To facilitate learning I’d appreciate if you could briefly summarise, how (which modules, masks etc.) you addressed the individual points above or in general your approach. Of course one can load all XMP files on duplicates and inspect the history, but it is easier with a little text.
This is the RAW + my XMP (again as archive as the uploader sees “zero bytes only”). Archive.zip (4.0 MB)
Thanks @priort, this is the right archive.
I have no idea why direct upload fails for me and I have to use archives. Others seem to have no problem to attach XMP files. Archive.zip (8.7 MB)
Here’s my take. I used:
two instances of tone EQ to tame the dynamic range
default inpaint opposed in highlight recovery which doesn’t do a perfect job on this image but I used…
sigmoid as tonemapper which desaturates the highlights.
Increased saturation in color balance rgb.
using the WB module I took a reading off a white chair. (no chromatic adaptation in color calibration).
color zones to reduce but not remove the blue lighting as I think it adds to the atmosphere, masked to the area with a path.
Diffuse and sharpen to sharpen.
I auto-corrected the vertical perspective but not the horizontal.
default noise reduction (profiled) and lens correction.
DSC_0497_03.NEF.xmp (13.2 KB)
Happy to answer any questions!
Edit - I forgot to say that re different light sources I don’t usually bother much about that - in fact retaining differences is beneficial (in my view I hasten to add!) to atmosphere. I just try to strike a balance that looks good in terms of white balance.
Here is an attempt by me. I am not really very happy with it. I did something that is not recommend. I used fusion in base curve to lift the details out of the shadows. But then I also used filmic to set the whites and blacks. These two modules should not be used together, but it seems to control the contrast. BTW, are the camera settings showing on my system correct? 6" f11 ISO200 was showing.
@123sg your edit has more punch. I had to retire from this edit because my desktop has a graphics card which for some reason doesn’t support Open CL so it was frustratingly slow. I had high hopes for this computer as it was a very reasonable computer when it was new. At work it run a computer resource hungry instrument and the replacement computer was billed at $18K. I was gifted this computer for $50 and then I replaced the graphics card with a NVIDIA Corporation GP108 [GeForce GT 1030]. I thought that would be sufficient to run DT. Evidently not.
I would have thought that would be fine too - maybe you can update drivers? I’n not sure how that works on linux, but I rather think Nvidia do make linux drivers…
Hopefully it can be sorted. Linux is a strange beast to work with when you have become used to working with Windows. I am hoping to get my head around linux as it is a nice operating system to breath life into older computers. Also Mint is a nice OS. I have posted a question about the graphic cards on the forum and hopefully someone can point me in the right direction.
I hope my Windows laptop just needs a new battery. It flashed a warning about going into powersaving mode and just shut down. Off to the computer shop tomorrow.
@123sg I think the most important lesson for me in this thread is your take on perspective correction. The slightly diagonal roof-line you left in your version is much more interesting than the frontal view I created from the RAW. Most of the other posters seem to agree.
@Terry In my initial attempt I did not use filmic at all, as I found it does not bring much improvement. Especially in its default settings right after enabling it the highlights looked terrible and I had to tweak the mask of the reconstruction tab.
It is interesting to see, to which degree the individual versions in this thread are able to reproduce the details of the crystal chandeliers that are next to the blow out lights. Some are more or less a bright blob, while others show some of the finer structures. I still need to go through the histories of the sidecars to see, what is different there.
