This repository implements the "hilbert transform" on a sound data stream in a faithful manner. The Hilbert transform is a phase shift in the frequency domain of all frequencies by 90 degrees.
The mechanism implemented here is a sliding window of Fourier transforms in which only the middle part contributes to the output of the transform. This buffering/windowing, at least according to our tests, provides much more accurate and also more efficient Hilbert transforms than say using an approximation by convolution.
Enjoy.
Various citation methods using zenodo DOI:
BibTeX:
@misc{scott_cotton_2019_2553680,
author = {Scott Cotton},
title = {wsc0/hilbert: Faithful},
month = jan,
year = 2019,
doi = {10.5281/zenodo.2553680},
url = {https://doi.org/10.5281/zenodo.2553680}
}