A C++ implementation of a transform that is kind of like the constant Q transform (CQT) except super fast.
Written by Matt C. McCallum.
This code depends on:
Help with installing IPP can be found here.
First you'll need to install IPP.
If you're on OSX the rest of the procedure is simple:
git clone https://github.com/MCMcCallum/FastApproxCQT
git submodule init
git submodule update
- run
./pull_thirdparty_osx.sh
Unfortunately no scripts have been written for other operating systems yet. Yet to come...
If you wish to build the C++ project and run tests or build the library via XCode or something:
gyp --depth ./FastApproxCQT.gyp
If you want to just build and install the Python module (only tested on system installed python 2.7 on OSX)
python setup.py build
python setup.py install
There is a simple demo for which you can view the results of the C++ module and an equivalent Python implementation.
Simply run python Demo.py
.
The python packages required for this demo are listed in ./requirements.txt
.