Skip to content

Latest commit

 

History

History
14 lines (11 loc) · 1015 Bytes

README.md

File metadata and controls

14 lines (11 loc) · 1015 Bytes

BrainPan

Panning audio to the ear with the relevant ear advantage according to Speech vs Melody

Theoretical Background

  • Neuroscience has shown that the right ear in most humans is better able to distinguish and relay information about speech information, and the left ear is better able to distinguish and relay information about melody.
  • This is because the left ear's received auditory input is modulated by processing in the right-hemisphere paralinguistic speech areas, and the right ear's auditory input is modulated by processing in the left-hemisphere linguistic areas, including Broca's areas, Superior Temporal Gyrus/Sulcus, and others.
  • I use machine-learning methods to identify the melody components of a song, and pan them to the left ear; identify the speech components of a song, and pan the to the right ear - for the optimal listening experience according to modern neuroscience.

Song

BADBADNOTGOOD - Time Moves Slow

Misc.

  • Must be run using Python2
  • Requirements: PyDub, Numpy