Get In Touch
Mouin Frada
Ph: +21699563061
Work Inquiries
[email protected]
Back
Architecture Design Generative Motion Visualization
December 27, 2024

Place Chelma – Unreal Engine Environment Reconstruction

I recreated my childhood fishing spot, “Place Chelma” in Hammamet, Tunisia, a small breach in the city’s ancient fortifications originally used for naval defense. I built the entire environment from vague personal memories, using no photo references. The goal was to practice new skills, keep up with Unreal Engine updates, and design a playable level for a future game.

  • Strategy

    Environment reconstruction from memory

  • Design

    Level Design, Art direction

  • Client

    Personal initiative

01 // Project Overview

What value
did i add ?

    • Designed a complete environment from memory – translated 20‑year‑old childhood recollections into a coherent 3D space with realistic lighting, water, and terrain.

    • Applied game level design structure – treated the project like a product: requirements, asset list, blockout, lighting pass, polish, and documentation.

    • Demonstrated software adaptability – learned and applied Unreal Engine 5 features (Lumen, Nanite, water system) within 4 weeks.

    • Created a portfolio piece that shows personality, creative thinking and initiative.

    “No references, just memories – built from scratch in Unreal Engine.”

02 // Challenges

The Problem
I was solving.

Personal: I wanted to preserve a place that no longer exists as I remember it. The actual “Chelma” breach has changed: modern construction, different access, eroded stone. My memory version is from age 8–12.

Technical: I needed to practice:

  • Level design workflow (blockout → assets → lighting → post‑process)

  • Unreal Engine 5’s latest tools (Lumen for dynamic lighting, Nanite for high‑detail geometry, Water system for coastal waves)

  • Environment storytelling without dialogue or UI

Professional carry‑over: The same skills used to build this scene are directly transferable to:

  • VR/AR design reviews (Safran Digital Transformation Award 2024)

  • Interactive product configurators (investor presentations)

  • 3D reconstruction workflows (80% faster reverse engineering scans)

06 // How i did it

Skills put
in action.

oh demo3 14 min

Real‑time 3D
& environment design.

  • Unreal Engine 5 (Lumen, Nanite, Water plugin)
  • Landscape sculpting, foliage placement
  • Megascans integration
  • Post‑processing and color grading
oh demo3 16 min

Level design &
creative workflow.

  • Blockout → asset placement → lighting → polish
  • Scale iteration (memory vs. reality)
  • Playable area definition
  • Sight line and composition testing
oh demo3 15 min

Lighting
& atmosphere.

  • Dynamic sky and clouds
  • Volumetric fog and light beams
  • Golden hour lighting design
  • Underwater refraction and fog
oh demo3 17 min

Self learning
& Early adoption.

  • Early adopter: used Unreal Engine 5.4 within weeks of release
  • Self‑directed learning: no tutorials, only Epic documentation and forums
  • No‑reference reconstruction: built from childhood memory
  • Experimental workflow: tested new features as they shipped
 
 
Next Project
Human digital twin.