Re: promoting LilyPond

[Carl Peterson]

On Tue, Dec 3, 2013 at 12:01 PM, Garrett McGilvray <address@hidden> wrote:

Now, to focus on a different point: the question as to whether a truly fair comparison can be made only by professionals. I have no doubt that an experienced professional in Finale (and I've ignored Sibelius because I've never used it, so I have no opinion) could produce a better score than the best that I can do in LilyPond. But I don't think each one's merit could be totally measured based on what one is able to achieve with the greatest skill and effort given at tweaks. I am fairly confident that if a person tried an experiment to make a sample score in Finale that looked like LilyPond's default output, he could nearly well achieve an identical look. He would alter stem, line, slur thickness. He could manually position each note to line up with LilyPond. He could develop a font that copycats LilyPond's default. In the end, the two results would be identical, and based on final output alone, the two options would therefore be judged comparable. Of course, default LilyPond is not the target goal, but my point is that it is not just about what one can do if he applies skill and time to tweaking output. I know that beautiful results can be had from either program with much tweaking on both sides, but default output should be at least part of the comparison.

Then we come to the fact that there are very many people who use either of these programs who are not professionals, or even professionals who do not have the time to tweak every score to perfection. In my case, I am very much aware of many of the tools to tweak just about everything in Finale. However, first, I don't want to have to fight with spacing at the minute level, and secondly, as I was trained to read the music, not write it, I won't know the finer rules of when and where I should override Finale's default. On the one hand, I look at Finale's default output, and on the whole I feel like it looks as it should. But then I look at LilyPond's output and see, "Oh yeah, that does look more correct." That's the best someone like me can do without knowing rules of engraving. So in my circumstance, a comparison of what a professional can do is irrelevant. I need to know rather what *I* can do or what I have time to do in one program or another. So my own comparison of my own work in one versus my own work in the other is exceedingly relevant and fair in helping me decide which is right for me. That is especially true since I am a hobbyist doing my own work for my own use. I'm the only one who needs to be pleased in that case.

And all of this is just to explain a comment I made about what aspect of LilyPond appealed to me that made me give it a second chance. That seemed to be the point of a thread about "promoting LilyPond."

Regarding what it takes to make a score look "right," I have some rather direct comparison between LP and Finale. When it comes to something as relatively-simple as an SATB hymn, my friends who use Finale have to do a number of things beyond note entry:

* They constantly have to go back and fix horizontal note offsets anytime they make a change to notes so that the treble and bass clef notes line up vertically.
* Because they work with shaped notes (and a custom shape note font at that), they have to do all sorts of tricks with stem lengths to avoid gaps between some of the noteheads and the base of the stem.
* Any number of other manual tweaks for slurs and ties and such things.

The bottom line is that I can transcribe a hymn note for note using direct text input into an LP template and fix any entry errors in the space of 20-30 minutes, with few, if any, of the problems my Finale counterparts encounter. The only manual tweak I use in the music is an override for the part combiner when I want three notes on a stem (such as a tenor and two bass notes, which happens rarely in the hymns I transcribe, and almost never in my own compositions). Everything else is handled by layout-block overrides, which are stored as a template. By comparison, one person (who does semi-professional Finale work and is quite proficient with Finale, from what I've seen) spends 2-3 times that time to get similar results.

The only major defect I tend to see in my output, relative to the same hymn in Finale, is lyric spacing, particularly horizontal spacing. There are two features which, if they do not exist, would make the LP settings much better:

1. Horizontal spacing priority for lyrics rather than note durations. In other words, can we tell the horizontal spacing engine to space lyric anchor points more or less equally rather than strictly going by note durations? The big issue is when there are significant differences in note durations, such as when a half note appears in the midst of eighth notes. This creates a huge gap in the spacing. This is particularly ugly at the ends of lines, since it leaves a huge gap at the end of lines, forcing the lyrics to be spaced more tightly in some instances.

2. Floating lyric spacing. Right now, lyrics are by default centered underneath the note they are attached to. This is fine in many circumstances, but when there are multiple stanzas with syllables of varying length, this can create some irregular spacing and general ugliness. Since I use manual line breaks, it also sometimes causes lines that I'd like to keep together overrun the margin, when this could be avoided if the syllables would be allowed to "float" horizontally to try and provide more even syllable spacing. Some hymnals take this to an extreme, where syllables may be completely separate from the note they are sung on, but it would be nice to be able to specify that syllables be centered, +/- 50%, for instance.

I realize that much of this can be done on a case-by-case basis through manual tweaks, but as I think many have said, the ideal is to be able to just put in music and let Lilypond be able to figure out what it needs to do make it look right from the start.