MMM: Mathematical Morphology & Music

Gonzalo Romero-García

This is the repository that accompanies my PhD Mathematical Morphology for Analysis and Generation of Time-Frequency Representations of Music.

Sounds from Chapter 3

They are stored in results/chapter_3 organized by instrument. Each instrument has

• input.wav: the input file
• sinusoids.wav: the sinusoidal component
• transient.wav: the transient component
• filtered_noise.wav: the stochastic component
• output.wav: the output; the sum of the previous three

Structure

The repository is organized as follows:

• data: the input data for the scripts, organized in:
  • audio: the .wav files
  • midi: the .mid files
  • musicxml: the .musicxml files
  • objects: the generic objects in pickle format
  • scorexml: the XSD for the ScoreTree object and the corresponding .xml files.
• mmm: the library itself, separated in piano rolls and spcetrograms
• phd: the folder where the images necessary for the PhD are saved, organized by chapter.
• results: some results of the PhD, in particular:
  • chapter_3: the audio files of each instrument. The audio files are the following:
    • input.wav: the input file
    • sinusoids.wav: the sinusoidal component
    • transient.wav: the transient component
    • filtered_noise.wav: the stochastic component
    • output.wav: the output; the sum of the previous three
  • chapter_4: some examples of the output of the compilation of a ScoreXML file into a .mid file
• scripts: all the scripts used.
  • graphs: the scripts used for creating graph figures for the PhD
  • introduction: the scripts used for creating the images of the introduction
  • pianorolls: the scripts used for operating on piano rolls;
    • analysis_harmonic_textures.py: a script corresponding to the figures and computations from the chapter 5, section 1
    • analysis_textures.py: another script for the same purpose as the precedent
    • compilation_score_tree.py: the script for compiling a ScoreTree into a midi file. The name parameter in line 8 serves to refer the name of the file, i.e.m, data/scorexml/[name].xml
    • dilation_harmonic_textures.py: a script for generating a figure of the PhD
    • dilation_harmonic_textures_default.py: another script for the same purpose as the precedent
    • find_minimal_activations: the script for finding finimal activations with the method proposed in the PhD (chspter 5, section 2)
    • graph_covering_activations: a script for generating some images from chapter 5, section 2
    • tonal_graph.py: the script for generating the tonal graph from chapter 5, section 3
  • spectrograms: the scripts for the chapter 3, namely:
    • analysis_synthesis_pipeline.py: the full pipeline (Mathematical Morphology and STN synthesis) of the chapter 3
    • transient_generation.py: the script for grnerating an image of chapter 3
• tests: a folder with tests