Ok. First I would like to demonstrate if I have understood the new approach.
When I use Fimic RGB, I have followed your advice and accordingly I get the best results if I first set the medium grey in the exposure as I want it (without consideration of the overexposure) and then use Filmic RGB.
I’m going to take an example now, so that there’s not too much abstract talk.
- In this photo I first adjusted the exposure so that the trees are well lit:
- Then I used Filmic RGB to adjust the dynamic range so that the details in the snow can still be recognized to some extent. And as the histogram shows, I still have enough space for further processing. However the image is very flat and diffuse which is normal as a result of the strong compression:
- Now comes the tricky task of emphasizing the details without affecting the dynamic range too much. Here you suggest to use a local contrast module. Although this will actually help to increase local contrast a bit it is not enough and there is a risk that you will either over-sharpen the image or get other unwanted effects. Note that the snow has not yet gained much in local contrast:
- So we have to bring the highlights back more to work on the details. For this purpose tone equalizer is a very good choice. The masking of guided filter can be a bit cumbersome, but after a little effort you get good results. Now you can slowly see the footsteps in the snow:
- Now we use the RGB curve to increase the details in the snow. I found this to be a difficult task, as it works differently than the normal curves in darktable. Even very small movements of the points can cause big changes in the image and it is not easy to get a reasonable contrast in the medium grey area. The presets show the way but it is quite tricky when you have to do something yourself. In this example I did not use the “compensate middle gray” option because it gives me more space. Now the footsteps are clearly visible:
- So, as far as details in the snow are concerned, we are almost there. I have found that in certain situations Haze removal module is very suitable for highlighting the details and is even much better than local contrast module. However, it tends to create the halos and you have to protect the shadows. In this case it has served well:
- Details are now very nice, but - as you can see in the histogram - the steps have also resulted in some shadow areas being a bit underexposed. I’ll be happy to brighten up the image a bit now without losing the details in the snow again. I found this very difficult. Both the exposure and the curves have caused the details to disappear. The best solution was to go back to Filmic RGB and move the middle grey luminance to the left and after that add a bit contrast with color balance:
After I increased the saturation with velvia, vibrance and color balance modules, the picture finally looks like I wanted it to. Now, not only the footprints are recognizable but also the fine texture, colors and diffuse shades in the snow:
I intentionally took this photo because you can clearly see how important it is - besides the dynamic range compression - to have enough space and possibilities to handle the details in shadows middle gray but also in highlight areas.
This article is much too long, in the next one I will make a few remarks.
Now few remarks:
Reuse of modules to avoid overloading the pixelpipe
In the course of my work, I have often noticed that the focus of my work changes during the course of processing. In the example above, it was important to me first that the trees (shadow area) are well illuminated. The actual processing later referred to the image areas with snow (highlights). The shadows were still considered but the priority has shifted to highlight areas. In the process of this work it is therefore sometimes necessary to go back to some modules from the beginning of the pixelpipe
But you can’t simply use a module at the beginning of the pixelpipe again, because these changes would have affected the results of all other modules that come later.
If I increased the contrast with color balance module after step 2 in the example above and want to use it again after step 6, it would not be recommended because the contrast focus was moved in between. So either i take a new instance of the module or use another one that has the same function (e.g. rgb curve).
The same applies to tone equalizer, for example. The mask for guided filter, which was made at the beginning, is no longer valid after a few steps in the pixelpipe. So it had to be readjusted, which would have changed the results of the modules that come after it. So here also applies either new instance or other modules for same job.
Linear, vs. perceptive workflow
It is still unclear which modules can (should) be combined with each other to avoid artefacts. As far as I have noticed, exposure, filmic RGB tone equalizer and rgb curve belong to the linear workflow.
You can combine them well. What about other modules? Should they be used later or not at all? In darktable blog, for example, there is a warning not to combine the base adjustments module with the new RGB workflow. Will the other modules be adapted to the RGB workflow at some point?
I ask this because besides the enormous improvements darktable 3.0 brings with it, this new factor of uncertainty has also arisen which modules to use when and the order of the modules in the pixelpipe has become more important.
This reflects a little bit my uncertainty. Contrast in color balance module is defined by contrast fulcrum. This did not give me the results I was hoping for, so I used the rgb curve in hope to better influence further increase of contrast. Since I am still lacking the confidence in handling (knowledge and experience), consider this as a child he wants to try it intuitively first.
I found it much easier to use the tone mapping module and limit its effect to the darkest areas of the image than to have to deal with the not so easy masking in tone equalizer again.
The question, of course, could be, does it make sense to integrate the tone mapping module into the RGB workflow and what would be the consequences for possible further processing steps?
I watch the movies in slow motion
On that step, it is quite obvious that your histogram still has some space on the right, so you could have been more aggressive on the white exposure. Also, since you later darken the snow to an actual middle grey-ish, you could have directly raised the middle grey in filmic to spare you some steps later.
Local contrast is good at emphasizing details around mid-tones. So again, putting the snow closer to middle-grey in filmic should help you there.
The contrast in color balance allows you to define a fulcrum very easily, so you can set whatever grey shade will be the reference for your contrast enhancement without having to worry about an off-balanced UI.
Again, local contrast would have been easier if the snow had been closer to middle grey. Haze removal is really grounded into physics (water droplets diffusion), it can backfire real quick if used outside its base assumptions.
That’s not true for all modules. Actually, the beauty of the scene-referred workflow is that any module doing linear operations on linear RGB can be reordered freely around other linear modules and still produce the exact same result (unless you use parametric masking).
Yeah, but that’s easily corrected by adjusting the contrast fulcrum accordingly.
That’s actually wrong. The mask of the tone equalizer is computed very early in the pipe, right after exposure module. Whatever you do later in the pipe has no effect on it.
See the answer I gave here: 3.0 How to get good results automatically?
The problem of tone mapping is it does a bilateral blur on log data, which is bound to produce halos. It’s just a bad algorithm.
New episode: color balance module only (my own challenge) :
Thank you! I just love the way that you use several instances
of the same module!
Claes in Lund, Schweden
Thanks!This is my own challenge to get to know the module better.
The nice (and at the same time over-straining) thing is that darktable offers countless possibilities to edit the image. Even with some modules separately, you can achieve a lot.
However, one must first get to know their potential. That’s why I sometimes experiment with single modules to see what you can achieve. If you know them, you know how to combine them with others to get good results quickly.
Sometimes I use Color Balance with Blend mode set to screen to create artificial light
Nice! That gave me an idea! Thanks for the tip!
New episode: Ruin in the snow
This is actually awesome idea! I’m trying to watch your series as a way to see how I can push some modules.
TBH @aurelienpierre’s 1h+ long videos and long articles full of science also help understand how module operate and thus be more “aware” of what influences what and what can be pushed where
Thanks again - very good videos! One bit of feedback for you to consider. In your older videos, I noticed you used the dark theme and a larger font while the new videos use the middle grey theme and smaller font. I find its much harder to see clearly smaller numbers on sliders and read small items like choosing a blend mode with the middle grey theme and smaller font.
Thanks for your feedback @Bbawt !
These changes have less to do with aesthetics and more with the processing of the images. When I used darker theme, I found that my processing was also darker and oversaturated. That’s why I decided to use a light grey theme. This gives me the best results.
As far as the font size is concerned, this is standard size for my screen resolution and I find it very pleasant because I have a better and tidier overview of the modules. For example, if I want to use color balance module with parametric and drawn masks, I don’t have to scroll up and down to see all values and settings. The same applies to the tab with the list of active modules, which I sometimes want to rearrange. This makes the work much easier.
I understand that this can be very tiring for the viewer, especially when the video resolution is low. I can try to take the darker theme when making videos.
By the way, I don’t prepare my processing for video, but process it “live” so to speak, and I need good conditions to do it properly.
I have also thought about making videos with commentary, so that it is easier for the viewer to understand why and how I do what.
That’s the purpose
Please don’t - in “cinema” mode it’s perfectly readable (IMO) and in full size it’s actually same as on mine screen But you could consider enabling subtitles - that could help since you don’t do commentary.
That’s awesome! I thought you do as most people I know doing tutorials - doing it at least twice, once beforehand and once “live”
Thanks @s7habo and @johnny-bit. I went back and retried viewing in theater mode and it was much better than standard mode or even full screen mode. So no need then to consider switching to the dark theme for recording the video.
I agree the new neutral grey theme is much better for getting correct saturation and the font size is more efficient.
I too would like to hear your comments as you edit. Especially why you might pick a certain module and editing technique over others to get the effect you want and any advantages/disadvantages it has over using a different method. I’ve been using darktable for about 5 months and still amazed at how many ways there are to do similar things - although the new linear RGB flow has been very helpful in simplifying.
This time with the voice commentary in German. I need your help for the translation into English!
IMHO there are quite a lot of videos in English, not so many in deutsch (and most focus more on modules and sliders than the image).
If it’s really a problem you could consider a no-speech version, like you did before. You used here some not used before (by you in other videos I’ve seen) modules, like Bloom. It’s certainly a good video.
Alternative: Add black titles as you did in the other videos. Even with no voice, one sees what you do (but the longer explanations will get lost)
Final word, very good video! Danke!
I love it - although my Deutsh education stopped in highschool with 2/6 score and teacher telling me that either I hate Deutch lang or her or both (she was right )
I find that subtitles in youtube are IMO the best option - I even watch one channel regarding 3d printing that sometimes uses Cantonese (or Mandarin?) but it’s subbed in english so anybody can watch it and also help translating to other langs.