My exploration into audio programming and DSP algorithms inspired by the writings of Julius Orson Smith III of Stanford’s CCRMA (Center for Computer Research in Music and Acoustics). I also use complex numbers for almost all the calculations.
I really started to learn programming in 2005, and this project inspired me to continue learning. I first started this project using Ruby and NArray, but in 2009 I started to use Python at work, so I ported it to Python and Numpy.
Of course I realised at some point that Python would never be suitable for realtime audio because of the GIL, and learned D, Common Lisp and Haskell. Some time later I also learned some Rust, Zig and Nim. Maybe some day I learn C++ properly.
Please be warned to only use Akasha for prototyping audio DSP algorithms and trying things out, and not expect to make anything playble using Python.
The functions in the module akasha.math.functions might be generally useful.
Akasha should work on macOS, Linux and FreeBSD.
See the /install directory for various installation instructions and pip
requirement files.
On macOS, it is recommended to use the framework versions of the SDL libraries.
In recent years, I have only used Macports to install the required dependencies instead of Homebrew, because I find the latters engineering quality to be dubious.
Basic usage with IPython:
source ./venv/py310/bin/activate
ipython
from akasha.lab import *
Then explore the akasha module and especially the audio, dsp and io
submodules.
In [1]: from akasha.lab import *
In [2]: bj = read('Bjork - Human Behaviour.wav', dur=120, sdir='../../Sounds/_Music samples/')
In [3]: snd = bj[..., 0] # Make the sound mono
In [4]: play(snd)
In [5]: anim(snd)
Run tests:
pytest -k akasha akasha
This work is licensed under the MPL 2.0 License.
© 2005–2025 Peter Hillerström