Usage
  • 3 views
  • 2 downloads

Percussive Sound Generation with Virtual Listeners and Modular Synthesizers

  • Author / Creator
    Salimi, Amir Saeid
  • This work focuses on the virtual generation of short percussive samples which can be used by electronic music artists in their compositions. Although recent advancements in digital synthesis, heuristic search, and neural networks have been utilized for the generation of a variety of sounds, the lack of access to large audio datasets, the problem of open set recognition, and high computational costs persist as barriers towards the expansion of digital sound libraries using these techniques. We present our approach towards the automatic generation of synthesizer programs which mimic one-shot percussive sounds. This work documents the implementation of an automatic system for generation of virtual percussive synthesizer programs using classical signal processing and machine learning. This system relies on virtual ears to find synthesizer programs which mimic percussive sounds, and to further categorize these programs into a number of common drum types. We demonstrate promising results in both detection and categorization of percussive sounds by representation of digital audio through Fourier transformations and autoencoder embeddings. Manual listening tests of the generated sounds indicate that the system can successfully generate drum synthesizers and categorize drum sounds. To facilitate future research, we share our curated datasets of free percussive sounds. These datasets can also be used for the replication of our work.

  • Subjects / Keywords
  • Graduation date
    Fall 2021
  • Type of Item
    Thesis
  • Degree
    Master of Science
  • DOI
    https://doi.org/10.7939/r3-ka30-pr47
  • License
    This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.