Guides

Layer Artist Guide: Exploring different art styles for your game

Emory Irpan

Aug 29, 2024 — 7 min read

Welcome back to our series on how Layer accelerates a game art pipeline. In the first blog, we covered an introduction of what game art pipelines are, and in the second we showed how Layer can save a ton of time during the ideation phase and exploring an art style for in game characters. In this blog, we’ll continue to outline the traditional ways that artists usually explore art styles, and then show how Layer can help.

Object style exploration

After exploring our character art style, frequently the next step is to create a style for in-game objects that complements the characters. However, all the style explorations can be done in parallel, so a game team could also make an object art style first before defining their characters.

For the sake of simplicity, let’s assume that we designed characters first. For our dragon game, we also need eggs for them to hatch from!

Similar to our style exploration with characters, an artist would first do some sketches, and then they or an art director would pick a few to take to full concept.

How Layer AI can help

Time savings

Similar to character style exploration, the same time savings apply to the object exploration step.

Layer can help game art teams save a massive amount of time at the style exploration step. For example, to create a single piece of concept art, an artist might have to do the following:

6 - 10 sketches at 5 to 10 minutes each (100 minutes total)
4 - 8 hours for a full render from a final selected sketch

They would then have to repeat this process for every piece of concept art. It’s easy to see how creating a style guide could take an art team weeks to do. Layer allows artists to maintain control with their sketches, but can save that 4 - 8 hours per final illustration.

Additionally, game teams can even skip the sketching stage and just generate concepts purely from text prompts. This saves time in the sketching process and allows a wider range of freedom when it comes to exploration.

The power of AI tools allows game developers to work both quickly and at high quality. For example, this is a new workflow for object style exploration:

Use Layer’s Forge tool and write prompts for various art styles for “a blue dragon egg on a white background”
Select the outputs that the teams like the best and use them as reference images to create sketches from (5-10 minutes per sketch)
Use Layer’s BW Sketch reference option to fully render out sketches in 30 seconds each.

During this step, we also showed how you can utilize style references in Layer and combine it with other guidance material like sketches to quickly execute on a specific idea or concept. This saves additional time when populating consistent concept artwork for a style guide.

Environment style exploration

The next step of the pre-production process is exploring your environment’s art style.

It’s important to make sure that your environment matches up with the art style of your characters and objects. In 2D games, the first step is to determine what the viewing angle will be, as that greatly affects your environment’s visual style.

Similar to how artists would approach developing character art, we now need to start sketching environmental shape language to complement our existing character art style. Let’s show some example sketches in each viewing angle.

As you create the environment style guide, some technical constraints come into play for 2D games. For example, in a lot of 2D games environment assets are tiled to make level / world creation easier. Your style guide should mention what size each tile is, which also influences the visual fidelity.

Side view games tend to have less restrictions, but isometric view games need to decide on a consistent viewing angle. Most isometric games tend to be around 30 degrees, which allows the x-axis (width) of the “tile” to be twice the length of the y-axis (height). This way artists can get an even 2:1 ratio when drawing the diagonals so the tiles line up neatly without any gaps. Topdown games likewise will usually reinforce an orthographic perspective, which means there is no perspective (things don’t get smaller as they move into the distance).

An example of orthographic perspective from the Liberated Pixel Cup styleguide

Once a viewing angle is decided, it’s up to artists to start creating concepts for it.

One of the biggest parts of creating environmental assets is the need to create variations, especially for tiled assets. For example if we needed to create a cute little mushroom for our game’s environment, we’d probably need more than 1 version, otherwise it’d get too repetitive.

After creating multiple assets, you can start getting a sense of how both your characters and objects look in the environment

How Layer AI can help

Similar to how Layer accelerated the character style exploration process, its generation tools can do the same thing for environments. For example, here’s a simple prompt to get us some grassy tiles to fit with our dragon game:

While one of the generations was generated at a top down viewing angle instead of isometric, you can see how one can quickly utilize Layer to build out environmental styles.

We can get even more detailed with Layer’s AI-powered canvas.

Open an existing generation in canvas
Highlight a specific tile with the lasso tool
Crop the selection as to a new layer
Use that portion and create variations (this is useful especially with environments because you need variations of the same tile to avoid too much repetition)

For environments in particular, you need multiple variations of similar assets to really flesh out how a game’s world feels. While we could use Layer’s variety of pre-loaded styles, at this point we might have a lot of concept art already from exploring different styles of characters and objects. We can utilize these existing generations or artwork to create our own custom style.

Gather previous generations from previous steps
Go to “Styles”
Upload artwork
Caption images (should include link to good captioning tips here)
Pick some default prompts relevant to our environment style exploration (ex: a flower with purple petals, a short round bush, a cracked stone)
Train
Wait for training
Select a variation
Show using the style to forge.

Here is a quick video showing how to train a style.

Time savings

Similar to exploring styles for characters and objects, Layer accelerates environmental style exploration. Let’s review how Layer saved time in previous steps as a reminder

Layer can help game art teams save a massive amount of time at the style exploration step. For example, to create a single piece of concept art, an artist might have to do the following:

6 - 10 sketches at 5 to 10 minutes each (100 minutes total)
4 - 8 hours for a full render from a final selected sketch

The power of AI tools allows game developers to work both quickly and at high quality. For example, this is a new workflow for environment style exploration:

Use Layer’s Forge tool and write prompts for various art styles for “a red mushroom with white spots”
Select the outputs that the teams like the best and use them as reference images to create sketches from (5-10 minutes per sketch)
Use Layer’s BW Sketch reference option to fully render out sketches in 30 seconds each.

By training a custom style, artists can streamline their generation speed, meaning that there’s less time hitting the generation “lottery”. While this narrows the amount of style exploration, it allows for deeper explorations within a specific art style.

Forging images within a style allows rapid creation of fully rendered concept images, each one saving 4-8 hours of work. With the amount of work that needs to go into creating environments (or even entire worlds), it’s easy to see how Layer can save multiple days of work hours for a project.

In our next blog we’ll explore how to bring these elements together to create a visually consistent professional UI for your game.