Noisemaker example output (CC0)

Noisemaker is an adaptation of classic procedural noise generation algorithms, for Python 3 and TensorFlow.


Noisemaker is intended for Python 3.5+.

Install Noisemaker in a new virtualenv:

python3 -m venv noisemaker

source noisemaker/bin/activate

pip install git+

Install TensorFlow in the virtualenv, using pip. See TensorFlow’s platform-specific docs for your $TF_BINARY_URL.

# Try to see if there is a wheel available for your platform.
pip install tensorflow  # or tensorflow-gpu, if you're all set up with cuda/cudnn

# pip install $TF_BINARY_URL
# e.g. Mac OS X CPU:


Activate the virtual environment, and run:

pip install –upgrade git+



See Noisemaker API documentation.


Noisemaker includes several CLI entrypoints. For usage summary, run with -h or –help.

  • noisemaker: Fully-featured noise generation pipeline
  • glitchmaker: Simple glitch art tool
  • collagemaker: Image collage tool

Indices and tables