Studying the Intervenience of Lyrics Prosody in Songs Melodies
José Fornari · 2013
ABSTRACT Songs are made of two intrinsically connected parts: poetry (in the form of songs lyrics) and music. The proper fitting between these parts seems to be made by acoustic features that encompass the relationship between them, representing two fields of sonic communication: musical and verbal communication. While lyrics convey semantic meaning, music enhances its emotional intention, filling informational gaps and enhancing its signification that otherwise would make the poetic meaning of lyrics incomplete of even misleading. This work presents an introductory research about the influence of lyrics on their accompanying melodies. The experiment here presented analyzes three famous popular songs. Computational predictions, given as time series of eight acoustic descriptors, were retrieved from pairs of audio files; one solely with the speech of the lyrics, and another solely with its corresponding melody. In order to avoid data tainting from human emotional interpretation, the audio files with the speech were generated by a text-to-speech voice synthesizer. For the same reason, melodies are generated by MIDI files. These pairs were analyzed by computational models of higher-level acoustic descriptors that output time series representing the development of a particular acoustic aspect on time. The correlation of each acoustic feature for each pair of audio file are here presented, in the form of the correlation coefficient. R The experimental results are here presented, explained and discussed, in order to introduce a study on the acoustic features that better describe the intervenience of lyrics prosody in song melodies.