3D Prop Modeling for Games

3D props are the objects that populate a game level — everything that is not a character, a weapon or the environment shell itself. Furniture, containers, tools, equipment, debris, signage. In production they are handled in tiers, because a hero console and a background crate need very different budgets.

Props are usually split into:

- Hero props — inspected close, highest detail budget
- Interactive props — used in gameplay, need correct pivots and collision
- Set dressing — placed in volume, share materials and atlases
- Clutter — lowest tier, exists to break up empty space

Hero and interactive objects, set dressing, destructibles, modular kits and kitbash sets, in realistic and stylized styles. We work from a single asset up to full sets for a level, and we take over existing backlogs when a project is mid-production.

Typical requests:

- Interior fixtures: furniture, equipment, consoles, medical and industrial props
- Exterior set dressing: street clutter, signage, debris, pipes, barriers
- Interactive and quest objects with gameplay-correct pivots and collision
- Modular kits with trim sheets and shared atlases
- Stylized prop sets for casual and mobile projects

Consistency is a production decision, not a review outcome. We fix the shared variables before the first asset: texel density, material palette, wear language, chamfer treatment, tier budgets. Then we review the set against itself in-engine, not asset by asset in isolation.

What we lock at the brief:

- Texel density across the whole set
- A shared material library and, where useful, trim sheets and atlases
- Wear and damage language — how much, where, and why
- Per-tier triangle and texture budgets
- Naming, pivot and folder conventions

Yes. Realistic props are built from reference with PBR texturing and material accuracy. Stylized props follow a defined shape language — chamfer sizes, colour ramps, and detail carried by the texture rather than the mesh. Most projects need us to match an existing style rather than invent one.

The practical difference:

- Realistic — photo reference, PBR metal/roughness, detail in the mesh
- Stylized — exaggerated silhouette, hand-painted or semi-PBR, detail in the texture
- Either way — your existing assets are the style guide

Yes, and it is the most common assignment. Send assets already in your build and we reverse-engineer the rules: texel density, material response, wear placement, chamfer treatment, poly budgets per tier. New props then extend the set instead of sitting beside it.

What we ask for:

- Two or three existing props, ideally source files
- Your technical spec and per-tier budgets
- Screenshots in your actual in-game lighting
- The naming, pivot and folder convention your project uses

A modular kit is a set of pieces designed to combine; a trim sheet is one texture shared across many assets. Together they turn a fixed asset count into an open-ended library and cut texture memory sharply. They pay off on large, repetitive sets — and add overhead on small unique ones.

Worth it when:

- The set is large and shares visual language across many objects
- Level designers need to build without an artist in the loop
- Texture memory is a hard constraint, especially on mobile
- The project will keep adding content after launch

Not worth it when the set is a handful of unique hero objects.

FBX and OBJ as standard exchange formats, with source files on request. PBR texture sets at the resolution each asset tier justifies, plus shared trim sheets and atlases where the set benefits. LODs, collision and pivots are part of the base scope, not an add-on.

Standard delivery package:

- Game mesh in FBX plus source file
- PBR texture set, resolution agreed per tier
- LOD meshes across the agreed chain
- Collision meshes and pivots set to your convention
- Shared material library where the set uses one

Props are usually where a level's performance budget goes, so optimization is planned at the brief rather than applied afterwards. We agree per-tier triangle budgets, share materials across set dressing to reduce draw calls, and build LOD chains sized to the distance each asset is actually seen from.

Levers we use:

- Material and atlas sharing across the set to cut draw calls
- Per-tier triangle budgets instead of one budget for everything
- LOD chains tuned to real placement distance
- Trim sheets and tileable materials to reduce texture memory
- Collision complexity matched to gameplay use, not to the visual mesh

Prop work has no meaningful per-unit price — a background crate and a hero console differ by an order of magnitude. Cost is driven by the tier mix, the style, the platform budget and how much is shared across the set. We quote per project after seeing the asset list.

What moves the number most:

- Tier mix — how many hero props versus set dressing
- Whether we start at concept or at finished concept art
- Realistic PBR versus stylized hand-painted
- Whether the set uses modular kits, trim sheets or shared atlases
- Target platform and the technical budget it imposes

Yes — this is a normal starting point rather than an exception. We begin by reading your existing assets and conventions, produce a small first batch for calibration, and only scale up once that batch matches what your team already has in the build.

How the handover works:

- We review your existing assets, spec and art bible
- A calibration batch is delivered and reviewed before scale-up
- Your naming, pivots and folder structure are adopted, not replaced
- Throughput ramps once the first batch is approved