Battle of the Heavyweights: Finale Versus Sibelius

This month’s Keyboard magazine contains a detailed and very welcome comparison of Sibelius and Finale notation software, called “Notation Nation” by composer Peter Kirn. I’m a Sibelius user myself, currently on version 2.11 (though 3.0 recently became available). Though Finale’s been around several years longer, I never bought it; I was daunted by its reputation of being difficult to learn. For years I used a stupid little program called Encore, and I don’t mean stupid entirely negatively: it would let me get away with things I wanted to do that were too far out for intelligent programs like Sibelius that think they know better than you do what a score should look like and hem you in with defaults. But the three things I love most about Sibelius are 1) its speed, 2) its speed, and 3) its speed. I can write a Sibelius score faster than I can one in pencil, and I often take a piece that one of my students brings in and, in a couple of minutes, key it into the computer so we can hear a MIDI version and experiment around with various changes. Among other things, it’s a great teaching tool.

No one seems to be monitoring the impact of notation software on composing, and it is sure to be vast - and homogenizing, for I can say from experience how reluctant one becomes to notate something that the software won’t do easily. (I’ll give examples below.) The worst pitfall of composing on a computer is, you never get to look at a whole piece of music at once while you’re working on it. At some point in composing, I like to take all my pages and spread them out, or hang them around the wall, so I can see the entire piece at once. Of course you can print everything out, but it was easier when I was writing on large, 40-line score paper. Inevitably, composing on notation software results in waxing creative while only seeing three or four measures at a time, like trying to acclimate yourself to a pitch-dark room with a narrow-beam flashlight. I compose first on paper when I have time, but I feel like a lifetime’s worth of having done so makes it possible for me to compose directly on the computer when I have to and still keep a good overall sense of the piece. I do urge students to start out with that trusty old composing machine, the pencil.

Still, notation software offers composers ENORMOUS savings in time and money. It used to be that a composer’s commission money for an orchestra piece was largely if not entirely eaten up by the cost of copying parts. My students, writing orchestra works, would spend four to eight weeks before the deadline copying out all the parts by hand. Sibelius (and I assume Finale too) can churn out all the parts to a decent-sized orchestra piece in half an hour or less, all perfectly correlated to the final score. The $300 (academic) price tag may seem steep at first, but the first time you write a piece with parts it far more than pays for itself. Whatever subtle disadvantage notation software turns out to have, the patent advantages easily seem to outweigh.

Kirn’s readable analysis pretty much confirms the usual Sibelius and Finale reputations. Sibelius, he says, is easier to learn, has a simpler interface, and takes fewer keystrokes to get things done. Finale is better graphically, for getting a score exactly the way you want it to look. Sibelius is more like word processing, while tool-based Finale is more of a graphics program, though as he also points out, their recent upgrades imitate each other, narrowing the gulf between them. On one hand, Sibelius likes to look a certain way, and I do feel that I give in to a lamentable impulse to avoid notations that Sibelius won’t handle easily. On the other, some of my students use Finale, and despite watching them operate it for years, I still can’t even manage to figure out how to so much as sharp a note in the program myself: I just don’t find it intuitively user-friendly.

For the record, here’s what I used to be able to do in Encore that I can’t in Sibelius: Henry Cowell, in his masterful 1930 book New Musical Resources, suggested that “tuplets,” as they’re now called in computerese (and we needed a word for that), don’t necessarily need to come in groups. For instance, say you have five quintuplet 8th-notes and three triplet quarter-notes in a 4/4 bar. Why not have two of the quintuplet 8th-notes, then one of the triplet quarter-notes, then another quintuplet 8th-note, and so on? Very difficult for humans to sort out, but the computer plays them beautifully. I wrote a 1999 piece called Folk Dance for Henry Cowell based on the idea, and I can’t renotate it into Sibelius, because too-smart-for-my-own-good Sibelius won’t let me insert a note from another tuplet in-between quarter-note triplets. Pokey-looking little Encore didn’t realize I was doing anything unusual, and didn’t raise any fuss. (Encore, of course, which went out of production years ago, didn’t work well with Mac OS9, and won’t print at all in OS X.)

A related issue: a few composers, myself included, have occasionally used meters in which the denominator is not a power of 2, and Cowell anticipated these as well. For example, a measure consisting of four triplet quarter-notes (and nothing else) would be 4/6 meter, since there are six quarter-note triplets to a 4/4 measure, therefore a triplet quarter-note can be called (pace Cowell) a 1/6th-note. In “Mars” from my The Planets I use 17/24 meter at one point, which is a measure the length of 17 triplet 16th-notes. Can’t do this in Sibelius, nor Finale as far as I know, at least with accurate playback: you can, with some trouble, move mountains to get it to look right. I’ve begged Sibelius representatives to change the program to allow such rhythmic oddities, but they only look at me as though I just stepped off a UFO. But there are similar examples in pieces as famous as Boulez’s Le Marteau san maitre, and Ives uses meters like 4-and-a-half/4.

Plus, no music software will let you write different meters or tempos at the same time without an incredible hassle. Too bad Cowell wasn't around to be a consultant for notation and sequencing software companies.

The other pain-in-the-neck with Sibelius (aside from the eternal ambiguity they’ve created by taking over the name of one of my favorite composers) is their greed. They refuse to retrofit versions so that, say, version 1 will open version 2 files, nor will version 2 save files as version 1. This meant that, when my students started bringing in 2.0 files, I couldn’t open them on my computer, which means that when a new version appears, pretty much everyone on the planet is forced to upgrade all at once. I know music copyists are especially bothered by this, since they sometimes need to give a score back to a client in an earlier version, and can’t. Sibelius’s tech support also used to be snobbishly condescending and arrogant at times (and their manual is a joke), but lately their help pages at www.sibelius.com have been greatly improved, and I’m finally impressed by the ease with which I’ve been able to solve problems using them. I won’t cavil too much about a software that has improved my quality of life so much - but I may quietly root for the next competitive notation software company to give Sibelius a good run for its money.

Kirn's article also lists, with prices and commentary, every notation software out there currently available. It's a good resource.