minimal_voice: Synthesize the sounds of a single voice in a given section

In shill1729/synthesizer: Synthesize sounds, melodies, chords, and stochastic compositions

Description

Synthesizes a single voice over a given section with model input

Usage

minimal_voice(
  initial_voice,
  P,
  W,
  n,
  nb,
  bpm,
  harmonicParams,
  state_lengths,
  melody_states,
  rhythm_states
)

Arguments

initial_voice	the index of the initial voice
P	the probability transition matrix determining the melodic transitions
W	the duration of the entire section
n	the number of steps to generate the DTMC(P) sample-path (make it large)
nb	the number of beats per measure
bpm	the beats per minute
harmonicParams	the parameters for the harmonics generation (see details)
state_lengths	the size of the state-spaces for each voice i.e. the number of melodies per voice
melody_states	the list of state-spaces containing the pitch-tones in midi numbering
rhythm_states	the list of state-spaces containing the rhythms of each melody in fractions of measures

Details

The algorithm generates a large DTMC and then uses the sample path up until the section duration is reached, this allows for state-spaces of melodies of varying lengths.

The argument harmonicParams must be a named list containing:

• lambdaRange the range of mean number of harmonics

• betaRange the range of the beta distribution parameters

Value

list containing the wave vector and the time-error between input section length and generated

shill1729/synthesizer documentation built on July 9, 2020, 12:52 a.m.