COSC363-Train-Environment

What is it?

Graphics project made with C++/OpenGL/glut/glew

Can I see a demo without spending 4 hours getting glut setup?

.webm demo hosted by me

TODO:

• Extra features

  • Add object collisions for barrier arms
  • Change dance floor to react to user's music

• Refactor

  • Break into logic and view sections
  • Split into multiple files with headers

• Fix bugs:

  • Investigate existing mouse bug where looking up too high creates mouse issues
  • Barrier arm lights don't glow. Only reflect from center light
  • Poor performance on OSX. Runs at a high FPS on Linux however

Refactor update (April 2026)

This project has been partially modernized into an OOP/module architecture to make future changes safer and faster.

New files

• include/world_state.h
• include/input_controller.h / src/input_controller.cpp
• include/animation_system.h / src/animation_system.cpp
• include/renderer.h / src/renderer.cpp
• docs/ARCHITECTURE.md

What changed and why

• State was centralized into WorldState so mutable runtime data is no longer spread across unrelated logic.
• Input behavior moved into InputController so key/mouse policy is isolated from rendering.
• Time-based movement moved into AnimationSystem so timer callbacks only trigger deterministic state updates.
• Camera transform setup moved into Renderer to decouple presentation from gameplay logic.

Build scripts

Both compile scripts now compile multiple source files instead of one monolithic cpp.

See docs/ARCHITECTURE.md for a teaching-oriented walkthrough.

Music Visualizer Mode (new)

The scene now reacts to four values read from data/visualizer_input.txt:

1. low in [0, 1]
2. mid in [0, 1]
3. high in [0, 1]
4. optional beat in [0, 1]

Format:

0.63 0.41 0.77 1.0

Reactive mapping:

• Speaker towers scale with low frequencies and pulse on beats.
• Speaker face rings expand and recolor with mid/high frequencies.
• Dance floor colors now track low/mid/high bands.
• A pulse sphere flashes around the dance floor on stronger beat energy.

If no external file updates are provided, the app uses a procedural fallback signal so the visualizer remains animated.

Quick demo signal

Run the app in one terminal, and in another terminal run:

chmod +x ./scripts/simulate_visualizer_input.sh
./scripts/simulate_visualizer_input.sh

This continuously updates data/visualizer_input.txt with synthetic band values.

Real system-output audio (Spotify / YouTube / Firefox)

This project now includes a Linux output-audio bridge that reads the default sink monitor (not microphone input) and converts it to low/mid/high/beat values.

Files:

• scripts/system_audio_to_visualizer.py
• scripts/run_system_audio_visualizer.sh

Run flow:

1. In terminal A:

./compile_on_linux.sh
./gl 2

2. In terminal B:

chmod +x ./scripts/run_system_audio_visualizer.sh
./scripts/run_system_audio_visualizer.sh

That bridge auto-selects your default sink monitor using pactl, so any desktop output (Spotify desktop/web, YouTube in Firefox, etc.) drives the scene.

Notes:

• Uses PulseAudio/PipeWire tools: pactl, parec.
• The wrapper creates .venv_audio and installs numpy automatically.
• Stop bridge with Ctrl+C.