The Charm of Impossibilities

I am sitting here trying to write microtonal polytempo music on Sibelius. I have found the most aggravating, patience-requiring method of composing in the history of music. I have spent the last two hours trying to fill three measures of music – largely because Sibelius will not allow a pitch bend command to be pasted onto a note in a tuplet. If Schoenberg really was trying to make it impossible for his students to compose, as Cage claimed he said, he would have made them do it this way. Future generations of composers will look at my music and say, “I’m sure glad he did it, because I certainly wouldn’t have tried.”


  1. says

    There often are a few work-arounds available to do such things. But then again, you are an experienced user, so maybe it is a very special case? If you post a screenshot, I’d be glad to have a look whether there is a way to de-charm this.

    KG replies: Well thank you. I do have work-arounds, but they’re a little time-consuming. Sometimes I place quarter-tone accidentals on every other note, apply the quarter-tone playback plug in, and then alter the pitch bend values. I can also use command-T to add text to each note and type in the pitch bend command by hand (~B 0,77, for example); and I can also use command-T and then command-V to paste in a pitch bend command as text. It would be so much easier if I could click on a note and command-V the pitch bend value. I’m talking about, say, an 8th-note in the middle of an 11-tuplet. Other ideas would be appreciated.

    • says

      Try selecting the pitch bend value (~B0,77) and then option-clicking near the note where you want it. That seems to work in 6.

      I can’t think of any workarounds for polytempo music though.

      KG replies: Thank you, that does seem to work. I have to be very careful where I click, because a half-millimeter right or left and I get a glissando on the note.

      • says

        In cases like that Kyle I increase the screen view – sometimes to 400% – so that there is more room for error.

        KG replies: I’ve often found myself doing that too – especially when trying to insert Johnston accidentals.

    • says

      Seems that I’m a bit late to the party that I started…

      Kyle, what you need is either Sibelius 7.5 (for which it is promised that you “can now copy and paste notes, text, lines, or lyrics directly into any type of tuplet”) or, in case that ending your ordeal isn’t worth the upgrade to you, Bob Zawalich’s Plug-In “Multicopy Object” (which will circumvent the tuplet barrier and will also allow you, as the name implies, to copy one object to multiple locations at once).

      KG replies: Arrrgh! Damn you, Sibelius 7! I hate the interface, but that certainly would be convenient. And that Multicopy Object sounds great! It would be so convenient to select every A on a page and give them all the same pitch bend with a click. I’ll check it out. But actually, I’m getting pretty fast at the command-T, command-V method.

  2. says

    I use Finale but have never written microtonal music. However in Finale you can input notes using a midi keyboard and type in a number for the note duration. I would guess this would mean you could put in the pitch bend information via the keyboard. Maybe Sibelius can do this – tedious but it wouldn’t take you three hours.
    My problem is different however,I don’t like equal temperament, but as I write mainly for saxophone or piano I am stuck. Living in Renaissance times wouldn’t have been any good as I like hot showers, bathrooms with flushing systems and psychedelic music.

  3. mclaren says

    I’m sorry, Kyle, but no current computer music notation program permits either serious microtonality or genuine neorhythmic music (you call it totalism).

    Example: if you want to notate a non-octave tuning using non-just non-equal-tempered pitches, there’s just no practical way to do it with any conventional computer music notation program. All such programs implicitly assume you have octaves, something like conventional pitches A B C D E F G, etc.

    Current computer music notation programs impose even more draconian limits on the notation of rhythm.

    Example: try to enter something like a repeating pattern of quarter notes with these tuplet values:
    5:4 5:4 3:5 3:5 3:5 9:7 9:7 7:11 7:11 7:11 5:3 5:3 17:13 17:13 4:9 4:9 4:9
    against another repeating pattern of quarter notes with these tuplet values:
    7:3 7:3 7:3 8:5 8:5 11:9 11:9 9:13 9:13 9:13

    and then try drawing tempo curves so that the first repeating pattern accelerates from 50% to 100% over the course of 5 repetitions, while the second repeating pattern decelerates from 100% to 60% over the course of 8 repetitions.

    Today’s supposedly “cutting edge” allegedly “sophisticated” music notation programs won’t even let your enter that stuff. They give you an error message. They try to tell you that you’re doing something wrong.

    The only way to get the above rhythmic values it to step-enter the tick values by hand one note at a time and then apply the spline tempo curves of Mark of the Unicorn’s Digital Performer to each melodic line. Then play back the melodic line into a second computer, and rinse, wash, repeat. Eventually you get what you want. But you have to fight computers every single step of the way (all puns intended).

    As with all cases where the triple-M (Manhattan Method of Marginalization) emerges like Godzilla from the waters of Tokyo Bay, a horde of New York composers will now rush forward to explain 1) I’m wrong (they’re lying); 2) what I’m talking about is trivial and Henry Cowell did it 100 years ago (they’re ignorant; Cowell proposed broken tuplets that add up to something nice and simple like 4/4 or 17/16 at the end of every barline. When you use sets of broken tuplets that don’t repeat, you no longer get a reasonable notatable time signature — for instance, the time signature of the first line above has a denominator of 180180, given by 4*5*7*11*13*9. I’d really like someone to show me how either Sibelius or Finale will allow you to enter a time signature with a denominator of 180180, given that both programs limit you rigidly to time signatures with no more than 100 in the numerator or denominator), 3) people in New York produce these kinds of rhythms all the time (once again, they’re lying); 4) yes, the New York musicians now suddenly admit they’re lying, but then they change their objection by now asserting that even though those these kinds of rhythms are genuinely new and do sound unique, they’re useless because humans can’t play them, and what really matters in contemporary music is virtuosity and the human touch (yet another lie, as Conlon Nancarrow’s music and all current computer music shows); 5) and as the final fallback position, the New York musicians now admit that this objection too is a lie but that doesn’t matter, since what really counts in contemporary music isn’t novelty or whether the music sounds interesting, but whether the music is “genuine” and “authentic”…and obviously music requiring such complex rhythmic notation can be neither (yet another lie and a final admission of the total failure of the Manhattan Method of Marginalization, since the real objection to such rhythmic practices is merely that the composer who uses them does not live in Manhattan — but of course the practitioners of the Manhattan Method of Marginalization don’t want to come right out and say this, because it reveals their utter musical and intellectual bankruptcy).

    Having digressed to debunk the usual triple-M sophistries, we now return to the main point.

    Which is: that a rhythmic example like the above represents mere baby steps in exploring new realms of rhythm. Consider broken tuplets made up of irrational or transcendental numbers: for example three quarter notes with a length of square root of 31 in the time of the cube root of 17 followed by two quarter notes with a length of the square root of 19 in the time of the cube root of 7, and so on. Notes with this duration are trivially simple to define: a note of duration 4:3 is just 3/4 or 75% of a quarter note, and if you’re using 480 ticks per quarter note, this boils down to 360 ticks. You can step-enter that easily into even an ancient DOS MIDI sequencer from 30 years ago. Likewise, a quarter note with a length of square root of 31 in the time of the cube root of 17 has a length given by
    2.6684/5.567764 = 0.47925 of a quarter note. At 480 ticks per quarter note, this equates to 230.04 ticks. Once again, you can step-enter a note with that duration even in an antique DOS-based MIDI sequencer from 1984, so you can certainly step-enter a note with that duration into a modern MIDI sequencer like Logic Studio or Digital Performer or Cubase.

    So you can step-enter even the most exotic tuplets into a MIDI sequencer and get ’em to play. But you have to blow right through deliberately crippled intentionally useless purposely frustrating designed-to-frustrate-composers-imagination computer programs like Sibelius or Finale. You must work directly with the MIDI stream by step entering the notes one at a time. If you have a composition with thousands of notes in it, that may take time.

    Well, Conlon showed us the value of patience. So buckle down, buckaroo, and get to work. Ignore the notation programs that are placed conveniently before us in an apparently deliberate effort to block any path toward creative and imaginative new music. To get these kinds of new pitches and new rhythmic values, you must bulldoze right through every digital wall the contemporary music world has erected to prevent us from escaping our prison of 12 Western pitches and powers-of-two and multiples-of-three rhythms. The jailbreak may even require high explosives. Forbidden dangerous highly-combustible materials like emotion and creativity, blockaded by Western contemporary music and strictly locked down under quarantine. There may be collateral damage from the resulting explosions of new pitches and new rhythms.

    Fortunately for composers interested in tempo curves, Mark of the Unicorn’s Digital Performer lets you draw graphic tempo curves using splines. What’s even better, the tempo curves can be automated to some extent. That is, you can specify a starting tempo and an ending tempo, and then enter a value from 1 through 99 which gives you either a negative or a positive curvature. A value of 50 is a straight line — simple straight acceleration. A value of 99 give syou an approximately exponential tempo increase with almost no speedup at the start and very rapid speedup toward the end. A value of 1 gives you a nearly logarithmic tempo curve, with a great deal of speedup at the start and very little toward the end.

    Step-enter the broken tuplets as above, then import ’em into MOTU’s Digital Performer and apply the desired tempo curves. Digital Performer displays ’em graphically in the MIDI graphical editor! Once you have the desired acceleration curve, you then sync the playback to SMPTE and record the MIDI output of one Mac into another Mac computer. To get a second tempo curve, you have to do the whole thing over again, because of course the MIDI spec only allows one tempo curve per MIDI file. But it’s trivial to get around that by playing MIDI files with various tempo curves into another Mac and overlaying ’em. By syncing to SMPTE, you assure that the melodic lines with different tempo curves are locked together precisely with no time slippage.

    You can even go farther, getting accordion acceleration curves like a set of 5 notes with acceleration 5% 6% 7% 8% 9% 10% against another set of 5 notes with acceleration 2% 5% 8% 11% 14%, and then go on to the next set of 5 notes with acceleration 6% 7% 8% 9% 10% 11% against another set of 5 notes with acceleration 5% 8% 11% 14% 17%, and so on. To do this, you have to step-enter the notes one at a time into the MIDI sequencer by hand, calculating their lengths. But once you get used to abandoning the crippling deliberately limited intentionally frustrating current music notation programs, the sky’s the limit. You can get xenharmonic pitches of any desired subtlety, if desired glissing from one microtonal pitch to another, and neorhythmic (you call them totalist) duration sequences of any desired complexity.

    The prerequisite if you want this freedom? You must abandon scriptism, the triple-M-designed (Manhattan Method of Marginalization) ideology that contemporary music is not “real” music and is not “serious” and is not “aesthetically valuable” or “historically relevant” or [insert vacuous Manhattan Method of Marginalization buzzword here] if the music is not notated using conventional Western common-practice-style music notation with a 5-line staff, conventional rhythmic flag and beam notation, and so on.

    In actuality, we now live in the 21st century (although composers in Manhattan are still stuck in the 19th century, and should ride horses to the concert hall while wearing top hats and cravats). Consequently, MIDI files can be parsed by arbitrary graphic programs to produce an unlimited range of visual notations. Lilypond offers one program for parsing MIDI files and generating conventional music notation, but many other programs can parse MIDI files and produce less conventional graphical output. The prejudice that only one type of graphic output represents “real” contemporary music remains as laughable as it is historically ignorant, since music notation has changed from neumes to six-line staffs to contemporary notations which don’t involve common practice techniques (Morton Feldman used lots of these, as did Earl Brown). This kind of narrowminded ignorance approximately equates to the absurd prejudice that only a hardback book with a linen sewn binding printed on acid-free paper represents a “real” book, while e-books or graphical novels or comic strips or blogs are not “serious” literature and do not represent “genuine literary works.”

    The triple-M will be around for quite a while and will work hard to stymie progress in American music. The Manhattan Method of Marginalization, with its scriptism ideology (all contemporary must be written down in conventional notation!) and its human performer ideology (any music not performable by humans given nothing but a sight-reading 3-minute opportunity to play the stuff is not serious music!) and its concert festishism (any contemporary music not performed in a Manhattan concert hall is not real music, so Mikel Rouse’s Quorum doesn’t count, and Pauline Oliveros’ “Deep listening” recordings in the Washington cistern with its 30-second reverb time doesn’t count and John Chowning’s “Turenas” computer music doesn’t count, and so on) represents the American musical equivalent of the Tea Party. Its job is to gum up the works, to crush progress, to prevent anyone from doing anything useful or worthwhile, to insure that if they cannot rule then the entire system must grind to a halt. It’s as essentially Trotskyite approach. If you can’t control the system, break it.

    But, as with the Tea Party, the triple-M will not prevail. The rest of us will simply grind over them and relegate ’em to the ash-heap of history. Even the most determined regressive backwards-looking reactionaries cannot prevent the future from happening forever. As Tolstoi noted, “These two warriors are the greatest of all: time and unconquerable patience.” As for the above methods being tedious and requiring a great deal of effort, well…as Michelangelo pointed out: “If you knew how much work went into it, you wouldn’t call it genius.”

    KG replies: It’s kind of an amazing tribute to Ben, Conlon, and Cowell that they could do with a pencil things we can’t do with a computer.

    • says


      As I understand it, Nancarrow switched to the player piano when he got frustrated with the limitations of human players. Since you are seemingly trying to use software specializing in standard notation to produce on the whole non-standard notations, may I suggest that you follow his example and choose tools and methods that are more suited to your needs?

    • says

      I think there are two different things here. One is notation, the other is playback. It seems that the notation is time consuming, but so is traditional engraving or copying out things by hand; the other is playback. I understand it is the playback that is causing the problems, not the basic notation.
      My view is that if the total engraving time is less than it would take me to copy out the score and parts by hand it is fine.

      KG replies: No, the playback works fine. I haven’t begun to spin out all the complications of my process. Eventually I’ll have to transfer all these notes into some pitch-bend-less MIDI notation, to play on retuned soft synths, so I end up notating every piece twice. But I compose with the pitch bends, because it’s just too hard to compose when I have to remember the *real* tuning of every note I put down because it’s not obvious. Finally, I will someday want to do a final score with Ben Johnston’s pitch notation so someone can look at a meaningful score, and that’s the most tedious and time-consuming task of the three. Essentially I make every microtonal score three times.

  4. Gerald Brennan says

    I find writing in Finale or Sibelius to be an exercise in hair-pulling if one ventures outside the box. Many times I have been told by their people that I simply could not do what I wanted to do with their software.
    Lilypad has a steep learning curve, and few of the extra-compositional niceties of the above-mentioned market leaders, but there is nothing I can’t do in it. I use Frescobaldi as a user interface.
    I am being forced to use it, essentially, because of the limitations of F and S.
    Brings up another point along the “media is the message” line. When using F or S, it is tempting to “solve” some problems by acquiescing to the limits of the software. Essentially saying, ‘ah the hell with it,’ and altering your vision to suit the limits of the tool. Very bad “solution.”

    KG replies: Admittedly. But I also have to admit there are some weird things I’ve started doing because Sibelius makes them so easy.

    • says

      I find myself doing that a lot, compromising with Sibelius when I could get exactly what I wanted by writing by hand… mostly because (a) my musical handwriting is pretty illegible, (b) my hand invariably cramps up after a couple of hours, (c) speed is of the essence because I have to have at least a rough draft of the piece written out before I get bored with it and decide to abandon it forever (which is usually about 2-3 weeks after I conceived the piece in the first place), (d) I am really bad at using rulers so my barlines always end up crooked. Excuses, I know.

      Someday I hope to reanimate Stravinsky’s corpse so he can give me a class on musical calligraphy. >.>