The IRAM digitisation  of the Firfov Collection was the first and the largest project for the systematic digitisation of the sound cultural  heritage in Macedonia until now (see BuzAr Collections). The digitised material created varieties of perspectives in the analysis and the understanding of the processes of the creation, reproduction and retention of the music folklore tradition.

During the formation of the secondary data for the audio material, we faced serious problems, since Firfov himself left very little written material regarding the participants, the time, the location and the conditions during the recording.

In order to collect and to reconstruct the missing data, we started to search for the singers and the other participants in the recording of this collection. Among the first to be approached were those whose names where known, or identifiable and living in Skopje, i.e. two singers from the female group Bapchorki – Leta Bardjieva and Dosta Donevska. The Bapchorki group alone was recorded with 64 songs in the Firfov Collection. Leta Bardjieva  was an especially interesting informant for us. She was the leading singer in the Bapchorki group with an extremely rich repertoire.  She was recorded with 17 songs as a soloist the collection, and there is a possibility that there are more recordings, since these data are preliminary at the moment.

We have defined several goals for these re-visits:

– to reconstruct the events which contributed to the recording and the creation of the Firfov Collection;

– to find out more about the background of the singers who participated in the recordings;

– to ask them to sing some of the songs (or other) which they recorded three decades ago.

We expected direct results in the improvement of the secondary database and thus better empirical material for comparison and analysis for the tertiary database.

In order to avoid discrimination of the answers or the singing, we tried to establish a very relaxed type of communication with the informants, without giving any special suggestions or interventions.

We visited Leta Bardjieva and Dosta Donevska in Donevski’s house in Skopje on January 19th, 2002. The video recording which was done with Sony miniDV camera and a stereo mike includes cca 120 minutes conversation and singing. We have edited a video documentary on Leta Bardjieva and Dosta Dimovska with their descriptions of the music folklore and its relations to the customs and life in the village of Bapchor, the formation of the vocal group Bapchorki and the recording conditions. Leta Bardjieva and Dosta Donevska sang 6 songs for us. Five of these songs are also recorded in the Firfov Collection. Some of them are present in the collection in several different versions and sang by different singers . The tune of the song “Chupche od onaka” is present in a version with different lyrics – Kinisav da odam. Here is the list of the songs and their versions:

1. Apsana me klade (Leta Bardjieva and Dosta Donevska, 2002)

Apsana me klade (Firfov Collection K20-2, Bapchorki)

2. Chupche od onaka (Leta Bardjieva and Dosta Donevska, 2002)

Chupche od onaka  (Firfov Collection K15-1/2 Bapchorki)

Kinisav da oda (Firfov Collection K7-1/2 Leta Bardjieva)

Kinisav da odam (Firfov Collection K20-1 Kosturchanki)

Kinisa da oda vo selo D’nbeni  (Firfov Collection K20-2/1 Kosturchanki)

3. Jordano mori zrno odbrano (Leta Bardjieva and Dosta Donevska, 2002)

Jordano mori zrno odbrano (Firfov Collection K26-1 Bapchorki)

4. Katinka bola legnala (Leta Bardjieva and Dosta Donevska, 2002)

Katinka bola legnala (Firfov Collection K26-2 Bapchorki)

5. Vecherajte ne chekajte (Leta Bardjieva, 2002)

Vecherajte mila mamo (Firfov Collection K1-2 Leta Bardjieva)

Vecherajte mila mamo (Firfov Collection K7-1/2 Leta Bardjieva)

Vecherajte mila mamo ne chekajte (Firfov Collection K10-1/2 Kosturchanki)

Vecherajte lele (Firfov Collection K21-1 Lefterija Shankovska)

This excellent and fascinating material gives enormous opportunity for comparisons and analysis both on a textual and contextual level. This was part of our definition of the mission during the digitisation of this collection.

Having in mind the complexity of the material, we decided to concentrate our research only on two segments - the pitch and the duration (tempo) of the songs.

At the very beginning we decided to determine the approximate pitch of each song and its versions, based on the nota initialis (or more precisely the first sustained tone, having in mind that some of the songs start with glissando), compared to A4 = 440Hz. We approximated, or rounded, each nota initialis to the closest pitch in our half step Western division, except in the case of “Kinisa da oda vo selo D’nbeni”, where there is a variation of the beginning. In this case we calculated the tone which is adequate to the nota initialis in the others (following the melodic line). We also did the comparison to the relative pitch of the newest recording, i.e. the recordings we did with Leta Bardjieva and Dosta Donevska in 2002. The following results were obtained:

1. Apsana me klade (Leta Bardjieva and Dosta Donevska) – F-sharp

Apsana me klade (K20-2, Bapchorki) – G

2. Chupche od onaka (Leta Bardjieva and Dosta Donevska) – C

Chupche od onaka  (K15-1/2 Bapchorki) – C

Kinisav da oda (K7-1/2 Leta Bardjieva) –  E-flat

Kinisav da odam (K20-1 Kosturchanki) – A (below C)

Kinisa da oda vo selo D’nbeni  (K20-2/1 Kosturchanki) –  A-flat (below C)

3. Jordano mori zrno odbrano (Leta Bardjieva and Dosta Donevska) – C

Jordano mori zrno odbrano (K26-1 Bapchorki) – B (below C)

4. Katinka bola legnala (Leta Bardjieva and Dosta Donevska) – A

Katinka bolna legnala (K26-2 Bapchorki) – C (above A)

5. All versions of  Vecherajte ne chekajte – F-sharp

At first glance, there was no consistency in the treatment of pitch by the folk singers except in the fifth example “Vecerajte ne chekajte” where all five performances, carried out at different times and by different singers, had the same pitch. There is a half step difference in the first and the third songs, and minor third difference in the forth song. The second song, is a special case because in the version “Chupce od onaka” we have the same pitch, while the other version “Kinisav da odam” is a minor third above in Bardjieva’s performance, and minor/major third below in the two recordings of the vocal group Bapchorki.

Having in mind that we rounded the pitch to a half step, the examples with the half step difference come closer in the 2nd and the 3rd example, or higher than a half step in the 1st example.

When we used the term consistency we did not mean only that there are differences in the pitch, but also that there are differences in the direction of the intervals  (up or down in comparison to the latest recording).

We considered the fact that the differences in the pitch could be a result of the speed of the cassette deck which was used for recording or playback. All recordings in the Firfov Collection where made with the same Philips machine, and we did the conversion using the same Marantz machine. The fact that some of the songs are higher and some lower is an argument against this possibility. An additional argument is that  there are differences in the pitch in the variants of the same song (number 3 and number 5) in the Firfov Collection.  Still we can not entirely deny  the possibility of creation of artificial difference, or similarity  in the pitch of the different songs due to the different speed of the recording and playback machines. This difference could be applied only to the comparison between the recordings we made from one side and those from the Firfov Collection on the other. The four versions of the song “Vecherajte ne chekajte” from the Firfov Collection are recorded and played back with same machines, thus there is no  possibility of an artifact in the pitch difference.

In the further course of our research we decided to concentrate on the analysis of the five versions of the song “Vecherajte ne chekajte”. There was an additional reason to select this song.

During the recording Leta Bardjieva started the song in another key, then stopped and asked to stop the recording because she was not satisfied with the pitch and she would like to start from more “appropriate” key. She “tuned” her voice and started the song with the identical pitch from her recordings three decades ago. There were no any tuning instruments in the room, or any other audio means (such as radio, TV etc.) which could have influenced her determination of the pitch. Leta Bardjieva does not have any music education, voice or other music training. In fact, this was the main incentive for us to start the special research of the pitch phenomenon.

In order to determine more precisely the pitch of each of the five performances of this song we approached them through the spectral analysis using Capybara hardware and Kyma software, from Symbolic sound corporation.

Figure 1. Spectral analysis of the five versions of “Vecherajte...” song

The spectral analysis was particularly applicable as all performances are vocal and monodic. Thus we avoided any additional influences in the distribution of the spectral envelopes which can be blurred by the use of instruments or polyphony.

Being interested only in the determination of the pitch differences, we limited the creation of the samples which will be used for the spectral analysis only to the first phrase of the song (the lyrics “Vecherajte lele”).We created the spectral files with the help of the spectral analysis tool. 44.1 kHz samples were used as basic and the spectral analyzer length which controls the length of the FFT, was positioned at 256 tracks (or sine waves, covering the area up to the Nyquist frequency).

Above we have presented the cuts up to approximately 5 kHz (having in mind that energy of the voice is concentrated mainly in these regions) of the five variants of the Vecherajte song.

Number 1 is the newest version, 2 and 3 Leta Bardjieva’s versions from Firfov Collection, 4 Kosturchanki’s version, and 5 Lefterija Shankovska’s version.

To determine the pitch of the fundamental sine wave we zoomed in the region of  the nota initialis and created new windows.

For this presentation we used a part of the spectrum analyzer window with centered 5th track. We inserted the pointer at the approximate position of the average pitch of the nota initialis (it is obvious that it can not be a fixed frequency, and is more “a number of frequences around the perceptive pitch”) and obtained the following results:

1st – at 387.1 Hz

2nd – at 375.7 Hz

3rd – at 376.6 Hz

4th – at 384.8 Hz

5th – at 378.3 Hz

For comparison the positions of the closest notes are: F-sharp 4 = 370 Hz and G4 = 392 Hz.

In general it means that all five versions are somewhere between these two pitches. Having in mind that most of the people find it difficult to catch the difference even between quarter steps, this pitch will be heard as the same. In fact, most of the musicians listening to the samples expressed their opinion that it was the same pitch.

Figure 2. Fundamentals of the spectral analysis of the five versions of “Vecherajte...” song

The first conclusion after the analysis of these examples, and especially after Leta’s attempt to locate the “correct” pitch, was that folk singers have internal or innate tuning instrument. The same pitch could not be mere coincidence as it was reiterated by different singers in different occasions, even after 30 years. The other examples confirmed the same conclusion as the majority of them are again either in the same or very close pitch.

Zivko Firfov transcribed this song for the book “Sokol mi leta visoko” (Gjorgji Donevski 1980) with nota initialis G. There is a simple explanation for this transcription – writing in G, was easer for reading.

As Leta Bardjieva stated during the interview, she has never recorded this song. The song became popular through the recording of the rock group “Synthesis” under the name “Me fatile”. When we compared the pitch of that version we found out that the comparative first note of the song is located between G-sharp 4  and A4.  We could assume that the pitch was determined due to the folk instruments included in the performance. The singers and the musicians from this ensemble learned the song through the score from the book “Sokol mi leta visoko”. Also, this group consists of educated musicians.

The question which followed our analysis was – if there is such a thing as “internal tuning instrument” is this a textual or contextual characteristic?

We could assume that the folk singers understand the pitch in a contextual manner. Their search for the most appropriate pitch is a search for the closest meaning they would like to convey.

As an argument pro this distinction we should quote the example of the song “Chupche od onaka”. Leta Bardjieva has recorded this song separately with other lyrics (Kinisav da oda) singing the same tune for a third interval higher. Singing the tune in another key, especially much higher, for a singer as Leta who has obviously a very developed sense for the pitch, was equivalent to singing another song, i.e. changing the context.

As far as the other element we were interested in are concerned – the duration (tempo), or the temporal dimension, we cannot say that we could find any similarity, and some of the songs might be very  different (much longer or shorter) in terms of the total duration. In fact, the small differences which appear at the phrase level are accumulated during the performance. Even the three versions of Leta Bardjieva’s performance of the song “Vecherajte” are very different in their duration. One explanation could be that the temporal dimension depends on lots of other factors - the mood, environment, breathing etc.

If  we compare these two characteristics – pitch and duration with their understanding in the classical music genre; we will see a completely opposite situation. Pitch is not a fixed category (the same song can be transposed depending on the voice, or the instruments performing it, for instance soprano or alto, violin or cello), while the tempi are precisely determined. The total duration of the composition could not differ substantially from performance to performance. If we follow the  globalization processes in music which aim to unify and adjust the local music systems, education, performance, training etc., to the Western dodecachordon culture,  can we say that humankind is losing this fascinating feature, i.e., the link between pitch and context?

* First published in 2003. Contemporary Trends in Musicology and Ethnomusicology, Second Struga Conference, ed. D. Buzarovski. Skopje: IRAM, pp. 41–52; also available at http://buzar.mk/IRAM/Conferences/IVConf.html (2003)