Population Marbles Data Animation

You are a physics simulation engineer building an audiovisual data-experience. Your task: model the populations of all UN-recognized countries (2024 data) as falling marbles in a realistic gravity simulation, then generate production-ready code that renders the visualization and synthesizes collision sound effects.

## Simulation Requirements

**Scale & Data**
- 1 marble = 1 million people (round each country's population to nearest million)
- Total: ~8,144 marbles representing ~8.144 billion people
- Use authentic 2024 population figures from UN or World Bank datasets (cite source in comments)
- Each marble labeled with ISO 3166-1 alpha-3 country code

**Physics Engine**
- Realistic gravity (9.81 m/s²), friction, restitution, and collision detection
- {{drop-pattern}} (examples: "simultaneous release from randomized heights" | "release in descending population order, 50ms intervals" | "regional clusters drop together by continent")
- Container: {{container-spec}} (examples: "tall cylindrical tube, 2m diameter" | "wide funnel narrowing to 0.5m at bottom" | "rectangular box 3m × 3m × 10m height")
- Simulation runs until all marbles settle (kinetic energy < threshold)

**Audio Synthesis**
- Generate distinct collision sound for *every* marble-to-marble and marble-to-container impact
- Sync audio triggers to exact collision timestamps from physics engine
- Sound properties vary by impact velocity (louder/higher-pitch for faster collisions)
- Output: multi-track audio or single mixed WAV/MP3, duration matches simulation length

## Code Deliverables

1. **Data loader**: fetch/parse 2024 population data, convert to marble count per country
2. **Physics simulation loop**: step-by-step integration, collision detection, event logging
3. **Audio synthesis module**: procedural sound generation or sample triggering tied to collision events
4. **Visualization** (optional but recommended): real-time or post-render 3D view of falling marbles
5. **Export pipeline**: save simulation data (positions over time) and final audio file

## Technical Constraints

- Language/framework: {{tech-stack}} (examples: "Python + Pymunk + pydub" | "JavaScript + Matter.js + Tone.js + Three.js" | "Unity C# with built-in physics and FMOD")
- Optimize for performance: spatial hashing or broad-phase collision culling for 8,144 bodies
- Deterministic: given the same seed, reproduce identical simulation and audio
- Include setup instructions, dependencies, and example command to run

## Output Format

Provide:
- Fully functional, well-commented source code (modular structure)
- README with: data source citation, installation steps, usage examples, and performance notes
- Brief explanation of physics parameters chosen (restitution, friction, marble mass/radius) and why
- Sample output description: "Simulation completes in X seconds (real-time/accelerated), generates Y-minute audio file with Z collision events"

Prioritize code clarity, accurate physics, and tight audio-visual synchronization. The result should feel like a mesmerizing data sculpture—every clink represents millions of lives.