The pre-production stage is where a majority of the brainstorming and ideation occurs. It’s also one of the most important steps in the game art pipeline, as it informs the production and post-production stages.

During the ideation step, a game team starts from a “blank slate” and begins to generate ideas on what they want their game to look like.

If a game team already has other ideation documents like game design doc or a pitch proposal, those are generally a great start to visual brainstorming. During this stage, moodboards might be created compiling similar ideas from other media franchises as inspiration points.

For example, let’s say our game designers have a cool new idea for a mobile match-3 game involving dragons. You play through levels and get resources to build up a dragon village. Since dragons fly, we want to make it a floating village in the clouds. This is a pretty common game concept since we’re able to find various examples in existing games.

Within our moodboard, we also pick a brighter color palette since we want the game to feel inviting to players of all ages and demographics.

Using Layer AI’s concepting and exploration tools, artists and art directors can turbocharge the ideation step by prompting AI models for ideas based off existing moodboards. They can even create completely brand new inspiration material by utilizing Layer’s forge tools.

These generations can then be incorporated into moodboards to help refine and define the visual style for a game. Here are examples to show how Layer can interpret same prompt in different art styles in a matter of seconds.

In many games, characters are the first thing to be defined, since they drive so much of the game’s mechanics and storytelling.

In the style exploration step, the team begins to solidify the art direction of their project. This involves exploration of multiple different art styles and how they would be visually represented in a game.

Character, object, and environments will start to be explored to figure out the overall aesthetic mood of a game. Even user interface (UI) elements will also have to be explored so that the mood of the game is communicated well to the player through elements they interact with.

In the case of our dragon game, artists would need to start creating style guides for characters. These would each be done individually, one at a time. When exploring styles, concept artists would have to work quickly to create something that has strong enough fidelity to be evaluated, but still be able to be done in a time-efficient manner.

For example, let’s say we’re trying to define how a dragon character would look in our game. We’d first do a bunch of different exploration sketches. For an average artist, each of these sketches might take 5-10 minutes. They’re meant to be low fidelity since investing a ton of time into each would not be efficient.

Our art director, designers, and maybe even marketing would then go through the sketches and see which ones most align with the game’s audience and overall feel. Once we’ve selected a few to refine, we begin to render out the concept artwork. For our little dragon here, we’ll draw out the rest of the body according to the cute style that we decided on.

Afterwards, we’d normally need to spend a few hours rendering out the dragon to something that feels like a vertical slice of our art style.

We’d need to do this with multiple characters, objects, environments, and more to create a cohesive style guide that represents our game.

Layer AI is ideal at exploring how a similar character looks in different art styles. With tons of pre-loaded styles to choose from, artists can greatly accelerate the envisioning process.

Instead of having to sketch and take those sketches to full renders for concept artwork, developers can quickly explore what something looks like without having to commit significant time. Simple prompts mixed with different art styles can produce a lot of high-fidelity results. Let’s see what kind of blue dragons we can make with Layer:

Here we use a wide-use model, DALL-E 3 from Open-AI. Because it’s a more generic model, we specified that we wanted a mobile game art style. DALLE-3 is only available with the Enterprise plan, if you are on lower plans you can use an equivalent base model like Stable Diffusion XL.

We can use the “Action game” style to see what a dragon in a action video game might look like.

Or can select from 40+ art styles to see what a dragon in a action video game might look like. Here are some examples:

You can see that it’s really easy to generate high-fidelity concept artwork with Layer. This allows much more rapid style exploration which ultimately gets your game team to production faster.

We could even take one of our earlier sketches and use it as a reference material, by hitting + sign on the prompt box and selecting reference type as B&W sketch or simply paste your sketch directly on the prompt box.

Let’s say we wanted something that was less cute. Normally an artist would have to render something completely new, but with Layer it’s easier to see what else could work for our game

If you feel the provided options aren't enough, remember Layer allows you to upload any custom models and still use the platform with full feature set. A good place to check out is Civit.Ai who offers a very large repository of custom trained AI models.

It’s easy to see how Layer can accelerate a game’s pre-production process. An artist could go from spending hours per final concept or art style exploration to mere minutes. This allows bolder, more varied exploration within the same amount of pre-production time, ultimately helping solidify a game’s art style more quickly and decisively.