The archiving of Macedonian music folklore in a digital format is undoubtedly the central and most important issue of modern Macedonian ethnomusicology. The successful conversion to a digital format should result in the formation of a multimedia data base with audio, video, graphic and textual data. The multimedia database should provide a complete overview of the different aspects of folk music and the possibility for various statistical and other types of analyses. Consequently, one should come to conclusions about the nature and laws that govern Macedonian folk music. Thus, two major goals should be achieved: the preservation of folklore in formats that would be useable or transferable in the future, and an unrestrained flow of communication between the different data formats.

    The lack of an appropriate methodology and defined standards together with the practical problems of the A/D conversion, determined our theoretical priorities - to develop the methodological aspects of digital archiving  of Macedonian music folklore first, and then apply the methodology in a given empirical corpus.

    Fortunately, the Firfov family held in their possession completely unprocessed Macedonian folk music material. This material was collected and recorded by the founder of Macedonian ethnomusicology, Zivko Firfov, in the period between the 1960s and middle of the 1970s, even though some of the recordings date as far back as the 1950s. Firfov has left behind him about 50 cassettes (the last cassette is numbered 50).

    Unfortunately,  Firfov himself has not provided us with sufficient information about the recorded material, and we are still in the process of surveying it. We roughly estimate that this collection consists of 1200 vocal and instrumental songs of Macedonian folk music. This repertoire is diverse and includes both the genres of rural and urban folklore. It includes well-known popular songs as well as completely unknown ones from the oldest rural tradition.  (Audio Samples (AS) 1,2,3,4)

(“Jano” - an old rural song from the village of Injevo near Radovis, performed by a vocal female group - Firfov Collection, Cassette No. 18MP3/Jano.mp3

 “Aj, donesete taja tenka puska”- a rural song with a “gajda” accompaniment - from the Kocani area, performed by the male duo “Stojanovi brothers” -  Firfov Collection,  Cassette No. 13  MP3/Ajdonesete.mp3

“Rucok ruca, more, vino pije” - a rural song with a “kemane” accompaniment - from the Kocani area, performed by the duo “Brother and sister Stojanovi” - Firfov Collection,  Cassette No. 13 MP3/Rucokruca.mp3

“Petko imal mnogu zeni” - an urban male song from the village of Gorno Nerezi near Skopje, performed by Angele Bocevski and Slave Cvetanovski with the RTS Orchestra of  folk instruments  -  Firfov Collection,  Cassette No. 2 MP3/Petko.mp3 )
 

     The quantity and nature of this material coincides with the goals and purposes of our project. The songs are recorded on audiocassettes,  which  is  unfortunately  an  unstable medium for the archiving of audio material. Consequently, there is an urgent call for these songs to be converted into a digital format.

    There have been three phases to our research project:

 
1. Defining the methodological aspects of the digital archiving of Macedonian folk music

2. Initial  testing of the methodology of the digital archiving of Macedonian folk music through one example (in our case the song "More zalajale" performed by Jana Janeva and Dana Mitova from the village of Chiflik, Kochani, 1972)
 

3. The formation of a pilot digital database through the conversion of the Firfov Collection.

    Therefore, our methodology should respond to the following demands and provide:
 

- Interactive browsing according to standard science and user parameters;

- Acoustical, statistical and other types of analysis

- Audio-graphic conversion and vice versa.

    If the well-defined instrument for analysis signals success in the further treatment of the empirical corpus,  it is evident that defining the structure of the database will be crucial in its use and  value. Thus, when defining the basic components of the digital database, we consider the following issues:
 

 -The characteristics of the material which is to be archived;

 -The technical possibilities and the capacity of the medium in which the archiving is conducted.

     As a result we concluded the following five essential components of the database for digital archiving of Macedonian folk music:

 1. Audio;
 2. Graphics;
 3. Text;
 4. Video;
 5. Animation.

    The audio part is the essential and central part of our database. In fact it refers to the main reason for the formation of the database. The remaining four components are mainly supplements, or a result of some analytical processes in the processing of the archived material.

     The graphic part of the base refers to music graphics, i.e. MIDI/ Score format.
 

     As a starting point for our textual component, we used the traditional  questionnaire  which  was  standard  procedure  for the archiving of Macedonian folk music in the past. This questionnaire was essential for the collection of ethnomusicological information in the Institute for folklore in Skopje - Macedonia. It consisted of general and specific information about the performer, the piece, the person who was recording, and the equipment and the conditions of the recording.

    The video component involves the photographic, video and electronic recordings of the events related to folk music (such as rituals, photographs of traditional clothing, instruments, and performers), and the remaining graphics (geographical maps, diagrams).

    The incorporation of an animation component in this database is somewhat unusual. Our idea was based on the opportunities offered by modern 3D-computer technology. We used this technology primarily for the purposes of an animation overview of the instruments.

    As we previously mentioned, the next phase involved operationalization of our methodology by the digitization of the song "More zalajale” .
 

(“More zalajale” - a  female  song  from the village of Chiflik near Kochani, performed by Jana Janeva and Dana Mitova - Firfov Collection,  Cassette No. 28 MP3/Morezalajale.mp3 )
 

    In the testing of this methodology, we used only the audio and the graphic component of our database, because in a way they reflect our central interest in the archiving of Macedonian folk music. On the other hand, because of the large scope of the project, we could only focus on these two segments, and leave the processing of the others to the remaining phases of our research.

    In the following phase our central goal was:

1. Conversion of the song to digital format;

2. Creation of the digital graphic of the same song.

    This conversion proved the advantages of the newly applied technology and also of the methodology that we chose for this project.

    One of our particularly interesting experiences was in using this technology to check the accuracy of our score transcription. Although Logic Audio (the basic hardware/software we used for conversion of our analog file) includes an option of audio to MIDI/score conversion, this part of the software has still not been developed to the level that it could read even simple monodic tunes. One of the features of this software is that it replaces all unclear  sounds with  the note C. When  we  applied this option to
our example, it converted the music to a lot of C notes, which made this technology practically unusable.

    Based on this experience, we decided that we would do the transcription of “More zalajale” in the traditional way.

    Our next step was to create a standard MIDI file, which we imported in Logic Audio. As a playback device we used a QuickTime instrument, which can be found as a standard component of all most recent system software. We decided to use a piano sound, because of its clear attack and pitch. However, when we compared the MIDI version  of  the  same  song  with  the  original audio through simultaneous playback, we noticed some discrepancies in our transcription.

(“More zalajale” (with MIDI piano) MP3/MorezalajaleMIDI.mp3 )
 

    This proved the advantages of our strategy in using both audio and MIDI files in our database, not only because of the score presence, but also as a control mechanism for the correctness of our transcriptions.

    We were aware that one song is insufficient to draw a final conclusion about the success of our methodology, and that the project's magnitude calls for its future realization in a greater empirical corpus. However, the development of the project was in complete accordance with our expectations. Therefore, the song was preserved in a format that provides:

 - a swift and easy conversion into other digital formats (the concept of data migration)
 - non-destructive analysis, i.e. a safe analysis of this material by other researchers in the future.

    Presently we are also concerned with the structure of the textual database, and (or) the initial steps in the definitions of the specific data we will be collecting. The other segments (video, music graphics and animation) will follow in accordance to the capacity of our human and technical resources.  It  is  certain that this is a complex issue that demands considerable human effort and technical engagement. Nonetheless, the promising results prove we are working in the right way and have encouraged us to continue to do so in the future.
 
 
 

       The categorisation of the ethnomusicological data in multimedia databases