Darktable 4.0
darktable.org
darktable.org
I've switched back and forth between RawTherapee and darktable for a few years now.
I prefer the UI and workflow or RT, but dt has better performance on my machine. And RT has this strange longstanding bug where the image is extremely blurry during editing, and only becomes clear after exporting. According to a github issue, a workaround is to disable lens correction (and anything else in the "transform" tab) but this didn't work for me, prompting my most recent switch to dt.
current biggest grip with dt... why are those damn arrows at the edges of the screen so tiny??? Expanding and retracting the side panels requires pixel-perfect accuracy on a ~20px target. I know I could use keyboard shortcuts, but my brain isn't wired for them yet.
You can make them bigger through some CSS tweaks: https://docs.darktable.org/usermanual/development/en/prefere...
I reported this issue on irc years ago and the dev response was so hostile I abandoned trying to use it and never looked back. I pay Adobe $500/year for software that had been tested by users with developers silently watching.
People have a finite slice of their life to give you for nothing. Putting specific suggestions on the issue tracker seems more respectful of their time.
Attention and stewardship are not universal.
1. "This one issue is what is stopping me from using this product. Is this current method preferred by your users or do you have any other reason to not want to change?"
2. "The proprietary solution costs $X per month. Is there any way to sponsor development? I am willing to give $X/n (where n is whatever number you are comfortable with) per month to get this issue merged and to have some sort of prioritization when triaging issues.
3. Failing those: write a blog post and propose a fork. This is usually the quickest way to get a reaction from narcissistic developers.
Depends on the PR. Bugfixes sure, but new features mean more work for the maintianer in the future.
Speaking as a project maintainer that is starting to see some traction: I'd love to see people coming with PRs for features. It would mean a big amount of validation.
I bought a Mac and use Lightroom and I’m happy to give them both my money.
I'm sorry you did not stumble into them, but instead in the few hostile ones. It's a pity that made you give up.
A general tip, I give to people wishing to contribute to OSS is to first watch from the sideline. Read issues, threads, subscribe to an ML etc. Then to ask. Don't just throw in PR or feature request, but ask: is this welcome? How do you work? Any particular details that you wish me to pay extra attention to (tabs, spaces, tests, documentation, design).
As a FLOSS maintainer myself, it can be very intimidating to have someone throw a PR at you that rewrites everything (it comes across as: you suck, your software sucks, but watch me fix all that), or that disregards things that I deem critical (tests, architecture) It is really hard to review it, without coming across as an arrogant bastard. And often software that I put on a back burner will require me significant effort to get back into. So merging a simple typo fix or dependency update might cost half a day, just to get the dev env back up.
Hey dudes, there's a pointer dereferencing crash here which occurs when I do X, Y then Z. Added one line change to fix it. SILENCE.
There are no test cases for most projects either.
My dad used to be a pro photographer, now retired, and gave me crash course in RAW processing. Besides some handling differences between lightroom and photoshop he nothing negative to say about darktable. And neither do I, it's not darktables fault that my photo library is a mess.
Making perfect AppImages is often possible, but the automated tooling isn't smart enough. A proper AppImage (this one is by me) look like this: https://github.com/Elv13/reclaimail/blob/master/docker-edito... . Obviously this doesn't scale very well to projects with 300 dependencies like Digikam. My NeoVIM appimage linked above "really, really" bundles all dependency and compile your NeoVIM config to luajit bytecode. It's 3.9mb compared to the upstream one which is 15mb without any config. Note than 0.7mb of that 3.9 is the spellcheck dictionary, 0.4 my enormous config, 0.5 the AppImage overhead and 0.7 all the legacy plugins still written in vimscript.
[0] https://packages.ubuntu.com/search?keywords=digikam
In clicking around, I've selected "add to library," and darktable is now beachballed. CaptureOne will continue to be my default.
(Before someone points this out, let me acknowledge that the apparent problems with darktable are undoubtedly my fault in some capacity and that I'm Using It Wrong, or that I can't properly read the manual, or that I should watch more videos on how to use darktable.)
>I'm also glad it exists but have never figured out how to use it on MacOS
* Open app, Select 'Add to Library' --> spinning beachball of doom
delete the ~/.config/darktable directory, to give me a clean start...
* Open app. App opens in a weird window the full width of my screen but only half the depth. I try to resize the window. It won't rezise, even though I get resize arrows when hovering over the corner. Finally manage to make it usable by hitting the fullscreen widget.
* Try 'Add to Library' again. This time it allows me to select my photos archive folder. I see there's a tickbox marked 'Recursive' so I click on that as I have subfolders in my photos archive. I can't select the box. Then the spinning beachball of doom starts again...
* Force quit
* Delete app.
Life's just too short to mess about with an app that hangs twice, requiring a force quit, before I've even managed to do a single thing with it.
This is the correct answer. I actually switched to Linux from macOS partly to be able to run darktable in a more native setting.
We're not talking non-MacOS [or non-Windows?] -like behaviour here. But the app not actually functioning. So why offer versions for other OSes, if it's Linux only software?
Here's a little list of the problems Pace can cause:
http://web.archive.org/web/20071009193327/http://www.paceap....
One contains your custom configurations like presets, the other contains info about file locations, ratings, film rolls, etc.
Deleting and reimport takes time when a lot of files are involved, but is my preferred way of handling the changes of large scale file operations.
But film rolls might not be the best way to manage files.
After six years, I found searching the database using tags and Exif data such as camera and date ranges with the collection tool works better for me.
What also helps me is never deleting or moving an image file.
That way the database stays current and I avoid the busy work of file and folder manager.
To put it another way, for me flat search is better than hierarchy navigation.
But my previous experience with editing Sony RAWs on Rawtherappe and Darktable wasn't good.
The color science is missed up in Darktable and in Raw therappe, I just couldn't get something decent out.
I would be editing for an hour and end up with a weird looking image, while in lightroom I would be done in 10 minutes.
People might say that I'm just used to lightroom but I'm not, I was using Capture one, then didn't shoot anything for multiple years, then just went pack to Lightroom and could edit photos without major frustrations and get decent results quickly.
Creative apps in general are a very weak point of the Linux desktop, and when they exist, that have many weird requests and problems that just using them on windows is easier(Davinci resolve requiring the crappy closed source AMD drivers)
I think your problem might be the base curve module and its default "Sony" preset, which looks like garbage. Try setting it to "neutral", it should look a lot better (presets are under the little arrow next to the "base curve" header).
The base curve presets, if I understand correctly, aim to make the photo look similar to the out of camera JPEGs. In my opinion, they look like garbage.
But these days I use Filmic for everything. My process is basically switch off base curve, switch on Filmic, adjust exposure so that mid exposure looks reasonable, adjust white and black exposure in Filmic to taste and done.
The important thing to me is that this works for every photo, not just the easy ones. ETTR photos of high dynamic range scenes work just the same.
I'll occasionally run the photo through DXO and export a DNG with DXO's noise reduction for further work in Darktable too. Makes the shadows noise-free no matter how hard you push them.
Out of curiosity what stops you from using dxo for everything?
[1]: https://docs.darktable.org/usermanual/4.0/en/module-referenc...
https://docs.darktable.org/usermanual/4.0/en/module-referenc...
If you want help you need to explain what you did or tried. Did you work in display referred or scene referred? I can recommend submitting one of your raw files to a Play Raw[0]. People will edit your raw file and send back an .xmp file with the edit metadata so that you can see what they did.
[0] https://discuss.pixls.us/c/processing/playraw/30In my opinion, the Lightroom edit in the OP is the best. Many of the replies with Darktable edits look extremely bad. Some have extreme color artifacts.
I never had an issue with the RAWs from an a6300 and another entry level Sony mirrorless. Though I never used anything other than dt, so maybe I don't know what I'm missing.
For anyone that wants to learn darktable I wholeheartedly recommended the forum at https://discuss.pixls.us/c/software/darktable/19
The darktable developers frequent this forum and it's an indispensable resource.
I did a studio shoot on Monday: clean strobe light, a tame white background, all settings manual (so no auto-lighting-optimizer), and fired up my raws in 4.0 on my calibrated monitor.
I loaded the roll into Light Table and things look fine (because, I suspect, that the CR3 file has an embedded JPEG with Canon's recipe cooked in).
Then I double clicked to bring it into Darkroom and it goes crazy: the backdrop goes a crazy purple hue, the exposure is wrong by about a third of a stop, and the skin is zombified.
It's utterly incomprehensible to me. Every year I check and every year I'm baffled at how dreadful it is. Cya next year I guess.
What you are looking at when entering the darkroom is an "undeveloped" raw. Darktable's philosophy is that developing this raw photo should be 100% the user's responsibility and that whatever the camera maker's JPEG engine does is irrelevant.
I still use DT because it is very powerful, but I do agree that it has a very steep learning curve, and therefore is frustrating for anyone who isnt fully committed to learning about digital color science, etc, myself included.
Both Lightroom and Photo Lab do exactly this. If Darktable cannot support reading that information and importing it as a preset, that's fine, but pretending it is an artist choice seems to misunderstand what is going on here. To get specific Picture Control settings (e.g. sharpness, contrast, saturation, hue, etc), High ISO noise reduction, White balance, and color space are typically important as defaults or presets.
You are right, but I'm very sorry to say that attempting to faithfully recreate the look of the manufacturers JPEGs will likely never be the goal of darktable (or any serious photographer for that matter). What happens when you buy a camera from a different brand? You will likely want to normalize the output between your cameras and not to have every other picture rendered with Canon's interpretation of the the scene and the rest with Nikon's.
- I've been shooting mirrorless, and consequently using EVFs, since the NEX 6 came out. When I take a photo, I have already seen it. I have a vague idea of what tweaks I want to make, so having to start from scratch just interrupts that. Of course, sometimes those tweaks will override basically every default setting, but that doesn't mean having them wasn't valuable
- I often shoot RAW not because I want to painstakingly develop each photo to perfection, but because I simply need more latitude to "fix things in post". Especially when shooting events in suboptimal lighting or when I need to raise the ISO to make sure a critical event gets captured.
Sorry, I don't understand. What is “simply not true”?
"Color space", by which I assume you mean input color profile aka the color matrix of the camera's sensor, has to be calibrated or reverse engineered by shooting standardized color targets under standard illuminants. DT does have this data and applies it automatically, but the calibration is provided by users and might not be ideal. You would probably get better results by doing your own calibration, but that is somewhat involved.
Picture control things like saturation etc that you select in camera are post processing steps that the camera applies to the raw in order to get a "developed" jpeg. They are proprietary to the camera maker and not published. Even if you can read settings like "saturation" and "portrait mode" from the metadata, you don't know which algorithm exactly that implies. Lightroom and other commercial software works with camera makers under NDA in order to reproduce the camera's processing. Could it be reverse engineered by open source software? Maybe, but the DT devs decided that they prefer to go in another direction.
Not handling that at the software level and leaving it to the user means that you would have to do a repetitive and complex work for every single picture when other softwares give you a "neutral" basis (even though any attempt at neutrality is always limited. True neutrality could not display the image with RBG at all...) to work from based on calibrated color profiles for the cameras (some are better at giving you a good basis than others of course).
This would seem like a weird choice as anyways you have to do a bunch of conversion, so not using camera profile just means the conversion is less faithful.
Sorry, not specific link, just stuff I picked up by following the pixls.us forum, where the DT and RT devs hang out.
The older I get the less time I want to be in front of a computer. I would prefer my photography software to do the workflow I want it to do, not what someone else thinks I should do -- regardless of whether it's free or not.
Besides, the more I'm behind a camera in nature, the better I usually feel. The more I'm behind a screen, the worse.
Rawtherapee I think has a feature that tries to apply curves in order to match the built in jpeg preview.
Otherwise you might be better served by commercial software such as Lightroom or Capture One, as they work with camera manufacturers in order to match the in-camera processing.
To me, I don't care really, but in 2022, photography is still an underserved segment in the open source software world. I feel like they can have their opinion, but then allow others to have a UI element to turn it on if they so choose.
There's a lot of high quality gear floating around ebay for fairly cheap compared to when they were new.
Oh my goodness this sounds amazing
Parametric masks, LAB curves, wavelet decomposition are very powerful ways to work with the raw stuff of images.
It has that linux philosophy - from nerds to nerds. But not for practical use unfortunatelly. No Fuji profiles, not many lens correction profiles, wired demosaic for X-Trans, not consistent UI, no native folder sync, no AI clone tool...
My other issue is that I have filters (custom applied adjustments) that I an loathed give us and can't be bothered trying to replicate.
So what it does for me is basically barely noticably adjusting the saturation/contrast values, fixing the horizon and applying lense correction.
It mostly sticks to standard industrial and scientific definitions instead of marketable names, and contains very little "magic" that is common to commercial photography software (such as saturation intentionally not being actual saturation, hidden curves, and so on). So you can use any good book on photography/videography and color science, and directly apply it to most of the stuff it has.
Additionally, the developers spend a lot of their time explaining their reasoning and writing about the theory in general, for example:
https://www.youtube.com/user/s7habo/videos - this channel is consistently great for both basic and non-obvious things
https://discuss.pixls.us/c/software/darktable/19 - main darktable forums, has very good discussions and explanations
https://eng.aurelienpierre.com/ Dr. Rant of darktable. Primarily technical stuff.
For a simpler interface that gets me (a naive user) half-way decent results pretty quickly, I'll drop back to LightZone. Whole bunch of ready-made presets that combine primitives (sharpness, contrast, saturation, curves etc), and easy to save new presets once you find a combination you like.
It's over hyped. Yeah, it's cool and all but anything else would do the same thing. The only thing that I actually found is better in Lightroom is the AI detection features. Same can go with Photoshop and paint.net (or gimp if you're into that kinda thing.)
You can't change the aperture after the image has been captured. I don't think you're the intended audience for darktable. Give https://filmulator.org/ a try!
I'd like to see far more work on the UX. I can imagine a scenario where you load it for the first time and you're presented with a choice for beginner or advanced UI. Beginner UI would drop you into a Lightroom-esque filtered set of modules, and take you through a tutorial of the UI.
I suspect this is because photos usually stand on their own. You edit a photo to something that pleases you and that is it.
With video, you have multiple pieces of footage recorded at different times, in different circumstances and locations that need to match in their look to make a video feel like a consistent whole. Getting the color right is much more critical for video.
Lightroom and Capture One (and heck, even Affinity Photo), make common photography tasks like highlight recovery easily accessed with a single slider. Sure, the Darktable UI may be "meaningful to colorists", but that doesn't necessarily result in a good UX for the main people that are likely to want to use it.
> "This is meaningful to videographers and colorists" is a not a great defense of UX for an app that is targeted towards photographers, not videographers or colorists.
I think you underestimate the amount of photographers who know the fundamentals, as well as those who are familiar with related fields. After getting better in painting, I now resent tools like Lightroom which obscure everything behind some magic. I almost always see what I want from the photo, and I need a specific result. Photography tools are generally inconsistent, and most are light years behind video tools in getting predictable results. I want pastel colors in Lightroom and I know exactly what makes them; it's a single operation. Yet in Lightroom and Photoshop it needs multiple tweaks which feel like a workaround. The saturation in photo software can mean arbitrary things except being an actual axis in the color space. It can use several intermediate curves without telling you, making you fiddle with the sliders yet still not get the result you want. And so on. I mean, I tried Luminar AI recently, and it has a slider called Punch; what the hell does it do exactly?!
There's a reason people are using things like Firegrade for Photoshop. Software like RT/ART and Darktable also does what it says on the tin and nothing more, this is a niche they fill, and I'm glad they exist. (although I'll admit RT still looks like an explosion of sliders to me, despite the fact I know what they are for)
It had a lot of momentum at the time as it was plugged a lot on HN, it was open source and had a similar feel to Atom.
But I have not heard of it since and I'm finding it hard to recall the name. Weird.
That video is a text book case of burnout, it looks like.
Darktable 3.8.0 Released - https://news.ycombinator.com/item?id=29672729 - Dec 2021 (2 comments)
Darktable: An open source photography workflow application - https://news.ycombinator.com/item?id=28281775 - Aug 2021 (36 comments)
Darktable 3.6 - https://news.ycombinator.com/item?id=27720908 - July 2021 (73 comments)
Darktable 3.2.1 - https://news.ycombinator.com/item?id=24156113 - Aug 2020 (47 comments)
Darktable 3.0 - https://news.ycombinator.com/item?id=21874528 - Dec 2019 (120 comments)
Darktable 2.4.0 released - https://news.ycombinator.com/item?id=16012499 - Dec 2017 (74 comments)
Darktable 2.2.0 released - https://news.ycombinator.com/item?id=13261849 - Dec 2016 (82 comments)
Darktable 2.0 released - https://news.ycombinator.com/item?id=10789390 - Dec 2015 (44 comments)
A look at Darktable 2.0 - https://news.ycombinator.com/item?id=10640753 - Nov 2015 (10 comments)
Why there are no darktable builds for Windows - https://news.ycombinator.com/item?id=9883018 - July 2015 (1 comment)
Colour is a bit more hands on and with negadoctor you really can get a nice starting base but it's much more manual. You need to make sure colour profiles and white balance match your scaner/dslr scan. The you need to set up the shadow casts and highlight colours. Much more manual just to then get a rather flat image with correct colours to your liking. Then it's more regular raw workflow so altering contrast and saturation but it's nothing too hard still. But it's 100% more hands on the using NLP (this is what I'm assuming you are using. Check YouTube see what you think.