In a recent episode of understanding darktable, Bruce Williams linked to this tutorial on color grading by Joanna Kustra. In it, she discusses using color grading creatively, such as using complementary colors in the subject vs background to provide more contrast and draw your eye to the subject.
Does anyone have a tutorial, guidance, or example processed raw+xmp files that showcases how to do this? The photos she uses in the video are portraits where she is able to control the entire scene, but for other situations like nature photography, you often can’t choose the colors that are present together. So, how can you adjust colors in post to achieve this effect? For example, taking a shot where the background and subject are not complementary colors and using color balance rgb to change the colors to be complementary? Or how to add color to a sky where you just missed the sunset?
Not a tutorial, guidance or anything, just want to share my thought, because it is an interesting topic.
This is quite common thinking, but in my opinion, most people (especially the ones that deals with the creative industries —photography included) already have the sense to choose the colour they want in their photograph when they are taking photo (through framing, composition, time of the shoot, set design etc.), even unconsciously. The question is, how to be more aware of it.
Rather than jumping to colour grading, the first one to be aware of is colour harmony (Joanna Kustra talk about in that video too), there is a lot of colour harmony that usually presents in nature (especially Analogous colour), but one thing to remember is that there are two types of the colour wheel, RYB and RGB for determining the colour harmony (personally I like using RYB more).
This is one of the earliest assignments in my photography class, after they learnt the basics of camera handling (I teach graphic design, photography is only a part of it), to make my student more aware of their surroundings when taking photographs. This is their result last semester (not the greatest photograph out there, but they learning from zero, and all of the photos must be SOOC, in this case, they have to take complementary, analogous, and triadic colour harmony based on the RYB wheel)
In my case, after I tried my best to take the photograph, and after the correction stage, is to “correct” some of the hues, using colour zones or something similar, even a small adjustment makes a big difference here, sometimes it is a personal colour preference, just like asking ten people which red is the Coca-Cola’s red. After that, colour balance RGB is used to set/add/enhance the overall mood/colour direction based on the image/story, sometimes multiple copies of the module with the mask are used if needed.
Channel mixer is helpful here. For instance, if you have green grass background and violet dress foreground, but you want direct complimentary colours (green-magenta or olive/lime-violet), then you can either turn violet into magenta, or green into olive/lime, using channel mixer. So the tutorials to seek out would be those on how to use channel mixer.
It’s a good point. I wonder which is considered more perceptually correct?
I can’t say much about which is more perceptually “correct”, but when it comes down to the history of the wheel itself, RYB colour wheel (sometimes called “artistic/traditional” colour wheel) is based on subtractive paint colour mixing, that’s why it’s commonly used in painting, and most people that used to seeing a painting feels like RYB wheel is more pleasing to the eyes (what’s pleasing to someone is most of the time based on their experience, after all, that’s why it varies from people to people, culture to culture).
On the other hand, the RGB wheel (sometimes called the “scientific” colour wheel) is based on additive colour mixing. And because it’s the colour model that is used in a screen, colour harmony based on the RGB wheel will feel more vibrant than the RYB ones on the screen.
Which one to used depends on what effect you want to achieve, and more importantly, always trust your eyes when tweaking colour, because colour, in reality, is rarely a solid colour. But be aware of the two wheel especially when you’re creating a mood board for a client/project, and if you used RYB, remember that colour mixing in the software will still be performed in RGB (i.e. if you want to neutralize some overlapping colour when grading, you must think in RGB, even when your harmony is based on RYB).
Sorry for the long post, I guess just saying that “this one” is more correct than other is really hard (at least for me)
To achieve this I would use Color Balance RGB. First use a parametric mask to isolate the color you want to change. Then use the Hue Shift slider to move that color to the desired color. Then use any of the global color sliders to dial in the exact hue that you want. Add another instance of the module to change another color.
@garibaldi interesting! I also watched Joanna Kustras video and wanted to try out what I learned in darktable.
I used the technique @paperdigits describes to somehow isolate color areas with a mask and changed hue in color balance RGB in combination with a look at the histogram in vectorscope mode to see the “color-wheel” and the distribution of colors. That way it worked quite good to get that interesting harmonies, at least for the complementary harmony.
Yes one way if you want to change just one color is using cb rgb to select the color range and change the hue, preserving tone perception.
If you dob’t want to change it everywhere, use a paint mask to mask parts not wanted.
Other way would be using a lut3d table to teawk colors and change them as you want and create a look or atmosphere.
@adrs Thanks for the detailed posts about RGB and RYB. I agree that you can (and should) use framing and composition to try and create the feel (and color contrasts) you’re looking for. However, some situations (e.g. fast moving subjects, an impromptu shoot where you can only shoot briefly, etc) don’t afford that amount of time beforehand. In those cases, it would be nice to be able to enhance the constrained scene you had to work with in post.
Yes thanks, I had watched this one a few days ago! This is definitely close to what I’m looking for, but I wasn’t sure how I would go about doing this without a reference image to work with (e.g. making up my own color harmony). I’m looking forward to @s7habo’s upcoming video!
Also you may be able to use a trick I have seen where people have essentially used the watermark module to create an svg file of a swatch or gradient to overlay on the edit to use for matching…there was an example some time ago here using this technique to match a desired skin tone…but it could be used in grading as well…
Thank you! I had to change the structure of the series about colors in darktable a little bit, because otherwise the episodes become much too long. Accordingly, I decided to introduce the color calibration module first in the first episode of the series. This episode will also be a little bit about color grading. So, it is worth to watch.
Yes I do agree with this, that’s why I always think that taking and post-processing a photograph cannot be separated to create an image I’m looking for (i.e. limitation of camera + limitation of the scene + limitation of editing stage), and make the most of it.
One of the website i often use to search for composition and colour grading inspiration from a movie is this one https://film-grab.com/, and https://artsandculture.google.com/ if need inspiration from an art/painting, their colour explorer is really helpful.
Yes, thanks for the good resources! I’ve used Palleton before, but after creating the wheel I want to use, I’m not sure how best to transfer it to my image. For example, you cannot just copy/paste an RGB hex value into Color Balance RGB. The watermark trick sounds neat - I’ll give that a try. I
need to figure out how to transfer each of the colors I’ve selected into the image - e.g. a different instance of Color Balance RGB for each color, each with a mask to select a different set of pixels?
)
