Chapter 2: Proportion Passes & A / B / C Iteration

Created by Sarah Choi (prompt writer using ChatGPT)

Proportion Passes & A/B/C Iteration (30% Deltas) for Weapon Concept Artists

From brief → ideation → iteration → finals → handoff. Written equally for concept‑phase artists and production‑minded artists.

Why proportion passes matter

Proportion is the quiet ruler of readability, plausibility, and appeal. Before materials, decals, or micro‑greebles, players form an opinion from mass ratios and axis relationships: barrel‑to‑receiver, grip‑to‑stock, mag‑to‑fore‑end, optic‑to‑eyebox. Proportion passes turn vague taste into a repeatable system and give production teams stable scaffolding for rig, animation, and LODs.

A/B/C iteration with 30% deltas is the discipline that keeps options meaningfully different while remaining buildable. Instead of drifting into near‑duplicates or swingy overhauls, you move the design in controlled leaps that stakeholders can evaluate quickly and downstream teams can estimate.

From brief to proportion targets

Translate the brief into proportion constraints that can be measured:

1) Class envelope. Define min/max overall length (OAL), weight cues, and stance. Example: SMG envelope OAL 420–620 mm; rifle envelope 780–980 mm; shotgun 650–920 mm.

2) Camera reads. First‑person (0.7 m) vs third‑person (8–40 m) demand different silhouette emphases. A DMR wants a long top line; a bullpup can compress OAL without losing barrel visual mass. Write which planes must read at icon size.

3) Ergonomics. Trigger reach, grip angle, stock line, fore support window, mag change clearance. Mark these as protected distances so proportion changes don’t break plausibility.

4) Mechanical story. Action family (gas/roller/lever/pump/coil/rail/mag‑tech) and reload choreography (tube, box, cylinder, belt, cell) impose proportion realities. Reserve space where cycles happen.

5) Faction language. Proportion can imply doctrine: monolithic slabs for imperial industry; delicate trusses for high‑tech modularity; overhung muzzles for brawler cultures.

Document these five as a “Proportion Intent” block at the top of your board.

The three proportion passes (blocking → tuning → locking)

Pass 1 — Blocking (coarse massing)

Goal: establish big three relationships with 60/30/10 thinking (primary/secondary/tertiary mass).

  • Primary axis: choose the dominant line (flat rifle top, arched shotgun, stepped SMG).
  • Receiver ↔ barrel: set the barrel visual weight vs receiver box (e.g., 60/40 for DMR; 40/60 for PDW with shroud).
  • Grip cluster ↔ stock: pick stance and balance (rear‑heavy for controllability; neutral for generalist; forward for brawler).
  • Magazine read: decide if the mag is a silhouette co‑star or subdued (PDW banana mag vs internal tube).
  • Produce 6–9 blocks per class, no interior lines, just value and contour. Annotate ratios numerically (e.g., OAL 880; receiver 360; barrel 420; stock 220).

Pass 2 — Tuning (functional windows)

Goal: carve windows for mechanics and animation without losing block intent.

  • Charge & bolt travel: ensure 1.5–2× bolt face travel clearance in receiver mass; mark charge handle arc.
  • Reload window: isolate negative space where the hero reload beat reads (magwell exposure, cylinder swing, cell swap plane).
  • Optic line: set eyebox height relative to stock comb to avoid chin‑weld looks; ensure attachment envelope.
  • Heat path & vents: reserve surface length where heat wants to escape; don’t put critical grips or optics over that.
  • Sling & carry: mark anchor planes that won’t collide with controls.

Output: 3–4 tuned variants of your best 3 blocks. Keep ratios labeled; note protected distances.

Pass 3 — Locking (metric commit)

Goal: commit to a proportional standard that modeling can trust.

  • Scale bars & datums: add 100 mm scale bars and centerlines; set world scale.
  • Critical dimensions: OAL, barrel visual length, receiver depth, magazine projection, optic rail length, stock extension range.
  • Sacred vs flexible: tag which values cannot change during skins/attachments (e.g., DMR top line sacred; fore‑end length ±5%).
  • LOD vows: annotate which curves/planes must survive at low poly to preserve proportion reads.

Deliver: 1–2 locked proportion sheets per candidate concept.

A/B/C iteration with 30% deltas

Definition. A/B/C iteration proposes three variants where each differs from the control by about 30% on selected axes—enough to feel different, not so much that class identity breaks. “30%” can be literal (dimension changes) or experiential (perceived mass shift), but you should make it measurable.

Pick 2–3 axes per round from this menu:

  1. OAL proportion (compact ↔ long)
  2. Receiver:barrel ratio (chunky core ↔ extended barrel)
  3. Mass distribution (rear‑biased ↔ forward‑biased)
  4. Top line (flat ↔ stepped ↔ arched)
  5. Undercarriage line (clean ↔ busy with devices)
  6. Grip angle & spacing (upright ↔ raked; close hands ↔ spread)
  7. Magazine emphasis (flush ↔ proud/curved/angled)
  8. Stock posture (inline ↔ drop‑heel ↔ skeleton)
  9. Fore‑end length (short grip‑centric ↔ long support‑centric)
  10. Front signature (stubby shroud ↔ long muzzle device)

How to quantify a 30% delta (examples):

  • OAL: 800 mm (A) → 560 mm (B) (−30%) → 1,040 mm (C) (+30%).
  • Barrel visual length: 400 mm → 280 mm (−30%) → 520 mm (+30%).
  • Receiver depth: 120 mm → 84 mm (−30%) → 156 mm (+30%).
  • Optic rail length: 240 mm → 168 mm (−30%) → 312 mm (+30%).
  • Mass distribution (perceived): shift the centroid forward by ~30% of fore‑aft span using added shroud mass and reduced stock bulk; verify via a centroid overlay box.

Round structure:

  • Round 1 (coarse): Change OAL + receiver:barrel. Keep grip cluster and top line constant.
  • Round 2 (mid): Freeze OAL; change top line + mass distribution.
  • Round 3 (fine): Freeze big axes; change magazine emphasis + front signature.

Each round delivers A/B/C sheets with delta callouts and mini thumbnail reads.

Readability & plausibility tests for each A/B/C set

Distance thumbnails. Shrink to 48 px, 96 px, 192 px. If testers can’t ID class ≥90% at 48 px, rework.

Flip & contrast. Mirror horizontally; test on dark/mid/light backgrounds. Proportion that only reads on mid‑tones is fragile.

Animation dry‑fit. Overlay proxy hands to run charge, fire, reload, inspect. Ensure the hero beat reads in silhouette for all three variants.

Attachment stress. Mount a suppressor + mid optic + fore device. If the variant loses its class read, integrate counter‑shapes (steps, shrouds) or adjust distribution.

Metric sanity. Check trigger reach, mag change room, ejection arc clearance. Flag failures on the sheet and in the matrix.

Visual development without proportion drift

  • Value scaffolding locks proportion. Use 3–4 values that echo your mass hierarchy so paintover doesn’t inflate/deflate volumes.
  • Material breakups must follow mass flows. Ribs and vents should underline, not contradict, your chosen axes.
  • Faction overlays can alter edge character without changing protected ratios. Apply chamfers/filigree as outline‑friendly moves.
  • Skin lanes should sit on planes sized during Pass 3; don’t invent planes later that force proportion edits.

For concept‑side artists: playbooks

  • Coverage fast, decisions slower. Use A/B/C to cover the space, then select with evidence (read tests, metrics) rather than taste alone.
  • Two bold axes, one safe. Most rounds, push two axes ~30% and keep one axis orthodox to avoid class drift.
  • Narrative ratios. Encode story in mass: a “sanctioned” rifle might have 60% receiver dominance (industrial heft); a “pilgrim’s” carbine might bias stock length for travel pragmatism.
  • Pre‑rig empathy. Draw charge arcs and mag swaps early. If they’re ugly now, they’ll be worse animated.

For production‑side artists: what to demand

  • Locked ratios in orthos. Receive front/side/top views with numeric proportions and scale bars; reject sheets without datums.
  • Attachment envelopes. Expect toleranced boxes for optics, cans, grips that reflect the proportion promise.
  • LOD vows. Require a list of sacred planes/curves that must remain at LOD1/2.
  • Traceability. Every final must reference its A/B/C parent and the matrix row; track why A beat B/C.

Indie vs AAA: scope the A/B/C game

Indie. Fewer rounds, higher leverage axes. One coarse (OAL + receiver:barrel) and one fine (mag + front signature) may be enough. Lean on modular attachments to expand variety post‑lock.

AAA. More classes, outsourcing, and skins mean stronger governance. Establish a central proportion bible: canonical ratios per class, examples of good/bad deltas, and test protocols. Vendors get a mini‑kit with a proportion checklist and pass/fail thumbnails.

Failure modes & fixes

Near‑duplicate variants. If A/B/C are indistinguishable in thumbnails, your deltas are <15%. Re‑run with numeric targets and freeze an axis to amplify contrast.

Over‑swing drift. If one variant leaves the class envelope, you exceeded 30% in the wrong axis (e.g., top line). Pull back, or change axes.

Animation collisions. If reload beats collide with furniture, you changed ratios without moving windows. Carve negative space; re‑place the magwell or cylinder.

Attachment erasure. If optics/devices erase the read, pre‑step the silhouette with counter‑masses; shorten rails or add shroud shoulders.

Value creep. Paintovers that change perceived mass indicate value scaffolding wasn’t defined. Re‑key the 60/30/10 blocks.

Deliverables & naming conventions

  • A/B/C proportion sheets with: three views, numeric deltas, 48/96/192 px thumbnails, attachment stress row, animation overlay, and a test checklist.
  • Versioning. Use stable IDs: CL‑AR‑P01‑A, CL‑AR‑P01‑B, CL‑AR‑P01‑C (class, platform, pass, variant). When selected, lock to CL‑AR‑P02‑A (next pass) to preserve lineage.
  • Matrix linkage. Each sheet references the class matrix row and logs pass/fail notes inline.

Handoff that actually works

Package the winner with:

1) Locked proportion orthos at world scale with datums and sacred/flexible annotations.

2) Animation & attachment envelopes matching the proportions; provide max/min positions (stock collapsed/extended) and optic eye line.

3) LOD vows listing planes/curves to preserve.

4) Mini retrospective explaining why A beat B/C, with read test thumbnails. This prevents future “why not C?” churn.

5) Skin & variant notes clarifying what can change later (materials, surface motif, accessory load) without violating proportions.

Closing: proportion is the contract

Great weapon designs feel inevitable because their proportions are intentional, tested, and documented. A/B/C iteration with 30% deltas creates meaningful choices; proportion passes give those choices structure that survives the journey from brief to package to game. Treat proportions like code—declare, test, refactor, and lock—and your concepts will ship faster, read cleaner, and scale further across variants and seasons.