the best of these turorials are never made by the company that makes the software, but by users of that software.
Isn’t this always what we are doing?
This is when you go to the manual, look up what they do, then play with those options until you understand it.
Here also plays a role sponsorship and competition between the beneficiaries (photographers or photo studios) who also offer services. Youtube is full of such sponsored content.
Good tutorials do not scratch the surface, but offer more in-depth knowledge that is independent of the tool used. This also means that the person offering the tutorial must have this knowledge and be well prepared.
But it also means that people who use darktble can also learn a lot from these tutorials. From topics that concern general knowledge such as color theory, to the use of concrete modules such as curves, local contrast, channel mixer, color balance etc. there are countless tutorials for another software that you can watch and then apply in darktable.
And also the topics that you have mentioned as examples are software-independent:
I will rather say, there are few professional photographers who use darktable, who are able to offer well-founded tutorials. This must change.
Precisely. I feel that I’m at a point where I’m comfortable with darktable and can translate other software features it’s modules, but all too often tutorials that you’ll find for a style wind up being advertisements for someone’s LUTs or presets.
And frankly, I’m not interested in a canned product that’ll deliver a look (although I understand that a professional might need something like that for consistency).
I don’t even like using the DT color mapping because I’m not interested in copying someone’s style. What I want to do is improve my skills behind the camera and at the computer screen to create the look I’m after at that moment in time.
I fully agree. And similarly, the “drag this slider until it looks good” approach is not really useful for me either. I can futz around with sliders all day until things look good, we don’t need tutorials for that. I really appreciate it when Bruce/Aurélien et al really try and explain what is going on with the slider, when it should it be used, when it shouldn’t, which module to use first, what particular technique it is good for, etc. It’s the old teaching how to fish rather than giving a fish analogy…
I’m not saying I have an answer. We just need more experts who have the will and the time to make these kinds of resources. Making great educational content is very time consuming.
I hope not, or at least that’s what I try not to do. I’ve been spending a lot of time recently studying pictures to try to decompose why a look has that look and then I use the modules as I understand them to approach that style.
I don’t pretend to have the answer either. I have the utmost respect for Bruce, Aurélien, Rico, and others who take the time to explain what’s under the hood… often over and over again. I’m not sure I have that patience. And of course the play raws and Boris’ editing moments are incredibly useful.
I would say that instead of asking if someone would kindly produce a whole new round of tutorials, I think more discussion here on this site on the creative side of editing would go a long way. At least that’s my thought.
Regarding your remarks on the luminance estimator description on the tone equaliser page, I agree it is quite terse. The reason we didn’t put more detail there was that luminance estimators are used in several places, and we didn’t want to be repeating information in the manual that can then become inconsistent.
The logical thing would be to place a reference to the description in the general Darkroom>Processing Midules>Curves section. However, that section itself refers to the “preserve chrominance” parameter description in the FilmicRGB reference page.
So, if you go to the FilmicRGB page, you’ll get some information about possible use cases for the different estimators.
@elstoc, maybe we need to try to organise this a bit better.
Yes perhaps so. Could you raise an issue please.
done:
Concerning the OP’s remarks about the preserve details description, I thought we’d already explained that the purpose of this is to preserve local contrast across edges where required, and the GF masks can be disabled if for some reason someone wants a simple tone mapping that doesn’t try to preserve local contrast.
I’m don’t know what more we can say in this section, as it is heavily dependent on the details of the image and what the user is trying to achieve. We’ve tried to communicate the key concepts behind what the module doing, but maybe this needs to be reinforced by modeling some actual examples. This would be getting beyond the scope of the user manual however. So, I’m not sure what to do…
While not at all wishing to appear rude, ungrateful or unreasonably demanding, I have to say that my personal feeling is that I don’t ‘see’ the key darktable module concepts being sufficiently well explained in a way that is compatible with my basic level of knowledge of the product, with the physics of light and how the human brain interprets that physics and with good photographic principles. Modelling some actual examples would be of considerable interest to me, except that I see videos on YouTube which can make me an expert on developing a fine image of a particular insect in a certain forest, in a certain part of the world, on a given day and time of the year, under quoted weather conditions, taken from a certain position with a given set of aperture, shutter, ISO, camera, lens and filter. But for other images with different parameters I’m still lost!
The last solution I want to see to that scenario is to make even more use of ‘AI sky replacements’. The logical conclusion to this wonderful marketing buzz-word - AI - is to have some device that uses AI to decide what photo I want, takes it and then does all the post-processing for me - hopefully including deleting the damn thing afterwards.
‘futzing around’ in image development (as an earlier post stated it) is the source of a significant part of the pleasure I get from photography. But I can’t experience that pleasure if I don’t know what I’m doing, and, emphatically, if somebody else is doing it for me with AI apps (the emphasis clearly on the ‘artificial’, rather than the ‘Intelligence’)
What steps would you take in an image editor that you’re comfortable with to make an image that you’re happy with?
Also if you do want to see your image edited, we have the Play Raws, where you’re welcome to describe what you think is missing or offer some challenge to the community to solve.
Just curious, when you watch these videos and become an expert in developing insect pictures, what does this mean? Does it mean you have memorised a procedure that you can then reproduce with similar pictures? Have you managed to get an understanding of why the procedure contains these specific steps, what was the rationale behind the steps? When you try to generalise this procedure to other images, what happens? Is the whole image messed up, or are there just some limited aspects of the new image that ha e come up disappointing? In that case, the fundamentals might be ok, but just some small set of additional specific knowledge or skill is required to deal with that specific problem.
I’m not sure how far a user manual can go as a learning resource. It’s certainly one input, but some more general learning in the domain of image processing may also be required.
Aurélien Pierre made some videos that dig into some of the more basic concepts when he was developing the filmic module. How did you find that material? Was it clear? Did it assume too much background knowledge? Did you understand it, but are still somehow unsure of how to apply that knowledge?
This is something that comes up a lot, so I just curious about your take on it.
Respectfully, I think you missed my cheap attempt at sarcasm. I found a notable example on line which attempts to teach me how to use a specific raw processor (It wasn’t darktable) but which was so specific in its application that it was only of value in developing that exact image. And even then its advice was of the form ‘move this slider to here’ without any information on what charcaterristics of the image, at that point in the development, required that adjustment to that variable. I recall being so irritated by this lack of understanding of the needs of the viewer that I wrote a rather hotly worded note to the video creator, who was big enough to agree with me!
I would like this lesson to be applied in future material which supports darktable - a raw processor which I am slowly coming to realise is, functionally, within the top 5 or so such solutions in the marketplace - certainly capable of producing significantly better results, in my hands, than my previous chosen solution (Lightroom, from Beta 1 to version 6.14).
You make my point entirely: with this type of, supposedly educational, videos I gain almost zero transferable skills: they allow me only to be able to develop an identical image of the identical insect, taken in circumstances which are identical in every way (in terms of their light implications). In fact the whole video could be replaced by a preset. I gain no insights, no understanding of why do this, now, in this amount, in this area of the image.
As I said in my original post ,which sparked this discussion, a really valuable guide should include some discussion on what characteristics of an image require what sort of raw processing response.
And, yes, I appreciate that this is probably not suitable material to go into the ‘function’ driven architecture of the typical, well established user manual. And yes, I see it as quite a difficult guide to write, because it covers such a huge ‘battleground’. But those really insightful and gifted people, like Mr. Pierre et al., can look at an image and almost immediately see what needs to be done to it. Of course their brain has done in a few seconds what mine could do in a few weeks (possibly), so doesn’t that mean that they could at least describe the process they go through, the assessments they make, the options they consider and the criteria for selection of specific settings of variables?
I now understand what you mean and the answer lies exactly here:
Maybe a comparison with music lessons will help. When you learn an instrument, there are two different parts of the lesson: Usage of the instrument (instrument manual) and music theory (key signatures, time signatures, melody, rhythm, harmony, tonal systems, scales, etc.).
Music theory is obligatory for all those who want to learn an instrument regardless of which one.
Translated to darktable - if you have general knowledge about image processing (let’s call it image theory) there is only the question how you can implement this knowledge with the help of darktable as an “instrument”. The problem you are addressing is that you don’t know the instrument nor the theory.
Well, good luck on your learning, I can’t spend more time on this if we aren’t going to take things seriously.
It sounds like you’re having trouble translating your successful experience with other image processors to darktable. Most of the controls found in other programs have analogous counterparts in DT.
Perhaps you would do best to start with the display referred workflow using the base curve and work with the most basic controls.
Then, once you’re comfortable with file management and some of the more intermediate modules such as levels and curves, then you can move to the scene referred workflow and begin to tackle Filmic, tone equalizer, color balance and others.
Then perhaps you’ll find the videos to be more useful for you, plus the discussion on this site can help fill in the gaps
I really like your analogy here. I am one of those people that if you give me a set of plans for a shed I can build “that” shed exactly. Lines shapes angles and attention to detail etc…all no problem. As I build it I could adapt and modify it to make it more robust or functional, however, sadly I am not creative. I wish sometimes I could draw or perhaps play an instrument but I just don’t seem wired for it. My weakness certainly gets exposed when developing an image. I understand for the most part all the tools and the impact on the image at the pixel level but as for the big picture I don’t possess that skill and quality that you have demonstrated for example in your recent work on the attempts to recreate stylized photos…(great by the way).
In a recent edit in playraw you suggested rightly so IMO to leave the foreground dark as to better highlight the glow of the sun on the subject of the image (a mountain) but because we have software and all kinds of tools most people felt the need to pull up strongly on the shadows which to me destroyed the overall light and tones of the image with no significant improvement. It is great for sure to have all these tools and everyone has a vision for an image but I can’t help but feel sometimes that many edits have been dictated by the tools and they have no editing to enhance the subject or compositional elements. People go through the tools and apply a bunch and in the end may have an edit with lots of color or shadows and highlights managed etc but often for me the results don’t match the potential of imagery provided by the photo.
To me this really highlights why the marriage of instruments and theory is so important however, I think many people overlook this relationship. They look at software out of the gate as a means to skip or simplify the theory or save time and this is not always the goal. The expectation is to push and pull a few sliders to get tone color and saturation changes to provide a pleasing result. Going forward I will learn to get nice pictures by iteration of this process which of course is part of any learning process but I think only in a more simplified environment. DT is far too complex with sliders that don’t always have a one to one relationship to learn it this way, ie I increase this and boom that is this part of the image corrected. All the modules have potential cross talk and nuances that can only be understood and mastered with understanding the theory so that when you do experiment you have context for the results.
I think the DT manual can and should provide that pixel level information but I think unless someone has the time and passion to write a book on color theory using DT throughout to highlight the marriage of instrument and theory then these questions will just keep coming. Even with a 10/10 manual if users are not willing to invest time outside of DT to gain a good grasp of the theory it will be difficult to go from a module to module description no matter how good it is to a finished edit IMO…
Thanks for sharing your ideas knowledge and techniques with us. I have learned a lot
Forget darktable. Analyze old masters. They spent a lot of time to find out what works. Learn from it and apply it…
…like Kubrick does, for example:
The scene looks hopelessly underexposed.
And then you’ve made a big step by learning that it doesn’t always make sense to use exposure compensation and filmic to see every part out of the shadows. Because it’s the darker shadows that frame the action. Reinforce the meaning of the scene.
But this is not part of the darktable manual. This has nothing to do with darktable!
This is just a tool like the paints and brushes for the painter.
Analyze old masters.
Yes, yes, yes, yes, yes.
Masters of photography, as well as painting. And drawing.
The tool is not relevant to aesthetics. Over the years, I have taken many courses in painting and photography. For me, the foundation is always, “What makes a good picture?”
For that, we need to know how pictures work. What are the elements of pictures, and how do they work together to create an effect, a feeling, an emotion?
We can learn a lot from books (and web sites), but taking a course (or evening class or whatever) is more immersive, with interaction between students and teachers, and we can test our ideas against other people.
Solid advice. I bet if you had posted a photo of the above scene on play raw…not many would have crafted the light on the the subject…I think in the end most edits are to contrasted too sharp and often there is what I call no decision on the light…often you have to “sacrifice” part of the image and make a decision on the subject and the light…at least IMO I don’t see that in too many edits…I can see it at times I just need to learn to craft it…ha ha…great example thank you…
