I had a bug in the original code for Tango, which I fixed today. It was occasionally choosing a sample file out of range, and as a result leaving out the notes. Hundreds of them I think. The fix was simple, and here is a new version. The keys are either otonalities (maj) or utonalities (min), chosen by picking one from the list of 33 keys, and moving up and down the list using the Markov Chain Drunkard’s Walk.
This is a playlist of some new versions of the twelve dance pieces for Prent’s Microtonal Slide Bosendorfers that I’ve been working on through 2017 & the start of 2018. In honor of the RPM challenge, I’ve redone all of them in the month of February. Actually, I redid them all today, February 27th, but the spirit is there, if the mechanics are vague. The language doesn’t say you can’t spend a year composing the songs that you intend to perform during the month of February, so that’s what I’ve done the last two times: compose all year, then produce realizations using my software during the month of February. All are based on classical piano works retuned or transformed on some major level from their original. Those at the end of the list were done first, and they have the most major transformations. Many stroll through their pieces using the Drunkard’s Walk Markov Chain, going forward or backward through the measures.
I picked this one because of all the trills and ornaments. Playing the piece on Prent’s Microtonal Slide Bosendorfer, I can take advantage of the subtle whammy bar on the instrument to replace some typical Mozart trills with slides and shakes. Never to leave a note with only a single ornament if you can give it a half dozen flourishes, as Wolfgang was known to do. Mozart published this when he was 19. I was playing high school rock band gigs featuring covers of “Smoke on the Water” when I was that age.
The tuning is based on two scales derived from the 31-limit tonality diamond. The primary scale is D major based on the utonality on 5:3 (A), in which there are very nice 9:8, 4:3, and 3:2, but the major third is 9:7, 435 cents above D. I call this a supermajor, and it has a sweet character. When the piece plays A major, D major, and E major, I switch to a scale based on the utonality of 5:4 (E), in which there are nice scales in those keys. I tried a total of 14 difference scales from the tonality diamond before settling for these two.
This is a final version of the Bach Adagio, transcribed for keyboard from an oboe concerto by Allessandro Marcello, first published in 1714 or so. The transcription is BWV 974.
Today’s version includes an additional D minor scale derived from the 31-limit tonality diamond. The scales used are the following:
minor 3rd ratio
Otonality on 16:9
Utonality on 5:3
Utonality on 9:8
Otonality on 1:1
Otonality on 32:35
Each of the scales also have several other triads that are just. For example, the D4 scale as an excellent just C major triad. D2 has a perfect F major triad, and D3 has a great G minor. Some others are close. D2 has a pretty good A major, except it’s 3rd is 24:19, 404 cents instead of 5:4 at 386 cents. Horseshoes and all that.
It’s performed on Prent’s Microtonal Slide Bosendorfer. I replace the trillos and mordants with slides. Previous versions are available here, here, and here.
I’ve modified some of the scales used in each measure to minimize wolf fifths where possible with the four scales I’m using. I think I can go further, but at this stage it has an interesting edge to it. I had been choosing scales by listening to all of them in a row, but this time I used Excel spreadsheet to reduce the wolf fifths, and choose the right major and minor thirds. There are still some loopy notes, but I think I can fix that by using more than one scale in a measure. At least that’s my intent for the next iteration. You can read more about the progress of this piece here, and here.
I’m nearing completion of this version of the piece. It now contains all 41 measures. I’m still not sure about the tuning and the trills. The tuning switches between four different D minor scales, all derived from the 31-limit tonality diamond. I chose each by listening for what seemed the most consonant scale for each measure. Some of the shifts from one measure are a little strange. I may need to look for some additional scales. None of the four have a decent A minor, which is used a lot in the last few measures of the piece. The dynamics are chosen using the drunkard walk Markov chain algorithm.
I’m just getting started on this piece. There’s a terrific version by Glenn Gould in one of his collection CDs here. It says that it is based on an oboe concerto by Alessandro Marcello. I’m not familiar with this composer, but I do love the piece, and especially the way Gould sings along with his piano, probably without realizing it.
Anyway, I thought it might sound good on my microtonal slide Bosendorfers. I use three different D minor scales, based on different otonality and utonality scales. I’ll document these after I’m certain that they will work together. Two are based on a D at 182 cents, and another with D at 204 cents. The other notes in the scale are also different. I have a fourth scale that uses a 7:6 for the minor 3rd on F, but it’s rather harsh, so I may not use it. The others have the F as 6:5, which is the standard minor third.
This version is just the first 14 measures. I am using Csound to replace the mordant and trills with slides. I need to work on them some more so they don’t stand out too much.
This version is a final cut of the short eight measure theme from Orlando Gibbons in the Fitzwilliam Virginal Book I song XL, “The woods so wilde” from approximately 1620. I play the eight measures straight, then Markov Chain Drunkard’s Walk across the eight measures, forward or backward. This continues for a few dozen measures, then the theme is restated at the end. More here and here.