Tuesday, September 27, 2016

This just popped up

As ever, this blog tends to feed itself once it's active. In my previous post, I wrote about happening on a Bach fugue and converting the notation from the ungainly key of C-sharp Major to the slightly more gainly key of D-flat Major. I mentioned discussing this fugue with pianist/blogger Erica Sipes, who's spent a lot of time with Book II of The Well-Tempered Clavier. In a Facebook conversation, she wrote: "The fugue reminds me of popcorn popping...starting with a kernel or two as the oil heats up and then speeding up as they all start popping. Profound, eh?"

Well, to me this actually seemed perfectly profound and on point (especially because of the way 32nd notes are gradually introduced to quicken the texture), so I started thinking about how to play with the concept. Since my piano tuner has advised against pouring hot oil into the piano, I decided to head back to the world of Scratch, MIT's marvelous graphical programming environment for kids (and kids at heart). I've actually learned enough "grown-up" programming since I last posted Scratch projects (see: here, here, and here) that I should really start learning how to do animations in JavaScript; but I wanted to see Bach popping right away, so I started scratching.

Scratch does have some limited MIDI capabilities (basically, you can tell it to play notes at a given pitch, for a given length, by a given "instrument"), but getting it to play a three-part fugue in which the beats stay synchronized, even as the tempo changes, meant I got to experience what it's like to build a MIDI system almost from the ground up. As I'll discuss more in a future post, this kind of problem-solving provides an interesting vantage point from which to think about how music works. Mundane functions like pausing, restarting, and jumping around within the music all required procedures that had to be thought through and executed. (In my programming experience, the latter is always much harder than the former!)

As for the popcorn animation, my first instinct turned out to be the best - nothing but a simple skillet in which each note "played" causes a kernel to jump up and then get randomly popped around the pan. I got to revisit my long dormant trigonometry skills as a way of ensuring that the randomly distributed kernels end up falling in a circular pattern. (Simply sending kernels to random values of X and Y resulted in a square distribution that looked silly with the round skillet.)



[The only nod to the actual musical structure is at 1:16 when the alto voice states the subject in augmentation (longer note values); if you look closely, you'll see especially emphatic pops for those four notes.]

Then, like kernels popping randomly in all directions, I found the project also started sprouting outwardly in unexpected ways. I spent a good bit of time trying to design a pop sequence that somehow shows the melodic shape of the three parts. This took a lot of work - and, frustratingly, resulted in something less elegant and realistic-looking than the simple random approach. The kernels that (for reasons unknown to reality) shoot popping kernels into the bowl look a bit too much like little bugs; perhaps a "bug" in the program. But I'm pleased with having met the challenge, although this version exists as an unfinished side curiosity for now. (Those are bass kernels in lower left, alto in upper left, and soprano in upper right.)


Other features crept in, including the fairly rudimentary functionality of being able to change instruments, tempo, volume balance, and key on the fly. Changing key (which, in a way, is how I got myself "into this fugue" in the first place) gave me the idea of allowing the user to change the key for each voice part individually. This led to the idea of a feature in which the music changes key randomly every beat, which creates a kind of 12-tone fantasy effect. (We all have different fantasies.) Then, it occurred to me that the integers I was inputting for pitches don't have to be integers, and suddenly a microtonal option was on the menu. (Unfortunately, "viola" is not one of the instrument options in Scratch's soundfont.) For the user who finds the popping kernels are inducing hunger, this microtonal option might help to erase any appetite.

Finally, because I already have the music entered into Lilypond, and because Lilypond is wonderfully flexible about output options, I converted the score into 35 one-bar systems which allow the user to view the score two bars at a time in the small window. It's not the most practical way to follow along, but it's nice for reference, and one can easily switch back and forth between staves and stovetop.

After I'd made it this far in writing this very post, I realized I needed to add a feature which inverts the music. Rather than simply invert each part within its own range, the parts are completely flipped so that the bass becomes soprano and the soprano becomes bass. Although it wouldn't have been so difficult to produce a nice elegant engraving of this new "score," I decided it's more fun to flip the existing measures upside down like so*:

Notice that the C-sharp which begins the subject in the bass is still a C-sharp two octaves up, and it even looks like a C-sharp in the inverted bass clef staff!

Here's a quick demo of most of these functions**:


That gets a little insane at the end as I flew to close to the sun in choosing a tempo that would get me to the end more quickly. All the more reason for you to try these fun features out on your own. Just go here or click the image below, which surrounds the embedded "game" within all the instructions you'll need.


Unfortunately, Scratch uses Flash technology which is becoming more and more outdated, so this program won't work on most mobile devices, and might not work on some computers, depending on what your browser thinks of Flash. This is why I'm not going to keep trying to iron out all the leftover kinks. (Sometimes, for reasons I don't feel like fixing, the kernels will suddenly grown enormous and overtake the entire screen - which is actually kind of a nice bug.) Time to move on and learn how to do this more elegantly and efficiently.

I have much more to say about "playing with" Bach in this way, and will post more soon. For now, I'll just say that the line between "playing Bach" (as in, performing it "respectfully" on a keyboard instrument) and "playing with Bach" perhaps shouldn't be so clear. Although there's no clear aesthetic purpose in inverting a fugue with microtonal distortions, I like to think of this program as a way of holding the music in virtual hands and turning it around to inspect like one would a snow globe.

In my last post, I quoted the great Charles Rosen about the satisfying way in which Bach's counterpoint fits the fingers. In the larger context, he's discussing the fact that this music was never intended to be performed publicly, but rather for more private encounters in which following the score (by performer and possibly by a few other listeners) is part of the experience of the music. Rosen writes:
Playing Bach for oneself or for a friend or pupil looking at the score...raises few problems; nothing had to be brought out, the harpsichordist experienced the different voices through the movement of the hands, the listener saw the score and followed all the contrapuntal complexity disentangling the sound visually while listening. Bach's art did not depend on hearing the different voices and separating them in the mind, but on appreciating the way what was separate on paper blended into a wonderful whole. [pp. 199-200]
Rosen goes on to suggest that modern performance in concert necessitates having the performer "bring out" the contrapuntal details more pointedly. I'm not sure I agree with that, especially because we know Bach loved secrets and hidden meanings, so he might enjoy the idea that various kinds of thematic connections are buried under a euphonious surface; but for now, I'm more interested in the idea that encountering this music can happen in lots of ways. And, yes, I'd agree that the best such encounters happen at the keyboard. Here again is my humble effort at playing this prelude and fugue shortly after learning them in my D-flat "version." (Fugue begins at 1:45).

 
 
Thanks again to Erica for her Bach/popcorn brilliant idea. Check out her brave and insightful online practice sessons and her lovely book on practicing. Come to think of it, practicing is something I should probably be doing! (I don't think it counts for me but, even as I type on a Tuesday night, I'm listening to my daughter zip through Bach's A Minor violin fugue, so I'm having a live fugue experience. Bach is the best!)



* Flipping the music upside-down distorts a lot of the tonal function, though the vague major-to-minor effect is cool.

** One feature I decided not to include, though I toyed around with it, is to have the notes played backwards. It's very easy to reverse the note/rhythm sets and that sounded cool in places, but running notation backwards also raises the question of whether notes should begin where they ended going forward (so that a whole note would begin right at the beginning of a backwards bar) or where they started going forward (so that a whole note wouldn't be played until the very, very end of a backwards bar), which aligns note attacks more often with the other voices but produces odd collisions going the other direction. If that didn't make sense, just think of how a record sounds played backwards. Assuming you're not trying to recreate that whooshing effect caused by reversing a note's sustain, then you're basically going to end up with significantly different rhythmic relationships among the parts.

Friday, September 9, 2016

The Joy of Engraving

There was a time in my musical life when I wanted nothing more than to know every corner of the repertoire in systematic fashion. I loved collecting complete sets of symphonies, sonatas, and string quartets on records, looking at them lining my shelf, listening while reading the score or liner notes, and imagining that some day I'd know each movement of everything that mattered. Perhaps not coincidentally, that was a time when finding something new took effort and diligence, especially if you lived in a small town in Arkansas with no classical record stores within a couple of hours.

Now that every corner of the repertoire is just a simple search away, I've come to enjoy much more the serendipitous encounter. So, for example, it would shock the younger me to know that I never did get to know all 48 preludes and fugues of Bach's Well-Tempered Clavier as well as I know Seinfeld episodes. (Incidentally, I also prefer just happening upon a random Seinfeld to choosing one on DVD). I once learned the first 12 preludes and fugues of Book I and performed them in an informal recital, but for some reason I've never wandered regularly into Book II - which means it can still surprise me!

The time when I'm most likely to open those pages is when I'm looking for prelude/postlude possibilities for Sunday morning. If no obvious chorale prelude or other hymn-based option is available, I'll often consider the more generic approach of finding a prelude in the key of the opening hymn or a postlude in the key of the closer. By definition, the two books of the Well-Tempered Clavier provide four options in each major and minor key. Still, most hymns tend to have no more than 2-3 sharps or flats, so that leaves a lot of well-tempered ground untrampled.

In planning for September 4, I realized our closing hymn was to be in the fairly unusual key of D-flat Major:


Technically, Bach doesn't use D-flat Major in his "cover all the keys" books, but that's because he chooses the enharmonic equivalent C-sharp Major. Basically, it's a different way of notating the exact same set of notes (at least for a keyboard instrument), though for many, the more commonly used D-flat, with its five flats, is easier to read than the seven sharps of C-sharp. Both key signatures would've been highly unusual in Bach's day (one of the reasons he created the books), and he may have enjoyed the intellectual game of pushing every note of C Major to the right. Sharp keys tend also to be associated more with music that's on the bright side. 

The C-sharp Major set from Book I is certainly very bright. The virtuosic fugue, one of Bach's giddiest creations, is one of my favorites, but I'll admit that when I learned it years ago, I was pleased to find that my Peters edition had both prelude and fugue printed in D-flat versions at the end of the book. However, neither of those pieces seemed like the postlude I was looking for (especially since getting the fugue back up to speed would be nontrivial), but I liked the fact that the more innocent-looking C-sharp Major fugue from Book II begins with the same two pitches (enharmonically) as the hymn tune above. I was away from a keyboard but chose it anyway, without playing a note or even listening to it. Bach is trustworthy that way.
As that Sunday morning approached, I was again away from home and piano, but I looked at the fugue I'd chosen a week before and grimaced at all the double-sharps and cancelled double sharps (which, by strange convention, are often printed with a natural AND a sharp next to the note to show that it's gone back to "normal"). Rather than say to myself, "I'd better start practicing," I thought, "I wonder if I can get this in D-flat Major." Well, neither of my editions had it this way, and none of the editions I could find online had flattened it, though Awadagin Pratt's recording curiously IDs it as D-flat.

With the wrong keyboard in front of me, I decided maybe I should just make my own D-flat edition on this humble Macbook. My ancient copy of Finale is only available to me on my Windows desktop, but I prefer the beautiful output of open-source Lilypond anyway, so I set to work entering the notes by name and number. 

I haven't written often about Lilypond here before, but though it does amazing things, the learning curve is steep. However, because I've been doing more and more programming anyway, the text-based interface has a certain appeal and I've slowly learned to love this way of interacting with musical ideas. I would go so far as to say that engraving music in this fairly arcane manner becomes another mode for encountering the music* itself (as opposed to the modes of listening or performing). The Lilypond experience is a bit paradoxical: the process is clunky and non-visual, but its algorithms do such a lovely job of spacing the elements of notation that the output can seem like magic. 

Actually, because I wasn't using a MIDI keyboard to enter pitches (generally a much faster approach), I chose to "cheat" by entering the music first in C Major. Although I'm just now learning that there are ways to enter pitches by scale degree, the default Lilypond input method means that a C is entered by "c" but a C-sharp is entered by "cis," a B-flat by "bes," etc, so it was much easier to "type in" the C Major version (which mostly just means ignoring the key signature, changing double-sharps to sharps, and naturals to flats). At this point, I realized that I could totally cheat and just play in C Major while turning the transpose button on our Rodgers digital organ up a step, but I wanted the tactile experience of playing as Bach might have felt it.** Plus, flat keys often sit more comfortably under the fingers. (Any pianist will tell you that D-flat Major scales are much more comfortable to play than C Major.)

Although the fugue hadn't made much of an impression on me when I listened to a few recordings, I found it thoroughly delightful to play and survived the Sunday morning postlude without too many slips. Meanwhile, the piece had slipped into my system, and I'd become intrigued by the possibility of making the Lilypond engraving look as good as possible, so I kept hacking away at it, while also starting work on the prelude.

The basic idea behind Lilypond is that it attempts to follow the principles of spacing that old-school music engravers would employ by hand. Its algorithms aren't perfect, so I've had to do some nudging here and there to get to where I am now, but the intricacies of laying out Bach's counterpoint on two staves create just the kind of challenges Lilypond is well-suited to solve. To be honest, I've never cared all that much about the editorial considerations among various editions, but I do care about elegantly laid out music which doesn't waste space. For my taste, most computer-generated modern editions use space very poorly, so one ends up with too many page turns and notes that don't really seem comfortable hanging out together. It takes time and craftsmanship (virtual or otherwise) to put notes, beams, accidentals, slurs, ties, etc. close together in ways that enhance readability.

My intention is to make my edition available on the IMSLP Petrucci Music Library (where I'd searched in vain for just such a thing). I'm pleased that the prelude and fugue each only occupy two pages, though I'll probably add larger font, three-page versions of each for those who love space to write in fingerings and phone numbers. I'll also likely upload the scanned version of the set from my old copy of Book I since that should also be easy for people to access.

I mentioned that engraving music is satisfying as a kind of artistic pursuit, but I do still enjoy playing, so it's been equally satisfying to learn these pieces while I worked, especially because playing through them is a good way to catch and fix mistakes. (I haven't yet figured out how to fix all the mistakes I make while playing!) I decided to try my hands at recording them so I could make a YouTube video of the score, and here you go:



Friend and fellow pianist blogger Erica Sipes has written and played eloquently about these pieces as part of her ongoing exploration of the entire Book II. I sent her my D-flat notation (not the recording) and she, having dutifully and appropriately learned the music in C-sharp, replied: "It's crazy how different that 'sounds' to me!!" I wrote back that, for the prelude, "I like the 'mellow' feel of D-flat," and she responded poetically, "I hear it as bright but mellow, if that makes any sense. It's like the sun behind the clouds."

This raises the fascinating question of how our musical perceptions are affected by key associations, and more generally by what we see on the page. One of the interesting truths about computer programming is that there are countless ways to write code that will produce the exact same output, but that's because the interpreter is a machine. Musical notation is a kind of programming, but we are both less reliable and, hopefully, more interesting as interpreters.*** (Another interesting difference between these types of script is that the slightest missing comma can disable an otherwise perfectly good computer program, whereas a human performer can overcome all sorts of musical notation errors.)

Beautifully engraved music can also help signal to the performer something about how a set of notes cohere. One reason I like more music per page is because I can more easily see larger shapes and structures, but I'm also convinced that the pure aesthetics of a good layout can have a positive effect. The visual becomes a part of the music for some performers. When I was searching online for information about the old Peters edition I own (with the D-flat Major appendix), I came across a message thread on pianoworld.com in which a poster asked about finding a copy of this out-of-print score (edited by Franz Kroll). Many of the responders wondered aloud why someone would want less-than-up-to-date Bach when so many more "reputable" versions are readily available. Here was the touching response from "Marie1":
All of your suggestions sound good, but what I need is the edition that comes closest to the Kroll. I am quite elderly, have been playing from the Kroll edition for years and find it difficult to contemplate a change....
I don't even know for sure if Marie1 is using the same edition I've used and loved, but I do know the feeling. I've annoyed many violinists by preferring my old International Edition Beethoven Sonatas to the Henle Urtexts they worship, but I might as well be as comfortable as possible while wrestling with Beethoven's demands.

As for this specific prelude and fugue, I'm amazed as always at what Bach can accomplish in two pages. The prelude, which at first glance/listen seemed uninteresting to me, is incredibly satisfying to play, in part because it invites so many different approaches. I like to think I'll never play it the same way twice, though the performance above is kind of on the neutral side since I was probably trying too hard not to mess up. It resembles on the surface the famous "Ave Maria" Prelude in C from Book I, but whereas that harmonic progression/pattern is so familiar that it simply cannot surprise, I find there are dozens of unexpected twists and turns in the C-sharp/D-flat prelude. Each time I play, the "sun" Erica mentions wants to peek out from behind different clouds. Honestly, it feels less fully realized than the famous C Major, and that appeals to me more than it once would have.

The fugue is just as satisfying under the hands. Although the simple subject is treated quite playfully, with entrances of various voices piling closely upon each other, I find the satisfaction in playing Bach is in how everything works together, both musically and physically. Obviously he knew how to write counterpoint, but just as importantly, he knew how to write counterpoint for two hands in a way that is not always about independence of voices but is always about a gratifying interplay among the fingers. Charles Rosen writes:
Bach's art did not depend on hearing the different voices and separating them in the mind, but on appreciating the way what was separate on paper blended into a wonderful whole. (p. 200)
And it's just a guess, but I suspect Bach also loved the act of putting notes down on paper.

Here's a link to my "D-flat" versions in their current state. Please let me know if you find errors! Note that the PDF includes both the two- and three-page versions of prelude and fugue.


 
* One of the fads of current music education is to teach children to refer to sheet music as "notation" and not "the music." While it's true that squiggles on paper don't produce sound, they are music to me! 

** Actually, it seems the original version of the Book II C-sharp Prelude was in C Major (see p.2 here), so Bach himself might've just ported C Major into C-sharp. Who knows, maybe he never even played it in C-sharp?!

*** Right after playing in church, before I'd fixed some notational errors, I tweeted and uploaded a video with Garage Band guitars performing. I actually like this version a lot, but we can safely say it would sound just the same if I'd sent the pitches over in C-sharp.

Thursday, September 8, 2016

Sunday Jukebox

In mid-August, my sleepy blog suddenly sprouted a three-part series focused on combining tunes in unlikely ways. The last of those posts explored the intersection of two hymn tunes - which apparently anticipated where I was headed this month as this week is also going to feature a series of three posts, all inspired by my experiences as a church musician. We began on Sunday with some adventures in improvisation. Today, I'm doing a bit of show'n'tell with one of my major summer projects. Tomorrow, we'll explore a new way I've found to avoid practicing!

While I've been figuring out what to do with my life these past couple of years, I've been learning more and more about JavaScript and programming in general. I debuted some magical musical listening guides a year ago today, also coming off a summer of digging into JavaScript, and then the programming lights went off for much of the school year. Until...as with so much of the creative work I end up doing, I kind of stumbled into this project, mostly due to procrastination and writer's block.

Each month I write a music director column for the church newsletter, and usually I'm at a loss for a topic until about 72 hours after the column is due. When it came time to write the last entry of the program year in late Spring, I found myself thinking about the months ahead, when the choir gets the summer off and the hymn-singing is left entirely to the smaller-than-usual congregation. As we all know, communal singing is less a part of our culture than it was a hundred years ago, so although hymn-singing is a wonderful and very accessible practice, being confronted with a new set of tunes each week can be a challenge for those who don't read music easily.

I decided I could use the column to discuss summer singing and to promise that I'd post the hymns for each week online (essentially writing a check I'd have to cash later) to help folks prepare. Back at the dawn of the World Wide Web in the mid 90's, I used to upload the weekly hymns for a different church job. In those days of low bandwidth dial-up, the only practical way to make music available was via MIDI files, which are very clever and versatile, but which often sounded awful in web browsers. Nowadays, it's pretty easy to find room to post .mp3 files, so I set up a system for inputting hymns as quickly and efficiently as possible and turning them into audio which highlights the various SATB parts as available.

The initial site design was relatively simple, but as I started to build up a stash of hymn recordings, it occurred to me that making this growing archive easily accessible and searchable would be fun - not least for me. Over a couple of fairly intensive weeks of work, the concept of an "organ console with stop knobs" design emerged, with a scrolling archive of hymn tunes at the ready, and a nifty search box to help find things quickly. There are separate stops for soprano, alto, tenor, bass, and descant (when available, as you'll see by entering "descant" into the search box), and an easy way to toggle the search by date or by tune name. Check it out!


For copyright reasons, the player does not show the musical notation, though each hymn title links to information about the hymn on Hymnary.org, which often shows the text and in some cases shows page scans of public domain tunes. I'm not 100% sure about the copyright implications of posting what are, essentially, my "performances" of the hymns as printed in our hymnal, but I think it's well within the spirit of supporting congregational use of the hymnals our church owns without providing direct access to copyrighted material.

The site is mostly designed from the ground up, though I did import the cool search feature using the DataTables plug-in. It's far from perfect, but I spent a good bit of time trying to set things up so that this works both on desktops/laptops and on mobile devices. That kind of thing gets complicated! Honestly, even a humble little feature like the "this week" playlist button took hours to get working properly, but learning to problem-solve in this way was a big motivation for doing this.

I should mention that in addition to the musical listening guides I wrote about last September, I did some experimenting last year with using that template to create practice pages for choirs. These pages allow the user to isolate a voice part, speed up or slow down, and jump instantaneously around within the music's structure. Depending on your tastes, you can try this out with Bach's Crucifixus or a jazz choir setting of No more blues.


As I'll discuss in my next post, I'm finding that I get a lot of musical satisfaction out of interacting with musical building blocks in ways that go beyond just listening and performing, though I still love those pastimes. The hymn-player above is a fairly straightforward idea/design, but I have other projects in mind that will expand on this idea of engaging musical materials in creatively interactive ways. What better way to avoid practicing?

Sunday, September 4, 2016

Sundays at the Improv

Improvisation and Composition played almost no part in my musical training, which is regrettable. I can blame this in part on the biases of classical music education in the late 20th century, in part on my own silly biases against non-classical types of music, but, probably most importantly, on a lack of confidence and the absence of context in which I thought my creative voice could make a difference.

As I've written before, my first real efforts at composition came from my experiences working as a church musician, where I wasn't being evaluated from an academic perspective. Whereas the "lesson and recital" context I knew well was defined almost entirely by the idea of performing existing works as nearly as possible to the way their composers intended, and given that there seemed to be an almost infinite repertoire of music better than I could imagine writing, it was striking to come to a Sunday morning and realize that Piece X by Great Composer Z didn't quite meet the liturgical needs. Meanwhile, I'd also come across a repertoire of chorale preludes (old and new) that were much more functional than great, so I started to see where I might get the best results by 1) creating my own functional pieces and, 2) tinkering with existing works. [I suppose that years of "faking" as an accompanist and adapting on the fly to whomever I'm accompanying has helped nudge me past the idea that my job is always and ever to play things exactly as written.]

The most basic kind of tinkering might just involve cutting and pasting and repeating some generic prelude by Bach or Handel, but pretty soon I'd also find it useful to flip something major into minor or, most fun of all, recompose the opening to a chorale prelude by the likes of Pachelbel so that it seemed like a postlude based on the closing hymn that had just been sung. In some cases, this would involve changing as few as 3-4 notes, which essentially created a bridge from hymn to a "piece actually based on a different hymn tune." In other cases, I'd try to keep the new tune going against Pachelbel's figuration as much as possible.

Perhaps in a future blog post, I'll provide a more subtle example, but I happen to have a recording of an extreme instance of this kind of segueway. A couple of years back, the morning service was to be followed that afternoon by the parish's biennial variety show, in this case a 50's themed "Rock Around the Clock" celebration of the church's 1950's founding. Rather than picking up from the final hymn, I decided to anticipate what was to follow with the cryptically listed "Concerto in D Major, Vivaldi, arr. by W. Haley." Clearly, I didn't spend a lot of time on it, but it basically begins with Haley in the style of Vivaldi, transitions into Haley in the L.H. with Vivaldi figuration above, and finally into Bach's keyboard transcription of Vivaldi.


That's hardly a real improvisation though it was very unrehearsed, but in the past couple of years, when I've been playing organ regularly for the first time, I've experimented more and more with improvising freely for preludes, simply taking phrases from a hymn tune and "playing with it" over various pedal tones (easy to do on organ!), re-harmonizations, etc. Sometimes it goes better than others. Playing softly and slowly almost always helps in this regard as it's easier to hide in a haze. I'll admit I can't always hear in my head exactly how things are going to come out, although every now and then I end up with a total accident that works out (the kind of thing that has long since stopped surprising me).

For the Independence Day-adjacent July 5th Sunday, the opening hymn was "O for a thousand tongues to sing," and as I didn't have handy any prelude based on the tune (Azmon), I thought first about freely improvising. Then I remembered that this tune is the basis for the first movement ("Old Folks Gatherin'") of Charles Ives'  Pulitzer-winning Symphony No. 3: "The Camp Meeting." A few months back, I had very loosely used the finale of Ives' Violin Sonata No. 2 as a prelude since "Come, thou fount of every blessing" was the opening hymn. On that occasion, I'd basically read more or less from beginning to end at the piano, bringing in the violin part as best I could, but also generally playing more slowly and softly than Ives had instructed. A sort of dream-like walk-through intended to minimize the revivalistic ecstasy of the original.

So, I went ahead and submitted "Improvisation on Azmon based on Charles Ives' Symphony No. 3" as the title for Sunday's prelude without really knowing what I'd do. Although I can do a passable job of reading from an open score (the Ives symphony is pretty lightly scored), I didn't want to rely on those skills for Sunday, especially as it would involve a lot of page-turning, a couple of transposing instrumental parts, and some kooky Ivesian accidentals.

I was actually on vacation, away from any keyboard for the end of the week, but I spent time when I could inputting parts of the score into the computer. When I got home Saturday night, I used Finale's quick and dirty "piano reduction" tool to squish Ives' ideas onto two staves, simplifying a bit by leaving out some octave doublings and the like. The "piano reduction" tends to over-simplify and obscure some of the voice-leading, but oh well, time was running out.

Sunday morning, I arrived nice and early and just started trying stuff. I went back and forth between piano and organ, and though I'm much more comfortable voicing unusual harmonies and contrapuntal textures on a piano, I decided the organ gave me more of the vibe I wanted. Part of my inexperience as an organist means I don't love making lots of registration changes on the fly when playing, so I knew I'd be limited in the range of colors available, but I'd pretty much decided to avoid the more harmonically adventurous sections of the movement where the problem of notes sticking out awkwardly is more likely. Here's the original:



And here's what one of my practice run-thrus sounded like. Unfortunately, by putting my phone right on top of the console, there are a lot of extraneous sounds from the keys, the pedals, the bench creaking, and the like.


Basically, what took shape is that I play through the first 41 seconds of the Ives, more slowly, repeating the progression introduced by brass/bassoon, and then (departing briefly from Ives) alternating the first few phrases of the tune against several repetitions of the last brass/bassoon bar. Then I skip over a more chromatic/dramatic bit to Ives' first clear intro of the tune against a walking bass line around 1:01. This also continues until things started getting too "interesting," at which point I jump (rather clumsily) to Ives' coda [5:51], once again bypassing a big climax. At the Adagio cantabile, Ives introduces a lovely, meandering, gently out-of-synch flute descant, which meanders a bit differently in my "version," (less off-kilter oboe material). To wind things down, I (sort of) repeat that descant over a pedal and then fade back to the brass/bassoon progression to end things.

My performance as rendered here is far from perfect, especially in terms of registration, but as a "piece," I think it works really well, with a satisfying structure that ends up being quite different from what Ives wrote, even though most of the musical material is his. So, it's hard to say exactly what this is. It's part cut-and-paste transcription, part improvisation, and part "cast your fate to Finale's algorithm." As it happens, Ives' symphony is actually based on three now-lost organ works, with the inspiration for this movement having been a prelude the composer played at New York City's Central Presbyterian Church around the turn of the century. Maybe I channeled Charles and found my way back to 1901....