Algorithmic Music – Nanda Vishnu
I started exploring algorithmic music as a result of lessons under Paul Botelho. Using Pure Data, I was able to see music differently, as lines of data, something that could easily be categorized and manipulated. This understanding is similar to total serialism, but I used less strict applications of structure.
Later, I became interested in the improvised aspects of Indian music, and after finding KKsongs, which had databases of tala and raga, and this book, it looked like one could “program” a raga of sorts.
The raga provides the probabilities for the weighting of the markov chain controlling note choice, as well as the frequencies used. I reference the gat, a recurring composed section, using the swarupa, the characteristic melody or phrase of the raga, in its place, occuring at 80, 100, 120, 200, and 340 seconds.
The tala provided an initial basis for structure, but later only derived the length of the piece. However, the tala does provided the probabilities of the markov chain controlling bol, or stroke, choice.
This piece of algorithmic music is controlled by a 10-second timer, acting as the general metronome. At the same time, a metronome, ranging from 3sec per beat to more than 2 beats per second, provides the “playing tempo” for the piece. This parallels the alap-jor-jhala structure found in Indian raga.
The melody line’s rhythm is derived from a hard gate applied to a pink noise signal. The threshold and envelope of the gate change to give the illusion of speeding tempo. This gate triggers a complex sine wave using info from the SHARC database, making a hybrid of a pizzicato violin, a Bb clarinet, a vibrato alto flute, and a martele stroke on a contrabass, accounting slightly for timbre change at higher registers.
The video is a procedurally generated visualizer, created with the GEM plugin for Pure Data. It shows the drone, a wavering tonic, fifth, and both a major and minor third, with abstract polygons representing the triggered events.
This work of algorithmic music is bolstered by street sounds found at The Freesound Project. This ties the work to a chaotic structure, growing to a frenzied pace by the end of the piece, and finally ending with the gat, adorned by the melody processor.