Hi there.
I’m relative new to RawTherapee, using it since December 2019. But studied the whole RawPedia and other forums over weeks, to get to know exactly how RawTherapee works.
Since Summer 2019 I’m working with a custom UniWB and ETTR (expose to the right) setting at my D7200. Because of that, I had to reduce exposure as a first step of RAW processing in my previous used converters.
But not so with RawTherapee. In rare cases I had to apply only -0.25 EV or less to get a proper result. (f.e. in Capture NX-D I had to apply -1.5 EV to get the same result)
Maybe someone could explain this to me?!
Is there some “hidden” compensation applied in RT or NX-D?
My settings in RT: Neutral Profile, automatic Camera Profile (input), only base table applied, no curve, no look… Working Profile: ProPhoto, Output Profile: RTv4_sRGB
Settings in NX-D: FLAT profile applied, Working Profile ProPhoto, nothing else…
Thanks for comments!
Regards
If you’re using RT’s neutral profile, no. Hard to say what NX-D does, as we can’t inspect the source code, but there is usually some behind the scenes stuff in proprietary raw processors.
Thank you. Of course there’s some behind the scenes stuff in NX-D. Can imagine that Nikon doesn’t expect that sombody is working with ETTR-method and so they’re adding some “extra” exposure, in order to prevent the user from questioning “Why is my photo so dark?” 
But thanks for confirmation that RT doesn’t apply such behind the scenes stuff. So I can be sure to see what the sensor saw, with no other influences. Btw, it’s the same what RawDigger shows, so it can’t be wrong 
🙋
CapNX and Adobe Lightroom both make background adjustments that try to match the embedded jpeg in your raw files and the camera profile and picture settings that are written to the raw file exif data.
This has been a massive bone of contention for me, and others, for years now; and I suspect it’s something to do with the ‘instant gratification brigade’.
RT is such a welcome addition to my workflow because its neutral is exactly what is says on the tin - NEUTRAL!
You’re right. Once upon a time, when I had detected this behaviour (shooting RAW, but only see the embedded Preview-JPG, with it’s own histogram, everywhere (camera’s LCD, Capture NX-D…)), I asked Nikon Support about the meaning of shooting RAW while judging, collecting, editing and exporting JPGs in fact. I’m still waiting for an answer, since summer 2019👌
Of course, there’s more space to adjust some sliders, you can change white-balance or some profiles in post… But you’ll never be able to see what the sensor saw, until you use RawDigger, FastRawViewer, RawTherapee or any other software which supports such a “feature”.
As I don’t like to be patronized from company’s, who are thinking that their users are children, I also went off… Surely, they mean it well - keep it simple or “attractive” for the consumer - but after all this should be a decission made by the user, if he wants to work with pure RAW (incl. Histogram) or not.
Regards
Yeah, totally frustrating. It’s a shame that in camera histograms are based on the jpeg data only. Why do camera manufacturers expect this to be useful if shooting in Raw only?
I second your praise for Rawtherapee being able to show true neutral raw processing, and even undemosaiced if you so desire!! This means that Rawtherapee is not just an excellent raw editor, but also is a good teaching tool for raw processing theory, as well.
That’s true. I think that’s because cameras are not manufactured by photographers rather by engineers who don’t care that much about the “product” . To them it doesn’t matter if creating a camera, a car, a laptop or toys… Same with comercial software engineers/developers.
In case of RawTherapee you can “feel” that the creators are knowing what they do - because they ARE photographers too. So they know about the users needs.
I wouldn’t wonder if some of Adobe’s Lightroom engineers are using RawTherapee undercover for their private pictures 
So, my way was like this:
Photographing in RAW - processing with Lightroom - export to JPGs - end of story.
No idea of color-management or what’s going on under the surface. Just typical consumer.
One day read about color-management - confusion - “calibrated” my DELL monitor - bought ColorChecker - made DCP profiles by myself, seperate one for each shooting or given light condition - used them with Xrite Lightroom Plugin - thought I’m done…
One day bought a new BenQ Photography Monitor and Xrite ColorMunki Photo Colorimeter - calibrated and profiled my new monitor with DisplayCal - looked at my past pictures - was shocked
Studied a lot about color management and digital photography in general - thought that Nikons Capture NX-D would render Nikon’s NEF the best - switched to NX-D - processed 10.000 photos again from scratch, with proper color-managed enviroment this time - was happy with results…
One day read about “ETTR” - tried out - but for exact use “UniWB” etc. is required - so studied UniWB and ETTR (thought that will be a neverending story, to get best possible quality photos out of my camera) - finished this challenge - thought I reached the end of the line (still used NX-D)
One day I was so annoyed from NX-D (rendering of the preview-JPG took half a minute, after just setting the exposure slider to -0,25EV…) - was looking for alternatives - tried around 10 other converters, but was never satisfied… until I found RawTherapee
Studied the whole RawPedia to become familiar with RT (which increased my knowledge furthermore) - did a lot of comparisons with “old” NX-D photos and “new” RT ones - was like awakening from the Matrix…
I was really surprised what RT is capable of to get out my NEFs! But was also sad that I didn’t found it 2 years earlier, before processing about 10.000 photos with NX-D (after they were ruined by Lightroom without CMS years ago)
So, here we are - today 
That’s quite the journey.
I don’t think I have quite gone down the color management rabbit hole just yet, though I did purchase a pro-sumer monitor based on color accuracy specs. I do a generally approximate version of ETTR, based on zebras, which I know is not accurate based on the jpeg processing shenanigans similar to what was discussed on this thread, but close enough for my purposes, and 2/3 stop underexposed isn’t that significant of a noise increase.
I have been using Rawtherapee for 5 years, and am still discovering things I can do with it.
I recently got into using the command line interface version of Ratherapee in Windows CMD scripts, for processing raw video sequences, and that was quite the nice discovery, in terms of workflow and computational efficiency. Raw video processing in RT went from very inconvenient and inefficient to only modestly so, and the processing speed is much faster, and I no longer have the huge slow down in computer performance observed when processing raw video in the GUI batch queue.