Sample based Bass Guitar Transcription

Jackman Brookes


Supervised by Dave Marshall; Moderated by Charith Perera

Transcriptions are the written form of music and are essential for the performance, study, and preservation of music over the years. Traditionally, the process of transcription is performed by a skilled musician who listens to a piece of music and transcribes what notes are played. For most people this takes years of training their ears to recognise notes but with the advent of automated transcription algorithms, transcribing music is more accessible than ever before. Current leading transcription algorithms are very proficient at note detection but there is more to an accurate transcription than just what notes are played. How those notes are performed is critical to understanding how a given piece of music should sound. This research focusses on trying to advance current transcription algorithms by developing a machine learning method for bass guitar articulation classification. To achieve this, various machine learning approaches are tested, along with different spectrogram types to find the most accurate model for articulation classification.

Initial Plan (06/02/2023) [Zip Archive]

Final Report (10/05/2023) [Zip Archive]

Publication Form