Search results
Nevertheless, automatic transcription of polyphonic music recordings still needs improvement (Holzapfel et al., 2019). Since many MIR algorithms depend on high-quality ground-truth training data, a joint effort of ethnomusicologists and MIR researchers can lead to technological improvements.
- 1 Introduction
- 2 Background
- 3 Method
- 4 Results
- 5 Conclusion
- Funding
Music transcription is a process of creating a notation of musical sounds, with music notation being the representation of musical sound through some other medium. This can take many forms, but this article, and the field of Music Information Retrieval (MIR) in general, is concerned with static, visual representations of sound, and specifically wit...
2.1 Music transcription in ethnomusicology
If they transcribe music at all, ethnomusicologists usually aim to document a specific performance on the basis of a recording (so-called ‘descriptive music-writing’), rather than to provide a model for performance (‘prescriptive music-writing’; Seeger, 1958). An ethnomusicologist may choose to make a ‘close transcription’ that includes details of playing style—melodic or rhythmic nuances, ornaments, timbral effects, etc.; or a ‘broad transcription’ in which such details are omitted in order...
2.2 Music transcription in MIR
In the MIR field, AMT is typically defined as the process of converting an audio recording or audio stream into some form of human- or machine-readable notation (Klapuri and Davy, 2006). The first approaches for automatic transcription of musical audio to machine-readable notation originated from the 1970s (e.g. Moorer, 1975), with the problem gaining attention from the early 2000s with the development of signal processing and pattern recognition methods for analysing audio signals. AMT is ge...
2.3 Music transcription in between the fields
In ethnomusicology, mechanical devices for automatic transcription and analysis have been used since before the advent of computer technology (Ellingson, 1992; Cooper and Sapiro, 2006), but their application has been limited mainly to the analysis of pitch (tonometry) and melodic contour (melography). The melograph devices developed and advocated by Seeger (1958) and Hood (1971; 1993) were a response to the limitations of staff notation and manual transcription: a melogram or spectrogram coul...
The methodology employed in this article consists of four main steps, which will be discussed in the following subsections. First, a previously conducted user study (Holzapfel and Benetos, 2019) has been extended, in which a group of musicology students with experience in transcription was asked to compile transcriptions for a series of short music...
In order to establish the basis for our analysis, we investigate the consistency between the two experts in the ratings of the transcriptions. As depicted in Fig. 1, there is a large consistency between the two experts in their ratings, with a correlation coefficient of 0.754. The ratings cover a large range of the overall scale, with about two-thi...
By comparing ratings of human experts with computational metrics through corpus and close analysis, we documented differences in how the quality of a transcription is assessed in ethnomusicology and in MIR. We revealed several aspects that the metrics seem to be ‘missing’ in Section 4.3. Computational metrics are only partially correlated with huma...
This work was supported by the Swedish Research Council (2019-03694 to A.H.); the Marianne and Marcus Wallenberg Foundation (MMW 2020.0102 to A. H.); E.B. is supported by a Turing Fellowship under the Engineering and Physical Sciences Research Council (grant EP/N510129/1).
The use of computers and other machines by ethnomusicologists is by no means new – Charles Seeger developed the Melograph for graphic transcription and the real-time analysis of pitch, dynamic, and timbre back in the 1950s – but a trend of using electronic tools for transcription and analysis did not follow for another 50 years.
- Leslie Tilley
- tilley@mit.edu
- 2018
Ethnomusicology (from Greek ἔθνος ethnos ‘nation’ and μουσική mousike ‘music’) is the multidisciplinary study of music in its cultural context, investigating social, cognitive, biological, comparative, and other dimensions involved other than sound.
the art in AMT may have to offer for (ethno)musicologists transcribing a piece of music. In the field of MIR, recent AMT research has mostly focused on automatic transcription of piano recordings in the context of Western/Eurogenetic music (see [3] for a recent overview).
Aug 25, 2021 · This paper argues for ethnomusicologists to begin using performance not just as a tool to understand the social and cultural field, but to use music and dance as methods in ‘translational’ ethnomusicology that focuses upon the translation and communication of artistic performance aesthetics and to theorise a space for research outcomes that ...
People also ask
Do ethnomusicologists use automatic transcription?
Are transcription and standard analysis necessary for Ethnomusicology?
Do ethnomusicologists transcribe music?
What is ethnomusicology?
Nov 4, 2019 · A user study on evaluating the usefulness of automatic music transcription in the context of ethnomusicology, which collects and analyzes quantitative measures, and receives a range of qualitative feedback from study participants, which includes user needs, criticisms of AMT technologies, and links between perceptual and quantitative ...
