I love how the results sounds, so I wanted to now how it works.
Very briefly glancing over the code and listening to the audio files in the samples folder:
The six articles correspond two six different voices or "instruments" which are pre-recorded samples of drone like sound and sequences of notes varying in pitch and speed. The letters in the article then trigger play and stop events for the samples according to a mapping.
How did you go about deciding on the mapping setting like "shouldPauseEventOnMatch", "isLoop" and "eventLength"? Is there a logic behind it or more testing and listing what works together?
I started with the composition (the music one). So these have been dictated by composition techniques: some samples should play in a loop, others should be played fully before the sequencer's cursor increments etc.
It has been a long-term ambition of mine to someday build an electronic aeolian harp[1], powered by weather data. This, I must say, is an even better idea.
> Each article functions as a step sequencer, where the letters are the sequencer steps and the track titles are regular expressions that switch the steps on and off.
Due to the album's concept there is no such a feature, sorry.
But you can fork the project on github and try to set attribute contenteditable to "true" for the div, which you'd like to fill up with your text. The buffers have ids "buffer-1", "buffer-2" etc.
Very briefly glancing over the code and listening to the audio files in the samples folder:
The six articles correspond two six different voices or "instruments" which are pre-recorded samples of drone like sound and sequences of notes varying in pitch and speed. The letters in the article then trigger play and stop events for the samples according to a mapping.
Example for 'u':
https://raw.githubusercontent.com/AlesTsurko/microscale/gh-p...
https://raw.githubusercontent.com/AlesTsurko/microscale/gh-p...
As the description indicates, there doesn't seem to be any tone generation or modulation going on in the browser.