Is darktable (or any photo soft) ever too complicated ?

I think I need to get that topic covered once and for all because I’m tired of repeating.

About learning an art

First of all, let’s be done with the so-called subjectivity of art (provided that, if everything was a subjective matter of taste or opinions, we wouldn’t even be able to communicate at all – so we need absolute definitions, even only arbitrary and locally valid ones).

Art is what happens when you stack personal expression on top of craftsmanship. Craftsmanship is the practical ability to manipulate some medium to shape it into something that matches a design. Then, personal expression is about creating that design to express something that seemed important to the artist at the time (whether it’s formulated at a conscious level or at an intuitive level, whatever).

Whether art should be original or innovative is not the topic of the current discussion, and doesn’t seem very important to me (how original is it possible to be anyway ?). Also bad art is still art, as bad/good is a property of the thing, and not its nature. And I don’t write art with a capital A as some people do, because it’s not a sacred or superior thing. It’s just stuff that people produce. Sometimes it’s good, sometimes it’s too expansive, whatever.

Learning art is therefore the combination of 2 things : mastering the medium, and learning how to open your creativity channel (which, given the current state of public education, is something you need to actively re-learn at adult age because that gets lost somewhere between 7 and 13 for the sake of conformation and academic success).

As it happens in this digital era, many arts have now converged to a digital practice involving a computer and, inevitably, some kind of software to interact with the medium. Computer science was mostly designed to handle data automatically, using it for artistic purposes is rather new and weird. But once you have created a digital representation of a physical medium, it’s only a matter of transporting the physical interactions you are used to in the physical world to a virtual space. (And here, you see why getting physically-accurate image processing matters to me : light should behave the same inside or outside the computer, and the craftsmanship is in how you shape it, not in how it behaves, so the artistic thing is no license to mess-up with light transport).

But don’t get mistaken, digital or analog, the craftsmanship is still about shaping a medium, being it a material or a virtual one. Any software you are using is merely a control panel over this medium that exists in the matrix. Photographers shape light, painters shape paint, etc.

How I learned art

The good thing about being a kid is you get sent to school to learn and you get no saying in that, so you shut the f*ck up and do your homework, or else… When people grow up, they get all kinds of grown-up excuses to avoid learning ever again (kids, taxes, overtime, old age, bad eyesight, etc.), but somehow they always manage to find time for the TV (explain that to me…).

Between 6 and 18, I had music lessons. When I was at the conservatory, we had 30 min of piano each week, but 45 min mandatory of choir practice (or orchestra for non-soloist instruments) and 2 h of solfeggio per week. At the end, at the conservatory, I was doing a solid 2h/day of piano (I was still in regular school on the day), which was a bit crazy since the music teachers kept asking for more while I was already at school 7h/day. I was 14, I got high scores in physics and piano, I wanted to pursue both but apparently, in this sad world, living a balanced life is a luxury reserved for Instagram, so it was a 3-way choice between music, regular academics or keep working more than 10h/day + transportation (until you forget why you are here and probably rage-quit anyway). Long story short, I continued music but out of this crazy place, with a teacher who happened to be burnt out from international half-rigged piano competitions.

Music started to be fun after 8 years. That’s many thousands hours of pain before it starts to pay off. Beside patience, it taught me something that would be very important in photography : the sense of nuance. Nuances are indications written on the score that tell you how loud (to simplify) you should play. But still, those “mezzo-forte” or “sforzando” are pretty much up to your own interpretation (and up to the health status of your piano), and the important thing is not in the absolute loudness but rather in the contrast you create between the loudest and less loud part, to bring life and spirit to a score. Contrast… Doesn’t that ring a bell ?

This is all very nice, except it’s perfectly useless if you don’t train everyday to force muscle memory to become reflexes in order to make very unnatural gestures feel intuitive, until you don’t even think about how you should put your fingers on the keyboard. As it happens, the only natural thing in one’s life is death and disease. Pretty much everything else needs to be learned and practiced, in a perpetual fight against your own mediocrity (and I’m only a musician, but ask dancers…). “Nature” is a misplaced feeling of easiness and obviousness, the proof beeing you will always find someone completely disabled at anything you find natural (not saying here that disability is a permanent state and should be the end of trying).

But that nuance thing is, I think, all what art is about. You have a message to convey, you need to make it clear enough to be understood or felt, but also avoid making it caricatural, otherwise people will start to only see the effects and stop caring about the content and its meaning. It’s not something you can learn from only pushing sliders in a GUI, you have too feel it. How many times have I heard “that musician is so good, such a virtuoso” while true virtuosity is the one you can’t hear (if you hear it, it means the guy is actually suffering). Letting the effect take precedence over the content is the issue I have with many artists, and also the side effect of a lot of image processing algorithms.

Art vs. software

Manipulating a non-digital medium makes it pretty obvious that you need skills and training to bend the matter to your wish. A piano, although enabled by a complex machinery, is nothing more than some felt hammers hitting a couple of strings. And yet, depending on how fast, deep and heavily you weigh on the key, the same string won’t sound the same at all. In fact, with some training, you can even identify pianists just by the way they sound. What is so great about acoustic instruments is they have a very large expressive palette, with a lot of nuances available.

Yamaha and Roland have very complex ways to try to recreate that palette in synthesizers, Yamaha using real-life sampling recorded on analog pianos, and Roland using a purely mathematical approach to create sound from scratch, with physical models. Their business is basically to give to the players the same level of richness and complexity they would expect from the real thing, by recreating them in digital. If you think about a digital piano, it’s pretty much a keyboard full of sensors which creates a lot of inputs (position, speed and acceleration of the keys) that will be treated by a software to match that inputs to a plausible sound. And even if your keyboard is not full of options, there are lots of ways to play it that should all be accounted for in the software.

Imaging software are different in the way they don’t have many sensors to record an user interaction, but only simple switches (mouse + dumb keyboard). But still, the medium they control is very complex, and there are also lots of ways to “play” it that should be accounted for. So, using more simple keys, they simply need more keys to allow the same amount of control.

When I look at current commercial photo editors, they don’t look like instruments to me, but rather like these auto-synthesizers that take a MIDI-encoded score in input and “play” it automatically and mechanically at the end. So you are not the musician in there, you are the listener of some arrangements the beat-box is doing on the score you feed it. And even if the interpretation is improved, and sounds more organic, that’s still not you who is playing.

Software vs. art

A digital medium is buried so deep under a graphical interface that you could forget it’s even there, and be mislead into thinking that the computer is doing most, if not all, the job for you. Plus, with Kodak having worked since the Bull’s eye (1895) to make photography accessible to the mass market (I insist on market, it had nothing to do with democratizing art), people kind of expect photography in general to be a simple button-pushing experience followed with immediate reward.

And I fear that 20+ years of GUIs and personal computers have made people computer-dumb, in the sense that they don’t know and don’t even care about what lives beneath the GUI, and what data they are actually manipulating. Then comes the silly analogy with driving a car without having to be a mechanic (or an engineer), as an excuse for not caring about what happens and how it’s made, as if having a sense of what is done was the same as knowing how to build it from scratch…

But remember… the medium we are manipulating, although virtual, is still only a medium, consistent with some analog reality, and the virtual medium is just as complex as the analog reality it represents (at least because you expect it to behave the same, and also because you don’t want to lose some bits). So the need for craftsmanship in the digital era hasn’t changed one inch since the analog era. Your image pixels represent a light emission. Playing with pixels should be the same as playing directly with light. Developing a raw should be the same as printing a film on paper. All that falls-back to exposure settings, either global or selective. The fun is, with digital, you get to build your own film stock from scratch, not being limited to film vendors aesthetics anymore. And you, my artistic friend, deserve to control your art, hence your medium, not to be cast away of your own work by an “intelligent” software that gives you instant gratification, because you are the artist here and not the spectator of a juke-box.

But how do you bring nuance into the control of something complex ? There are not many solutions, aren’t they ? There is a catch somewhere if your software looks simple while the medium you are working on is not : you either get a control reduced to dumb stuff, or some things are done behind your GUI without asking for your permission.

Simplicity vs. simplicity

A simple software interface doesn’t imply automatically that it will be simple to get a good result. Actually, it’s often the opposite. 2 years ago, I got the chance to “pilot” an Airbus A320 in a full-featured flight simulator (the kind used for pilots training, suspended on actuators, and tilted accordingly to simulate acceleration), for fun, since I was working for the manufacturer of that simulator. A commercial aircraft dashboard is… intimidating, to say the least. But the beauty is everything is directly under the tips of your fingers, directly accessible with dials, knobs and screens.

Setting up the autopilot (altitude, target airport and radio frequency) was much easier and faster on an Airbus A320 dash than setting up a destination in a Tomtom car GPS, where you have to navigate through several menus, wait for the laggy screen to respond each time, miss the confirmation button on the touch screen, retry, and such. But that autopilot was surrounded by dozens of brothers and sisters, and I sure needed guidance to find it.

So the relationship between the number of controls in an interface, and the easiness of getting a good and fast result is a very counter-intuitive thing. Sure, you need to learn what all these options mean and do. But once you learned, it’s quite difficult to live without them. Nobody ever complained that there are 104 keys on a keyboard, didn’t they ? On the contrary, having to choose between key combinations and more native keys to get things done, the choice is pretty easy (until your keyboard takes all your desktop, but that’s another issue). Therefore, simplifying is not always making things more simple. Quite often, it makes them more awful, since a GUI is merely a control for something underneath, and controls are there for a reason. Simplifying the GUI would actually suppose to simplify the underlying algorithms first, which is not always possible and quite often not desirable. But people always focus on the tip of the iceberg, don’t they ?

The sensible thing to do is rather to make the complexity more understandable, and more logical, instead of denying and hiding it. But again, “logical” is assuming some background from the user. And that’s where the concept of intuitivity crashes. Intuitive is the property of something you don’t know to resembles something you already know, so you can reapply the same paradigm and spare the learning. Hitting the Escape key to end something is quite intuitive to me, because every OS I have ever used since Windows 95 worked like that. But intuitivity is deeply grounded into your cognitive referential, which depends on your culture, education, experience, OS of choice, and such.

In image processing, “obvious” things can be very unintuitive. For example people complain a lot about filmic desaturating the highlights near white (Y = 100%), making them lose their blue skies, because they just don’t realize white is supposed to be… achromatic, hence have no saturation, although that’s the very definition of white. So that thing that every other software lets you do, where you end-up with highly saturated highlights that will produce gamut clipping later (whatever colour space you are using, by the way), is just a nasty stabbing in the back that degenerated into bad habits for a whole generation of photographers.


One could object there is a distinction between complex and complicated, but let’s avoid word play here : no software will ever be complicated enough to represent accurately the medium it manipulates in a way that allows as much nuance as the physical world allows. This has nothing to do with willingly obfuscating things for the fun of it (it’s really not fun, by the way), it has to do with acknowledging the complexity of the reality in which we leave, embracing it, and giving options to users to control it so artists can actually be the masters of their own work.

Will it be difficult to master digital arts ? Not as much as it is to master analog arts. Remember the analog world doesn’t have Ctrl+Z. It took me 2 months to learn the C programming language to start hacking darktable, almost 5 years to learn photography from theory to practice, but it took me bloody 12 years to learn how to play the piano. There are shortcuts in C, but not in art. You have to suffer through all the necessary bits to learn your trade, there is no other way around. But anyway, what’s the rush ? Some things just take time, and you need to accept that.

So, if you are not up to the task, there are still iPhones and OOC JPEG. No choice is wrong, you just have to live up to your ambitions. I just don’t want to read people complaining about having to go though 12h of darktable video tutorials before being able to use it, because that 12h are bloody worth 5 years of my life, so you actually saved 14588 hours instead (full disclosure : I learned darktable with Lightroom tutorials, around 2012).

EDIT: about beginner-friendly software

In the photography hobbyists world, there is this belief and expectation that everything should be built for and around beginners. Many websites, Youtube channels and publications are just about that : teaching beginners.

While everyone was a beginner at some point, the fact is “beginner” is, by definition, a transient state that is meant to be overcome. So, the fundamental fallacy of this is we can’t design a software as a tutorial to get you started during your first 6-12 months, when this is supposed to be a full-fledged work horse for actual use. Designing something for beginners is a nonsense, how do you deal with users as soon as they are not beginners anymore ?

As it happens, if you search the internet for “advanced photographers” education or even “expert” level one (yes, continuous education is a thing even for experts), there is zero ressource. Isn’t it a clue ?

EDIT 2: about photo-software for geeks

Geeks are a valid audience too, who deserves to have his needs fulfilled too. Disregarding a software because it’s “made for geeks” is just as stupid as disregarding a software because it’s “made for common folks” (whatever “common folks” even mean). Ansel Adams was a huge geek, who trained as a pianist (another one), and later invented a very clever method to handle HDR scenes with film. His work now sits in the MoMA. So… Geek is not the opposite of artist, it’s just another flavor thereof.

As a matter of fact, Adams compared the film negative to a music sheet, and the print as the interpretation, which is very much how I see the digital RAW development, and why I compare darktable to a music instrument and Lightroom as a juke-box.


Excellent post. Should be pinned!


I am a bit baffled. Are you angry about something?

This could be something that gets lost in translation (from a non native speaker) either on the sender side, the receiver side (meaning myself, I also am not native english speaker), or both.

I am asking, hopefully in a polite way, because a lot of the phrasing seems rather harsh or dare I even say, too harsh?!

1 Like

Not sure if being angry or not changes anything to the validity of the point, but what is it you find harsh ? I’m not used to sugarcoat stuff, but I don’t think I have been offensive. Sometimes, the best way to advise people to pull their fingers out of their arses is by saying just that…


Although I too am not a native English speaker, I don’t see any harshness in the article. None at all. Quite the opposite! It’s full of kind words and encouragement (at least for me) and is great at setting expectations.

I’ve never used “pull finger out of arse” but rather “head out of arse” :wink: Idiom aside I fully agree with you. I do believe it’s necessary to do so. Some might not take it lightly or even rebel to it, but I still believe it’s necessary.


I think it’s perfectly fine to say things not sugarcoated.

But your headline is a rethorical question that you answer yourself ultimately with a no after a rather long elaboration that, in my mind, misses the point of your ‘frustration’. It’s not a question for an open discussion or at least it doesn’t seem like it for me. (It’s not really a no, it’s a ‘click the EULA or don’t use it’-black or white choice)

Let me demonstrate by giving that last paragraph of yours a different spin. I think then it becomes obvious why to me it seems like you’re ranting (for me at least it sounds a bit like ranting).

‘‘So, if you are not up to the task of teaching people, don’t. No choice is wrong, you just have to live up to your ambitions. I just don’t want to read developers complaining about how users don’t get their UI, before being able to think about designing a good UI, because that 2h of fiddling around with parameters is ultimately wasting millions of manhours on the userside. I can’t imagine the amout of time being saved by a good UI implementation when I started.’’

I think darktable is great. I think your contributions to darktable are great. But that doesn’t mean that darktable is perfect. This is not about simplicity/complexity, darktable IS complex and always will be. To me the question is: how to deal with users who are MUCH MUCH MUCH lazier than you and me. My best answer is: take this real life measurement of yours that people keep coming back not understanding how stuff works, and try to make darktable (or just your modules) better. Not just the algorithms, but maybe also the userinterface or even the whole userexperience. dt3.0 is such a leap forward!!!..just improve on whatever feedback you get. done. and you do, that’s why you made videos. not everyone will watch those, not everyone will know they exist, not everyone will understand them. Is it okay for those people to still want to use darktable? yes! The more selfexplanatory the software gets, the less you have to deal with those people.

now, before someone misunderstands: I don’t want parameters to be hidden. I like them. I also think not everything can be done with UI/UX, so good resources need to exist. Even 12h long youtube videos! I personally love getting into the details how I play my software so I can get the best result out of it. But I also see that some instruments (often FLOSS instruments) are needlessly hard to handle while they sound great. What you see is that people want to play the instrument but don’t understand why it’s not 12 keys per octave but 48, and sometimes they are arranged like this and sometimes like that.

Did I sugercoat this too much? :wink:

1 Like

I completely agree with this statement! I don’t like to advertise here, but I have in some ways touched on this topic in my last episode about darktable, at the end of the video:


I think you miss the point here. I have said nothing about how the current dt’s UI is right now. In fact the whole point of me learning C to hack darktable was because of the poor workflow it got me at that time, with all the unpredictable colour shifts. Plus the GUI is full of rough edges for lack of developer time to finish the product in its details.

This post is about choosing some photo editing software for its control power while, at the same time, expecting magic right out of the box without opening a manual. Plus the annoying default assumption that photography is/should be easy.

To every photographer out there, I just would like to ask this : did you choose photography because you were too lazy to learn how to draw ?

Very often, I feel like photography is the entry-level art picked as a fall-back because everything else felt too difficult. Which has consequences in terms of user expectations at the software level, because then, software should be easy as well. But there is really no ground for that, except maybe Kodak business, which is one very particular market (even if probably the most massive and lucrative one). And if anyone has clever ideas to make UI easier without hiding thinks or degrading quality, I would be very curious to hear them, but unfortunately, as of now, it’s all just removing things with no replacements (also, I’m very curious how we will control algo once every clutter will have been removed… AI is not on the table).

And there, there are political decisions to be made : we can’t make everyone happy at once. So here is only my harsh political position on the matter.

darktable will be self-explanatory the same day quantum physics will be. Until then, we have universities.


“Undisputed opinion is scary”.

1 Like

That’s good actually! This means I misunderstood you.

But from what do you deduct this? Don’t people choose dt not because finally they have 500 parameters exposed but because as a tool it’s algorithms can do stuff that other software can’t? And who expects magic? That is a rethorical question from me, pointing at what is to me a strawman argument.

But people wanting ‘easier’ is not the same as ‘easy’, one is relative, the other is absolute.

Well, someone with such an attitude, even if he could master dt, still would not be what I would call an artist. But that’s not a problem, is it? That’s just how it is.

Again, ‘easy’ vs. ‘easier’.

Yes, this discussion is to be had, it will be a long discussion, it will be difficult, but this is not complicated to solve! I remember a thread which could be resurrected for this.

I agree! But the danger with this sentence is that it can be used against all criticism, it’s a razorblade.

I didn’t say self-explanatory, I said more self explanatory. You want to learn quantum physics with a good book, or with a bad book? The differences in pain and suffering are VAST and there is no bonus for learning it with a bad book.

1 Like

Once you start talking to people about their tooling, you come to find the answer is “a lot of people.” A lot of people want a tool with one slider, a la lightroom’s clarity slider.


I did analogue photography for 30 years, took a break until digital hardware caught up, then did digital photography for 10 years.

In the bad old days, making a mask for unsharp masking took me 30 minutes (expose, develop, stop, fix, dry). Then a few more minutes to make a print. Bother, it’s not quite right. Make another mask … And that was just black-and-white. I never attempted USM in colour.

Digital is a different ballgame. When I tweak the USM slider, I complain if it takes more than a couple of seconds. If it takes me 30 minutes to figure out and explain how to do a composite animation, I’m going slowly. Oh, and that includes publishing to the entire world.

The bad old days was mostly craft. Making a decent B&W print took a certain skill, knowledge and experience. The kids today … instant gratification … never had it so good … blah blah.

At their hearts, analogue and digital photography are the same. We make pictures with light. But the technology has changed, and keeps changing, more quickly than we can keep up or write the documentation. And sometimes it isn’t perfect, perhaps because someone hasn’t yet found the perfect solution to a problem, and we want that problem solved today. If not yesterday. C’est la vie.


By definition, RAW development does not has to be one-click easy. Cameras have JPG output for that.


There’s a famous talk, “Simple Made Easy” that talks about the distinction between simple interfaces (which expose their parameters without fuss) and easy interfaces (which require little learning).

It uses the archaic term “complected” (as in “braided”, “convoluted”) to describe easy processes, as they hide their complexity and are therefore harder to actually understand. Contrast that to “complex”–the Airbus cockpit is complex, but not complected–it is very clear what each button and dial does.

I like Darktable for that. It’s complex, but simple. Whereas Lightroom always does more than one thing per slider, with color shifts and halos where there shouldn’t be any. Sure, it’s “easier” (it has fewer sliders), but it confuses the heck out of me. So there.


A good read and some good points are made:

Rather, he wants them to realize that by understanding the components that actually create different looks, artists can become “authors” of their looks, instead of “shoppers” picking from limited off-the-shelf options.

He admits that this is a difficult process for anyone who either doesn’t have extensive technical knowledge, or access to a post-production house with people that do. But he believes that even just changing our mental model is a start. Instead of seeing cameras as paintbrushes, he wants them to be seen as measuring tools that can be shaped into anything with proper display prep


It’s nice to have some philosophy, rather than disputing which version of a certain parameter is appropriate :slight_smile:

Several years ago, I wrote ground-up code for the fuel-injection system on my motorcycle. There was existing code for the hardware, but it originated from American V8 cars. It was compromised. So I started again. Along the way, I learnt that a lot of what was said to be necessary was not… it was simply that people didn’t understand the system, they were trying to adapt things that had been written by other people who didn’t understand the system, and they looked at OEM Japanese manufacturer code, which is enormous. I also learnt that when a chip manufacturer gives away the compiler for the modified C that is ported to their µ-processor, it’s not because it works perfectly. Even the bits of assembler that are certified as “verified” clearly were not. After pushing the bike home a few times, I finally got it all working.

So by the way of that story, I’m trying to say that there is an aesthetic side to programming, which can manifest in very efficient code. On the other hand, that aesthetic may depend on the user’s deep understanding of how the code works, how fuel injection works, how a motorcycle engine works. Some of that is objective, some is user specific, some is an interaction of the two: I may give more weight to smooth throttle response, another may want a sharper reaction or simply more power.

darktable suffers from trying to please everyone. I don’t expect to ever touch velvia, I consider detailed masking to be an emergency fix for a bad photograph, and I don’t like the tone equalizer, because it seems to be telling me what is a “correct” way to improve a photo. I avoid sharpening except at the minimal level of tuning the underlying frequency response around the Nyquist frequencies (of the different rgb channels). I work 99% in B&W. I add weird kinks in the tone cure to create globally unrealistic lighting despite my dislike of (local) masking.

But this is all very subjective. I may change my opinion. I have a deep knowledge of FFT sampling and less understanding of wavelets. My knowledge of colour theory is patchy as hell. It’s entirely possible that changes in my technical competence will modify my subjective response to a certain type of image… it all comes back to emotional engagement, and that isn’t independent of emotional engagement in the technology.

We’d all like a software tool that reflects our own habits of thought… even if they are wrong!

A quote from André Lichnérowitz borrowed from a physics text:
“On a besoin que le mathématique devienne un instrument de pensée”
But we don’t have the luxury of setting the bar that high for users of software.
A quote from Nellie Melba, a soprano of great repute from the end of the 19th century:
“Give 'em muck!”
Presumably, we find our individual truth in between the two.

Maybe the answer comes back to music, or photos: do we get more pleasure from playing alone, from playing with a few friends, or the possibility of performing in front of a concert hall? From making photos, from the appreciation of a few friends, or being a “famous photographer”?

In the end, we all do what we want to do…


Recent film and photography suffer imho from the designed look as much as from the off the shelf look. It’s natural I guess but there’s a self conciousness and lack of immediacy that has come with the over engineering of looks. I guess scripts suffer the same. The film can be beautiful but theres often this layer of “look” to obviously sitting over the film. I can barely watch another orange teal movie.

I find it interesting that this is less of an issue with older films despite the amount of effort spent on “look” even in the 70’s. Perhaps I’m objecting to craftsmanship reaching levels so high that it becomes a barrier. It becomes overly self aware. Contemporary movies captured on analogue suffer this as much as digital due to the processes involved in making the movie.

I know though that someone should be able to engineer a look that feels immediate and not designed. I know that the error probably lies with the people designing the looks rather than the looks themselves but at the same time I feel the level of design ability and skill is a huge hurdle in itself.

The above is not philosophical. About purity. Or some such. It’s that the results are not great i my subjective judgement.

I think the big difference with analogue art/craft is the breadth of resources you can call upon in the learning journey and the type of learning required. There are books, courses, tutors that can take you from being a relative novice to playing Beethoven, painting landscapes etc. Most people understand that learning the analogue arts is hard but they can see the path in front of them and it’s often one that’s as suitable for four-year-olds (at least to start off with) as it is for adults.

The digital learning curve is harder because it always, to some extent, is software-dependent, and the learning material varies significantly in quality. To understand it in the depth that you do and want others to (no criticism here - I agree) requires a lot of fundamentals to be in place, some of which can’t be fully appreciated until you’ve progressed a fair way through school (and further). And some of which (the sciency-mathsy bits) haven’t previously been part of the traditional analogue art/craft world making it difficult for people to move from that world to this.

Things like Photoshop/Lightroom, with their single sliders hiding lots of complexity are much closer, I suspect (having never delved into the analogue or Adobe photography worlds) to the sort of things that were done in analogue photography (waving bits of paper about over enlargers and the like). For example, you can dodge/burn but doing so without halos is hard because it’s imprecise and you don’t always have complete control. darktable is a whole different (in my opinion, better) approach that forces you to learn. Learning the tool helps you understand the theory; learning the theory helps you understand the tool.

Personally, I’m somewhat of a magpie - flitting to the next shiny subject to learn before I’ve fully mastered the last - and photography is the first ‘hobby’ that has held my attention for a significant period of time. The sheer breadth of the subject has kept me fascinated for years - the mix of maths, physics, art, psychology, computing, and just getting out and experiencing the world in a different way. And I’m still nowhere near close to understanding it to the extent I want to. I very much consider myself an amateur.

I like that this is hard and I like that it is technical. It forces me to understand what’s going on under the hood. And the way darktable is put together is a big part of why I use it and haven’t moved on to another application - because I can use it to learn some of the physics and that learning makes my craft better in a way I can understand.


There are good resources (covering the theory and practice) for digital cinema professionals (especially for colourists), but their content never made it to photography it seems. That’s one of the things that baffle me, to be honest: I don’t get how two sister disciplines (cinema/photography) are so far away in terms of culture and education.

You could easily learn colour theory in painting books, then look at Youtube painting sessions to see how painters build a picture from scratch (what colours they use, how they build their palette). Then learn about colour timing in cinema books. Then, looking back at old darkroom technics helps giving context to modern light and colour processing, at least it becomes more real.

Just caring about photography and nothing else, because what you do is photography, is doomed I think.

The maths bits are my personal added value, but you can honestly make without them and there is no reason to get impressed. It’s useful for those who want to run the extra mile and go in-depth, but it’s not mandatory. Just like you don’t need to solve Schrödinger’s equations to get a sense of what an atom is. I think people make it sound worse than it is. Understanding how things interact with each other doesn’t necessarily imply you should know if they follow a log or an exponential rule.