pyquenzer

pyquenzer is a Python module to sequenze and retune short audio samples. Its main purpose is to offer an easy and simple way to generate melodies and chords with different tunings. pyquenzer uses the music computer system csound to sequenze and retune audio samples.

dependencies

• music computing system csound
• Python library mu

usage (see example)

For generating WAV - files, its necessary to define an Instrument first. This Instrument contains details about the synthetisation process. In the next step one may pass arguments about the content (pitch and rhythm) that shall be synthesized to the previously defined Instrument.

1. Defining an instrument.

Load the respective module.

import pyquenzer

Define the concert pitch in Hertz. The concert pitch is indicating the frequency of the first pitch of a scale.

concert_pitch = 260

Define the path of the scale that you want to use. The scale has to be written in the popular scl format. In example/scales you can find several scl files. They are taken from the scl files archive of the Huygens-Fokker instituion.

scale = "scales/example-gamelan.scl"

Make the samples dictionary. This dictionary contains the paths of all audio samples that shall be used to synthesise WAV files later. Every entry contains two elements:

1. an estimated frequency in Hertz
2. a list containing the paths of all corresponding samples

samples = {
    261.626: ["samples/c0.wav", "samples/c1.wav"],
    349.228: ["samples/f0.wav", "samples/f1.wav"],
    440: ["samples/a0.wav", "samples/a1.wav"],
}

Define (optionally) the scale_decodex argument. The scale_decodex argument has to contain as many entries as there are pitches in the used scale. Every entry represents one scale degree. If no scale_decodex is specified by the user, every scale degree is represented by its position (the scale-decodex of a hepatatonic scale would automatically be defined as (1, 2, 3, 4, 5, 6, 7) and the scale_decodex of a pentatonic scale as (1, 2, 3, 4, 5)).

scale_decodex = (1, 2, 3, 5, 6)  # notation style similar to the Javanese 'notasi Kepatihan' thats skips the 4

Initialise an Instrument - object. Besides

• concert_pitch
• scale
• samples
• scale_decodex

the Instrument class also provides additional optional arguments:

• nchannels: Shall be 1 if the used samples are mono and 2 if the used samples are stereo
• overlap: determines how long two following notes of a melody are overlapping (in seconds)
• volume: factor to adjust the volume of every sample
• release: indicates the duration of a linear envelope that releases the sound of each sample (in seconds)
• reverb_volume: volume of the reverb channel

my_instrument = pyquenzer.Instrument(
    concert_pitch, scale, samples, scale_decodex=scale_decodex, nchannels=2, overlap=0.5, volume=1.2
)

2. Rendering WAV - files.

To render WAV - files call the Instrument object with:

1. The path/name of the resulting WAV file

2. The path of the pitch file

Pitch files are text files that describe the pitch of each event. For more details about their form study the examples here.

3. The path to the rhythm file

Rhythm files are text files that describe the duration of each event. For more details about their form study the examples here.

my_instrument("output/example0", "pitches/example0", "rhythms/example0")

A rhythm file can also be replaced by a floating point number. In this case the rhythmic duration of every resulting tone/chord is equal (0.25 seconds in the example below).

my_instrument("output/example1", "pitches/example1", 0.25)

installation:

First install all necessary dependencies. Then download pyquenzer via git and install it with pip:

  $ git clone "https://github.com/levinericzimmermann/pyquenzer"
  $ cd pyquenzer/
  $ pip3 install -r requirements.txt
  $ pip3 install .

To test if everything works well one may try (inside the example - directory):

  $ python3 example.py