one, named dummy for mask, is placed before the original instance, and does nothing but define a brightness-based mask, as close to the original image as possible;
the other, which I named contrast boost, I moved to the end of the pipeline. It also does nothing (does not really adjust the exposure); however, it is merged in multiply mode, which increases contrast. That adjustment would be rather extreme, but I tell it to use a raster mask from the dummy for mask instance, so it affects highlights mostly, but those apply multiplier levels close to 1, so they don’t cause a huge contrast boost. If you zoom into the image, you’ll see that the clouds are there.
This boosted saturation too much to my liking, so I reset color balance rgb to use the add basic colorfulness preset.
Yes sorry for not using darktable previously. Here a darktable version with the intent of making it look like your lightroom file. Used filmic rgb, shadows and highlight and saturation
I don’t have the nice local contrast tools of the “regular” softwares, so there was only so much lift i could give to the shadows before the sky went white, and even then I had to introduce a small bit of HSL saturation to keep the sky from going too purple-ish when I tone-curved.
My current metering strategy sometimes allows very high-energy highlights to blow through sensor saturation, and I’ve found the RT recovery algorithm to mitigate those quite nicely.
Here is my take, DT 3.8.
DT is not that good (yet, but WIP for 4.0) at blown highlights recovery.
However for all the rest I dislike the LR rendition, is too “digital” and unnatural.
This was more interesting than what I expected. I played around with Filmic highlight reconstruction and the tone equalizer and started getting some results:
But I couldn’t seem to shake the magenta border between the red and blue sky and the vertical structure in the horizon without impacting the clouds. So I brought down the red and blue brightness in Color Calibration and found this:
3.9.0+25~gc4e00c2c2-dirty
In the end I was not happy with what I got…so I pull out the old watermark trick and warmed up the foreground a bit…just another way around the blown highlights in this one but not a recipe to recover them…
The ‘vertical structure’ is the result of using highlight reconstruction with the method reconstruct color. Creating such artefacts is a known issue with that method.
Was not so much interested what we could get out of dt by using any fancy tricks. So this is no filmic and underexposed to allow best impression about the blown out sections. Using WIP highlights recovery mode in default settings.
