Skip to main content ↑ ↓
Topic: Lyrics-packing: oddities, inexplicabilities -- and praise. (Read 8000 times) previous topic - next topic

Lyrics-packing: oddities, inexplicabilities -- and praise.

After passionately expressing my frustration at the inefficient way in which NWC was packing lyrics, I've been busy for some time just transcribing notes. But I'm back working with lyrics, and I must say that the "standard rules" for lyrics placement (introduced sometime during that period) represent a significant improvement -- about 10% shorter lines overall. Couple this with the "force system break" feature, and I think the result is more like an average 15% reduction in total number of, say, 4" systems on typical hymns.

And, of course, ESPECIALLY considering the cost of the upgrade which provided this feature, I think public appreciation is the least it deserves. [imagine appreciatory gesticulations here].

Now, I'm not, um, fully satisfied yet. And I'm seeing some _very_ wierd behavior.

1) Item: the construct "quarter note in alto, two slurred eighths in soprano, one syllable in texts" occurs on average, more than once a line. Usually it's handled fairly well--syllable begins under the first eighth note and continues under the second as needed. But, occasionally it decides the second eighth note cannot be placed until after the end of the syllable. Why?

2) Item: I carefully set margins and system breaks on my development machine (which has my registered copy of NWC, upgrated to [almost?] the latest version; but has no printer access.) I go to another machine (which has printer and internet access), download an evaluation NWC (leaving it unregistered in case someone else gets to it), and print from there. Some lines appear substantially longer: as much as 1.5" on a 4" line -- massive gratuitous white space (which I can't see as anything but a wierd error).
Most lines are no longer -- and some may even be a bit shorter (which might arguably be variation between releases). Has anyone else seen this kind of thing?
3) I'm still seeing some unaccountable space -- occasional unnecessarily large gaps around a long syllable.

[I can provide example files for all of these.]

4) Perfectionist that I am, I still think more space can be sqeezed out. I really wish the beginning of long syllables would float all the way back to a vertical bar line, rather than wasting almost a 16th of an inch at the beginning and end of every measure. It looks to me like successive syllables on different lyric lines never overlap (although I do see they will abut the same vertical line.) I wish I could squeeze out some of that blank space before a double bar. I wish the syllable-divider dash would disappear to save space when the all the space around it was compacted out. But we're talking 10-15% space wasted here, not the 30-40% that it used to be. As I said, a big improvement, and perhaps the difference between "being able to generate usable, bindable hymns in the standard hymnal page format" and "not really being able to do more than single-pages".

Re: Lyrics-packing: oddities, inexplicabilities -- and praise.

Reply #1 –
Would you send me the files, zipped? My e-mail address is directly under my photo at http://www.icogitate.com/~ergosum/index.htm

Since I do SATB all the time, and am familiar with what you describe, maybe I can help.

I assume you know that lines can be "squeezed" by reducing the point size of the lyrics font. Since I don't mind post-processing the music (church hymns are short enough), sometimes I enter the lyrics at a tiny size to squeeze the measures, then change the lyrics text size when I drop the WMFs into documents.

But I'll try to help without post-processing.

Re: Lyrics-packing: oddities, inexplicabilities -- and praise.

Reply #2 –
The NWC evaluation does not have all of the latest features. It is only there for evaluation.

Long syllables will always end up taking more space. There are some occasions where a syllable cannot use all of the space that you see, such as when a bar line intercedes, or the next note indicates that it also has a syllable.

Re: Lyrics-packing: oddities, inexplicabilities -- and praise.

Reply #3 –
...which is also true of professionally-printed sheet music.
 

Re: Lyrics-packing: oddities, inexplicabilities -- and praise.

Reply #4 –
I think I understand the intrinsic spacing limitations, and (with a number of years of experience in optimizing compiler development) I appreciate that a program simply can't match what a careful human can do. But I'm seeing behavior inexplicable even by what I am deducing of the NWC algorithm, even on the registered version.

Let me prepare some exhibits by Robert for a sanity check.

To clarify:

My understanding of the evaluation version was that it was identical to the released version, with merely some editing, saving, and printing functionality disabled. I thought I was avoiding all of that, since I was printing files saved by a registered version, and doing nothing but tweaking margins (and even that I had expected not to have to do.)

Are you saying that the evaluation version may also be missing some of the fixes and enhancements (to layout) available in the latest downloadable upgrade?

I have a broader concern, in that these texts are being prepared for (small-scale) distribution. I was hoping that people receiving copies could print them without difficulty on the latest evaluation version. Perhaps I should be trying to print from the browser plugin instead.

Re: Lyrics-packing: oddities, inexplicabilities -- and praise.

Reply #5 –
Aside from possible differences between the evaluation version and the web update, there will be much significant (or even extreme) differences between different users' systems.

Specifically, how your page layout will look will depend on the properties of the default printer on any given system. It is the default printer's properties that NWC looks at in determining the available space for each system. In addition to the margin setting within the NWC program, there will be an "unprintable area" defined for the printer itself.

The above is true even when using "virtual" printers. For instance, if I use the "Applewriter PostScript" device to print to a file, the result is often very different than if I print to "5-D PDF" (aka "Jaws PDF") -- even if I select the same paper size.

The only way I know of insuring that every user sees exactly the same page layout is to send them files in a "predigested" document format, such as PDF, Postscript, or even image files (TIFF, GIF, PNG, etc.)

Re: Lyrics-packing: oddities, inexplicabilities -- and praise.

Reply #6 –
Good catch, Fred. It is indeed the case that different printer drivers (even on the same system) may produce different results. This is particularly true if a measure just barely fits in a system. If printed by another driver that requires more clearance (whitespace, margins), the measure might no longer fit on that system. If so, it will be placed in the next system. That might not be the desired result!

Dependence on printer driver is not particular to NWC.

Re: Lyrics-packing: oddities, inexplicabilities -- and praise.

Reply #7 –
Different printer drivers were involved, and that probably accounts for some of the oddities I've been seeing. I hadn't run into that on any other program, but most of my work has involved content development for scrolling windows, not fixed page sizes.

It seems I need to force system breaks at the right places, then at the worst, all other things working, all another user would have to do is tweak the margins.

While preparing a set of bug exhibits, I note these patterns: most of them involve some combination of: one syllable on two notes (or one-over-two notes); or system extention oddities.

Re: Lyrics-packing: oddities, inexplicabilities -- and praise.

Reply #8 –
Stephen, that sounds like something that can be routinely solved. Doesn't sound like a program problem. Get me some files to look at.

Don't forget that you can select a note, then change its properties so that a lyric syllable won't be placed there. So, for example, if you have Soprano holding on 1/2 note while Alto moves on two 1/4 notes, you can select the second 1/4 note and tell NWC that it should not place a syllable on that note (because it was already placed at the point where S started the 1/2 and A sang the first 1/4).

Re: Lyrics-packing: oddities, inexplicabilities -- and praise.

Reply #9 –
The files are in the mail. Thanks.

Re: Lyrics-packing: oddities, inexplicabilities -- and praise.

Reply #10 –
Indeed, I got Stephen's e-mail and files. I'll reply to him directly. If it's something of general interest, I'll also reply here.

Re: Lyrics-packing: oddities, inexplicabilities -- and praise.

Reply #11 –
I replied. It was general stuff. Some tips:

(1) If you have aVeryLongOneSyllableWord taking up too much space in the lyrics, you can try removing it from the lyrics (use underscore as placeholder), placing it as text using the lyrics font, and fooling around with the alignment.

(2) If you wish to place numerals in front of each lyrics stanza, and if your music happens to begin on the offbeat (that is, the leading measure is not complete), try this: Place a rest, of appropriate duration, to start the first measure. Change its properties to hidden, lyrics syllable always. In the lyrics, use a space between the stanza numeral and the first word. This will treat the numeral as a "syllable" and attach its position to the hidden rest.

If you cannot place a hidden rest to start you music, then you can start your stanzas like this, with an underscore:

1._Deck the halls with bought of hol-ly, (etc.)

(3) If your final staff system has enough "stuff" in it to fill the width of the page, but yet it isn't shown as filling all the way to the right: Change staff property to hidden system ending. Manually place system close bar at each staff. On the top staff, force system break.