Goodboylove Nmeme hehe

A graph of 3 trophic levels for 1e7 cycles, created with R using the lotka-volterra equations and one for density dependent growth. Pitch-shifted and normalised in audacity.

44.1 million years of population dynamics

Kraeftspad VII outro

Submersion

Randomly-ish generated in R. Sung by vocaloid.

Vibes in Am9

Quadratic anxieties

Nice shirt, d00d

The caves are alive with the sound of fear

Hide

heherminator

Made using a a python script (chatgpt helped me heh) that generates midi data using markov chains, monte carlo methods and evolutionary/genetic algorithms.

script: https://colab.research.google.com/drive/1qiO-fhM4wcxwmmUG7SLV3Nt42PoaD3ef
Basically like this:

Each chord is an individual.
First, an initial population of randomly assigned chords is generated. The most ”fit” chord is selected and output as midi data. A number of offspring are generated from the parent chord. Each position of the chord can mutate into a different note. The probability of a mutation occurring and the actual result of that mutation is determined by a matrix of probabilities, so that some mutations are more likely than others.

The fitness of each offspring chord is evaluated using a set of criteria. 

There are certain intervals that are defined as more fit, thirds, fourths, fifths and octaves for example. Each ”good” interval adds to the fitness value of the chord. In addition, there is a large fitness disadvantage added if the offspring chord is identical to the parent chord or if the offspring chord has several repeating notes (as in both pitch and octave). 

After all offspring chords have been evaluated, the most fit chord is selected and output as midi. This goes on for t generations and then the whole sequence is converted to a midi file.

Onaturlig selektion

Quadrapop Phantasms

3333

Djaumá

Popular tracks