Setting

At the start of 2024, the internet was actively churning with topics like "neural networks are killing artists," "machines will soon replace humans," and so on.

There exists an institution of concept art—whether for film or games, every idea requires wrapping in some sort of package to then pass the images to production. Before sets and props are created, they pass through several hands, starting from the art director's or creative director's brief and ultimately to the executor. It's been more or less like this always: an artist drew visuals, they were transmitted further. Whether it's Giger or Mike Hill—each artist had a certain vision that found its point of realization in film or games. Sometimes they set the tone for an entire franchise, sometimes the artist combined the role of art director as well, as in the case of Disco Elysium and Rostov. Sometimes the artist's work was so complete in itself that it got adapted—as with graphic novels or Stålenhag's work.

Concept art isn't about pretty pictures. In the vast majority of cases, it's about the idea, design, the "feel," if you will. Whether an image is drawn "beautifully," in a general sense, doesn't bring any substantial benefit to further production, but of course, the more appealing the work, the easier it is to "sell" it. This is a significant difference from, say, illustration, where the characteristic of "beauty" is important, the artist's "style" is valued, and the artist themselves is the main protagonist. When creating a commissioned illustration, the artist must also hit the brief, but the work's value isn't primarily in the idea-design.

Most of a concept artist's work consists of faceless sketches, top-front-side views, doodles, diagrams. No matter how much ArtStation wants to convince us otherwise, in reality we draw "digital canvases" and key art rarely—maybe 5% of the total workload if we're lucky.

We solve practical tasks: what the room layout will look like, where objects will be placed, what those objects are, what the character will wear, etc. Atmosphere is certainly part of this consideration too, and sometimes you need to push a certain mood for the hero, emphasize the direction, give space for other parts of production to work.

Conflict

The main thing that's changed over 30-40 years in this field: speed and volume.

Two important points to keep in mind when comparing yourself to masters of the past. Yes, I already said "concept art isn't about pretty pictures," but why?

Every show, game, film has become significantly larger. Cinema can no longer make do with one room and give it 30-40 minutes of screen time. Modern games have dozens of full locations broken down into hundreds of small stories, where this or that part of the environment immerses the player in lore.

It's hard to compare games like the first Doom or Wolfenstein with modern shooters. In the nineties, you showed the player was "in a new place" by tint-recoloring a texture or adding new monsters. The entire volume could be covered by one or two artists, and storytelling was handled by a single illustration from a professional—the game's cover art printed on the box.

Even with tight development deadlines, there was no problem. One illustration for the cover vs dozens of atmospheric key arts. One recolored texture vs hundreds of detailed locations. But everything changed, games became more complex and interesting, requiring significantly more content.

Entire concept art departments appeared on staff, separate people for props, separate for environments, separate for characters. Two artists (surprise!) stopped being able to handle such volume. The speed at which new content needed to be delivered also increased, not least thanks to technical progress.

If you look at how Syd Mead worked—whether on major projects of the 80s or in the early 2010s—not much changed there. Mead came from engineering, though he possessed excellent technique and a recognizable style, but worked for many years as an artist within industrial design. He used Photoshop, yes, but still remained a proponent of technically simple solutions: gouache, liners, markers, rulers and compasses. Given that at the start of the technical leap he was already famous and, I assume, quite wealthy, he didn't need to "keep up," increase speed or output volume. People didn't come to him for a hundred locations 🙂

In contrast to Mead: the ability to lay out paints, draw sketches, figure out how forms and volumes will work better, work out color on a "small" A3 format, and then move to a "full" format and paint a 50×70 cm picture—modern concept artists simply don't have that. And Mead worked exactly like that.

And the main thing that evaporates when we change medium and transition to a fully digital version, and besides that add speed and set larger volumes—isn't the "beauty" of the work. No, what evaporates is the ability to think properly. The original Blade Runner looks naive compared to its modern sequel, it was filmed long ago, productions were smaller. But it's alive, you believe in it, it's a cult work.

Remember the joke about "counterfeit Christmas ornaments"? Same situation. Modern stuff—does it shine? It shines. But it doesn't bring joy at all.

In terms of technical equipment, the profession has received what Giger and Mead couldn't even dream of—3D blockouts, photobashing, tons of free models, textures, 3D scans, the ability to correct work digitally, neural networks, after all. But none of this matters if you don't have time to think about what exactly you want to show, what design will work effectively.

Until you can live a life inside your work—it won't be convincing.

Resolution

I don't have a recipe for "how to do it right."

I'm not afraid of technical progress and what will happen to the profession next. Neural network options will certainly spoil someone's life, but the beauty of pictures isn't the end goal of concept art, and it's precisely "beauty" that neural nets push.

I believe that the main enemies for us aren't progress and new tools, but the constant production race, the demand to deliver more in tight deadlines, and the shift in focus to "beautiful and sellable."

[UPD] Two months later, at the end of April 2024, a couple of important videos came out from Mr. Feng, founder of FZD School. They address exactly these questions, here are the links:

Imagine that they brought you a box. Plain, cardboard, unremarkable, like many others. The top is sealed with tape, on the sides there are small cutouts - handles.

Think - what's in it?

It could be anything, it could be a new transforming robot, it could be a medieval character or a tuned Subaru, it could be a Lego kit, a stack of photographs, or a 25th century Space Marine self-help medical kit.

There may also be a futuristic city, a huge volcano shrouded in fog, a whole story about how an expedition team makes their way through a blizzard to find a polar station that disappeared three years ago under strange circumstances.

The box is all your fantasy, imagination and creativity. It is fueled by the visual library accumulated over the years. It also includes all the films that you watched, the books that you read, the random landscapes that you saw from the car window.

We, concept artists, refer to this box very often, almost every day. All ideas and solutions come from her, she influences most of the work.

Feel the horror when you open the box and there is nothing in it . Nothing at all, not an empty bag, not a spider, not even a wall.
All-consuming emptiness, universal nothing , it feels like you are falling headlong into a bottomless well.

This is a creative block, "fear of a white sheet", with the handbrake pulled on, with which we will not go anywhere further.

Today I will tell you a little how to let go of this handbrake, overcome fear and show a number of tools that were invented for this from the point of view of a concept artist. Go!

"Headlights and Belt"

When you get behind the wheel of a car, even before you start and go somewhere, you must at least buckle up, plan a route on the navigator, turn on the headlights, move in a chair, turn on the air conditioner or stove and check the mirrors.

Such a routine does not cause bewilderment, it cannot be neglected, because it is part of the trip. Similarly, such rituals cannot be neglected in work, almost everyone has a number of actions that are necessary to tune in to a working mood: drink coffee, check mail, turn on the appropriate music.

But what if the headlights are working, the belt is on, and the car is not going anywhere? Exactly! You need to start!

The creative process is really very similar to starting the engine, after turning the key, the car that was previously in the cold from the battery begins to rotate the starter and only after a couple of seconds (at best) the engine starts, warms up and is ready to go. Likewise, the brain needs the first impulse to start generating ideas.

For most artists, the most difficult period is precisely the sketching phase, when the main idea has not yet emerged in the head, or the moment of transition from the sketch to the first real strokes. The problem of " what to draw" is the most common problem , so let's look at it today.

Here is a list of methods that work for me personally, I will tell you about each in more detail:

1. Mr. Feng's method
2. Alchemy method
3. Constructor method
4. Neural network method

1. Mr. Feng's method

You are probably aware of Feng Zhu , a veteran concept artist with vast experience. He hosts the YouTube channel FZD School , named after his school in Singapore.

His method can be described in nine points:

• Collection of references
• Start with a good base
• Defining the main function
• Human factors and life support
• Functionality shapes the details
• Adding light
• Adding different materials
• Adding animations and sounds
• Personal detail

Feng says that in most cases 80-90% of what you draw can be safely taken from real life and references, adding only 10-20% gag.

Did you see a cool house in the photo? Draw it, and as soon as you draw it, try to combine it with that other house. Perhaps you need arches-transitions between them? Let's add a river and a water mill here, by the way, and here's a ref!

This approach greatly simplifies life, you do not bother about trifles, you only follow the general form and logic of the structure.

This approach "from reference" develops further in all points - the definition of the main function of the structure allows you to specify which references to look for, the human factor adds humanityidea, placement of the necessary attributes of a person's life - doors, windows, beds, tables, chairs, etc. The formation of parts is based on the function assigned to the building (tavern? Forge? Medical compartment of a spaceship?). Light, both artificial and natural, is determined by additional elements that need to be present in the image - lamps, chandeliers, windows, candlesticks, fireplaces, monitors or a fire. As soon as you get to this point, most of the work has already been done, it remains to add variability - think over what materials different parts can be made from, what points of interest can make sounds and what additional animation can be present: perhaps there will be a cat sitting here, but here flutter the curtain? After that, a small personal touch of the inhabitant and now you have the concept art ready. And despite the fact

Balancing the workload of the complex part of the work on the brain is another good help. Decide on the main idea as early as possible, but let your head rest. Throwing sketches according to the references and spending half a day on Monday, you automatically unload your head for the rest of the week, when you can just do a clean drawing of an already formed idea.

2. Method of "Alchemy"

If the situation is sadder and you need to invent something completely new and unusual, then the search for forms comes to the rescue. I would compare this method to the Rorschach test.

Alchemy is a small program that many concept artists know about. Unfortunately, it has not been updated for a long time, but the existing functionality is still sufficient. She is good at generating quick shapes, in her you can release your hand and watch what the subconscious mind throws out this time.

You can compare this process with looking at clouds in the sky - it looks like a fish, but a dinosaur. Of course, you can set yourself some more specific task and draw with some scope, but it is important to remember that a critical assessment should be given only after an hour spent in a relaxed state. Instant self-criticism kills creativity, and this is exactly what is required at the sketch stage. The brainstorm rule is that we do not criticize during a storm, only after .

This is the case when you don't have to worry about drawing details at all, you don't have to think about the function - only general shapes, a contour. After some time, having selected the options you especially like, you can start thinking about what exactly you want to draw and convey to the beholder. By this time, the brain will already be included in the operating mode, the motor will start and it will be possible to continue the journey.

3. Method "Constructor"

The structure of the "constructor" approach is quite simple and the simplest analogy is cooking vinaigrette, only instead of neatly chopped vegetables we will have various pieces of art. This is "~ bashing" in its true sense, but it can also be divided into separate parts: kitbashing, sketchbashing, 3dbashing and others, the meaning remains the same and is well described in the famous pilaf recipe (rough translate: "we take a bunch of dill, then a cat's ass, 25 potatoes, 15 pubic lice, a bucket of water and dick there, an armful of firewood and pilaf is ready!").

To understand how "sketchbushing" works, it's best to watch Nick Carver's performance as part of Trojan Horse was a Unicorn, link to the video. But I'll attach the work of Nick to make it clearer.

To begin with, Nick draws several houses from the reference, which he later uses as parts of one, something larger.

Such a constructor allows you to relax your brains, you do not have to come up with all the elements in aggregate and immediately have a ready-made image of the final image, but work a little bit, step by step, bringing yourself closer to the final goal. All that needs to be done in the end is to combine the pieces so that the perspective works correctly and catch the cool shape.

The same is possible in three-dimensional graphics, instead of sketches, a ready-made set of models is taken and something new is already drawn on the basis of it.

I highly recommend checking out the Bill Making Stuff channel , where the host makes cool robots from natural trash and garbage and more (English humor included).

Another follower of this technique is Thomas Scholes.

As you can see, this method works with any style and can be used in a variety of ways. From the point of view of creativity, it is always easier to use ready-made cubes, you see, the fantasy will wake up from sleep.

4. Method of neural networks

So we come to the final option to stir up the engine and see something in an empty box. Taking into account the current attitude towards this (in 2021, neural networks are still perceived by artists as "cheating"), I will immediately make a reservation that this method should be used carefully.

Nevertheless, being able to turn on such a machine is a sin not to use it. At the moment, there are at least three strong neural nets, sharpened for image generation. I'll skim through two of them: the first is Nvidia Canvas (requires an RTX card), the second is Artbreeder , which is free.

With the first one, everything is more or less clear, you draw with different "entities" and you get a picture that at the very least resembles a landscape. Not so densely, of course, the resolution is fairly small, but in order to turn on inventoring enough poke fifteen minutes and to understand in which direction you can develop what is given on the screen.

The most famous concept artist Scott Robertson has been unable to tear himself away from the canvas in recent days, it is very interesting to watch him on his instagram .

It seems to me that it makes sense to resort to a breeder in a situation when nothing helped at all. At the moment, this is an actively developing online software, and so actively that it becomes a little scary that it will be able to generate in five years.

At the moment, he knows how to create extremely good landscape solutions, portraits and characters, which, of course, greatly simplifies life at the start. For my part, I can say that as a person who regularly digs into a completely empty box with completely black walls , it was a therapeutic decision to use it.

For three days I sat and poked buttons, watched the results, and the brain itself began to slowly finish drawing the scene in my head: what could be happening here, what would I change, what would I fix.

Of course, the result from it cannot be used directly, but as a starting point for your imagination - why not?

So what's in the box?

It would be strange to end without some instructive thought. You can probably say that "everything comes with experience and time", "the more practice, the easier", but all these are empty words. Yes, skills will develop over time, but the paralyzing fear will remain.

It's scary that it won't start - for everyone, without exception. Both beginners and professionals. Everyone is afraid that it will not work out and today there will be “that very day” when I will not be able to drive at speed to top-artstation. Something will break on the way, a part of the brain will be taken away, I will not be able to make another smear.

In order for your box to produce interesting things, first of all, you need to put more in it.

Good luck!