Revelations #5: Building the Unseen

Hello! My name is Adrian and at Pixelated Milk I build levels for SacriFire. This isn’t my first rodeo with this kind of work - some of you may have already enjoyed some of my creations, for example in Hard West and a few other AA, AAA, and mobile titles. In this devlog, I’ll briefly walk you through how that process works in our studio: starting from design, through whiteboxing, and all the way to the visuals.

Before the First Block

Because SacriFire is a story-rich game, before we can start designing anything level-related, we need to know the story first - mostly so that we can break it down into the smallest design-relevant chunks possible. As such, the game’s script becomes our launching pad into the whole process - it serves as a guidepost for what should appear in a given segment of the game, what locations will need to be built, and where the player will ultimately end up. You can read more about creating the story in one of our previous devlogs, but of course, that’s not what we’re here for today!

The Shape of the Journey

When the script has been thoroughly dissected for information, it’s time to get cracking. One way or another, my main design work starts with a blank sheet, be it paper or its digital equivalent. This is where the foundation is laid: a good framework plan that makes designing easier and becomes the seed of the eventual documentation.

One of the main pillars of level design in SacriFire is theming. Within the team, we commonly call these splashes of a theme “gimmicks.” Effectively, this means that apart from their core platforming and combat sequence, each dungeon should include an extra element that is characteristic only of that specific location. The goal here is to add variety to the gameplay and avoid repetition; at the same time, any such gimmick should be simple enough to remain an addition to the main gameplay, not something that overtakes it to the detriment of the general experience.

To choose a good theme, you need to carry out proper research. After all, SacriFire is a retro-style game, so reaching back to the roots of the genre makes perfect sense. It’s obvious that thousands of games came out before ours, similar or otherwise, and there is a lot we can learn from their experience. Thanks to them blazing the trail, we don’t need to reinvent the wheel - just learn from their examples.

Another important element of a level is its context. Context influences form and structure. It dictates the elements that need to be taken into account in the shape of the whitebox. A prototype of a temple interior, an underground railway station, or a bandit base will all be drawn differently. Context defines the locations and sections of the game level, as well as the sequence of transitions between them. Combined with story assumptions, it builds a flowgraph: a point-by-point overview of the actions and events that follow one another and affect the player.

All of this is then enclosed in the level documentation. That documentation describes all the elements I’ve mentioned, supplemented by pertinent information provided by gameplay designers: the player’s available abilities, equipment, enemy descriptions, event order, quest objectives, and character/level interactions. Documentation is an essential production tool. Video game levels don’t exist in a vacuum, and almost every member of the team contributes in some way to their eventual creation. Programmers create components and interaction mechanics, artists make models, the animator handles cutscenes, and our sound specialist takes care of SFX and ambience. Needless to say, all of this would be extremely difficult to coordinate and integrate without a detailed description. Once the theory has been put on paper, the prototyping stage begins - we move on to whiteboxing.

Before the World Has a Face

Essentially, whiteboxing means building a playable prototype game space using simple shapes that define the framework of the gameplay. At this stage, visual content is mostly ignored. We do this because making changes is much faster in this form than in a version already “armed” with level art. Trying to modify a more advanced scene usually ends in a domino effect: one change leads to another, then a third, then another, and before you know it, the work has to be done all over again. Some level designers like to start this stage with a sheet of paper and a pencil, but I personally prefer working directly in the editor, mostly because it makes live testing, iteration, and establishing scale much easier.

Scale defines the length of time it takes to traverse the stage, the amount of space available for gameplay, and the amount of content that needs to be built. In SacriFire, we have standardized our level sizes. We developed that standard through trial, error, and testing. In the end, we decided that the demo dungeon should serve as our benchmark for both gameplay time and physical size.

Once the level is enclosed in space, choice points need to be placed within. The ability to make decisions that affect the course of a level is crucial to the feeling of agency. The player shouldn’t feel as if they’re simply running forward on rails, ticking off one action after another. These decisions don’t have to be matters of life and death. Oftentimes, simply placing a platform or a staircase on the player’s path is more than enough. Actions themselves can be easily tied to controls and measured through metrics such as APM.- This can suggest when something is too boring, too dense, too easy, or too difficult.

Overcoming obstacles, in turn, is responsible for building satisfaction. The simplest obstacle is, of course, an enemy to defeat. Here I could go into SacriFire’s unique combat system revolving around time stopping around you, and how that factors into the overall challenge... but I believe ... Krzysiek already mentioned a bit of that in his own devlog. ;-) Apart from combat sequences, SacriFire’s levels also employ platforming sections and “light” puzzles as obstacles.

Interaction with the game world connects the gameplay layer with the visual layer. As such, the game world is never static, which means every level contains places where the player can interact with their surroundings. Usually, these interactions are based on switching, moving, collecting, and unlocking objects. Connected by a network of dependencies, these interactions create puzzles to solve, which we also treat as obstacles to overcome.

To further refine the flow of sequences, individual sections of a level often need to be framed. Consider, for instance, the concept of gates. Gates always give us confidence that the player will not skip any key point on the map, that they will be in the right place for story events to be handled, and that - from the content side - we will have the ability to switch things on and off, which is necessary for technical reasons. Such gates don’t always need to take the physical form of doors, mind you; more often than not, level sections can be separated by narrowing the available space or, for example, by adding an elevator.

Finally, as the whitebox nears completion, it still needs to be properly entwined with game logic. In level design, nothing happens by itself: every collision, every ledge the player can grab, every trigger, and every checkpoint is placed and set up by hand. Each game logic object has to be checked to make sure it works properly, that the dependencies between objects are correct, and that they don’t generate unexpected results. This is not the hardest part of the job, but at the same time, it can be the source of many problems.

Pieces of the World

When it comes to environmental art, the work begins long before there is even an idea for any specific level. At Pixelated Milk, we have three people working simultaneously on level design and level art, one 3D artist responsible for models, one pixel artist who also handles the textures, and one VFX specialist. Of course, every particle, every model, and every texture is hand-crafted and made from scratch. Our two artists have created over 2,000 assets in total, which translate into 4,183 objects available for use. These assets are combined into thematic collections, and their consistency is carefully curated by our Art Director - the man whose devlog you have also already had a chance to read. Those collections are then used to create level art. Broadly speaking, work on level art begins only once the necessary collections are complete and the whitebox is playable.

Where the World Takes Shape

Level art is, ultimately, the art of compromise. Skillfully building the visual layer requires balancing gameplay, artistic, and technical requirements while taking into account the available collections of 3D models, textures, and sprites. It is never the case that I can simply build whatever I dream up. The gameplay has to remain readable, the visuals should support the player (not to mention be built from the assets we already have so there is no need to increase their number), and the amount of stuff on the screen should not cause the hardware rendering it to spontaneously combust. :-)

The visual layer is built around prototype locations where artistic composition can be applied. If you think back to your favorite games, you can probably come up with some places that gave you that “wow” effect and made you step back from the heart of the action for a moment. If so, you owe that to carefully built frames that use tricks as old as art itself. Attention inside a frame is guided through the skillful use of shapes, directing the viewer’s eye toward points that were carefully and intentionally planned by the artist. Frequently, building just a few large forms is enough to make an image interesting, as smaller objects only serve to fill it with detail. Becoming too attached to detail is a common mistake among beginner artists. Filling the view with too much of it makes it harder for the eye to focus, causing it to wander, and ultimately renders the image chaotic and noisy. The key to make it all work is balance.

Naturally, the use of color is treated with similar care. Color theory is a field that describes - with almost mathematical precision - the many ways of combining colors into beautiful harmonies. It is color that defines the mood: the golden-white districts of the Spire obviously suggest wealth, but they also bring a sense of energy, passion, and joy. The green-blue Erebus is an oasis of calm, longing, and melancholy, while the deep blue-violet tunnels of the mines are meant to evoke mystery, darkness, and fear. On top of that, color space should be subject to gradation, whether on individual models or progressing across an entire scene. This makes the colors even more interesting.

It is hard to talk about color without touching on lighting. Lighting adds a sense of depth, which is especially important for us because SacriFire is a 2.5D game. Without light, it would be easy to forget that the game operates in three dimensions, and that is definitely not what we want. Both light and shadow also serve as a great complement to composition, because they can be used to complete shapes.

Additionally, it’s worth mentioning there’s also the magic of post-processing, something that only really kicks in during the final stages of creating level art.Players most often associate it with effects such as bloom, blur, or DoF. Adding cool effects, however, is not its sole function, as we employ post-processing to finalize work on color through color grading.

Through Fresh Eyes

When designing levels, it is difficult to look at them from the point of view of a new player. After all, when you are building them, you see everything laid out in front of you. Working on a scene encourages tunnel vision: focusing on some aspects at the expense of others.. For one, there’s the eternal tight rope of properly balancing the experience you’re creating. After all, a game’s difficulty is largely subjective and tends to vary with one’s familiarity with the game or genre in question. That is why, at regular intervals, we organize internal testing sessions after which every team member can share their impressions, observations, and criticism. On top of that, our team is supported by a QA duo who tirelessly hunts for issues to report. Their work is worth its weight in gold to us, and they always give us something extra to do. ;-)

One Last Door Gate

As you can see, the process of creating playable levels requires something more than simply making models and placing them around a scene. It is long, iterative work that requires difficult decisions at every stage, and each of those decisions in turns comes with a whole chain of consequences.

In SacriFire, we currently have 270 level scenes. Among them one could enumerate sprawling cities, intimate interiors, combat-dedicated arenas, minigames, cinematic sequences, and dungeons. Each scene requires an individual approach, but all of them are connected by the tissue of the same creative process.

If you have any questions about any of the stages discussed here - or if you are interested in other aspects of SacriFire’s production - let us know in the comments. I’ll see what can be done. ;-)

And while you’re here, don’t forget to add SacriFire to your wishlist and follow us on social media. We’ll be sharing more details about the upcoming playtest soon.

https://store.steampowered.com/app/1661330/SacriFire/