Monday, October 09, 2006

Music Notation: Sins and Fibs

Too often, talk about music notation gets reduced to a question of using Fin(ale) or Sib(elius), the two leading commercial notation packages. I happen to use Finale, but not for all my notational needs, and -- this probably sounds like a broken record -- I am a serious advocate for the maintenance of a diversity of products with alternative features and functions.

First of all, many of us still do some or all of our work by hand, and there's no reason to change if it works for you. For the most gifted calligraphers, a switch to computer notation, will save some time, but if you're picky about the layout, probably not as much time as you'd wish. I felt that I had to switch for the sake of my eyesight, but the jury's still out on that one. Moreover, the composer's own, hand-drawn, manuscript, will always have an aesthetic caché of its own, an aura of authenticity, if you will.

Some of us will decide to roll our own notation programs. I tried this, myself, writing in a mix of postscript and forth on my old Atari, but the result wasn't especially pretty, and I am now, officially, a SWIRP, someone-who-is-retired-from-programming. David Feldman, a composer who does a lot of algorithmic works, remarkably, is able to program in Postscript so that he can go directly from algorithm to finished score.

(Many more composers design their own fonts, or at least tweak the designs of existing fonts. The font which composer and copyist James Ingram made for Stockhausen's scores is one good example. Clarence Barlow did all of his own text and music fonts for his Atari environment. Having your own house fonts as well as layout style can be a cool thing.)

Some notation programs are purely graphic -- play-back of a score is not part of their schtick -- but, and in-part precisely because they are not obliged to come up with a realistic play-back -- they can be extremely flexible on the graphic front. Score, the granddaddy of notation programs, still has a small loyal user-base, and can do almost anything in standard notation. Graphire Music Studio can do even more -- circular scores, exploding sections etc. -- with a more contemporary interface, while Lilypond, from the free/open source software world, aspires to reproduce the results of classical engraving techniques through some very clever algorithms. Lilypond has less capacity and is less flexible than either Score or Music Studio, but as open source, it's a moving object -- if you need or don't like a feature, you can add your own.

Notion and Harmony Assistant emphasize playback, but both have respectable out-of-the-box notation faculties, and both, I believe have features that may point to the next generation of notation and production programs. I bought Harmony Assistant myself, and have learned a lot from it, but I am still not fluent enough with the interface to really test its limits. Writing additional "rules" scripts for Harmony Assistant is very easy, and I am especially excited about its capacity to render microtones live directly from notation into digital audio. In the Linux world, Rosegarden has native notation capacity, but sophisticated users will probably want to export from Rosegarden into Lilypond.

Finale and Sibelius are the two best-sellers in the music notation world, and they can both keep most users perfectly satisfied, or perfectly frustrated, which ever comes first. I believe that there is some consensus that Sibelius will give most users an acceptable output straight out of the box, as the house styles and the entry techniques for newbies are well-designed for newbies, but more complicated things get rapidly more difficult to do. Finale, on the other hand, has a steeper learning curve, partially due to a long development history and backwards compatability, and the out-of-the-box house styles are notoriously inadequate, but once you've oriented yourself, the power of the program becomes quite evident, and I believe that more elements in a score are tweakable than in Sibelius. This makes it easier, in the long run, to attain a specific house style, whether you want to imitate Henle or the best Broadway arranger, or invent your own house style. (I believe that Finale also offers more flexibility as to entry, but am not certain. I like to use command-line entry in Finale, and, as far as I can tell , this is not possible in Sibelius, but I'm happy to be corrected on this point).

But not every element is tweakable... for example, barlines in Finale are present at a low level in the data structure, so they remain fairly rigid elements in the program. If you want to have non-alligned barlines, as in Ockeghem or Ives, you're in for some elaborate kludges. Or, you could try, as an alternative, the program Lime, which has some serious limitations on other details, but makes non-alligned barlines in a snap (and has a built-in-tuning table function for ready microtonality). Lime only costs US$65, and for many users, it may very well be sufficient for all their other needs.

And there are a number of other programs in the landscape which are less powerful, but also much less expensive than Fin or Sib. Finale offers its own scaled-down version. Capella, Igor, Berlioz, Mozart, and Turandot have each a loyal, often local, user community, and each can produce attractive scores. While in Hungary, I purchased Turandot, the local notation product. While certainly nothing as radical as Harmony Assistant, it has an interesting interface, and for common practice era-style scores, the out-of-the-box-styled output can be excellent, very close in fact, to that of the Editio Musica Budapest.

No comments: