Skip to main content

Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
TagsGame Engines

What kinds of 3D assets do you actually use, and what does your workflow?

A topic by EmaceArt created 22 days ago Views: 239 Replies: 5
Viewing posts 1 to 3
(11 edits)

Hi!

I’m a 3D environment artist – for 15 years I’ve been designing worlds in Unity and creating modular 3D asset packs. For most of my career I’ve been on the “environment” side of development: from the first blockouts, through level design iterations, to the final art pass and real-time optimization.


I don’t churn out generic “blocks” – I build hand-crafted locations with personality. I push low-poly beyond simple cubes, sculpting and refining shapes so they stay lightweight but feel almost like photogrammetry, catching light in interesting ways instead of looking like random boxes. I design both natural environments (forests, swamps, cliffs, fields, cemeteries) and urban ones (small towns, ports, industrial zones, retro cities) for devs who want worlds with a distinct vibe, not just filler geometry.

Here’s where you can see this in action: my ArtStation portfolio. I post cross-sections of worlds I build as personal/hobby projects – necropolises, swamp towns, retro sci-fi spaces and stylized settlements. I also handled full worldbuilding on Survival Machine (Steam): from blockout and choosing modules/props, through set dressing, mood and lighting, to gameplay-ready optimization. For me it’s a key reference case that shows how I work in real production, not just in a single “beauty render” scene.


Unity is my main engine – I know its quirks, workarounds and traps around modular environments. I target them with carefully chosen add-ons and my own AI-assisted tools that strip away dry technical work, kill repetitive clicking, and let me spend ~90% of my time on art… but that’s a story for another thread ;)

Godot, Unreal and Roblox are engines I want to explore deeper from a pipeline angle: how their prefab/Blueprint equivalents behave, how to handle LODs and collisions, and what asset formats devs actually expect. Long term I don’t want to just drop “raw models” there, but curated sets tailored to each runtime.

I’m publishing more and more stuff on itch.io, and instead of guessing based only on my own research, I’d rather ask developers directly what you actually use and what really helps you ship. I keep an eye on tags, trends and devlogs, but nothing beats talking to people who are knee-deep in their engine, juggling sprints and pushing builds.

Over the next quarter I’ll be posting only free packs here – from small “single shots” (test scenes, single biomes, mini locations) to trimmed slices of my bigger Unity Asset Store packs, reworked as self-contained modules, not scraps. The goal isn’t to dump full commercial sets, but to share ecosystem fragments you can actually build with, test your pipeline on, and see if this world-building approach fits your project.

So now’s a good time to drop your needs and gaps: solid sci-fi corridors on a modular grid, stylized fantasy villages, low-poly graveyards or swamps, or specific “utility” scenes for lighting and optimization. Treat this thread as an open line to an environment artist – ja bring experience and asset-market insight, you help steer priorities so what lands here serves real games, not just pretty screenshots. And yeah, comments under my devlogs are very welcome.

1. Which engines are you currently using?

Unity / Godot / Unreal / Roblox / other?

For which engine do you most often look for ready-made support?

1. modules for level art (walls, floors, ceilings, terrains, biomes),
2. models/props (filling the world, environmental storytelling),
3. lighting/post-process (test scenes, profiles, presets),
4. optimization (LODs, proxies, collisions, lightmap-ready meshes),
5. prototype scenes (ready “playable spaces” to immediately test gameplay)?

2. What kinds of projects do you work on most often?

More like:

  • large, continuous worlds (RPG, exploration, survival, base-building, immersive sim), where biome consistency, sane modularity and performance on bigger maps really matter, or

  • smaller productions: mobile games, arcade titles, roguelites, short levels/puzzles, where clarity, readable flow and fast iteration matter more?

You can also clarify whether your projects are mostly:

  • singleplayer (more cinematic pacing, heavier focus on mood),

  • co-op / multiplayer (more emphasis on readability and performance),

  • jam/experimental (lots of prototypes, little time to polish assets).

All of this has a big impact on how you design modules, density of detail and performance budgets in practice.

3. Setting and visual style

What kinds of worlds do you most often place your games in: fantasy, sci-fi, post-apoc, cozy, horror, realistic contemporary settings (cities, suburbs, industrial), immersive sim, retro-futuristic vibes, or something very niche?

And what kind of visual style do you usually work in?

1. realistic (closer to photorealism, more focus on detail and materials),
2. stylized (strong shapes, simplified textures, more graphic look),
3. low-poly (deliberately simplified geometry + clear silhouettes),
4. retro-pixel / PSX (limited palettes, “noise” and aesthetics of older generations),a mix of those approaches?

4. Where does environment work “hurt” the most?

  • time – scope gets slashed as sprints and deadlines take over, and environment slides to the bottom of the priority list

  • lack of experience in level art/composition – it’s hard to find a good rhythm of space, readable paths and strong POIs without solid examples and a modular base;

  • performance – polycount, LODs, draw calls, lightmaps; it’s easy to overshoot detail and then fight frame-drops;

  • lighting / post-process – everything is “kind of okay”, but the mood just doesn’t land, or the scene looks good from exactly one angle,

  • motivation drop – the original vision was ambitious, but lack of time/tools/assets turns the world into a simplified version of the idea.

If you have your own classic environment “pain points”, like “I always get stuck on X when doing environments”, please add them. That’s 

5. What kind of 3D assets do you want to see on itch.io?

Do you prefer engine-specific packs:

1. ready prefabs / Blueprints,
2. test scenes with correct pivots, collisions and LODs,
3. post-process profiles / lighting setups,

or more engine-agnostic packs:

4. FBX + textures + a few variants,
5. neutral materials for further tweaking,
6. sets built with a “engine-agnostic” mindset you then plug into your pipeline yourself?

Are you more into modular building blocks for your own worlds (walls, corners, junctions, biomes, terrains), or ready scenes/biomes you can tear apart and adapt to your project while preserving their rhythm and composition?

If this forum section allows polls, I’ll add a simple one for: engine / game scale / setting.
If not, you can answer in a quick format like:

Engine: Unity / Godot / Unreal / Roblox / other
Scale: large, continuous world / smaller game / arcade
Setting + style: e.g. “low-poly fantasy”, “retro sci-fi”, “dark post-apoc”
Main pain point: time / experience / performance / motivation / other
Asset type: engine-specific packs / universal packs / mix

Based on your answers I want to better steer future environment packs on itch.io so they actually support production and your daily workflow – not just look nice in a thumbnail. Thanks in advance for any concrete replies, from detailed breakdowns to short, to-the-point answers.

www.emaceart.dev
https://discord.gg/HUYU5nWC

(+1)

Cool art. I use Unreal, and I do think that for me personally, I prefer just models, not Blueprints with built in functionality. But, that's mostly because I enjoy doing things like adding functionality to models, such as making a door system. It also helps because I can better integrate it into already existing systems in my game. I don't really need any specific assets, but I would suggest trying to make a complete set of art for a game to allow for better style continuity. (Weapons, characters, etc.) I do get that some of that is very different, but it probably would be best. Also, don't forget to upload your art to engine specific marketplaces as well as itch, since that's where most of the traffic for assets goes.

Thanks for the detailed reply and for mentioning you prefer “just models” without built-in logic / Blueprints – that’s exactly how I like to structure my packs: clean modular meshes, a few example prefabs, no forced gameplay systems so they’re easy to plug into your own logic.

I’m more of an environment / level artist than an “everything” generalist, so I aim for complete environment sets (architecture, interiors, props, foliage, a consistent style language). I’d rather push quality in one area than spread it thin across weapons or characters.

I’m also seriously thinking about going deeper into Unreal – next year I want to build a tool for migrating whole Unity projects to Unreal without having to redo everything by hand. If that works out, I’m planning to drop several big packs on that engine as well.

What sort of functionality would prefabs have? I mean, from art packs I've gotten, it has been things like a particle effect in a model or a light source in a lamp. Is that the sort of thing you do? Also, good luck making a project migrater. That would be quite a sight.

Show post...

"I mainly use a mix of ready-made 3D assets along with some custom ones I create myself. Most of the time, I rely on stylized or low-poly packs from marketplaces, then tweak them so they fit the game’s look. My workflow is pretty simple: I import the models, clean up the materials, adjust textures if needed, and then optimize everything for performance. After that, I set them up in the engine with proper lighting and animations. Nothing too fancy—just a practical workflow that keeps things smooth and consistent."

That part about keeping the process “smooth and consistent” got my attention – it sounds like working with assets, for you, is closer to editing than sculpting. It’s an interesting choice, because it shifts the focus from pure modelling to how the scene is put together: what rhythm it has, where it gets dense, and where you leave intentional gaps instead of adding more “pretty” models.

I get the sense that handling a mix of store-bought assets and custom models without obvious stylistic seams is, in practice, harder than building everything from scratch. That’s the moment when the mix stops being a compromise and starts reading as an authorial decision – instead of adapting to the assets, you assign them their own hierarchy and role in the frame. The player doesn’t have to name it, but they immediately feel the difference between a scene that’s been designed and one that’s just been assembled from pieces.

I like this kind of approach, where the “simplicity” of the pipeline doesn’t pretend to be modesty, but is there to keep the space readable and give it its own accent.