In the early days of Pegataur, we weren't sure if we were going to have vocals or not.  We came up with the idea that, to stay true to the esoteric conceptual basis of the band, we should come up with a way to analyze the music, on a song-by-song basis, using Qabalistic techniques.  The challenge was to come up with a way to translate musical notes into Hebrew letters, and to let our songs tell their stories by spelling Hebrew "words" and "sentences" which could then be studied using gematria to figure out what the songs were telling us.  We could then use these jumbles of words and numbers to help us write lyrics and choose titles for our songs.  This presented a series of puzzles to solve and questions to answer and decisions to make, let me assure you! 

General Approach/Background

I decided that I wanted to write some software to do this for me, a decision that I would not exactly come to regret, but would make me realize not too far into the process that it was going to be a pretty serious undertaking (I think this is a common trap I find myself in as a developer, by the way -- "that should be easy" -- famous last words!).  I made the decision to start writing this software in C#.NET as a Windows Form (i.e., a standard desktop GUI-based executable) application in maybe April 2007; I had a working prototype to analyze our songs with by the end of March 2008.  The heat was definitely on, as we had finished recording our album and were having it mastered.  I couldn't really drag my heels on this anymore, as we needed song titles for the manufacture of our album and an easier way to refer to the songs for our set list.  By the way, it works, and I used it to generate the song titles for our upcoming release, Eternal Flight, as well as for the creation of the sigils featured on the album art that correspond to each song.

Problems/Questions

Before I could start writing the code, I had to answer a lot of questions and figure some stuff out that I hadn't really been forced to think about before in any great detail:

• How do you map musical notes to Hebrew letters?  Should I use an absolute system (e.g., "the note C always equals the letter Aleph") or devise my own relativistic system (e.g., "the note C will have a different Hebrew letter correspondence depending on context")?
• If a musical note is the base "atomic" or lowest level of musical "input" to be used to generate song data, how is that data going to be structured in any meaningful way?  How to break-up the "words" generated by combination of notes into phrases, sentences, paragraphs?  Where do you draw the line!? 
• How are you going to deal with the fact that in music, a single note might repeat itself several times, or you might lean towards a tonal center or key?  Won't you wind up with nonsense and a strong bias towards certain things based on frequency?  Is this good or bad?  What if you default to writing songs in the same key all the time, more or less?  Will all your songs "mean" the same thing?
• How do you model music theory in a programming language? 
• How do you model or represent instruments in a programming language?
• How do you model the Hebrew language in a programming language?
• How do you model the Qabala in a programming language?
• Where are you going to get the data for your gematria?  Are there any databases out there that you can use already for this, or do you have to create your own from scratch? 
• How are you going to capture music/song data?  What kind of interface?
• How are you going to persist the data? 

I will be answering these questions and taking you on my own personal descent into madness in subsequent blog posts, but let me just give you the teaser: I got it to work, and I did use my application to analyze the music on our album.  The song titles were based on what the songs are really trying to tell us!

Solutions/Outcome

Here are some of the key features and aspects of the software used as a part of the Pegataur Working:

Data Entry

Musical "data" is captured at the level of the "measure" using a virtual guitar fretboard.  You just click on the fretboard and notes are entered into the measure.  Each note gets a Hebrew letter assigned to it based on the key of the measure.  "Words" are built up by consecutive notes, and spaces break up the words into phrases as you create chords.  This form allows you to tune the virtual guitar strings, or choose from one of two presets.  The Hebrew is analyzed using gematria, temura, and notarikon algorithms.  It also allows you to export the tab that is created during this process.

Song Project Data

A Pegataur "song" is comprised of "parts," which in turn are comprised of "measures," which are in turn comprised of "notes."  The main interface allows you to create and manage a song.  As you create parts, you can add measures to these parts.  Double-clicking on the measure opens the measure form (as shown in the preceding section).  As more songs and measures are added, the song data is rolled up in a structured manner and analysis using gematria, temura, and notarikon is once more done at the level of the parts, and at the level of the song as a whole.  Once you have finished your song's data entry, you can proceed to the song analysis tab (next section).

Song Analysis

The heart of the software is quite possible the analysis engine.  Once you have song data to analyze, it goes through each song's measure, part, and the song as a whole, and builds a list of all the generated "words."  It then checks to see if these are "real" words in Hebrew, and sets these aside.  Then it goes through the different cypher algorithms for temura and notarikon, and sees if those are real words.  Then it does random permutations of the words and checks to see if these anagrams are real words.  Finally, it takes the calculated gematria values of all of the words, and finds all real Hebrew words that also correspond to that number. 

At the end of this, you have a huge list of real Hebrew words and phrases that can be used to generate lyrics or song titles - there's even a random lyric/title generator functionality to aid in this process (like pulling words out of hat!).  In addition, each Hebrew letter has its own set of attributes (i.e., astrological, elemental, planetary, and tarot correspondence), so it does a frequency analysis and determines the dominant attribute for the song as a whole.  For some esoteric icing on the cake, it also generates a list of "spirit names" that allow you to choose a ruling spirit for your song, and potentially create a sigil for that spirit -- this is how I arrived at the sigils used to represent the songs in the album art.

Other Tools: Sigil Generator and Hebrew Keyboard with Hebrew-English Dictionary and Qabala Analyzer

As if that wasn't enough, I also created a couple of stand-alone utilities within the application.  Namely, a Sigil Generator and a Hebrew Keyboard with Hebrew-English Dictionary and Qabala Analyzer. 

The Sigil Generator (above) allows you to pass a Hebrew word into it and get a sigil representation based on one of the classical 7 magick squares.   

The Hebrew Keyboard with Hebrew-English Dictionary and Qabala Analyzer (above) allows you to peck out Hebrew words using a virtual keyboard, or you can paste Hebrew directly into it OR transliterate English into Hebrew, and then get an automatic gematria, temura, and notarikon analysis.  From here, you can also look up to the dictionary by words or gematria values.