Open source and professional photography : lies and wishes

There is one thing you will find on the home page of pretty much any open source (call it libre or free if you will, those lines are blurred) image editing software : the promise that it is, somehow, suitable for professionals. Marketing has abused that word for decades, it is only natural that it should affect non-commercial and non-profit projects as well, just to try to buy some cheap credibility.

The meaning of “professional” is a complicated matter because it plays – on purpose — on two different levels that it tries to blur:

• the lucrative aspect, meaning working for clients to produce turnover, which has to pay for wages, expenses and taxes,
• the skillful aspect, meaning mastership of a craft and quality of the result.

Since we all know that one amateur that is far more skilled than this particular professional, let the mastership aspect rest and focus on the lucrative aspect. This is taboo in some countries, since it is such an artistic endeavour, and it’s not well regarded everywhere to mix money with art or passion (hello, France). Apparently, true artists should be starving and suffering.

The essential characteristic of a professional is therefore to own a business that needs to generate profits to stay sustainable. And since the amount of money needed each month to pay wages is pretty much set up by housing fees and cost of life, and since the amount of workable hours is pretty much capped at 40 h/week ¹, there is this thing called “efficiency” that comes into play : being able to do enough work in a limited amount of time as to reach hourly market prices while still making enough at the end of the month. But the photography business has been seriously hit over the past decade and this is started to become very difficult, because market prices have dropped a lot:

• Newspapers have flushed their photo-reporter staff to publish “readers contributions” or hiring freelance photographers (and therefore cutting on social insurances, taxes and camera gear),
• Magazines have made an habit to publish unpaid pictures (playing on the vanity of amateurs photographers “buying” this way the title of “published photographer” and not needing the income — hello Vogue Italia),
• Stock archives platforms are promoting competition between photographers from Western countries (who have to pay Western taxes and insurances) and from emerging countries where a couple of dollars/euros is already a significant amount,
• Customers who now own smartphones doing pictures (optically) as good as professional images and don’t really see the difference between a well-framed and well-lit picture, and what their iPhone produces…

Lots of photographers I know are now working 2 jobs or more, and/or working for shady businesses like “massages” parlours and escorts services. In the wedding industry (which a lot of photographers avoid for the amount of stress it creates), the requirements have simply grown unreasonable (hundreds of edited pictures in short amounts of times) and the general knowledge of new “AI” tools among clients creates more and more unrealistic demands, since those pretty tools have only marginally improved efficiency (but they sure are spectacular). Some photographers have simply sold all their equipment and moved on to another life.

But that was even before COVID lockdowns. Since then, lots of commercial photographers have had to give up their studio because of rents and taxes to pay while not being allowed to work. Studios are not the only recurring expense commercial photographers incur, since Adobe Cloud and most software have gone subscription-based.

So, it pretty much sucks to be a photographer right now, and it’s not going to be any better with time. But we still need photographers, at least on the news reporting side, but also for books, calendars, magazines, events, posters, advertising and then… well, art. It’s just that we don’t think they are worth the money they need to keep working.

You might think that the libre/free/open source world would care, since it is all about joining tech with political matters like freedom, privacy or fair/affordable access to knowledge and education. Think again.

See, the problem of open source photography is it is moved by and for IT-engineering-tech bros who buy expensive gear out of their salary and only for fun. Those people enjoy amateur photography : they don’t have to deliver results, so can just afford to trash the pictures if nothing good comes out, work small batches and take breaks whenever they want, because they have little to no time constraints. They also have an above-average tolerance to GUI glitches and usability quirks, as long as it’s open source, because they can typically work their way through a command-line terminal. It is also a surprisingly homogeneous social group of 50-something upper-middle-class white men.

But the IT-engineering-tech bro is not just disconnected from the utter shite in which commercial photographers are right now, he is actually all smug and condescending about all of those who need to make money from what he considers a hobby. The idea that, if your hobby is someone else’s job, then you will never be really good at it doesn’t even seem to brush them either. “I’m just happy to not have to rely on photography to make a living” is what you will hear them say.

In this context, Lightroom and the likes are considered with scorn as under-software for the muggle, while Linux and free/libre/open source (FLOSS) is for the real wizards. While there is some truth in that, and I’m the first one to acknowledge that Adobe’s approach on colour is outdated and yields bad results (or that Windows/Mac users are too easily deterred when things get technical), let’s not get carried away and examine what FLOSS image processing apps can achieve before we call every Lightroom’s user a stupid button pusher.

FLOSS image processing sucks on the matter of productivity. It goes on the workflow and usability side, as much as on the computational efficiency and runtimes, with app using CPU pipelines and GUI toolkits out of the 1990-2000’s. The main reason is there are close to no professional photographers on Github, where the FLOSS development happens, or in IRC chats, where the decisions are taken. And if you have no idea what an IRC chat is or how to join one, well, that’s my point. So, who is able to say what commercial photographers need to be productive ? Nobody. But who wants to know ? Same nobody. Why ? Because commercial photographers are stupid push-button users of Lightroom and that’s all they want to use.

See the problem ? This loops into a self-fulfilling prophecy where, however you put it, commercial photographers are not welcome.

But why should FLOSS, which is produced for free out of the free time of the 2% of the world population’s who can code, care about commercial photographers anyway ?

Well, you know how car brakes are tested at high speed and tires on wet roads ? Or how bridges are designed for the worst traffic conditions ? Design always has to account for the most demanding use case. Otherwise, you just design systems that may or may not kill people. If you can’t do it all, then you put a legible sign in that elevator saying how many people it can carry at once (don’t repeat it, but the regulatory safety factor on elevators is 6, so it’s actually designed to withstand 6 times the load written on the sign, just to account for those idiots who find clever to jump in it). In any case, designing for the least challenging case may actually be a challenge since it will let too many options open, which is generally not good when it comes to choosing one.

But do you think that FLOSS image processing apps have a “designed for lightweight recreational use by command-line-able people only” mention on their home page ? Nope. It’s all pro-grade, or so we are told. How ridiculous ? You tell me.

Building apps that are efficient in a time-constrained and result-bound workflow makes them simply efficient for everybody. It’s really just a way to prevent mediocrity from becoming the common denominator. I know commercial photographers who would like to be able to do more than what Lightroom offers while still being able to be fast and efficient. But nothing serious is available to them, so they simply fall back to the typical.

It’s all the more nonsensical that the IT-engineering-tech bro loves to buy “professional” cameras and lenses. So when it comes to gadgets, no compromise is done, nothing is spared, we strive for the utmost quality. But when it comes to software, then, no… Amateurish and option-bloated soft is the rule. Why ?

How to make FLOSS image apps better for commercial photographers ?

The zeroth step would be to stop jumping on solutions before the problem is fully defined. That’s the number one issue of all open source projects I know of : people think they will spare time by getting things done ASAP. Actual design starts with defining what problem you want to solve, for whom (who is the user ? what kind of education/time spent learning do we expect from them ?) and in which context (covering the olympic games or the war in Ukraine is not exactly like editing a picture of your dog on Saturday evenings). Yet devs seem to want to jump head-first on their IDE to reinvent everything in the exact same way as it was done before : half-assed, with no perspective, and by the mighty power of the code snippet. Putting “how” before “why” is the surest way to hit a wall. The thing is, the longer you think about problems, the more chances you have to find the actual problem beneath the apparent problem and maybe factorize problems in a way that will make the solution much easier and much cleaner.

The first step is to acknowledge that commercial photographers exist. From where we are starting, it’s already an improvement. That includes having communication channels that are not so geek-oriented or where discussion is not entirely technical and low-level. Basically, stop with all those Linux-centric forums that happen to be about image processing, and where every next discussion is about your preferred Linux distribution.

The second step is to make the editing process as uniform as possible. In a scene-referred workflow, a lot of the early picture normalization (color profiling, exposure and white balance) is just linear scaling. These settings are fairly easy to transfer between pictures shot in the same lighting conditions and building tools to individually normalize pictures to the same base is easy too (it’s actually a new feature introduced in darktable 4.0). The flaw of the scene-referred workflow, though, is that the white level is variable, so there should be an unified way to define it pipeline-wise instead of having to set it for each filter in the pipeline.

The third step is to make editing apps use GUI paradigms that are consistent with the rest of the OS or desktop environment. I don’t know what it is with Linux apps, but it seems the dev teams make it a priority to re-invent every GUI interaction on their own, and make it forcefully different from the rest of the OS. Double-click opens things. Click or arrow keys select things. Escape… escapes from the current view. Stop being clever for the sake of being clever, you are not helping anybody.

The fourth step would be full support of Wacom pens and tablets. Meaning stopping with binding essential commands to keyboard “shortcuts”. Actually, the very name “shortcut” implies that it is an alternative way of doing the same thing faster, if there is no GUI way of triggering the same command with a pointer device, it’s not a shortcut anymore, it’s an annoyance for anybody using pens and tablets. A couple of brush strokes straight on the picture are much faster to dodge and burn than any slider-based brightening & masking, and that’s the beginning of a fast and efficient individual editing while still working in batch.

The fifth step is called inter-operability. Say what you will about evil Adobe, but you can open an Adobe image file in pretty much all Adobe imaging apps. Then finish working it elsewhere. From a right-click -> contextual menu -> open in “…”. While you can technically open most file formats in most FLOSS apps, you still have to juggle between apps, file browser, conversions and such. Again, hard to explain how just a couple of clicks more or less make all the difference when you process hundreds of pictures in your typical week, to people who mostly edit a couple of them on week-ends.

The sixth and last step would be for all the white collars and similarly-well-paid but entirely-replaceable staff who use art to make their boring life bearable to stop despising the ones who got the balls to go all in and made it a career. Not sure if it’s jealousy or if they are simply high on the capital’s fumes, but if the very thing they all advise their children against is their only joy and principal hobby, at some point, there is a chat to be had with a therapist.

This post is the long-form version of a forum post  that got censored by the very tech bros called out on their bullshit.