Improvement Suggestions

But we can see here in the forum that there are people out there steadily using old versions of darktable, so the old versions of the manual have to be available.

I can see, though, that it is possible to mark such older manual version with info that there are newer ones available, (or to equip them with a date that more correctly indicate their age). But is it reasonable that people who already do a massive voluntary effort to develop and maintain dt and the documentation shall spend even more time on serving you who for some reason donā€™t bother to go to the source to check if what google throws at you is the newest version available?

You should know that if you just rely on google, you take a risk, and I posit that by not searching up the pages of the darktable project you are somewhat lazy and perhaps should blame yourself rather than demand more of others. Am I right?

You mentioned the trouble of needing several module instances for several subjects (thatā€™s why youā€™d need 5 instances of color balance rgb, as I understand). But now you say you want different effects for each of the subjects. How would you do that having a single module instance?

If I understand you correctly, youā€™d want to define a single mask (potentially using params and several shapes), and add several ā€˜effectsā€™ (modules) to it. As you already know, the darktable way of doing that is the opposite, defining the mask on the first module, and then using the mask as a raster mask on the other modules dealing with the same selected subject. To keep track of them, you may name each module instance. Probably not optimal for your use case, but it is possible to achieve (I rarely need to do that, though, but I understand that your editing needs are more demanding than my adjustments, which are usually global, maybe adding custom processing for the main subject or the background only). Changing this would be a huge undertaking, I think.

I was not suggesting to take them downā€¦ but hide them from googleā€¦ of course they need to be available for older version.

No, even if you use the search function on darktable (like search | darktable) prevent old information, even going back to DarkTable 2.6.

I must admit that I found your message a bit harsh. Itā€™s important to remember that everyone has different ways of seeking information and support. And I believe that fostering a friendly and understanding dialogue on these forums is more productive. Iā€™m always open to constructive feedback, but I would appreciate it if we could keep our conversations more positive towards new users.

3 Likes

I think the thing is that the information base for DT is all over the place due to its long history and many and varied contributorsā€¦ Some current information is not of great quality etc etc so really I donā€™t think there is any magic fix. Already its is a struggle to keep up the documentation for some recent changes. I donā€™t have the time currently in my life but maybe soonā€¦in any case as an example there are not (I dontā€™ believe) a ton of videoā€™s on the contrast eq and specifically for the edges tabā€¦ so still this 11 year old video when the contrast eq was just called the equalizer to me is still one of the best videos on the application of this moduleā€¦ There are also some old blog posts on the DT site that are very informative in a "conceptual " way if nothing elseā€¦ so any effort to stream line the information would I suppose be useful but if something like this video was ignored due to age it would be a shameā€¦

I havenā€™t meant to be harsh, but written communication has its drawbacks.

I have responded to you in this thread, by greeting you welcome, and agreed with you that there are things that may improve the situation for beginners and developed further on that. But now you are asking for even more from people who work for you gratis, for a problem you have created for yourself.

Yes, we may have different search strategies, but we chose what strategies we follow, and in this world there are actually sometimes consequences from choices we make. You chose, in my eyes, an unsafe search strategy by not doing any QA on the information you got. Searching for darktable ought to give you darktable.org high up on the list, and under Resources you would see it all.

Itā€™s easy for us when we are new to FOSS to put the same evaluation scales on it as we may do for a commercial offering, but it doesnā€™t work that way.

EDIT: I followed your link to search on darktable.org. The search result that follows from this link gives as the second item on the list: ā€œHow do I ā€¦ in darktable 5.0?ā€ shouldnā€™t that give some indication that a 3.6 manual is not up to date?

And youā€™ll say that the module is too complex :slight_smile: I love people having always a simple view of things that seems to work for THEM.

But let me say:

  • If you remove old modules you break all current edits done by users since years.
  • So the only solution is to create new modules with all controls for the merged modulesā€¦ sure, a very good idea. Youā€™ll end up with one more module which was not the goal AFAIU and worst the module will be very complex.

Let me state it another time, if you donā€™t want to use a module fine. Most of them are not visible anyway. And if in a workflow you find that there is still too many modules then remove them create your own display with just the module you want. What is so wrong with that?

And let me bet, your layout wonā€™t be the same as other users. At the end youā€™ll understand that darktable is very customizable and can be setup for your liking.

My dream is to stop having this fallacious argument about too many modulesā€¦ but thatā€™s a dream I fear :frowning:

4 Likes

Thanks for the clarification. I did do QA on the information I got, I was just pointing out that it can be confusing that the old documentation pop up in Google. Yes, it might be an unsafe search strategy in your eyes. But it is not a very unreasonable one it would

You are assuming that I am new to FOSS, but I am not. And it is different. I am very grateful for people to develop this kind of software for free. And I also know - by experience - that you often as a FOSS developer donā€™t get much appreciation for that. And thatā€™s what I have been trying to do in all my posts, giving appreciation for what has been done.

To be fair, Pascal, people need to already be familiar with darktable to find which modules they like, and to even discover that the UI is customisable. Not many start by reading the manual front to back, and even if they do, they may not know what modules are worth adding / keeping, exactly because they lack the experience. I think a newcomer / complete beginner will have a hard time ā€“ thatā€™s where the recommended videos are a great help; maybe we should agree on a few good introductory videos, which should be linked directly. Many of us in the community have been using darktable for a decade or longer, and have tweaked it to just the way we like it, but I understand that for someone completely new, the learning curve can be steep, and the UI confusing.

This is something I struggled with at first, and I started keeping a document titled ā€œAn Idiotā€™s Guide To Darktable (Me, Iā€™m the Idiot)ā€ where I try to map photography concepts to darktable processes. I havenā€™t worked on it in a while but if I can get some motivation Iā€™d be happy to work on it some more and publish it.

This is something I saw multiple folks talk about when I was researching ā€œlightroom alternativesā€. One specific complaint was that darktable is clearly written by programmers, for programmers. And I get the complaint, but I donā€™t think thereā€™s anything wrong with that. For me, darktable was overwhelming at first because I was trying to translate how to do things in Lightroom with how to do things in darktable. Once darktable clicked for me, I realized that darktable does make a lot more sense for the way my brain works precisely because I like seeing the nuts and bolts. Thereā€™s a lot of editing software out there that abstracts away the complexities of image editing, but thereā€™s not a lot (or any?) like darktable that expect you to put in some leg work to understand how things work. Having options is great!

This discussion does highlight how much more there is to FOSS development than pure coding. Documentation, training, best practices, etc, are all important things that people can contribute to without the need for intense programming knowledge.

1 Like

I think one thing weā€™ve missed here is that darktable has the same unspoken premise as a lot of other free software: we trade initial ease of use for a lot of control. In the beginning, the amount of control isnā€™t beneficial because one does not know how to harness any of it. But as you work with it, the control becomes more apparent; one will grow to appreciate it.

I think its widely accepted that you will have to work a bit to get there with DT and I think the devs and community at large accept that and think it is a decent trade off to make.

To help offset some of the learning curve, we operate this forum, we have chatrooms, we have docs, we have videos.

Lastly I think we all accept that darktable isnā€™t for everyone. If you like Lightroom or DxO because i t gives you something DT doesnā€™t, then you should use those tools. We have something powerful and unique, and that is lovely.

7 Likes

To be fair, Pascal, people need to already be familiar with darktable to find which modules they like, and to even discover that the UI is customisable.

Sure, but there is a bit more than 35 modules when starting dt the first time. 9 in first group Tone tab, 4 in Color tab and 10 in Retouch tab. I do not count the ones in Effets tab which are really optional. In those selected the first time there is no or very few overlap in terms of feature.

And I agree the learning curve is steep because darktable is a complete toolbox for RAW development. And from experience I find darktable complexity a toy compared to Photoshop. No one would start using Photoshop without reading doc or following tutorials and/or paying sessions.

But maybe Iā€™m too much involved to see the complexity of darktable at this stage, for me it is simple as every module have the same blending for example. So when youā€™ve understood the 1 to 10 sliders and how they work together or which preset to use youā€™ll done. But again Iā€™m probably biased :slight_smile:

6 Likes

@Pascal_Obry
It certainly seems to be the never ending story :slight_smile:

I thinkā€¦
taking the time to read this section:

https://docs.darktable.org/usermanual/development/en/overview/workflow/process/

and this one

https://docs.darktable.org/usermanual/development/en/special-topics/color-pipeline/

in the manual give some good starting tips and direction for new folksā€¦ Maybe some suggestions by new users after reading some of these sections for what information was found to be lacking in the eyes of a ā€œnewā€ user would be helpful as that is harder for an experienced person to sometimes pick up onā€¦

Of course there is always room for improvement in any project but DT by its nature and by design in many instances is to expose as much of the editing process as possible. This will not always jive with a simple interface. As well, there will always be a spectrum of users and given the nature, nuance and intricacy of the DT pipeline its not a trivial thing to understand.

@martinus many of the recent tools were introduced and propagated by a dev that went on a color science crusadeā€¦ just a sample of the context and framework for DTā€™s inner color workingsā€¦and this sort of depth and detail was carried over to several other key modulesā€¦

Without a bit of history and a deep dive its hard to just land in to DT and get context for why things are the way they areā€¦it takes time for sure.

Once I started to follow the developement a bit on Github I was much better able to understand why things were the way they were and what suggestions/requests were currently out there along with which ones were being worked onā€¦ I know right out of the gate that all of this is beyond what some people are looking for to just ā€œdevelopā€ there photoā€™s but this is the nature of DT and why it actually appeals to many people so its a circular argument in many cases because people come in with another frame of referenceā€¦

Case in point for me during the pandemic I had lots of extra time and I was on my computer and I thought I would organize years of photoā€™s. I didnā€™t know what raw file really was. I have owned a series of compact camera that captured jpg. I had 15 years worth of vacation and family photos. I stumbled on to rawtherapee and got a bit overwhelmed just trying to fix up my jpg images. I then found DT but back in 2018 I really struggled with the interface. I was lucky enough that a great chap @Bruce_Williams made an epic series of videos that got me past the UI hurdles and the basics of using the programā€¦from there I learned about raw captures and I just spent a lot of time and followed the development odf DT which was massive for a few years with new features and modules appearing and being modifed in rapid succession, and it was only through participation that I was able to hang in thereā€¦

Where has this left meā€¦Well I have Affinity photo, Gimp etc etc. I one time bought ON1 photo raw when it was like 40 bucks as was Affinity photo. ON1 has some fancy tools and the UI is pretty good but you know what, I get in to it and begin editing and then I just miss the access to all the tools and presets that I have learned to use in DT. It has taken time but I have a way of working that I am locked intoā€¦donā€™t get me wrong I experiment a lot and will also work in RT and ART because interesting things go on there. However, when presented with the more modern choice with fancy AI selection masking and tone adjustment I still come back to DTā€¦

If I was a pro photographer or had other needs maybe my decision would be different ā€¦ I canā€™t say.

Anyway its some random commentsā€¦ @martinus If you can find a way to immerse yourself you will find that you can get some nice edits pretty quickly and if not the only cost to you is your time so thatā€™s a bonusā€¦I wish you luck and I hope you will have your ah ha momentā€¦ :slight_smile:

2 Likes

These discussions often comes up when someone hopes that darktable can serve as an alternative to their preferred tool, especially when pricing becomes a concern (or when they have switched to Linux).
It is not reasonable to expect a free and open-source software project to alter its core philosophy just to accommodate the preferences of individuals who believe they speak for a majority.
If someone is seeking a free alternative to Lightroom, it would be more productive for them to support or initiate a project with that specific focus.

to be honest - raw image processing is all about math and algorithms. And thereā€™s no one size fits it all approach if quality is a target. If you need a simple solution - why not just use the out of camera jpegs? That the ultimatve way to hide all that mathematical complexity behind a couple of presets or picture styles ā€¦

I think thereā€™s a clear pattern: those who put effort into learning the nuts and bolts of Darktable ultimately enjoy it and understand the way it is; those who donā€™t want to do a lot of learning tend to get frustrated with it and think itā€™s overly complicated.

This isnā€™t intended to be a slight on the latter group. Some people just want to get to the end result with minimal effort. But I think itā€™s a useful reminder to newcomers that the software really starts to make sense after you put the effort in to learning it fully.

3 Likes
  • als er een bericht binnen komt op een groep die open staat dat word dit bericht niet automatisch weergegeven

Wowā€¦ I am sorry for unleashing such a discussionā€¦ I do want to sayā€¦ Much of what I said was a suggestion (besides the brush) but not a complaint. So I am fine that it is written for programmers :slight_smile: My intension was to provide feedback that may or may not be useful.

But, maybe if it such a never ending story - people commenting on the number of modules that seem to do the same thing maybe there is a point there :wink:

I do both agree and not agree with this. Yes raw image processing is all about math and algorithms. But I would argue that not user needs to know the algorithms, if cleaver software engineer provide the tools to help other artistic people unleash their creativity on their work. Because developing a raw file is just one part of the equation, taking a good photo is completely different one, is also takes considerable skillā€¦

That would be awesome! That would really help new comers in DarkTable, if that is directly in the documentation.

This is a generic problem, whether you are looking up manuals, news articles, programming information etc.

I donā€™t know which search engine you use. I use Duck Duck Go, this has a drop-down for the search duration. If you only want the latest information, you can specify ā€œthe last yearā€ or ā€œthe last monthā€

(There are a variety of different search engines, this EU page gives alternatives for search and other digital services)

This is something I have noticed as well - and not just here, but other sites too. Thereā€™s always someone ready to criticise DT, but rarely do they offer any solution more elaborate than ā€œmake it like Lightroomā€. And when they do, like here, almost invariably fail to consider DTā€™s goals as a raw developer and its ā€œtargetā€ users, or if the proposed idea even fits into DTā€™s technical architecture.

Thereā€™s currently someone putting actions behind their words (!), by working on a new tone eq UI. Looks very promising too.

I assume you mean denoise (profiled) and astrophoto denoise. How would you combine them? Technically they work in very different ways, so itā€™s not like you can reuse any of the sliders. Without completely rethinking how they work, you would essentially end up with two modules in one, at which point you may as well just keep them separate.

Hereā€™s a little exercise for you (and anyone else with a clever idea): Sit down with pen and paper, and sketch out the UI/UX. Make sure you cover all current use cases. If you can come up with something that is an actual improvement on the current situation, Iā€™ll be very impressed indeed.

You just provided a good example of ignoring the technical realities.

Have a look at this:


The two yellow dots are not there for fun. Those modules dictate where many other modules can be placed in the pixelpipe. HL reconstruction and demosaic both depend on a somewhat correct white balance to function correctly, so WB must come before. Color calibration CAT works with RGB pixels and so must, at a minimum, come after demosaic. CC CAT is also not just a new UI on top of the old WB algorithm - itā€™s completely different. If DT was like LR, where the processing pipeline is hidden, this wouldnā€™t be a problem and they could indeed be combined in some way, but since the module order is the processing order, they have to be separate.

And nothing wrong with that. They just shouldnā€™t be using darktable then. If they want something free, both RawTherapee and ART are mostly like that. Or they could go with just about any of the commercial options. And lets not forget that there are plenty of former LR and C1 users that prefer the DT way (hi @bastibe), so should they then be forced to use something they would consider worse?

But thatā€™s the thing, they mostly donā€™t do the same thing, even if it superficially may seem like they do. I count six different modules that can denoise - but they all have they pros and cons, while also working in different ways that mean they canā€™t easily be combined. Which is of course one of the things that makes DT a raw toolbox and not an ā€œon-railsā€ editor like LR. Short of making DT exactly like LR, I donā€™t think thereā€™s a way to placate the users that get confused by that. Now, there are some modules that could be removed without losing any functionality, such as curves where tone eq does the same and more, but I donā€™t even want to think of the amount of complaining that would trigger. Even something as replacing the old Lab levels with RGB levels managed to piss several users off, despite there being no functional difference.

3 Likes

The keyword here is ā€œseemā€. There are always differences in the trade-offs, and those can be important enough to prefer one over the other (e.g. sigmoid vs filmic: they are both tone mappersā€¦).

Once you have a bit of experience, you find what works best for you, and you can hide unused modules from the UI (see the manual)

Iā€™m sure someoneā€™s already said it but this threadā€™s way too long for me to check every post (41 new comments since I last looked!) so Iā€™ll make it even longer by saying I never see any of these ā€œduplicateā€ or unused modules because Iā€™ve set up my own panel of favourite modules that follow my normal workflow.

And if I do need something else, I can always search for it in the search bar. I think it took me, like, 30 mins to set that up how I like it.

I understand someone who wants a free LR clone off the bat might be put off at first glance, but if youā€™ve made it this far, I donā€™t think itā€™s beyond anyone to customise DT in terms of what modules you see, fwiw.

On the complexity of some individual modules, Iā€™m happy to use presets and tweak. I think @nwinspeare did a pretty good job of intuitively explaining diffuse and sharpen with practical as well explanatory videos on YouTube.

Could it be better? Maybe. Proposing actionable solutions are always welcome, I believe.

1 Like

Its a never ending story because there are two options ans neither of them are good:

  1. Remove modules. Then you get your edits broken.
  2. Make a new module combining the old modules. Now you have even more modules and that just exacerbates the problem.

So we hide them or depricate them and that is about the best that can happen.

2 Likes