Spv: Torbjörn Gulz
Using single tracks of double bass in a jazz-ensemble-recording to improve the estimation of the beat and use the measured positions in a score-notation of the performance
The aim is to find methods to adjust improvised jazz music to a grid built on onsets from the double bass, the drums, or tapping using Score Cloud.
In this project, you will work with experts at KMH Royal College of Music and Doremir (the company behind ScoreCloud).
Objective:
You will try to find a better way to present a score (Western notation) when more than one instrument should use the same beat-mapping grid. The meter-grid is created from one of the more beat-stable instruments such as bass or drums. In cases where none of the included instruments can be used for beat-mapping, annotated tapping is another alternative.
You will be provided with sound files from jazz recordings. The music will be in separate tracks. You will collaborate with jazz musicians while trying different levels of difficulties.
Background:
In the last twenty years, the possibilities of transcribing acoustic music using technology have increased. Still, one of the difficulties is to determine the beat of the music creating the meter-grid for the score. The problem could be changing tempo, inexact tempo, or even to find the beats. Jazz is often rhythmically complicated, but in most cases, the score is simplified, which creates a gap between the computer's detailed annotation and the expected, more readable notation.
Challenges:
- To process annotated data from the recordings and get a better presentation of all the instruments using a score
- Present easier-to-use score interpreter for music with a more "unsteady" beat. This interpreter will partly be used for transcriptions.
Technical environment:
Score Cloud, Sonic Visualiser, Matlab. (Logic)
Outcome:
Present a usable score interpreter for improvised music with a "floating" beat.
Supervisors: main-supervisor: Torbjörn Gulz (Jazz performance expertise), co-supervisors: Andre Holzapfel (beat tracking, onset detection), Sven Ahlbäck (CEO of Doremir, the company behind ScoreCloud)