Hearing from Within a Sound

Max patch and python codebase used for exploring and demonstrating the work described in the AES paper Hearing from Within a Sound: A Series of Techniques for Deconstructing and Spatialising Timbre.

Max Project

Dependencies

• IRCAM Spat~ v.5+
• kac_maxmsp

Python Resynthesis

Demo video 🎥

cd python

Dependencies

• libsndfile
• pipenv
• PyTorch with CUDA enabled

If you're using conda:

conda env create -f environment.yml
conda activate aes

Install

pipenv install

Run

You will need to specify a directory of audio files to resynthesise.

pipenv run python resynthesise.py --audio_dir </absolute/path/to/audio/files/>

Testing

pipenv install -d
pipenv run test

How to Cite

@inproceedings{wolstanholmeHearingSoundSeries2023,
  title = 	{Hearing from within a Sound: {{A}} Series of Techniques for Deconstructing and Spatialising Timbre},
  booktitle =	{International {{Conference}} on {{Spatial}} and {{Immersive Audio}} ({{AES}})},
  author = 	{Wolstanholme, Lewis and Vahidi, Cyrus and McPherson, Andrew},
  year = 	{2023},
  month = 	aug,
  address = 	{{Huddersfield, UK}},
  copyright = 	{All rights reserved}
}