Transcript bulgarian academic music portal

Dr. Lubomir Kavaldjiev,
Associated Professor at
Bulgarian Academy of Sciences – Institute for Art Studies
Digitalization of the Bulgarian folk music
and its presentation on the Internet
This presentation is based on our more than 15 years
experience in the digital rewriting, manipulations and creation
of music and the representation of the received results locally
and in a computer networks, includingly on the Internet.
Bulgarian Academy of Sciences
Institute of Art Studies
Section: INFORMATION SYSTEMS IN MUSICAL
CULTURE
and its Web-orientated multimedia databases
presented worldwide today on
INTERNET
You can see here some of the results from our work in the
last 10 years on Internet address:
http://musicart.imbm.bas.bg
It is the BULGARIAN ACADEMIC MUSIC PORTAL
created in 2003 in Institute of Art Studies and fully
available since mart 2004
BAMP
Here are some “screens to video”
capture records. You can see how
everyone can easily navigate and search
in
the Bulgarian Academic Musical Portal.
Overview
Buenek dance
Djamal na Nova godina
We have started working on the audio
digitalization at the Institute of Art Studies in the
early 90s of the 20-th century. In our first attempts
we have used very primitive means (under
operating system DOS, an 8-bit soundcard and
software designed by Bulgarian programmers amateur).
Because of the humanitarian orientation of the
Institute, the attitude to our work was not
appropriate and we have no technical experts.
On the other hand there was a priceless
and quite big records collection of
authentic Bulgarian folklore (that is to
say, folk music untouched by any
modern composers). More than 170000
records have been created for the last
75 years. The records are on different
types of storage: “soft” shellack
records, vinyl records, magnetic tapes,
Umatic video cassettes.
All of them were in bad condition and needed
urgent restoration and re-recording. But we haven’t
any money and technical experts in audio and
recording. The only solution was to transfer the
most valuable records to PC and to process the
sound (noise reduction, spectral correction,
normalization etc.)
I’m not folklore researcher (ethno-musicologist). I am
specialist in music theory (systematic and cognitive
musicology, aesthetics, psychology), as well as in
electronic music. I have created a lot of compositions in
this sphere since 1985; some of them are in a folkelectronic style. The music that you can listen in the
beginning and as a background on the slides is based
on my electronic compositions. I am a FEM founder and
one of its organizers. (link to FEM homepage)
This is the first electronic and computer music festival in Bulgaria
(Gotze Delchev 1989), which has already gone through 13
performances till 2002. I also made the first 35-minutes television
Bulgarian film Underground mysteries in 1988.
These activities and the established contacts
helped me a lot in my work on the collection of folk
music for the Institute of Art Studies.
In 1992 I have already had some main scientific
objectives:
-Digitalization of the collection and computer
processing of the obtained audio results, as well as
their storage on CD’s.
-Cataloguing of the huge mass of text information
connected with this collection, effective means for
data input by operators who are humanitarian
experts (folklore educated) and creating of
computer applications for search.
-To make the results from the 2 activities above
accessible by the means of electronic publications
on the Internet or on CD-s.
Nowadays, 12 years later I can be
satisfied because these objectives
were successfully achieved to a greater
degree.
The first multimedia
database for musical
folklore started to work
already in 1993 with
operating system MAGIC
in DOS, but which has
Windows shell. The next
year we have already had
a short extract from it on
the Internet
After this internet orientated base
made an impression In Bulgaria and
abroad as the first base of this kind in
Europe and it was world
acknowledged and it received an
award from the GIP-competition
(organized with the support of G-7,
the European commission and
Japan) in 1997.
(GIP = The Global Inventory Project
is a G7 Pilot Project coordinated by
the European Commission and
Japan.)
PROJECT DETAILS
Summary
As a result next year we have been financed by a
the foundation Open Society .
The base was converted to MS Access.
The local application was put into practice with DELFI
programming, and by ASP scripts it was possible multimedia pages
to be made dynamically generated on a customer request. A year
later we were financed in addition by the same foundation for semiautomatic English translation development (we made a thesaurus
with folklore terminology) and transcription of the rest of the texts
data by international standard Cyrillic/Latin. In this way the Web data
were generated bilingual. Such is our base today as you have seen
in the beginning of our presentation.
In 2003, with the help of the Bulgarian Culture National Fund, we
have developed the project.
The BULGARIAN ACADEMIC MUSIC PORTAL was found which
already includes:
-The music publications Database of Institute’s associates
-an electronic edition of a magazine "Bulgarsko muzikoznanie"
(Bulgarian Musicology)-quarterly, online version & archive (which
shows issues’ summaries and contents )
-The Database for music performance. It is multimedia too, but at
that moment the audio is not enough, because of the lack of
money for technical equipments and specialists for recording
original performances from the professional players. Of course,
the Multimedia database for authentic Bulgarian musical folklore
mentioned above was integrated here.
BAMP
I have to call your attention to other two
problems that haven’t been solved yet.
They concern not only our collections, but
all digital collections of musical audio
information at all.
The first one concerns the way of
database organizing and searching. This
task is very difficult when you work with so
huge mass of integrated text, audio, video
and graphic information (In general we
work with scanned music texts which are
music decodes hand - written, like pages
from musical score).
The factual material can be catalogued
easily in alphabetical order, by names,
years and so on. But databases concerned
with clear musicology and culturology
require strong and undoubted system of
organization, which is accessible and easy
understandable from the user without any
special musical education. It demands wide
known terms databases about genres and
cultural functions. We are planning a
project which will show in the WEB space
glossary of the common – using musical
terms in Bulgaria and some other words,
concerned with the whole musical and
cultural practice, includingly the newest
achievements
So far we mainly based on traditional experience in
the musical theory and ethno – musicology, and
also some, created by me cognitive models
(synchronous and diachronous) for description of
universal recursive relations in the musical culture.
Especially this kind of models is TIEM for basic
functions of the Art from the point of view of its
user:
In the folk base – in one of many systems for searching is
exactly used this model, where the magic (ritual) function is
divided into two parts. This is due to the specificity of the
folklore rituals – customs based on concrete events
(weddings, births, funerals and so on) and other activities
which are annual. The Traditional (the day's routine)
function is divided into two parts too.
In a different way this model was used in our
Database for music performance. Moreover, it
integrates the search devices, more of them
created by me.
(Article)
Another important problem is called by me
“the second degree of digitalization”
on musical audio objects.
As far as I know, this problem about the systems for
musical searching is not solved worldwide too.
What I mean is that so far everyone who is searching a song has to know
the name of the author, or the name of the singer, the concrete song and
sometime even the name of the file (usually in MP3 format).
For example, most of the users of our folk database are Bulgarian
emigrants, who are home – sick or foreigners who knows some Bulgarian
song by ear and try to search it in this way, from our site or from somewhere
else.
Up to now it is not possible to satisfy such order for information from
distance. This would be possible only if they (he or she) are in contact with
a specialist – musician in front of whom the user could perform the melody
and the specialist, if he knows it could tell them the right name and the
place where it could be found. Often the musical specialist knows the
melody if he work in the same genre but he doesn’t want to share the
sources from where it is possible to be found via Internet for free or even by
paid access.
At the same time, we are offering free internet
access to more than 10 000 records of this kind.
But, we cannot offer search by singing a melody.
So far our search devices are text orientated or are
working with indexes describing presents of audio
multimedia or video-file, integrated to right info –
card for every object in bases.
But the sing – search is possible, too.
This is not only theory. It could be done with
software applications which already exist and we
are doing it manual.
Maybe somebody was impressed by complex
searching in our database about folklore, where
there is a field of MIDI. Now, however, if it is
searched by this parameter the database answers
that there are no existing records.
The MIDI protocol actually can be compared to electronic score. It is
remarkable approximation of the music audio. MIDI shows in a simple
numerical data almost everything which happens in music and in this
means that it follow the conventions of the writing down in musical notes
which add some things which can not be written in notes in traditional
way and which are described in score with words and special signs
(about temp, slow change of pitch, dynamic and so on). The MIDI
protocol is used in electronic music still from “analogous” period of
synthesizers for realization on connection between sound databases
and synchronous playing in a group. Today it is wide used in music, in
computer audio and video techniques, as in a light spectacular DJ’s
shows and in many other places.
All songs in a collection can be converted in
MIDI files.
If one performed song can be converted in MIDI
file the pitches and rhythm can be easily
extracted from that MIDI song and can be
presented with numerical strings. The same can
be done with received from the user audio file
where he is trying to sing the melody looking for.
Then the strings are compared on interval
relations between notes with all in database of
MIDI converted melodies and we see on a list roll
coincidences probability degree. The user has
also a possibility for trial hearing of all found in
this way audio examples. In this way the user
may decide which one is the melody looking for.
But the difference is that now it is not necessary
to download thousand records from Internet and
to listen all them. Now, the search is reduced to
some records, sometime to only one.
With that task the search by singing the melody is solved as a
scientific problem for us. We would like to underline once again
that till now we don’t know such search engines for popular music
in the world, nevertheless of the enormous searching especially of
music (MP3) in the last decade.
At the end of my presentation I will demonstrate exactly the same
converting. I will convert a folk song from our database into MIDI file.
The same song has a music transcription that was done at the time of
recording by the explorer ethno musicologist. As a result we can
compare: 1.the original audio recorded performance with 2.the ear
decoded performance written in notes by a specialist and 3.with the
MIDI decoded generated in the computer.
•
From now on, there is only one
step to solve the big problem
with automatic searching by
melody. In the Institute of Art
Studies at the Bulgarian
Academy of Sciences we are
sure that we can solve
successfully this problem in
collaboration with colleagues
from European Union and
other countries. Of course, we
need sponsorship and
proposals to our future
projects.
We hope that cultural
integration of explorers from
different scientific fields and
countries will help to achieve
useful final results. But, it will
be better if we could solve
successfully advanced
problems and if we work
perspective fields