Research trends in music categories
This research effort was originally used to find out what average music looks like from a signal processing standpoint.
Research Paper: http://raphaelelspas.com/averageMusic/.
Clone this repo.
Install youtube-dl: sudo apt-get install youtube-dl
Install ffmpeg: sudo apt-get install ffmpeg
Jump into git folder: cd Average-Music
Get midi-csv converter: wget http://www.fourmilab.ch/webtools/midicsv/midicsv-1.1.tar.gz.
If this command produces a failure, go to (http://www.fourmilab.ch/webtools/midicsv/).
extract folder: tar -xzvf midicsv-1.1.tar.gz
Jump into midi-1.1: cd midicsv-1.1
make: make
jump back to main folder: cd ..
get wav-midi converter: git clone https://github.com/kichiki/WaoN.git
jump into WaoN:cd WaoN
make: make
All scripts are in scripts folder:
cd scripts
More man/options info at the top of each script file.
Put mp3 in Average-Music/mp3/
python notePresence.py <csv_file> [options]
chmod +x multipleYoutubeNotePresenceGraph.sh
./youtubeNotePresenceGraph.sh <youtube_URL> <file_output_name>
Shell script reads from text file which needs to be formatted like this:
<youtube_URL> <file_name>
<youtube_URL> <file_name>
Example:
https://www.youtube.com/watch?v=s0f_go8vQOY river_flows_in_you
https://www.youtube.com/watch?v=pFZ5_q86u0E sherlock
Code:
chmod +x youtubeNotePresenceGraph.sh
./youtubeNotePresenceGraph.sh <youtube_list.txt>
Will fill in soon
mp3 to wav coversion is done with the ffmpeg package that can be installed on the command line with the following command:
sudo apt-get install ffmpeg
wav to midi is done using the program waon found here.
midi to csv is done by the included midicsv.exe found here. This file is included in the source here so it does not need to be downloaded again.
This program first collects an mp3 from youtube-dl for ease of collecting mp3s:
- mp3 --> wav
- wav --> midi
- midi --> csv
There is a batch script that performs all of these tasks and places each file that it creates in it's respective folder.