Barney
is a tool to be used to analyze audio data, providing a high performance method of evaluating waveforms, spectrograms, and log-energy plots.
Barney
is the open sourced version of Fred
, a speech audio analysis tool developed within Sensory. The open source version will in time have Sensory specific workflows be phased out, and more analysis capability added in.
Barney
It is written using Python, the Qt GUI framework and libraries common to the PyData ecosystem (numpy
and scipy
). For Plotting, the pyqtgraph
library is used very extensively. In addition, soundfile
, sounddevice
libraries are used to read in audio files, and handle playback of audio from numpy arrays in a low-latency fashion.
Testing is done via pytest
library with the pytest-qt
plugin, documentation is generated using the sphinx
library.
Quality checks are done with mypy
and flake8
linters. black
is used for formatting, and pre-commit
is used to enforce their respective usage.
From a python 3.9 virtual environment run
pip install -r requirements.txt
python -m barney
To run Barney, you need python 3.9 installed on your system.
First, read the contributing guide which explains the basics of the development environment, the style guides and so on.
MIT License
Thank you to Sensory Inc. for open sourcing this tool to support its continued development.