sndtrck 0.5.0

Sound manipulation and representation via partial-tracking

SNDTRCK

Partial tracking in python 3

Dependencies

Mandatory

• python >= 3.7
• numpy
• bpf4: interpolation curves
• sndfileio: simple API for reading and writing sound-files
• loristrck: partial tracking analysis based on Loris. See the project for its own set of dependencies

Optional but HIGHLY recommended

• pyqtgraph: efficient plotting via Qt
• h5py: allows saving partial-tracking data as HDF5. You need to install hdf5.

Installation

We assume that pip points to your current python >= 3.7 installation. Otherwise use pip3 and python3

Linux

$ sudo apt-get install fftw-dev
$ pip install -r requirements.txt [--user]
$ git clone https://github.com/gesellkammer/sndtrck
$ cd sndtrck
$ python setup.py install [--user]

OSX

$ brew install fftw
$ pip install -r requirements.txt [--user]
$ git clone https://github.com/gesellkammer/sndtrck
$ cd sndtrck
$ python setup.py install

Windows

Follow instructions here)

$ pip install -r requirements.txt
$ git clone https://github.com/gesellkammer/sndtrck
$ cd sndtrck
$ python setup.py install

---

Basic Usage

Example 1

import sndtrck
spectrum = sndtrck.analyze("/path/to/sndfile", resolution=50)

# Get the chord at 500ms, but only the partials louder than -30 dB
print(spectrum.chord_at(0.5, minamp=-30))
# [A3+, C5+10, E5-13]
spectrum.plot()
spectrum.show()  # this will show you the spectrum in the default applicatio for your system
spectrum.write("spectrum.sdif")

Features

• analysis of sound-files in many formats
• automatic configuration of analysis parameters
• filtering of partials based on multiple criteria
• resynthesis
• plotting
• export the spectrum to many different formats (sdif, hdf5, midi)

Transcription

Go to sndscribe for automatic transcription of spectra into musical notation

License

GNU