Python for Sequencing

Data

This example needs tab-separated text files as input, accepting multiple time-series in one file. It needs to be in formatted as follows:

• Each column represents a single time series.

• Each row corresponds to a sample inside a time series.

All time series in the data have the same number of samples. An example for a file with two time-series, each with 10 data points (samples):

0.12,  1.03
0.15,  0.98
0.22,  0.91
0.35,  0.87
0.48,  0.82
0.62,  0.79
0.71,  0.74
0.83,  0.69
0.94,  0.63
1.02,  0.58

Python

The Python script is minimalistic, consisting of 3 code blocks:

• 1: Argument Processing

• 2: Read Data

• 3: Playback Loop

It relies on numpy and python-osc - the installation is covered in the repository. The script can be executed like this:

$ python3 sequencer.py PATH_TO_FILE --col 1 --interval 100

The arguments explained:

• --col: Determines which column will be used.

• --interval: Determines the sampling rate of the playback (how long does the script wait to send the next sample).

• --port: This optional argument can change the port the script is sending the OSC data to.

Pure Data

This example builds on the introduction patches on Using OSC in Pure Data. It receives incoming OSC data and uses the message with the path /val. The incoming values are scaled and then turned into a tone with changing pitch, using the osc~ object, which is sent to the audio output with dac~:

It might be necessary to review+ the following sections:

• https://ringbuffer.org/computer_music_basics/Puredata/getting-started-with-puredata/

• https://ringbuffer.org/computer_music_basics/Puredata/pd-audio-io/

• https://ringbuffer.org/computer_music_basics/Puredata/control-rate-and-audio-rate/