’t exists.
my credentials are well proven by explaining the bug and answering the question in the first post by the topic starter… remember it ?
I really understood your point.
Your method is right (I have a scientific mind too), your conclusion is right.
Nevertheless you insult people which, with years (a life) of practice give a very good solution which is probably better than yours and pretend they don’t know what they are talking about.
Think about it.
There are several ways to go to Rome.
Don’t put your ego so high that you think only your path is good.
1 Like
Mostly finding a bug in PL.
I still don’t know what that bug is. And why focussing on PL and not LR.
I’m open for everything.
George
1 Like
Not much one can do with that sky. DSCF5499.RAF.dop (16.3 KB)
Or, directly in Lightroom:
DSCF5499.xmp (11.1 KB)
Note: all of the above is a proof of concept. For future reference: Learn how to expose for the highlights and/or do some bracketing. 5498 can be used to set a UniWB though - if Fuji accepts that image for WB.
Didn’t thought about it.
Interresting.
I may have fallen into the trap of talking about ego.
He means that highest values in exported DNG sould be 65535 instead of 57837 (16bits) in this case since highlights are totally burned. (equals to 255 vs 225 converted to 8bits).
Or that this highest 57837 value should be noticed in the DNG for other softwares to interpret it right.
(This is its solution by modifying metadatas).
And Joanna showed a way to get this 65535 value in exported DNG by scaling values with curve in photolab.
Which keeps a higher dynamic in the DNG than reducing it to 57837.
The highest values in that dng are 255 and 254. Normal clipping values. I don’t know where he got the 57837 from. I’ve asked but no reply.
George
255 is max value in 8bits converted in decimal (we can better understand decimal numbers than binary).
65535 is max value in 16 bits converted in decimal.
DNG is 16 bits depht (8 bit depht is not enough to encode all RAW extracted values).
So possible values are from 0 to 65535 in decimal.
8 bit depht is ok for exporting pixel images which won’t be processed later (jpg or tiff-8bits in photolab) - with some exceptions sometime (absolutly no noise and very fine gradients at big scale in the image which can lead to banding for example).
16 bit depht is ok for exporting pixel image which will be processed later (tiff-16bits in photolab).
@noname
Your knowledge is very appreciated (at least by me).
But you should better consider other people, including those who have a big background in photography, even if their approch and way to explain is not as scientifical as yours and lead to other methods.
Scientifical knowledge is not required to shot good photographs even if software engines are only scientific computations in background.
And if you are an AI bot, your developpers should train your statistic algorythms with more consideration to people.
You have inspired me and, using a Control Line over the sky, lowering the colour temperature and highlight slider, like you did, I was able to get a “blue” and white sort of cloudy sky…
Oh, and a curve…
Oh brilliant. I’ve just learned something new.
You don’t have to explain me what binair counting is. I was just asking where he got the number 57837 from. It’s an essential part of his reasoning.
George
Sorry if I offended you.
Just wanted to give an explanation to your question.
This topic is a bit nervous, isn’t it ?
Hmmm. Questionable 
it comes from reading datas in rawdigger.
That’s maybe the answer you wanted.
Writing the facts:
1 overexposed areas in a rawfile of a relative short supported rawfile (x-trans files) by DxO.
So a hickup isn’t a leep of fait.
2 inside DxOPL the purple glow isn’t a factor thus it seems they have solved the demosiacing to endfile balance of WB black and whitepoint.
3 not only LR shows the purpleglow faststoneviewer did it too with the rawdng.
Has DxOPL a flaw? Does it go wrong/bad?
Not inside PL itself apparently.
Is adobe to blame?
Don’t know, don’t have the application nor the knowledge to have a fundated opinion to state that.
My aproach would be run that DNG through serveral applications to see if it’s behaves consistance in purpleglow.
If it does then i could dare to state it isn’t go wrong at the adobe part.
(it would be very odd, far leep, to expect that several different raw/dng developer give the same flaw on a perfect DNG.)
Second test would be take an secondairy similair camera and use that to create the same files to see if it’s behaving the same.(so you know this isn’t a random error from that camera.) (we tested both dxo applications). (i must think about from which version is Fuij supported? And can Wide Gamut be a factor? Both are relative new parts of DxO PL’s toolbox so some finetuning could be necessary)
All this testing doesn’t need any university education level of mind just common sense and logic thinking.
Most of this is already done and i expect that dxo technicalstaff is doing the last steps.
So i just wait to see the outcome of there investigation.
About the way of talking/writing on this matter, well stay normal within respect of each point of view and reason with fact’s (even when those fact’s seems to merge in personal opinions and remove them selfs from the fact definition after a wile of debating) and examples without any stabbing on the persons believe or speak to hand and wait for the last finger standing. I have zero respect for anonymous swearing and ranting and ridiculing someone. I say ring my frontdoor bel and say it in my face.
And live with the consequences. I know i do. 
Aldoh i do like a good discusion because it widens your view on things and there is always something to learn from it. (Even when you need to clarify and proof your knowledge/statement.) i don’t like to be ridiculed by a remote bully safely in the dark but i am a grown up so when i am wrong i am wrong an i can take that without any beef or lasting anger.
So i move on and wait for the result of the dxo ticket.
And what data are that? One can’t just mention some data without saying what they mean and where they come from.
George