Conclusion
By identifying these false problems, we shall be able to develop a transcription program that will also be applicable to the abstract tablature of Hans Gerle. We also plan to develop, on a microcomputer, an interactive transcription software package applicable to the other types of tablatures. The use of computer science should thus considerably facilitate the work of the musicologist.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
A tablature represents a series of practical references that apply directly to the performance of the music instrument. Alphabetic or numeric characters indicate which string and which stop on the neck must be used to produce the string pitch. They are completed by proportional music signs: indication of duration, subdivisions of tactus, repeat marks and pauses. The tablature presentation varies according to the country of origin, the period, and the instrument.
The German notation differs from French or Italian in that it is synthetic, that is, symbolized by one character a double reference, the string designation and its stop on the neck. It is also abstract in two ways: alphabetic and numerical references do not follow the progression of the sound pitches, and the typographical presentation of the characters gives greater importance to the global visual perception of the player, instead of trying to reproduce the polyphonic structure of a work.
For example, to determine the nature of the note corresponding to “s,” which is C# or D flat, one looks in the vertical part for the character “i,” “r” or “4” (corresponding to E, A or D), then in the preceding and the following one. The presence of only one of them leads to the identification of a C#, otherwise it is a D flat. Moreover the identity of the ambiguous character, as soon as it has been determined, is extended to all octaves and for the full length of the piece. Such a test is not very general, and consequently the corresponding internal proportional value is used instead.
Automatic transcription by the E.R.A.T.T.O.-C.N.R.S. team. Musicological studies by Hélène Charnassé and Raymond Meylan. Computer implementation Henri Ducasse (Paris: French Society of Musicology, 1975–1978).
Automatic transcription by the E.R.A.T.T.O.-C.N.R.S. team. part 1: Musicologist: Hélène Charnassé and Raymond Meylan; part 2, 3, 4: Hélène Charnassé. Computer implementation: Henri Ducasse (Ivry, E.L.M.E.R.A.T.T.O., 1977–1978).
In fact, the collection of German tablature on which we are working contains mostly transcriptions of pre-existing vocal works. If they are profanes (chansons) or religious (spiritual melodies, psalms, motets), they are all written in counterpoint and the vocal as well as the instrumental compositions are organized on the level of the “part.” It is the conception and the superposition of the different parts which structure the organization of the piece. Their identification is thus necessary for determining the graphics of the proportional notation transcription (materialized by the orientation of the note stems) and calculating the duration of the sounds.
The suppression of the “cumulus” intervenes in certain cases, as when another note is played on the same string (the initial vibration is then suspended); when the duration of the sound is longer than a whole note it is restricted to a whole note. The carryover of rhythmic values over bars or beats will be studied later.
At first, the pitches were the only factor but we soon realized that we had to include duration as well.
Printing in proportional notation requires dealing with voice representation techniques. This has not been adopted yet.
As far back as 1976, Raymond Meylan, a team member, noticed that melodic continuity could be used to restructure the voices in a “Hans Gerle” tablature system. He did not, however, see the determining factor of the ending saturated vertical as a prime contiguity generator, and consequently difficult problems arose. They were mainly due to the absence of a fast proportional representation plotter, which prevented walking through a large sample of transcription to discover general rules, etc. Today, our microcomputer software enables us to screen a transcription of an average of 60 bars in less than three minutes and to try out several alternate algorithms in less than a day.
By convention, the authors of German tablatures respected the logical superposition of the heights of the notes in chords without considering the crossing of voices, which they could have easily reproduced. Ochsenkun himself, despite his constant reference to the vocal model, respects them rarely.
Additional information
Hélène Charnassé, musicologist, is the head of Equipe de Recherche sur l'Analyse et transcription des tablatures par Ordinateur (CNRS-E.R.A.T.T.O.) in Ivry-sur-Seine. Bernard Stépien, computer-scientist, is also a member of this organization.
Rights and permissions
About this article
Cite this article
Charnassé, H., Stépien, B. Automatic transcription of sixteenth century musical notations. Comput Hum 20, 179–190 (1986). https://doi.org/10.1007/BF02404458
Issue Date:
DOI: https://doi.org/10.1007/BF02404458