I think it is intuitive - as long as you know what you want to do, and as long as youve used it enough to get used to the way the pipeline behaves.
There are some quirks (the style/presets saving the values, instead of the actions - like the auto-adjust bit in tone equalizer), there are some areas that feel like they want to fight you (the masks 
), but other than that…
I have never felt this amount of freedom combined with the amount of things that can be achieved in a raw developer software before.
1 Like
darktable is great but it’s not easy to use learn.
Concerning your initial wishlist:
tone mapping: using one (like default) automatically hides the others so you only ever see the active one. (ofc the old behaviour can be turned on but imo just confusing for most users) and you can find them in search.
With the default workflow using sigmoid this is setup like you are suggesting?
white balance: a clean interface with normal temperature and tint sliders that handles everything in the right place without 2 modules where the obvious to use one shouldn’t be used
Thats one area thats especially hard to learn. @kofa is working on something to make it easier but the technical background of the current solution (it has its reasons) and darktables limitations make that harder then it seems.
I find TE very hard to get right, too. There was a big push some time ago to make TE better but that PR crashed and burned unfortunately… I have high hopes my shadow/highlight-management-needs will be fulfilled by @jandren 's excellent new contrast&texture
selective edits
For one darktables pipeline-structure introduces some limitations as to how masks can be reused by other modules. To properly re-use (and link) masks you have to use raster-masks. For that to work the origin of the mask has to be earlier in the pipeline then it’s “children”.
I like the idea of a dynamic module filter to show modules connected by the same mask. I don’t know how this would fit into the current UI however.
2 Likes
But it also wouldn’t hurt your workflow I think maybe there are even ways to improve your workflow, like selecting part of an image and that’ll be the selected range ofc keeping the ability to ability to adjust the sliders.
We shouldn’t look at only reasons to avoid change we need to look at ways to improve things for everyone
1 Like
I think it is fine to use the term “intuitive” in this case, since within all forms of photography, analog and digital, the concept of white balance being tied to kelvin is universal (e.g. daylight = 5600k). For the vast majority of people who have does any amount of shooting in RAW, they understand the concept of white balance in kelvin values.
The theory/math behind Color Calibration’s “White Balance” is very different in comparison. I’m not saying that Color calibration is not good at what it does, just that it obfuscates an obvious use case. For example, if I want to use a standard daylight kelvin across all the photos in a set, my understanding is that I have to figure out what cct value is closest to 5600k and then put that in. Or, I can do something that seems to go against the intended workflow and disable the CC module and use the White Balance Module, but it’s controls also don’t work well (the tint slider is crazy sensitive.
Now maybe I just don’t understand how the Color Calibration module works… But that seems to be a real design flaw when so many other people are confused or have questions about how the CC module works.
1 Like
That’s exactly why I don’t like Darktable. I’ve been using it for fun for a few years, but after a while its complicated tools always start to annoy me.
I usually use Rawtherapee, which suits my needs perfectly. It’s simple, fast and does a brilliant job of sharpening photos – in that respect, DT doesn’t stand a chance.
3 Likes
You dont have to do that. Just pick “daylight” and dial in the temperature?
2 Likes
Have you even used darktable recently? Capture sharpening was ported from RT and afaik there isn’t any other sharpening algorithm that’s lacking in dt compared to RT
1 Like
But 5600cct doesn’t equal 5600k right?
My 2-cents:
Tone Equalizer is excellent the way it is, because you can separately mask different tone ranges in separate instances and treat them differently. This was demonstrated clearly with examples in some of Boris Hajdukovic’s Youtube videos a few years ago. I don’t do this often, but sometimes it is a life saver.
PS - Here is a great example. I need to do this more often.
DSC00636_01.ARW.xmp (9.4 KB)
3 Likes
If I am honest, my memories of LIghtroom are that it is relatively simple and easy to use. But I abandoned LR6 and settled on DT. It took some learning but it gave so much more creative power. It is fast and simple to use once I have learnt to use it.
Now occasionally I want to do something in Affinity and that program is in my view not intuitive and difficult to use. But I guess that is because my brain is wired for DT’s workflow. So intuitive is a relative term in my view.
DT is also designed for the photographic enthusiast who is willing to learn. There are heaps of subscription programs available for those who want simple or even AI edits.
What’s so hard about Darktable? Nothing to those who want to take the time to master it.
EDIT: Maybe keeping up with all the great new improvements to DT and the new modules every stable release is difficult. I see a new saturation curve model got added to the DT master a few hours ago. DT is anything but stagnant.
4 Likes
I agree that shouldn’t be changed but I don’t think that giving it a good default is bad there will still be the same sliders for adjusting it and maybe we can even improve that workflow making it easier and faster but never sacrificing control
AFAIK those cct numbers are somewhat arbitrary anyway. IMHO its either calibrating to a known neutral grey and/or color grading by eyeball 
.
Am I missing something about the current mask workflow that people seem to have trouble with? In darkroom, on the left side you can open mask manager and create yourself a shape. Or you can use some module to get yourself a shape or a group of shapes. Then in any module in the pipeline you can choose this shape or group under drawn mask (just under opacity). Changes to the shape or group will be linked between all modules that use this shape or group. The only thing missing is being able to form a group inside the mask manager itself.
2 Likes
Parametric masks can be be included in cut&pasted modules, or used in presets that can be applied to other images.
1 Like
Ignore CCT…if you want the rationale read Aureliens WB document on the Ansel site…it makes things clear. Look at the isobars on the standard daylight or plankian curves…several combinations of values will resolve to the same CCT. WIth the CAT you think of an illuminant and correction between the scene and display to preserve neutral and colors…WB is a partial CAT only …they are not exactly the same. Its a different approach… and yes more complex. The WB module is coupled to the CAT computation process but from there its a correction in a perceptual colorspace…Rather than everyone thinking of CC as the new “WB” maybe we need to think of it as the CAT, maybe that is the issue then people wont be trying to make them or use them exactly the same way. You are either choosing to WB oru image or instead you are going to use a chromatic adaptation transform to deal with scene, white, neutrals and colors to perceptually correct them to display conditions…
I think @kofa is working on some things that will help to smooth this out a little…
1 Like
Yes, but they will now hold a new position in the pipe and thus a different input which can invalidate them…raster masks are fixed, parametric masks are not and they are influenced by the pipeline…This is also the major issue with the mask centric module…the pipeline processing and the order.
First it would break all edits most likely for previous versions…Second and maybe I should give this more thought but when you create mask two and add a new or different local contrast instance to that, if the modules are processed in the current pipeline what data is that going to use and what happens if you delete a mask down the line … anyway I think it would require an entirely new processing pipeline to accomodate that and there is a nice rewrite of masking in the works that people can try out that might be to some folks liking that have issues with the current situation
Take a look at this and have a play with it… if you haven’t already…
deleting a mask would obviously delete the effect as well so removing those modules from the pipeline.
and it should just process the modules in the current pipeline i think and allowing to reorder it like it is right now. the only real issue i see is development time and how to deal with parametric masks i think that one solution could potentially be to ignore them altough a proper solution would probably be to have 1 module that acts as a master for it (default probably the lowest one in the pipeline )
i truly think that there’s some solution that doesn’t break previous edits
They are a valuable and frequent part of my workflow.
For instance, some gold ribbon on the stage backdrop at my regular venue. I have an exposure preset to calm it down.
Is it enough just to apply the preset? Sometimes yes, but it’s effect has to be assessed. It often interferes with other things. Sometimes that’s fine, sometimes I use some quick brush strokes to limit its effect.
Sometimes, one of my presets may be nothing more than a starting point, which I adjust or change completely, but then store as a preset which I can re-use during that session.
I didn’t know what CCT was. If I found the correct article on Wikipedia, I have learned the following:
Correlated color temperature (CCT) refers to the “temperature of a Planckian radiator whose perceived color most closely resembles that of a given stimulus at the same brightness and under specified viewing conditions”.
Truth be told, I have learned nothing. 
3 Likes