32dafd94e0
- Fix compose.py to select different samples per section instead of one per role
- Add select_many() to SampleSelector for diverse sample selection
- Migrate 862 samples from scattered dirs to libreria/samples/{role}/
- Rename files with consistent convention: {role}_{key}_{bpm}_{character}_{hash}.wav
- Add migrate_library.py script with dry-run and verification
- Backup original index as sample_index_pre_migration.json
- 72 tests passing
FL Control — Reggaeton Production System
Python system for generating complete reggaeton .flp projects for FL Studio from the command line, using intelligent sample selection and algorithmic composition.
Features
- Forensic Sample Analyzer — 4-layer audio analysis (signal, perceptual, musical, timbre) using aubio
- Intelligent Sample Selector — scores samples by key compatibility (circle of fifths), BPM proximity, and character
- Melodic Generators — reggaeton-idiomatic patterns: bass tresillo, melodic hooks, chord blocks, sustained pads
- FLP Builder — assembles valid FL Studio project files from a JSON song definition
- MCP Server — 28-tool Model Context Protocol server for AI-assisted production
Quick Start
pip install -r requirements.txt
# Analyze your sample library
1_ANALIZAR.bat # or: python src/analyzer/__init__.py
# Compose a track
python scripts/compose_track.py --key Am --bpm 95 --bars 8 --output output/track.flp
# or double-click:
COMPONER.bat
Project Structure
fl_control/
├── src/
│ ├── analyzer/ # Forensic audio feature extraction
│ ├── composer/ # Pattern generators (rhythm + melodic)
│ ├── flp_builder/ # FL Studio .flp binary assembly
│ └── selector/ # Intelligent sample scoring & selection
├── mcp/ # MCP server (28 tools for AI integration)
├── scripts/ # CLI entry points
├── knowledge/ # Musical domain knowledge (progressions, templates)
├── data/ # Generated indexes (gitignored)
├── .sdd/ # Spec-Driven Development artifacts
└── COMPONER.bat # Quick-compose launcher
System Requirements
- Python 3.10+
- FL Studio (for opening generated
.flpfiles) - ~4GB disk space for sample library (not included)
Workflow
- Drop your sample library into
librerias/ - Run
1_ANALIZAR.batto build the sample index - Run
COMPONER.batto generate a track
Architecture
The system uses a JSON SongDefinition as the single source of truth decoupling composition logic from FLP binary rendering. See .sdd/ for full technical specs and design documents.