A device for generating midi sequences based on music theory and statistical rules

Abstract

The invention discloses a MIDI sequence generation device based on music theory and statistical rules, comprising a computer system configured to: receive chord information sequence, rhythm pattern and key; initialize an octave according to the received key The initial pitch probability of the 12 tones; chord correction, melody motion correction, repetitive tone correction and tone threshold correction are performed on the initial pitch probability of the 12 tones according to the chord constraint, melody motion constraint, repeated tone constraint and tone threshold constraint. Obtain the pitch probability distribution of 12 tones; select a note from the pitch probability distribution in order of pitch from high to low for conditional screening, and add the notes that meet the screening conditions to the generated melody; integrate the generated melody and Rhythm patterns generate MIDI files. The MIDI sequence generating device can generate various MIDI sequences according to a given melody.

Description

technical field [0001] The invention relates to the technical field of music technology, in particular to a MIDI sequence generation device based on music theory and statistical rules. Background technique [0002] MIDI is an industry standard electronic communication protocol that can define different notes and control signals for various electronic musical instruments and other performance equipment. Therefore, MIDI is a method of symbolic representation of music, which is widely used in music production. [0003] With the development and popularization of artificial intelligence technology, computer composition has become a new branch of music production, and symbolic music creation, that is, MIDI sequence generation, has become an important part of it. Many previous work on MIDI sequence generation contains different approaches, often with their own generational backgrounds and strengths and weaknesses. [0004] The current mainstream MIDI sequence methods can usually ...

AI Technical Summary

Problems solved by technology

However, this kind of work currently has many defects, such as the quality of music generation is generally not high, the generated single-track MIDI notes will have many mispronunciation, and the harmony between the generated multi-track MIDI tracks is not high

[0006] To sum up, existing MIDI sequence generation methods cannot guarantee that the diversity of MIDI generation can still be achieved in the case of controllable melody generation.

And the singleness of the existing MIDI generation framework has also become a bottleneck of the MIDI sequence generation problem

Images