Blog

The People – Variation 12 #67

This is a final version of variation #12 with a few modifications. I changed some parts to ensure that when playing any one tetrad, the chances of another tetrad slipping through are greatly reduced. Previously the bass part was always in the “a” tetrad even when the upper voices were in the “b”, “c”, or “d” tetrad. That’s no longer the case, as the whole band plays in the same tetrad at the same time. I also modified the volume settings to boost the audio on the extremely quiet parts. The same basic structure is used in this version, with a few more tetrad combinations. Here is the part of the source code where the macros are called to make the music:

@ play variation 12
@ 10765
.8th 28*12
.abcd ba
.flat aa
p35&loudness-m*.&v12-start-*.
&loudness-*.p22&v12-vamp-Dnu-3*. p35&v12-vamp-Bbo-3*. &v12-chor-1-*.
&loudness-*.&v12-vamp-Dnu-3*. &v12-vamp-Bbo-3*. &v12-chor-2-*.
.8th 24*14
.abcd ab
.flat bb
&loudness-*.p22&v12-vamp-Dnu-3*. p35&v12-vamp-Bbo-3*. &v12-chor-1-*.
&loudness-*.&v12-vamp-Dnu-336*4. &v12-vamp-Bbo-336*4. &v12-chor-1-*.
.8th 21*16
.abcd ca
.flat aa
&loudness-*.p22&v12-vamp-Dnu-3*. p35&v12-vamp-Bbo-3*. &v12-chor-1-*.
&loudness-*.&v12-vamp-Dnu-3*. &v12-vamp-Bbo-3*. &v12-chor-2-*.
.8th 24*14
&loudness-*.p22&v12-vamp-Dnu-3*. p35&v12-vamp-Bbo-3*. &v12-chor-1-*.
&loudness-*.&v12-vamp-Dnu-3*. &v12-vamp-Bbo-3*. &v12-chor-2-*.
.8th 28*12
.abcd bc
.flat cc
&loudness-*.p22&v12-vamp-Dnu-336*4. p35&v12-vamp-Bbo-336*4. &v12-chor-1-*.
.8th 24*14
.abcd cd
.flat dd
&loudness-*.p22&v12-vamp-Dnu-336*4. p35&v12-vamp-Bbo-336*4. &v12-chor-1-*.
.8th 21*16
.abcd da
.flat aa
&loudness-*.p22&v12-vamp-Dnu-336*4. p35&v12-vamp-Bbo-336*4. &v12-chor-1-*.
.8th 28*12
.abcd ba
.flat aa
&loudness-*.p22&v12-vamp-Dnu-3*. p35&v12-vamp-Bbo-3*. &v12-chor-1-*.
@ maybe a crescendo at this point? redefine vol4?
.all-ran2 r16
.all-ran3 r16
.all-ran4 r16
.all-ran5 r16
.all-ran6 r16
&loudness-fff.&v12-end-*.


or download here:
People var #12 – 67

The People – Variation 12 #50


This is a new variation that uses the 16 note otonal and utonal scales of the 31-limit tonality diamond differently that the previous ones. In the first 11 variations, I choose pairs of triads to make a hexad for each key. For example, if I was playing in B♭ major, I’d select the otonality on 16:9, and pick one triad based on 4:5:6 from the 16:9, and another based on 7:9:11. The same sort of thing was done for an utonal scale, in this case based on 9:8.

For this variation, I’ve been exploring four note chords. There are four in a 16 note otonal or utonal scale based on the 31-limit tonality diamond. For example, in this variation is still use the 16:9 otonal 16 note scale, but instead of contrasting two triads, I chose to contrast four tetrads (4-note chords). The most consonant is one based on 4:5:6:7/9, shown in tan shading. I refer to that as the “a” tetrad in the chart below. The “b” tetrad, in pink, is based on 9:11:13:15/18, much more challenging to the ear accustomed to 12 tone equal. Keep going up the harmonic series, is the “c” tetrad in blue uses the 17:21:25:29/32, and the “d” tetrad in purple is based on 19:23:27:31/32. These intervals are more challenging.

Here is the 16:9 otonal 16-note scale with the different ratios for each of the four tetrads based on the 16:9 otonality, abc&d.

And the utonal scale based on the 9:8 utonality.

Variation #12 starts out with the “a” tetrad, plus glides and trills starting on the “b” tetrad and moving to the “a”. The next section has “a” tetrads moving to the “b” tetrad. There are more combinations of “c” and “d” that follow, with a return the the b to a tetrads towards the end.

There are sections in between that quickly move through the chords in measures 34 and 36, then back to iterations of the 16:9 otonality and 9:8 utonality.

The piece ends with a legato rendition of measures 33 – 36 as written by Rzewski.

Most of the selections of the lists of lists uses the Markov Chain Drunkard’s Walk algorithm, but I sometimes mix things up by choosing a repetition-heavy randomization, that attempts to repeat the previous choice fairly hard, but not absolutely. It basically picks a random number, and if it’s not the same as the previous one it picked for a particular list, it tries another for a fixed number of iterations. The result is repetition of many but not all selections for a few beats.


or download here:
People var #12 – 50

The People – variation 11 – #45

This is new version of variation #11. For this one, I revised the hexad for A major and E major. These are derived from overtone series of the note a 4:5 ratio below the root of the chord. That is, 1:1 otonality provides a major chord that starts on the 5:4 and includes a standard just major third triad plus another triad that is not. The 4:3 otonality provides a major third chord on the 5:3. The result is a less jarring sound than the previous scales. The tuning uses 9:8 utonality for G minor and D minor, 4:3 otonality for A major, and 1:1 otonality for E major.

I generate the source code for Csound using a text preprocessor I wrote. The preprocessor takes care of sorting out which sample files to use and other tasks. It includes a macro processor component that stores and retrieves text elements that end up as lines of Csound source code. A macro is defined by a period followed by text. What follows is a fragment that defines three macros of varying duration, kind of a vamp in the key of G minor that might last 1680 beats or only 720 beats. An eighth note is either 30, 24 or 20 beats long:

.v11-gmin-vamp-1680 p35&all-480-v11-n66hgd. &all-480-v11-n66hgd. &all-480-v11-n66hgd. &all-240-v11-n66gd. &pick-v11-loud*.&vel.d1680
.v11-gmin-vamp-1200 p35&all-480-v11-n66hgd. &all-480-v11-n66hgd. &all-240-v11-n66gd. &pick-v11-loud*.&vel.d1200
.v11-gmin-vamp-0720 p28&all-480-v11-n66hgd. &all-240-v11-n66gd. &pick-v11-loud*.&vel.d0720
@
.v11-chorus-0960 p35&all-240-v11-n66gd. &all-240-v11-n66EC. &all-240-v11-n66dd. &all-240-v11-n66AF. &pick-v11-loud*.&vel.d0960
.v11-chorus-1920 p35&all-480-v11-n66-gd. &all-480-v11-n66-EC. &all-480-v11-n66-dd. &all-480-v11-n66-AF. &pick-v11-loud*.&vel.d1920

When I want to call the v11-gmin-vamp macro, I append an ampersand to the front of the macro name. Here I make a new macro called v11-middle made up of other macros:

.v11-middle &v11-gmin-vamp-*. &v11-chorus-*. &v11-Emaj-vamp-*. &v11-chorus-*. &v11-dmin-vamp-*. &v11-chorus-*. &v11-Amaj-vamp-*. &v11-chorus-*.

The name of the macro is important, because in the v11-middle macro, I can call any of the three macros that match the first few characters of the macro name, up to the asterisk.

The structure of the piece goes like this:

@ 4*6 * 6 = 120 30 beat eighth note - this is a comment. the @ sign means comment.
@ 5*4 * 6 = 120 24 beat eighth note
@ 6*4 * 5 = 120 20 beat eighth note
@ variation #11 combined
&v11-start.
.8th 4*6
&v11-middle.
.8th 5*4
&v11-middle.
.8th 6*4
&v11-middle.
.8th 5*4
&v11-middle.
.8th 4*6
&v11-middle.&loudness-pp.
&v11-ending.

The v-11-middle parts are about minute or two each, depending on what macros the randomizer has chosen to select.

I’ve been playing around with the randomness selections. This version allows some of the middle parts to deliberately play the same parts over and over again, instead of following the drunkard’s walk Markov chain. Repeats as variation, so to speak. I also made each voice have it’s own density shifts from always playing a note, to almost never playing, which allows the voices to shift focus more.

The .8th 4*6 variable changes the duration of the quarter note. 4*6 uses fits 4 eighth notes to a measure. The 5*4 fits 5, and the 6*4 fits 6. The duration of the measure stays the same, but it speeds up to fit more notes in the measure, then slows down to fit fewer.


or download here:
People var #11 – 45

The People Variation #11 – 42

This is a new variation #11 derived from measures 29-32. Here is the original version of these measures as written by Rzewski. This variation has several different tempi like variation #10. The tuning uses 9:8 utonality for G minor and D minor, 4:3 otonality for A major, and 1:1 otonality for E major. Note that only D minor uses the classic utonal scale, and the Major scales are chosen from a different otonal source scale that does not have the traditional overtone series as scale degrees. I’ll write more as I work on the variation later.


or download here:
People var #11 – 42

The People – Variation #10 – take 33

This is another version of my variation #10 of People. It is based on the harmonies of measures 25-28, which I play on scales derived from several different otonality and utonality scales. I’ve been adjusting the relative volumes of the different piano lines to get a better balance. There is a great deal of indeterminacy in the making of this variation, with lists of lists of lists, selected using the Markov Chain Drunkard’s Walk.

The specific 16 note otonality or utonality used are listed below the keys in the table and score fragments below. The source of the tones is the tonality diamond to the 31 limit.

Here is the original version of these measures as written by Rzewski.


or download here:
People var #10 – 33

The People – Variation #9 take 28

This is variation #9 of People, composed by me. It is based on the harmonies of measures 21-24, which I play on scales derived from several different otonality and utonality scales. The following table shows the keys that are in the material for measures 21-24, followed by the scales from the 31-limit tonality diamond that fit the tuning best, followed by the six most important tones in those 16 note scales. For example, the first line says the key is E minor, derived from the 5:4 utonality 16 note scale. From those 16 notes, the most important are E, G, & B, and they are contrasted with another triad higher up the overtone (undertone in this case) series. I spent a lot of time with Scala and a keyboard to determine the most harmonious six notes to use, listening to the best, and discarding the rest.

variation #9 measures 21-24
key scales notes
e min E uton E G B D Gb A
D maj Bb oton D F+ A Db G+ B
d min E uton D F A E G B
C maj C oton C E G A# D F+
Bb maj Bb oton Bb D F G+ C Eb
A maj F oton A C# E C G D
g min C oton G Bb D F A C
F maj Bb oton F A C Eb G Bb
A min E uton A C E


Here are the four Scala tunings that I played with while composing.

  1. F oton 4:3 otonality
  2. C oton 1:1 otonality
  3. Bb oton 16:9 otonality
  4. E uton 5:4 utonality

And here is the original measures 21 – 24 as written by Rzewski.


or download here:
People var #9 – 28

The People – variation #8 take 24

This is variation #8 of People. It is based on the harmonies of measures 17-20, which I play on scales derived from the 4:3 otonality and the 9:8 utonality. From those 31-limit 16 note scales I derive A major and G minor from 4:3 otonality, and D minor and Bb major from the 9:8 utonality. There are lots of slides and trills that explore other notes in the scales. I should mention that these variations involve a lot of indeterminacy. There are a few dozen alternatives at any given time, and I allow the preprocessor program to choose which one to play. Each choice is collected in a list, and the algorithm picks either the next or previous choice from the list. There are lists of lists of lists several layers deep. When I post a variation, the variation is taken from two-four measures of the theme. I run it through the processor three times and pick the one that is most interesting, and discard the rest.


or download here:
People var #8 – 24

The People – variation 7 take 20

This is variation #7 of People. It is based on the harmonies of measures 15 & 16, which I play on scales derived from the 5:4 utonality, the 4:3 otonality, and the 9:8 utonality. From those 31-limit 16 note scales I derive E minor from the 5:4 utonality, A major from 4:3 otonality, and D minor and Bb major from the 9:8 utonality. There are lots of slides and trills that explore other notes in the scales.


or download here:
People var #7 – 20

#39 Cuernavaca 1975 CN

This is a story about a piece I created in 2001 while listening to the music of Colon Nancarrow. The story is fake, but the music is “fake but accurate”.

I recently found a large music box at an antique store in the Fremont neighborhood in Seattle. Fremont is a district that welcomes visitors with the sign, “Welcome to Fremont, Center of the Universe. Throw your watch away”. The only two industries are beer and antique stores. The music box is about the size of an old washing machine, made of wood. The only markings on the box are a small label on the bottom: “F. Nicole 014751”. It has a crank on the right side, which connects to a spring-loaded drive for the cylinder. Inside is a long row of about 100 metal teeth, from about 1/4 inch at the top to almost 8 inches long at the bottom. The big ones make the whole case vibrate when they play. Inside is a large metal cylinder, with tiny pins that pluck the metal teeth. It appears to be removable, but there weren’t any more cylinders at the store. It runs for about 6 1/2 minutes on a full crank, and has a delightful gradual deceleration as the spring runs down. I’m not sure where to start the song, as it doesn’t appear to have a beginning or an end mark. For this recording I start and end at around the same place. The tuning of the keys is what attracted me to it in the first place. I have no way of knowing if it has drifted from its original tuning, but it is a charming sound, nonetheless. It appears to shift from a utonality-type sad minor key into a harsh super-major, and sounds a bit like some of the changes in a Philip Glass composition. (I know because I have been listening to a lot of Glass lately.) There is a cryptic notation on the drum: “#38 Cuernavaca 1975 CN”. This must be the title of the piece I guess, but it doesn’t sound very Mexican to me. Perhaps there are some more cylinders somewhere that I could try out. Sounds kind of like circus calliope music, for a Javanese circus troupe. Boogie-woogie on Saturn. If anyone has any idea about the music box, let me know.

The source code for the music box is below. How can you have source code for a physical instrument? The actual samples are taken from my finger piano, an amplified kalimba-like instrument I made in 1978. The samples were cleaned up with CoolEdit and programmed using Csound to simulate a music box. Then convolved with Teatro Alcorcon to simulate reverb. Tricky.

Not the instrument, but what it might be like, if it existed.

Listen here:

or download here:
#39 Cuernavaca 1975 CN

Source: #39 Cuernavaca 1975 NC