Bayesic: Algorithmic Music using Bayesian Networks скачать в хорошем качестве

Bayesic: Algorithmic Music using Bayesian Networks

Bayesic is a program that makes algorithmic music using a Bayesian network. The user is prompted to specify nodes and probabilities for the network tier by tier. The program then generates a given number of samples from the Bayesian network created. If a node has state 0 in the sample, no note is appended for it. If it has state 1, it will have a volume that decreases as a function of its distance from the first tier, and a frequency that is a weighted average of the frequencies of its parents. If a parent has state 0, its weight in the average will be lower than if it has state 1. The notes are appended to a stream in the order in which their associated nodes were first specified. A short rest is appended to the stream in between samples to distinguish their individual contributions to the music. When all the samples have been processed, the result is saved as a .midi file, which can be listened to using any digital instrument. This video (as I realize) sticks to one screen, and does not display either the visualization of the tree (which the code provides!) or the DAW on which the MIDI files are played. It also distorts the output by enough that it becomes close to intolerable. The prettier music generated by the program can be found at the following link:    • Bayesian Algorithmic Music Demo