Darktable is my Photoshop (GAS in software)

My raw processing goes in cycles. I spend a few months in darktable, until it all becomes too much, and I check out a commercial alternative. Then that becomes too infuriating, I go back to darktable, and the cycle repeats.

I am currently in a Lightroom cycle. I do this because I want to cut down my editing time, and a quick tuck on the “Shadows” slider is simply faster than than the equivalent set of motions I’d go through in darktable. To say nothing of AI masking.

But when I recently prepared a wall print, I realized I needed to jumped back to darktable. Nice as AI selections are, a properly configured combination of parametric and drawn masks is just way more powerful, as are the multiple fantastic tools for sharpening and contrast that just go so much farther than Lightroom’s “Clarity”. To say nothing of the limitations Lightroom places on the edits of masked layers.

At this point it struck me: Darktable is my Photoshop. Much like how Adobe users use Lightroom for the daily grind, but bring out the Big Guns with Photoshop, I use darktable for the detailed jobs.

Perhaps that is a path to wisdom in my mess of a workflows after all. Use Lightroom or Capture One or DxO for the dailies, and darktable for the important things. And come to think of it, perhaps cull the riff-raff more vigorously, and spend more time on important pictures. That sounds like it’d suit me well, and satisfy my fomo at the same time.

Do you ever get the urge to try other software? I seem to slowly be winnowing off camera-induced GAS, but software is another matter entirely for me. Perhaps a professional affliction as a software developer?


With a combination of some presets and keyboard shortcuts, my per-image editing time is pretty short with darktable. I spend more time deciding what I want than I do achieving it. Commercial options are rather few and far between on Linux but I can’t say I’ve ever really been tempted to try anything else after I found darktable.


I find the line (or point) between what’s done in my raw editor and then my bitmap editor moves based on the image and purpose. The choice of both has changed but not as much.

Three years ago it was RawTherapee and GIMP. Then it became darktable and GIMP. Then surprisingly ART grabbed my attention and mostly replaced darktable. I say surprisingly not because ART is inferior (or superior) or anything else, but because I wasn’t expecting it. I just seemed to “mesh” nicely with ART. Then Affinity had a sale on Photo and I tried it. Built-in panos and stacks, adjustment layers, inpainting (despite not being nearly as powerful as Ps’s content-aware fill nor certainly its AI tools) and other things hooked me.

So now I’m using ART and Photo. But I still eye Lr and Ps at times for other reasons. So far I’ve not jumped.

So yeah, I seesaw around, too.


I did for a while, jumped between C1, darktable and RT. This went on for a few months until the whole scene referred discussions started in this forum. Reading " Darktable 3:RGB or Lab? Which Modules? Help!" was an eye opener and during those few months/years darktable got a lot better in my opinion, which was when it finally clicked and I set on it. D&S also helped since I was obsessed with sharpness, at least when it came to being as sharp as C1 or OOC jpegs.

For other software I’ve jumped a lot, window managers, text editors, terminal emulators, shells, music players, image viewers etc. In the end I now have other things to entertain myself with so I mostly use what allows me to jump start a new system as fast as possible without much configuration. That’s KDE, nvim(with my configs), konsole or kitty, zsh, jellyfin for media(centralized in my home server), my own image viewer, gimp + gmic-qt, logseq, etc. I still want to setup sway but with an nvidia gpu I had some drawbacks, will try it again once I get an intel cpu with an integrated gpu so I can have some sort of multi gpu system(nvidia optimus), hopefully that works.


Funny that you’d mention these. I find it puzzling that I am rather stuck in my ways with these (you can pry Firefox and Emacs from my cold, dead, hands, if you want me to let go), but not with raw development. Despite pouring many an hour into making darktable a cozy home, and automating things beyond reason.

Perhaps it has something to do with photography being about the end product, and the path to get there is somewhat arbitrary. Perhaps I’m merely not good enough of a photographer to have found my way, in contrast to my more rigid stance in software development. Or perhaps I’m just more interested in play in photography, whereas I have to deliver consistent market value when programming. Interesting observations, at any rate, and good food for thought.

1 Like

Me, too. I’m not on Linux for the time being but I’m still always installing and trying things. I like variation, which perversely enough was an aspect of (particularly) old MS-DOS software years ago that was simultaneously frustrating and interesting. There were some really poor UI decisions / paradigms in those days as the evolutionary process plodded “forward”, but there were occasional gems as well (and I don’t mean GEM).

It was always interesting to see how someone approached a problem unencumbered by past solutions. Software now is more consistent and “expected” but generally far less interesting and fun. I guess it’s been commoditized.


I don’t switch much at all.

Back in my Windows days, I used Paint Shop Pro for a long time. I can’t even remember what I may have gone through before that.

Then, switching to Linux, I used The GIMP for a long time. But a few years ago, I started dabbling with darktable, and here I am. I don’t even try other alternatives, because I just focus on doing better with darktable.


I was on the other thread talking about LR etc… saying I was considering buying DXO pureraw but I ended up trying LR’s new denoise and ended up subscribing again. I was looking through all the images I’ve got and there’s many that I don’t want to keep the raw files, I just want a real quick edit. So renting LR for a year or so to get my files thinned down to just my best/favourite images makes sense.

I do like to try all different photography software but do try to spend the most time learning DT. I’ll be able focus better on that when I’ve sorted my images out I feel.

1 Like

I believe photography also has a lot of marketing around the “what if” and a sense that things could be much better across the road. Where as a browser or a text editor/operating system(emacs :stuck_out_tongue:) is more or less stable and fit for you once you get down into it.


It’s funny because there were never as much programmers as there are now, or at least, as much people who code. I guess you could argue a great bit of idiosyncratic software came from educated people who had more creative means to achieve what they wanted, whereas nowadays there’s a lot more people basing their work on eachother, or having less formal education in programming. Web has also thinned things down at least when it comes to GUI’s in my opinion.

There’s also a lot more that’s already done for the programmer. That is, libraries, frameworks, plugins, etc. When you didn’t have a UI framework ready to go, you came up with your own, which reflected your thinking, preferences, approach, etc. For better or worse (often both).

You don’t have to go back too far to reach a point where “if you need X, you write X” was quite often the rule. Not good for standardization, but IMO more programmers knew more about more stuff as a result. Less programming by dragging and dropping, so to speak.

I’m in no way up to speed on development these days, but I have to wonder just how many “average programmers” (whatever that’s become) would know what to do if / when they encountered something like:

asm (  "incl x
        movl x, %eax
        imull y
        movl %eax,x"

Not that I would, but I wonder… :smiley:

1 Like

If x and y are global variables,
that would equal x = y*(x+1).

I pride myself on being able to code the entire stack, from manual memory allocations at the bottom, all the way up to Javascript, interpreted in a website, as part of a web view, in a UI control (which is today’s predominant an ludicrous way of building desktop apps). Not that I usually do much UI dev.

But I draw the line at assembly language. Even in the embedded world, that stuff is just no longer needed, beyond a passing reading comprehension for the gnarliest of debugging sessions.

Overall, however, I feel that complexity levels in programming haven’t changed all that much. It’s just that we replaced our erstwhile biggest headaches (assembly) with higher-level ones (build systems). In turn, we now solve bigger issues, thanks to the giants whose shoulders we stand upon. But we keep pushing against the limits of complexity our tools allow, so effort stays constant.

There’s a thin parallel to the original discussion here. I started this thread out of a struggle with attention allocation. Should I optimize for speed (Lightroom), or double down on skill instead (darktable), or indeed stand up for idealism (FOSS)? I know, I know, first world problems.


You betcha… if you say so! LOL :smiley:

1 Like

Abends, gnädige Hr Doktor!

Should I optimize for speed (Lightroom)

Optimize for speed? Then use vkdt :-)))


Yeah, I’m not a developer (I’ve only written a little C, Perl, etc. long ago on an occasional lightweight basis) but even I can see that HTML was never intended as an application framework. But market pressures and non-programmers shoehorned it into that. Then they added the layers of Java… It’s just like lots of other technologies. I hate most web apps ('cept PIXLS.US of course! :stuck_out_tongue:),

So I take it you’ve not bought many Steve Gibson, Peter Norton or John Socha books, eh? :slight_smile:

1 Like

I admit wouldn’t and I program for a living, but since I never received formal education, I never needed to learn assembly :smiley:

My love of details makes it look interesting,* but it’s a little like building a house from toothpicks. But first you have to build the hammer and nails… :slight_smile:

* I say that while sitting here safely at my desk. But if i were facing 400k lines of ASM at work and was told to document, debug and speed it up, I’d probably change my opinion.

1 Like

I don’t jump around much. Darktable’s masks won me over. Also I love the control from the various modules including RGB color balance and diffuse or sharpen. I don’t get too hung up on the scene referred discussions as I just care about the look, so that is not DT’s selling point to me. I use Microsoft’s Image Composite Editor for panorama stitching of tiff files created in DT. I have pretty much retired my LR program and used GIMP in preference to PS for years. However, DT’s masks mean I rarely need GIMP to edit photos any more.

1 Like

That reminded me of a young man I worked with back in the mid-90s. Our project was switching from a procedural programming paradigm to object-oriented. My corporation brought this man in to coach us in making the switch to a completely different way of thinking. And, of all things, his main job was as a professor of music! He used object-oriented programming in the Smalltalk programming language to compose music and control music synthesizers. He was amazing!

PS to @bastibe - I used to do systems programming of IBM mainframes in assembler language. Those were the days! It was way back in the 80s.