01:43 ━━━━●───── 03:50
======= ⇆ㅤ ㅤ◁ㅤ ❚❚ ㅤ▷ ㅤㅤ↻
Endless creative explorations
In my journey to blend music and mathematics—my two greatest passions—this repository hosts my personal experiments. From ignorance I explore, try, fail, and occasionally succeed in coding tools and techniques that mix music, linear algebra, signal processing, numerical methods, and machine learning.
• Classification – Algorithms that identify musical genres or patterns. • SVD – Audio separation and compression. • Filters – Equalizers and audio-effects design.
Explore the infinite possibilities where music meets science. Whether it is compressing signals, classifying genres, or designing filters that reshape sonic perception, this repository is conceived as a creative laboratory for innovation and learning.
Language: Python
Libraries: numpy, scipy, librosa, matplotlib, scikit-learn, tensorflow, pydub, soundfile, joblib
Workspace: Jupyter Notebooks for interactive visualization & experimentation.
| Folder / File | What it does | How to try it |
|---|---|---|
CLASIFICAR/ |
Traditional machine-learning genre classifiers. | python CLASIFICAR/clasificar.py (Random Forest). Make sure path inside the script points to the GTZAN dataset or any folder with *.wav files organised by genre. |
CLASIFICAR/REDES_NEURONALES?/ |
Neural-network based genre classifier using Keras/TensorFlow. | python CLASIFICAR/REDES_NEURONALES?/clasificar_redes_neuronales.py after setting the same path variable. |
SVD_SepararAudio/ |
Vocal / instruments separation via Short-Time Fourier Transform (STFT) + Singular Value Decomposition (SVD). Generates separate .wav files with adjustable gain. |
1) Place target audio inside the folder. 2) Edit the filename in SVDmusica.py. 3) python SVD_SepararAudio/SVDmusica.py. |
FILTROS/EQ/ |
A toolbox of creative audio effects: distortion, reverb, echo, and low-pass cleaning. Demonstrated in EQ1.py, it loads a guitar riff and writes processed variants. |
python FILTROS/EQ/EQ1.py; tweak parameters such as gain, room_size, delay, or filter cutoff to taste. |
Notebook SVDmusicaGRAFICO.ipynb |
Visual explanation of the SVD separation process with plots. | Open with Jupyter Lab / VSCode and run all cells. |
- Create a Python virtual environment (optional but recommended):
python3 -m venv venv source venv/bin/activate - Install dependencies:
pip install numpy scipy librosa matplotlib scikit-learn tensorflow pydub soundfile joblib
- Execute the desired script as indicated in the table above.
Pull-requests, issues, and ideas are welcome! If you find a bug or have an improvement in mind, feel free to open an issue.
“Mathematics is the music of reason, and music is the mathematics of emotion.”
_░▒███████
░██▓▒░░▒▓██
██▓▒░__░▒▓██___██████
██▓▒░____░▓███▓__░▒▓██
██▓▒░___░▓██▓_____░▒▓██
██▓▒░_______________░▒▓██
_██▓▒░______________░▒▓██
__██▓▒░____________░▒▓██
___██▓▒░__________░▒▓██
____██▓▒░________░▒▓██
_____██▓▒░_____░▒▓██
______██▓▒░__░▒▓██
_______█▓▒░░▒▓██
_________░▒▓██
_______░▒▓██
_____░▒▓██
▒▒▒▒▒▒▒▒▒▒ 100% ᴄᴏᴍᴘʟᴇᴛᴇ!
Exploraciones Creativas sin fin alguno
En la búsqueda de fusionar la música y las matemáticas, mis dos pasiones principales, presento en este código mis experimentos creativos como motivación personal para no dejar la carrera e ir a dedicarme solamente a la música :) Desde mi ignorancia exploro, intento desarrollar cosas y varias veces fracaso, pero muy cada tanto logro codear herramientas y técnicas que combinan música, álgebra lineal, procesamiento de señales, métodos numéricos y aprendizaje automático.
▶︎·၊၊||၊|။||||၊၊||၊|။||||၊၊||၊|။||||။·၊၊||၊|။||||။||၊|။||||၊၊||၊|။||||၊၊||၊|။||||။·၊၊||၊|။||||။||၊|။||||၊၊||၊|။||||၊၊||၊|။|
Clasificación: Algoritmos para identificar géneros o patrones musicales. SVD: Separación y compresión de audio. Filtros: Creación de ecualizadores y efectos de sonido.
Explorar las posibilidades infinitas donde la música y la ciencia se encuentran. Ya sea comprimiendo señales, clasificando géneros o diseñando filtros que alteren la percepción sonora, este repositorio está pensado como un laboratorio creativo para innovar y aprender.
Lenguaje: Python Bibliotecas: numpy, scipy, librosa, matplotlib, sklearn, tensorflow Frameworks: Jupyter Notebooks para la visualización y experimentación interactiva.