I'm going to tell you here just a few important things that will help anyone who is ready to try it to get started. Actually, I think the best way is to try it out, so I'm attaching the Gimp file. You can download it, just see how it's done, and even experiment with it. Gimp is a legal free Open-source application: https://www.gimp.org/ Basically everything in it is organized very intuitively, I'll talk about some things that are not so obvious right away, and they are really worth telling you about.
For example, people often underestimate how convenient keyboard shortcuts are. It may be too difficult to use them right from the start, but we get used to them very quickly and then do everything as before :)
Clone stamp is also not an immediately intuitive tool. It transfers pixels from one place to another. But besides just cloning, it is often used for face retouching. If you set the opacity to 30% and turn on alingnment, it's very convenient to soften wrinkles, pimples, or other skin imperfections, or lighten the skin. My sample-file has a separate layer with such manual retouching. The layer names tell you where it is. You can try adjusting its opacity and get different results.
We also have a wide range of filters for different types of images. There's no point in telling you much about them — you just have to try them. When it comes to face retouching, we mainly use different types of blur and sharpening. In addition, I'll show you the incredible G'MIC, which can be used separately or as a plug-in in Gimp. It will be in the filters menu if you install it. It contains a collection of extremely fancy, but classically algorithmic filters. I hope that I have managed to provoke you to try it yourself, and I wish you success and maximum satisfaction. There are also earlier posts on this blog that deal with 2d pixel/vector graphics (few basic steps, required to make various graphic compositions from small flyers, labels and booklets to big calendars and posters), you can check them out later, they are generally still quite relevant: http://blotproject.blogspot.com/2013/12/cddvd-cover-tut.html
You can also see a few examples here, all done in the ways I've already told you about, I just have to say something quite obvious, and in this case it really works — skill is achieved by practice.
Does it make sense to do this at all when we have incredible AI capabilities, or, more precisely, machine learning (artificial neural networks)? I think in some cases, yes, there is. I am over 60 years old, and for half my life I have been doing what I am doing now in a virtual (digital) form, in a "molecular" form, so to speak. And has "molecular" art disappeared during this time? And there is another important reason why I think so. Sometimes people still care about authenticity. If there are their loved ones in a picture, people often want to have a picture that may be worse in quality, but it's genuine. The fact is that nowadays, artificial generative networks are engaged in improving images or videos. It will probably remain so. That is, what they produce is a some fantasy, they are hallucinating, as data scientists call it. In the original, we have some pixels, and nothing more. What is there is there, the rest has to be generated, and this is a limitation that will remain for a very long time. Perhaps at some point in the development of civilization, the surviving civilizations will learn to recover information that is inevitably lost from our field of vision. It doesn't disappear completely, at least according to physicists 🤔. But first, we need to survive. We have an incredible, promising future, but only if we manage to preserve the achievements of our civilization and multiply them.
