Arts and Humanities Commons^™

Generating Audio Using Recurrent Neural Networks, Andrew Pfalz

LSU Doctoral Dissertations

Long Short Term Memory cells are a type of recurrent neural network that perform well when predicting sequence data. This works presents four approaches to modeling audio data. The models are trained to predict either raw audio samples or magnitude spectrum windows based on prior input audio. In a process called sampling, the models can then be employed to generate new audio using what they learned about the data they were trained on.

Four methods for sampling are presented. The first has the model predict a vector for each vector in the input. The second has the model predict one …

A Rhythmic Transcription And Spectral Analysis Of Luciano Berio's Thema (Omaggio A Joyce), Christopher C. Mccardle

LSU Doctoral Dissertations

Thema (Omaggio A Joyce), composed by Luciano Berio resides within a tempo of 80 beats per minute. The discovery of a constant tempo allowed for a precise rhythmic transcription to be created which found musical structures including large formal sections, subsections, phrases, rhythmic and melodic motives, layers, and dialogue. Analysis of the rhythmic phrases shows that electroacoustic music can have a controlled structure and that these structures earned Thema an enduring legacy of masterpiece. The transcription provides a road-map to compare and contrast different elements that reoccur throughout Thema. The work was composed with a well thought-out and …