I don’t want to throw all the questions about Boris’s lessons into one topic.
I have a pressing question about Darktable tools.
Very often I see that when working with color balance, the author (and not only Boris) works either with the color mixer or with the rgb color balance tool.
Often it is not accompanied by any comments like, here we will use such-and-such tool, and here such-and-such because-and-such. The author simply takes a certain tool and starts telling you how he will use it to change the temperature and tint of the photo.
The man wants to emphasize the greenery on the leaves. Moreover, he uses a mask. That is, it is easy to process such details in the color mixer, but no - he uses the color balance tool. He does it without explaining why he uses this tool and not the color mixer.
It’s often possible to achieve similar results with different tools. But in this case he used color balance to tone shadows and midtones differently what is quite easy to do in this tool. Also the channel mixer would affect all green subpixels in the image, the masked color balance does not.
However you can also ask him in the topic he created for his videos.
The color balance tool is more like a color compressor, when you don’t want to emphasize a color, but rather “cover” it with another color. What is more convenient (sampling by lights, midtones and shadows) is not always correct. You can also go with masks.
And then. If “green pixels are affected,” should we avoid the color calibration tool altogether?
I am not all for the color mixer, but I need a clear understanding of where one tool is used and where the other is used.
Yeah - I’ll ask him there. It’s just, on youtube comments have a habit of sinking in if the clip is already old. Often no one responds. Already asked his other topics there, sort of. That’s why I asked the question here. He himself said that you can ask questions about his videos here
You don’t have to ask on YouTube. There’s a topic on this website where Boris often responds to questions in depth. Also @nwinspeare has two videos on the channel mixer in color calibration that go into the maths behind its use and practical application:
Also, this particular video from Boris goes into a little more detail into the reasoning:
I know what a channel mixer is.
It’s just that people can spend hours explaining the theory and practice of working with this tool, and then bam! - “and now we will paint the picture with color balance rgb” Without explaining why with the color balance rgb tool.
Of course, this should be understood as "use the tools of your choice, analyzing the behavior of each tool - you have a head and you’re not a robot. But then, I beg your pardon, couldn’t a man himself have guessed how to work with the tools? Without lessons?
I’m getting a little emotional here. I apologize, because I am reasoning like some kind of consumer.
Well, okay, I’m not going to look good. But others will read the answers and it will be useful for them, if the topic is discussed. I apologize.
When you write color mixer (which, I think, does not exist in darktable), do you mean the color calibration module with the simple channel mixer preset?
So for me the “channel mixer” is a way to impact color by directly impacting the value of pixels in a chosen color channel. The result on the color and if not balanced the brightness of the image depend on what changes you make and how many channels you manipulate. Unless you mask it in some way the changes are global.
The color calibration module targets tonal ranges. It has its own internal mask to define the boundaries and roll offs for those regions. You can also use the traditional masking to further limit this like you would in any module.
Traditionally the CB module had a version of the 4 way tab. It had pickers that would automatically add -180 hue to offset color casts. It still does. You can also select the color picker and add 180 degrees to enhance the hue present in each tonal range.
You also have a luma slider …so basically you have a set of LCh sliders for the tonal ranges as they are defined in the CB mask tab.
The newer tab has added the extra features to allow the user to target chroma saturation and brilliance again partitioned to impact the ranges in the tonal mask.
So its up to the user to decide what tool modifies the color in the most efficient way for them. You can also make very good use f the other tabs in color calibration ie the brightness and colorfulness … sometimes these can be easier for people to understand and modify than the channel mixer…
EDIT…you will see Boris also use the color lut module from time to time. I use it often as well. It can be one of the best and quickest ways to tweak skies and foliage…you can either select your sky or foliage with the picker and modify the chosen patch or just modify a couple of the green and or blue patches…you need to be careful as you can get the image to degrade but for these simple changes to say deepen a blue sky it also works nicely
Most likely, the tools are just used impromptu. Through experience, one simply chooses more of the instrument one likes.
For myself, I made a conclusion. Color calibration is more aggressive than the (conditional) hue/saturation tool. It should be used very carefully, clearly understanding what you do. As an example of such an understanding, so as not to be demagogic, it can be used in two cases:
If we think that the camera matrix of the shooting device has somehow cheated during shooting and the hue has shifted by some degree, or we need to adjust the saturation of each direction r,g and b. To do this manually - here you need to have a gigantic experience, because just a “good” impression of the changes made after an hour, not to mention a day, can seem terrible. Here the tool acts as a code lock of 9 wheels, and to use it purely on the basis of logic - the same will be the result. Photography is not logic. It is art (imho).
Decor. artificial color shift. used in trend approaches - tilorange etc. Purely commercial processing for the user who likes this style.
On the plus side, work in a linear gamut, which causes headaches, as the tool will still have to learn how to use in this case.
RGB color balancing. This is a more human tool. If you take the tab of the operation with shades, it reminds me of the white balance tool. Only here you can adjust the white balance separately for highlights, midtones and shadows. The disadvantages are that it doesn’t work in a linear gamut, which leads me to think that the tool is not so much for white balance correction as for image stylization. The same white balance tool, for example, can work in a linear gamut.
And this is how things are in Davinci Resolve. You can switch the node with the lift, gamma and gain tools to the linear gamma mode. In this case, we still have the ability to operate with shades in different brightness ranges, and you can only in the light. There is also a layer overlay mode - color.
In general, everything is very complicated and not clear. We need to digest and digest the darktable view
His lessons are good - the style, the calmness in his speech, but his own experience decides everything.
Perhaps not always but many times he will have explained it in a previous video and he tends then not to reinvent the wheel…it can be a matter of going back far enough…but as others have said if you have a specific question he will usually do his best to clarify…
Oh, that’s a lot of questions. I am very happy that you are interested in my videos. And I appreciate the questions, of course.
By the way, you can ask the questions directly to me.
Let’s get started.
There are two different approaches to how you can address a topic. One can
first explain the tool, and then demonstrate how it works with some examples.
you can start editing and when the time comes to use a tool, first take the time to explain the tool and then do this step with the tool.
"2 "takes much more time and breaks the flow of editing.
I rather choose “1”, because otherwise I will need much more time for each video than usual - because every time I take a tool I have to explain it as well - which would be boring for those who already understood it from previous examples.
This approach is of course more difficult for some users, especially the beginners, because they have to watch the explanation video first to be able to follow other episodes. Besides, they have to be able to remember things.
However, this method also has some significant advantages:
If someone is really interested, he will make the effort to watch the explainer videos several times,
he will ask the questions and clarify what he did not understand
He will also study the documentation much more intensively.
he is also much more motivated to try the things himself instead of consuming served solutions. This is also important to understand the principle behind it, so that it is easier to transfer it to other examples instead of just blindly copying the steps.
The explanation of how the color balance works can be seen in the previous episode…
…If you understand that, then it becomes clear why I chose the color balance:
I didn’t want to change the hue of the area (which you will do with Channel Mixer) but to give the leaves a “bluish” tone. I used only color sliders in “4ways” tabs for shadows and middle gray area to adjust only these areas color-wise exactly for this reason from your quote…
…you don’t change the original color, but only “emphasize” the specific area with a different color. You can control the strength of the effect with the “chroma” slider (the color that covers) in the “4ways” tab, combined with the saturation functions in the “Master” tab (color intensity of the original colors).
Color calibration is used primarily for color corrections related to white balance. It affects three primary color channels (red, green and blue) - their combination, color intensity, brightness and also serves as a conversion to black and white with “gray” tab.
Color balance RGB is used to control contrasts, saturation and create a “color mood”.
But that doesn’t mean that you can’t use both tools creatively. That’s why, for example, in my videos about Channel Mixer, I showed how to use it for both white balance and color grading.
A good explanation about these two tools was also done by @priort
But before we get into too general discussions, I am very willing to answer concrete questions and for that - as far as my videos are concerned - it is indeed better to use the already existing thread, to which @paperdigits and @kofa also referred to. Otherwise we will lose ourselves in articles.
Again I also put a link here:
