During his visit to Montreal in April 1993, U Ko Ko, at my request, improvised a series of short compositions illustrating each of the tonalities or modes of Burmese music. He played these on a MIDI keyboard and the data was entered in Finale 2.0, a notational software program for the Macintosh. Finale is primarily a music notation program, but it does also allow direct entry of digital information from a MIDI device, such as a MIDI keyboard. The data thus entered is stored as a digital recording of the performance. In order then to get this digital information into notation, the recording is played back and the player must manually mark the beginnings of each time unit, or measure, usually by striking a key at the appropriate point in the performance. U Ko Ko, both played these examples and entered the time unit indications. The program then created a rough equivalent in Western notation of the performance. Although each key struck during the performance was entered in the transcription, even to keys which were lightly or mistakenly touched, the rhythmic accuracy of the performance had to be edited in order to more closely resemble the recorded sound. In spite of the effort invloved in editing notational inaccuracies, this method was able to preserve with great accuracy the important performance characteristics such as tempo changes and variations in velocity, e.g. key pressure.
Ensuring a high level of accuracy in the retained performance is a matter of juggling between setting the definitition at a high rate resulting potentially, in too much information to manage, or at a lower definition, possibly resulting in the loss of rhythmic detail. The important advantage which this method provides, is that every note struck in the performance is retained. The disadvantage is that refining the rythmic structure so that it better reflects the sound of the actual performance can be a painstaking process.
After obtaining a set of files which contained both sound and rough notation, these were then transferred by means of a Finale ETF (Enigma Transfer File) to a PC and imported into a Windows version of Finale 3.0 where the detailed editing was carried out. The transfer process retained all of the original performance characteristics. The automatically produced Finale version of the notation was not precisely what he played. I revised the computer generated score to better express in Western notation what I heard in the music. Numerous difficulties present themselves with this method. For one, the break point between the two hands is something which was not accurately relayed from the MIDI keyboard to the computer program. The computer decided upon an arbitrary break point between the two staves but it was necessary for me to re-edit these manually, according to my best understanding of Burmese piano technique. One difficulty I noted was that, as the notation was edited the original performance characteristics became replaced by the literal notated version. Thus, the playback now sounded like a literal and mechanical interpretation of the notation. Editing of the notation had to be done from a copy of the file in order to preserve the subtleties of the original performance.
What is particularly useful about this computer aided method of transcription is that the recording of his performance as he played it is retained in the file while at the same time, albeit with some rhythmic inaccuracies, every note originally played was retained in the notated version. Something close to the sound of the original performance was preserved in the performance characteristics in the Finale file, and this could be slowly and painstakingly reworked until the notation more closely resembled what I heard in the original performance.
In all of this the question must be asked, is the final transciption a description of the original performance, or a prescriptive notation for later analysis or performance. I compromised. A purely descriptive notation is not feasible within the limitations of the traditional Western notation system. What I sought was a notation which would show the fundamental elements of pitch and rhythm and make visible the two line structure of the music and Ko Ko's characteristic performance of it, while the recorded version of the performance characteristics in the file would serve as a more accurate record of the subtle changes in tempo, (key) velocity, and rhythm.
email to rgarfias@uci.edu
last updated 6 November 1995