Transient detection is a common rst step in a large numbers of audio software features, including time-stretching, elastic audio, pitch-shifting, compression and many more.
This dissertation presents an overview on uses, methods, techniques for transient detection, and two algorithms which have been developed during a four months stage experience at n-Track Software in Rome, with the supervision of Dr Marco Bertola and the n-Track development team.
These algorithms have been implemented in C++ programming language, based on papers B. Thoshkahna, F. X. Nsabimana, K. R. Ramakrishnan, Transient Detection Algorithm for Audio Using IterativeAnalysis of STFT", 12th International Society for Music Information Retrieval Conference, 2011, and F. Patin,Beat detective algorithms", 2003. The focus of this work is an analysis and a comparison between chosen time-based and frequency-based techniques and their nal extension and implementation to be a tool for n-Track Studio DAW.
Archivio tesi