+0.5 is the current setting and this is correct.
There was a discussion about this. I remember someone mentioning that it should be 0.83 (including a rationale why), but I cannot find it any longer. This discussion might even have been on IRC (#darktable.)
Anyway, maybe Aurélien can give you a bit more details about the why.
Thanks.
I’ve no need of the details. Just checking there wasn’t some kind of error.
The +1 EV matched (more or less) the look of OOC JPEG with high-dynamic compression enabled (on Nikon, it’s called Active D-lighting, or Auto Lighting Optimizer on Canon). This is suitable for landscapes and night scenes.
But then, studio shots (under controlled lighting) were over-brightened compared to the camera base curve (OOC or darktable’s). So, after discussions, we chose to tweak the defaults to match the most basic case (hence +0.5 EV), assuming the people doing HDR-scenes will most likely be technical enough to fix the brightness themselves, while the people shooting with basic settings are the ones who most likely need their hand held.
That’s odd, and 0.83 is very specific. I would imagine the exposure factor would depend on the camera and its jpeg settings (and how they fudge iso values)
Ah, aurelienpierre beat me to it. That explanation makes sense 
Gave my search for this 0.83 that’s in my head something extra and found it again:
The last paragraph:
That means that for a shot metered by an average camera (as of 2014), to get 100
%reflectance at 100%in the RAW and 18%reflectance at 18 %, we need to apply +0.83 EV to the RAW values.
As always, Aurélien’s above reply is spot-on.
Note that Fuji owners will have to add an additional +.7 EV, according to
So instead of +.5 EV (which is the present value automatically added), Fuji owners are recommended to use +1.2 EV.
Have fun!
Claes in Lund, Sweden
With my Sony A7Riv, I find 0.5 EV is not quite enough for most of my images, but it is trivial to set the exposure slider to whatever default value you want (I find +1EV seems to work well with the Sony), and then on the exposure module select “create preset”, give it a name like “Sony default” and check “auto apply”. You can tell it to apply that new default only for raw pictures shot by your specific camera model, or tell it to apply that default blindly to any picture. As long as my picture was well-exposed in camera, this often gives an excellent starting point. often only requiring some minor tweaks to develop the image. It really doesn’t get any easier than that 
The measurements from DxOMark seem to confirm that +1 EV is roughly the correct value for the α7R IV, except if you meter at ISO 50 (and shoot at 50 or 100), in which case no adjustment is needed.
For the past 2 years, I have held when users complained about filmic being more complex than base curve and such. My politics is first to do the right thing at the right time in the pipeline, then make the right thing easier through the UI and workflow, rather than indulging into user’s bad habits and nasty shortcuts.
It’s good to see users now realizing that rigor pays on the long run, because the step-by-step decomposed workflow (exposure / color balance / filmic) is easier to scale to whatever dynamic range than the all-in-one bundle that base curve and the like are.
It only took several iterations to bend the whole workflow around that logic to make it as usable as the previous nasty tricks. We are getting there.
The mapping graph looks great.
Awesome! These UI-improvements are just what filmic needs!
Just after looking at pics in the PR (especially the legends in the “look only”, I feel like I understand the filmic controls much better. The “Dynamic range mapping” view also nicely visualizes what’s happening.
(disclaimer: while I have read Aurelien’s posts on filmic here in the forum and watched some of his shorter YouTube videos, I haven’t (yet) watched the longer ones - at least not completely)
I’d try putting the mapping graph the other way up: with the scene at the top and the display at the bottom. I tend to think of data flowing down the page, so the current graph looks like the inverse of the filmic transform.
But the pipe runs from bottom to top so input at the bottom and output at the top is the correct way for darktable
Depends on your perspective. Darktable takes a raw image as a base foundation, then it applies on top of that various filters, each building on what the filters preceding it have done. It is like a tower of building blocks, where each additional processing step is applied on top of the previous steps
I am a Fuji XT2 user and +1.2EV is a bit heavy. Not sure if Aurélien hasn’t ment +0.7EV for Fuji instead of +1.0EV for all other brands. I prefer +0.7 to +1.0 EV depending on the image.
Do we have other Fuji users here?
Yes, he explicitly stated 0.5 + 0.7 = 1.2 for Fuji users.
But of course you are free to use whatever values you prefer.
Ok, you are maybe right, I interpreted it differently.
@anon41087856, can you please confirm +1.2EV for Fuji as “default”
Thx in advance
daddy confirms. 