Tuesday, February 4, 2014

The results of a test of an automated transcription of Psalm 1

Westminster Leningrad codex Asheri -
happy 
I can see the Westminster Leningrad Codex (WLC) here in Unicode - If you go here and type in Psalms 1, you will get this result as the first word:
1:1אַ֥שְֽׁרֵי־
The coding on this site is very clean but that first word bothers me. It has an F (merkha) at the beginning of the verse.  And where I would expect the F, I see the missing tonic. The Letteris edition has this:

אַֽשְׁ֥רֵי.


Aleppo happy
If one then looks at the Aleppo codex, it is ambiguous. Both are problematical as far as understanding where Max Letteris got his interpretation. David Mitchell says Aleppo is superior to Letteris - so I should use it to verify the WLC Unicode - but... what will it lead to? This first note is just the beginning.

Here is the music based on the WLC - and here is a translation or two. The music was written without intervention from me using an automatic transcription algorithm which I wrote this week. The algorithm translates the Hebrew text into music XML. What can you do with the Bible? You can sing it - but with what melody? And in what style? This I hope to explore in the next five years and I hope many will share and continue the exploration - once I get my routine into a cloud offering so anyone can do it. The Letteris music is what I am used to through the work of Suzanne Haik Vantoura. The exploration will take some detailed work but at least the initial coding is now reduced by several hours.
The Letteris text for the psalm is here. The music above is not the same as my earlier transcription. This is a problem to me and uncovers several musical assumptions from Suzanne Haik Vantoura that I can only wonder about at this time. Did she impose cadential figures on the text? Did Max Letteris impose them? What is going on here? Prior work on this psalm is clustered here from some time ago