How can I recreate the filmic example described in the 2.6 release notes (darktable 2.6 | darktable)?
I download the PlayRaw image excluding the xmp-file, disable the base curve and adjust the exposure according to the description so there should be no overexposed areas in the sky. But a large overexposed area is still remaining. I can proceed and get an ok result but this requires a big safety factor adjustment. See screen prints below.
I can also remove the overexposed area by a further exposure adjustment, but then the image becomes very dark. What is wrong? It doesn’t help to download and use the xmp-file.
Halløjsa Olaf,
Here is my attempt 
Did you start with the original .arw file
or what?
Med venlig hilsen,
Claes fra Lund i Sverige
Hej Claes
Hyggeligt at høre fra dig igen…ok I will switch to English…
I followed the link in the release dokument:
Let’s take an example of such image (taken from the [pixls.us PlayRaw contest “Backlit”]([PlayRaw] RAW Challenge: Backlit
and downloaded DSC02168.ARW uploaded by PkmX. The red colour is of course the overexposure indicator. Why is the sky still overexposed after having followed the instructions and adjusted the exposure?
Mojn, Olaf!
Yes, I believe that it is better to keep the most parts in English
a) What is your opinion on my swift version (above?). Is it in the right direction or still too blown?
b) I am using dt 2.7.0.
c) I cheated (a little): I killed most of the blown highlights in RawTherapee before I lifted the image into dt, i.e. I cross-processed the image.
d) First thing I did in dt was to use filmic preset 14 EV backlighting.
e) By all means: I am not an expert; I am just a friendly tech experimenter.
Med venlig hilsen,
Claes fra Lund i Sverige
Hi’ Claes
Thank you for a fast response as always.
I’m exploring all the exiting new features in DT 2.6.0 as this is the newest release available (to me) and the release described in the release document.
I think it’s a great idea to explain the new features by examples in such a way that DT users can reproduce the results by following the instructions in the release document, which I failed to do in this case. So no cheating and cross-processing please!
Since my first post I have discovered, that I am (almost) able to reproduce the results presented in the release document by changing my preset for highlight reconstruction from “reconstruct” to “clipping” which is the default. I also assume that the over/under exposure indicator is turned off in the screen prints shown.
Have a nice weekend…:o)!