melody
Play melodies on any scale and any mode
Play anything!
A software suite leveraging machine learning and music information retrieval algorithms for use in the music class. This is a research project developed in the department of Digital Media and Communication of the Ionian University by Dimitri Smailis aka Roblokie Aflo. In its current version the system is developing for use in music education. Relevant papers can be found in Research gate
Tools
This research project is about developing a set of tools to support music educators and creative practitioners. The basic idea is that you can play anything with anything i.e. play melodies, chords and drums of any timbre by striking a pen or a bottle of water on a table using a contact mic. No need for tablets or MIDI controllers of any kind. Turn your student tables into musical instruments that can play any sound from any key/scale. At this stage, three distinct applications have been developed, each addressing a core element of music: melody, harmony, and rhythm. An additional application is currently in development, allowing users to create an instrument with the exact number of notes comprising a tune's melody. Many thanks to Rodrigo Constanzo for his Data Knot tools and to the FluCoMa team for the library
hamony
Play any diatonic chord
anything
create an instrument with any number of triggers and play it
specific
Create an instrument with the exact number of notes of a piece of music
apps
video demo
a short video demonstration of the capabilities of the system. In this example 5 people play a samba tune (Voodoo by Rhythms of Resistance) with sticks hit on desks, every stick has a different material on the tip hence producing different timbre/mfcc's. Every stick produces a different class which is then mapped to a different sample
apps
video demo
Another demonstration of the capabilities of the system. The first part is the "Varkarola" by Offenbach where Melody, Harmony and Rhythm are all in use. The orchestration was made by the researcher Dimitrios Smailis. The second part is an idea of the students, can we play jingle bells' melody as a team? The third part is experimentation with several sounds provided through Ableton's samplers and synthesizers. What followed was a spontaneous rap like improvisation.
Machine learning techniques in the music class
To date, Artificial Intelligence based music generators have not thoroughly convinced musicophiles in terms of appreciation. This premise is amplified by the fact that humanity has a long way ahead in morally accepting the creative endeavors of machines. Therefore, articulating intent when using ML&AI in music practice seems to be an acceptable practice and this explains the fact that AI music generators are being promoted as creation assistants. Likewise, we suggest that ML algorithms have the potential to be used effectively as teaching assistants.
