Modular planning is the bridge between greybox and production meshes. You are not modeling yet—you are deciding which pieces repeat, how they snap, and what stays unique so Lesson 4’s low-poly pass does not turn into sixteen one-off walls nobody can reuse.

Lesson Objective

By the end of this lesson you will have:

1. A module grid (unit size in meters) written down and applied in Blender with snapping
2. A kit list table covering straights, corners, posts, floor tiles, trims, and hero props
3. Pivot rules (floor at bottom center, wall bases at floor line, trims on edge) documented in one note
4. A clear split between instanced kit pieces and bespoke meshes that are allowed to break the grid

Step 1: Lock the grid from your blockout

1. Open your Lesson 2 file. Measure a typical wall span in Edit Mode or with the Measure tool.
2. Pick a module length that divides most spans with at most one cut piece per run. Common starting points: 1 m, 2 m, or 0.5 m for tight interiors.
3. Enable Snap → Increment with Absolute Grid Snap if you like hard steps, or Edge/Vertex snap if your blockout already sits on clean increments.
4. Write the rule in a text block in Blender or a readme: MODULE = X meters, FLOOR_THICKNESS = Y.

Pro tip: If your game uses tile-based movement, align MODULE to tile size now. Retrofitting later costs UV and collision pain.

Step 2: Tag blockout masses as Straight, Corner, or Hero

Walk the ARCH_Shell collection from Lesson 2:

• Straight - wall runs you can tile with W_Straight_1M style names
• Corner - inside and outside corners (often two different meshes)
• T-junction / post - places three walls meet; sometimes a column module is cleaner than boolean corners
• Hero - anything that breaks repetition (gate, statue, custom roof chunk)

Mini challenge: Rename at least ten blockout objects to include KIT_ or HERO_ prefix so the outliner tells the story without opening the spreadsheet.

Step 3: Build the kit table (one page)

Create a simple table (sheet, Notion, or markdown in repo). Columns that save teams:

Common mistake: Listing twenty modules before you prove one straight and one corner snap in the blockout. Start with five rows, then expand.

Step 4: Floor and trim rules

Floors

• Decide one tile footprint for 90% of walkable space.
• Note stairs and platforms as exceptions with their own IDs (ST_STEP_025 for a 25 cm rise, for example).

Trims

• Trims usually want length = MODULE or MODULE/2 so they cap wall ends without scaling.
• Put trim pivots on the back edge that kisses the wall so array or snap placement is boring and reliable.

Step 5: Pivot conventions (write them before modeling)

Agree on:

• Walls - origin at bottom center of the back face (outward normal is +local Y or +X—pick one and never flip mid-project)
• Floors - origin at center of bottom face
• Props - origin at ground contact center unless the prop hangs (lamps, chains)

Pro tip: Duplicate one empty named PIVOT_WallStandard and snap it where the first real wall module will live—engineers can use that as a reference when importing tests.

Step 6: Instance vs unique (scope control)

Instanced kit - anything you expect more than three copies of in the level.
Unique - hero silhouettes, story doors, broken variants that sell the world.

If everything is unique, you do not have a kit—you have a set dress list. That is valid for tiny scenes, but this course assumes reuse for portfolio scale.

Troubleshooting

• Corners gap by a few cm - Your module does not divide the angle bisector; adjust corner piece width or nudge blockout back to grid.
• Trims float - Wall thickness in blockout does not match kit spec; fix the table, not random Z moves.
• Too many module IDs - Merge left and right straights if they are mirrors; use mirror modifier or negative scale policy in engine docs later.

FAQ

Do I model kit pieces in this lesson?
No. Planning only. Greybox stays the stand-in until Lesson 4.

What if my art style needs organic curves?
Keep kit pieces for 90% of structure; use hero meshes for curved silhouettes and blend them into straights with overlap skirts planned in Lesson 6.

How tight should the grid be?
Tight enough that another artist could place a room from your table without DMing you.

Recap

• You chose a module size grounded in the blockout
• You classified shell geometry into kit vs hero
• You wrote a pivot and naming contract before mesh detail

Next Lesson Teaser

Lesson 4: Low-poly modeling pass replaces blockout cubes with real silhouettes while keeping edge loops friendly for UVs in Lesson 5.

Cross-link: when you need export naming early, skim our Blender for Game Assets guide export chapters and the Blender to Unity and Godot export pipeline article.

Bookmark this track if you are building a portfolio environment across several sessions, and share it with a teammate who skips planning and models straight into texture pain.