First of all, we would like to apologise to the real specialists in this sector, the Music Lecturers and Teachers, who first had to study very important and complicated rules regarding Harmony (that is, the art of combining and linking simultaneous sounds) and then put them into practice in order to allow their students to understand harmony fully, allowing them to apply musical structures correctly.
We would also like to apologise to those who, down the centuries, have worked hard at finding a fabric suited to musical scores, which succeeded in providing "almost" fixed parameters lending some regularity to the concept of "music as the art of sounds", at least in the Western World.
This programme has no intention nor desire to present itself as some type of universal panacea, in a setting that is already well defined and established.
Our aim is simply to help all those who wish to become more involved in this wonderful world of sound, despite their shortcomings or having a level of knowledge that is almost limited to musical notions, to create their own music in a way that may perhaps be atypical, but that is certainly highly personal .
Machiavelli said that the end justifies the means...
It was one of those dull, tiresome afternoons with the distant sound of thunder foreboding impending bad weather, and interrupting the silence from time to time.
Awaiting a deluge (or something very much like it) the only familiar sound was that produced by a pounding head.
"Is it possible", the owner of that head asked themselves, "that despite my musical studies, I am incapable of creating any combination of sounds that could, to all effects and purposes, be defined as music"?
This was the dilemma being faced when, going back in the mind to a science-fiction story written recently, the question was formulated cautiously:
"If it were music, what type of music would it be?"
The reply, unleashed from the depths of some sensation, was:
"Music beyond music!"
This gave rise to a profound conflict between Reason and Creativity.
Reason wished to go into serious matters, or at least thoughtful ones.
First of all, the person in question knew absolutely nothing about Harmony.
She had only learned to read music and reproduce the sounds on the piano, but not to compose.
Creativity on the other hand proposed (almost student-like) a pure and simple challenge.
"What if we tried converting words into sounds?"
So she started thinking and writing, producing graphics, erasing and changing what she'd already done - layouts and anything at all that can be said.
Finally, with a sufficiently legible manuscript she turned to a friend of hers, a lecturer in computer science, in an attempt to involve her in this game/challenge.
The strange thing is that she was so successful that the result of this immense effort was "SOUNDWORD"!
"Soundword" is an innovative program that allows anyone who so wishes, to generate pieces of music automatically, starting from a written text.
This brief report describes the basic idea behind the way it works.
We will also list a series of possible practical, worthwhile and interesting applications.
In addition we will outline what refinements and extensions we believe could be made to the program.
"Soundword" is a program that allows you to produce musical pieces completely automatically, starting from a text written in any language for which a finite range of symbols is used to write its words.
The program does not so much change the assigned text into music, as much as it uses the sequence of symbols making up the text (the words) to generate a particular succession of sounds.
Music produced by "Soundword", called "Metamusic" only partially applies the traditions used for centuries now in the West for composing music. Nevertheless, it remains sweet to the ear.
The instrument that produces the sound is a piano, since it is ideal for this purpose.
It sounds good, does not provide unpleasant disparity, and offers the greatest range of registers and dynamics. In fact, it is almost without defect.
The pieces of music are read in the G key (also called the Violin key) (fig. 1) and can be reproduced on the piano or on a keyboard (central section), using the right hand (fig. 2).
They are structured on the basis of regular 4/4 beats to the bar. This time has been used as it allows a greater number of notes to be inserted more easily into the bar, that is, the bigger or smaller space on the line occupied by musical symbols (fig. 3).
Looking at the keyboard on a piano, we take seven white and five black keys on it into account. This range allows us to obtain twelve semitones.
The "alterations" are not indicated at the beginning of the piece (sharps, flats - that is, the coded signs that are used to modify the notes in the piece, making them become semitones) because, not knowing what result will be obtained from the entire phrase, an expert would only be able to say at the end of reading the musical piece whether it was in "C major" (no change) or in "E major" (four alterations relating to the notes F, C, G, D) or anything else.
The only "alteration" considered by the program thus far is the sharp, which is indicated as normal to the left of each relevant note, always only within the bar itself (fig. 4).
As far as the difference between sharps and flats is concerned, in practice, one proceeds as follows:
This means that, although two different names are used, the same note is played (fig. 5).
To make things easier for anyone using the program, we have therefore decided to use only one of the two alterations (the sharp) at this stage.
Pauses are not indicated in the program, that is, the signs that are used to indicate a silent space that corresponds to the value of each individual note, nor other annotations such as extensions marked to the right of the note increasing its value by half, a corona, etc.
This is so because words have neither pauses, nor extensions of sound.
Each"Soundword" transforms each word into a 4/4 musical bar.
Each letter that makes up the word produces a note within the bar.
Irrespective of the number of letters it contains, each word accounts for a 4/4 bar. This means that longer words produce bars that have very short notes, whereas shorter words produce bars with fewer notes that are longer.
The possible duration for each single note is 4/4, 2/4, 1/4, and 1/8.
The program also allows for pairs of notes and quatrains, with a combined time of 1/4 and 1/8.
In addition, the program uses a sophism for excess notes (fig. 6) - that is, a number of notes that, when grouped, exceeds the number allowed by the time.
A few examples:
When the word "I" is used a bar is produced that has only one 4/4 note. The word "my" produces a bar with two notes, 2/4 each, while the word "experimentation" produces a bar with fifteen notes, the total time for all fifteen amounting to 4/4.
For words that have more than two letters, the total time (4/4) is broken down into single notes, the duration of which is not unequivocally determined.
If a word is made up of three letters (e.g. "our"), the 4/4 time can be obtained by assigning a timing of 2/4, 1/4, 1/4 , or 1/4, 2/4, 1/4, or 1/4, 1/4, 2/4 (fig. 7) to the three notes.
Fig. 7
The number of combinations does not always multiply geometrically with an increase in the length of a word.
For four letters (e.g. "from"), only one combination is possible (4 notes, 1/4 each).
For six letters (e.g. "colour") five combinations are possible, while for seven letters, eighteen combinations are possible!
"Soundword" provides an element of unpredictability (or creativity) to the translation, choosing the timing for the notes in a bar completely autonomously.
In addition to the time for the note arising from the translation of a symbol (letter), the name of the note, which may be any one of the twelve possible notes, must also be defined (fig. 8).
From this point of view, "Soundword" is parametric, and uses a translation table as the configuring parameter, which associates a note to each symbol in the alphabet used to write the text that is to be translated.
This association can be done in a number of ways. For our program we have used what we believe to be the most suitable from a wide range of possibilities.
"Soundword" can also work in the opposite manner.
It can take a piece of music with a 4/4 beat and turn it into a text with "words" written in symbols used in the alphabet chosen.
Each bar of the piece of music produces a word with as many letters as there are notes to the bar, with the letter corresponding to each note being determined by applying the translation table in reverse.
For example: the note "C" may be turned into the letter "a" if the translation table links the letter "a" to the note "C".
Naturally, the words obtained are normally not words that exist in the language whose alphabet was used for drawing up the translation table.
This is not only true for just any piece of music, but also for a piece of music obtained by running a text through "Soundword".
Contrary to what you may expect, it is not true that if a musical piece is retranslated, the original text will be obtained again, even if the same translation table is used.
The reason for this is that there are only twelve notes, while generally a language's alphabet has more than twelve letters (twenty-six in English).
Then consider that in Chinese there are 3,000 to 5,000 ideograms depending on the individual's educational level ....
Thus, if the translation table attributes a single note to each letter, the same note may be linked to a number of letters.
Thus when translating back from notes to letters, the result is not uniquely determined.
This means that, when a piece of music obtained from a text is retranslated, the original text is not obtained again, but a text in which some letters have been replaced by others.
For example when the phrase "The small lost planet ..." is translated, retranslating the music into words, one may obtain, "kyb jgoyy yejk zyohbk ...." - a completely new, original language.
A piece of music produced by "Soundword" from a sentence that makes sense may be seen as a cryptographic code, from which a "scrambled" version of the original may be obtained.
"Soundword" provides the possibility of associating a sequence of guide codes to a piece of music, to force the choice of letters, when each note is retranslated.
This sequence of codes is generated by "Soundword" at the time the text is translated into music, and it is indispensable in order to recover the same text by retranslating the music.
"Soundword's" applications are widespread, covering leisure activities, teaching and communication.
A few examples are given below.
An aspiring amateur composer may easily produce his first draft of a melody using "Soundword" and then alter it as he wishes, following the indications provided by the program.
An expert composer can extract a piece of music from the rough draught obtained, changing it by using precise harmonic rules, in order to create a song or musical score, because sometimes when the beginnings of a melody are heard, useful starting points are found for developing the work.
"Soundword" can be used as an enlightening method for introducing children to music and writing music, allowing them to discover the analogy between the structure of a sentence in natural language (words, letters) and the structure of a piece of music (bars, notes).
It can also be useful for introducing children to other types of language (mathematical expressions, etc.) while at play.
The word / sound and mathematics / sound links allow some of the barriers to be removed as well as embryonic preconceptions leading to rejection on the part of children.
In addition to this, using the computer makes one used to operating some of the most advanced instruments available.
It is also interesting to see how different languages produce music that has differing characteristics (which could increase interest in studying foreign languages).
For example, the different average length of words (comparing texts in Italian, English, French, German, Spanish and Portuguese) means that the resulting melodies have a different rhythm.
Similar considerations can be made as regards the prevalence of certain letters in the words of one language compared to another.
This could also give rise to new interesting starting points for those studying or teaching philology.
"Soundword" can be used for transmitting messages within pieces of music, as J.S. Bach did when he wrote numeric messages, such as his date of birth and other important indications regarding his work, into his music.
For those with computers, the sender and receiver need only have the same "Soundword" configuration, and along with the music a sequence of guide codes is transmitted to make reverse translation unequivocal.
The basic idea behind "Soundword" lends itself to many refinements, variations and extensions.
A short list of possibilities is given below:
As has been stated, when a piece of music created from a sentence using "Soundword" is retranslated, one does not generally obtain the same sentence, unless a code guide sequence is used.
This is a result of the lack of unequivocal re-conversion of musical notes into letters of the alphabet.
The code guide solution is not entirely satisfactory as it creates a superstructure that is extraneous to the piece of music.
One solution that may possibly prove better is to make use of a dictionary for the language in which the text is assumed to have been written.
Given a musical bar, "Soundword" will examine all the possible sequences of letters that could be generated by the bar, taking all possible reverse translations for the notes in the piece into account.
"Soundword" does not provide pauses, that is, moments of no sound in the piece of music, since the letters in a word fill the corresponding bar completely.
However, the possibility of inserting pauses of differing length may be added, based on the punctuation in the text.
The time used (4/4) was chosen due to its practicality.
Naturally, the same idea could be applied to all the other times commonly used, given the wide choice of time available for each individual note: 4/4, 2/4, 1/4, 1/8, 1/16, 1/32, 1/64 (fig. 9).
Taking the same basic idea as that used for "Soundword", a translation may also be done syllable by syllable instead of letter by letter, thereby generating a note for each syllable.
This reasoning can be extended to quantitive meters (such as the Latin meter for example).
If the text is written in a poetic style, one can take the verse as the basic element rather than individual words, thereby assigning a preset time to each verse.
Leaving a bar empty, one can also indicate a "poetic pause" where, in reading the text, one takes a deep breath so as to emphasis the phrase.
The addition of the left hand chords, that is, the emission of a number of notes simultaneously that can always be reconstructed as superimposed thirds would offer a greater, more complete sound impact.
This program is covered by copyright worldwide. All rights reserved.