Towards Automated Processing of Folk Song Recordings

Abstract

Folk music is closely related to the musical culture of a
specific nation or region. Even though folk songs have been
passed down mainly by oral tradition, most musicologists study
the relation between folk songs on the basis of symbolic music
descriptions, which are obtained by transcribing recorded tunes
into a score-like representation. Due to the complexity of
audio recordings, once having the transcriptions, the original
recorded tunes are often no longer used in the actual folk song
research even though they still may contain valuable
information. In this paper, we present various techniques for
making audio recordings more easily accessible for music
researchers. In particular, we show how one can use
synchronization techniques to automatically segment and
annotate the recorded songs. The processed audio recordings can
then be made accessible along with a symbolic transcript by
means of suitable visualization, searching, and navigation
interfaces to assist folk song researchers to conduct large
scale investigations comprising the audio material.