Left Justified Lyrics Fix

• Nov 14, 2018 - 01:01

Hi there.

A small thing I've noticed that's actually surprisingly distracting when I set lyrics to left justified is that the left-most tip of the left-most letter doesn't actually go right under the note as it ought to - but it does if I give it a horizontal offset of -1.00. Without this offset, it seems oddly out of place and doesn't really work. Unfortunately, the horizontal offset can't necessarily be globally applied because not all the lyrics are necessarily left justified in a single score. I was wondering if the default left justified could be set to what is presently -1.00 horizontal offset?

Thanks,
-Gidoza

Comments

• Nov 15, 2018 - 19:44

Can you explain why you are setting lyrics to be left justified? This is not correct by any conventional engraving standards, except for lyrics that are part of a melisma, and MuseScore left-justifies those automatically.

Anyhow, I can confirm a very slight (nowhere near 1sp) discrepancy with left-aligned lyrics, that I don't see with staff text (which is left aligned by default). My guess is that lyrics might have a slight "fudge factor" built in so that centering looks the way we might expect over a variety of different head widths (whole notes are noticeably wider than quarter notes), and that this same fudge factor is applied unnecessarily for left-aligned lyrics.

Reply

• Nov 15, 2018 - 21:19

Hi Marc,

Actually, it would be quite the opposite of a melisma...rather than many notes for a syllable, it's liturgical chant with many syllables for a note. Since the chant isn't necessarily metric (actually almost never metric), we use a single note to indicate that all the syllables are sung smoothly and usually at a different tempo than everything else.

As for the reason I'm doing it - I'm only doing it for these specific stretchy notes, for two reasons: 1. The singers find it much easier to read, and 2. It looks far nicer. Seeing as convention is concerned with engaging with both these things and is doing neither of them in this situation, the proper response is to break convention.

Here are some examples to demonstrate the problem.

Example 1 - You will see the aforementioned stretchy notes marked as breves in bar 3 and bar 4. In this example, they are set to centre justified as usual. It doesn't really work, as it isn't clear where the previous note in line 3 ends for the lyrics and the new note begins.

Example 2 - One attempt to solve the situation is to add a new note; the first syllable goes on a bonus note and the other syllables go to a phantom note, which could be made visible as well, but quite frankly just looks like more of a mess than anything if made visible. This solution sort of works, but becomes awkward with the use of dashes to another note that the print-out doesn't record, and also the spacing becomes unusual as well.

Example 3 - Here, I have set the original note to left justified. It is now perfectly clear what words are associated with what note and not possible to mix them up. However, the words under the note are strangely to the right a little.

Example 4 - Finally, in this example, all is the same as Example 3, except that I have given the left-justified lyrics under the stretchy notes a -1.00 horizontal offset. Now it looks great!

Attachment Size
Example 4.mscz 20.88 KB
Example 1.mscz 21 KB
Example 2.mscz 21.08 KB
Example 3.mscz 21.04 KB

Reply

• Nov 15, 2018 - 21:59

OK, that's true, I forgot about that other special situation where left-aligned lyrics are sometimes used. And your example also helps clarify why you are seeing a different offset than we are - you are using the widest possible notehead, and also it seems playing some tricks, using the breve notehead on what is actually only a quarter note, and the same for the notes in the other voices. Change those noteheads back to quarter and things line up much better.

Yes, I think this is a side effect of the fudging we do to attempt to center the lyric under the center of the notehead. Probably some code is being hit that shouldn't because it is left aligned instead. Seems to be off in 3.0 as well, although differently so. Feel free to try a development of 3.0, play around, and report your findings, so we can look at fixing it!

Reply

• Nov 16, 2018 - 00:56

Thanks Marc, I appreciate this reply.

It helps to know that it's more of what I'm doing in particular and not the program globally that is creating this odd situation, nevertheless I'm glad there's potential to fix it! I'll play around with 3.0 sooner or later and see what happens. :)

As for the trick - it seems to me that there's no better notehead to use for a stretchy note that a Breve note (if you have other suggestions - I'm open, but this is the only kind of note I've ever seen for this; this, or a quarter note with double-lines on both sides instead of a whole note), but I only use the head with a quarter note instead of a real 8-beat note for two reasons.

  1. Calculating the beats I need for each bar is a pain when it's an 8-beat note.

  2. More importantly - if I want to play back the whole score to listen to it, an 8-beat note would be a disaster for hearing it in any smooth way.

Hence my solution, and the consequent problem. :P At least for now, the -1.00 offset is a temporary fix.

Reply

• Nov 16, 2018 - 03:35

Well, what you are doing is unusual, but it should work. If nothing else, a melisma starting on a breve would be left aligned and should align better.

As for using a breve notehead on a quarter note - if it works for you, fine, but you shouldn't need to "calculate" anything in order to use real breves. Just use Edit / Measures / Join Selected Measures to combine measures until they are big enough. If it's not supposed to last 8 beats, no good reaosn to sue a breve. Eg, if it only lasts four beats, just use a whole note.

Reply

• Nov 16, 2018 - 12:26

There are a couple good reasons to use a breve notehead, anyways.

  1. I want to specifically indicate that it's a stretchy note, and this requires something other than a regular notehead.

  2. As previously mentioned, anything more than 1 beat (from the perspective of music playback) makes the playback very dull and hard to follow. I can't imagine using anything other than a quarter note for these because it would make assessing the work and the musical flow very difficult.

Reply

• Nov 16, 2018 - 16:36

I agree there are good reasons to use a breve, my point is that you really use a breve, not a quarter note with a breve notehead. As explained previously, there is no such thing as a "stretchy note". There are quarter notes, there are half notes, there are whole notes, and there are breves. Each has its own specific duration. It is true that in some non-metric / chant music, breves are often used as a placeholder for any long duration. Still, it is a long note no matter how you slice it, so having it playback as a quarter note is just wrong, not sure why you'd want to hear it played wrong. But if you do want to hear it played back wrong, I'd still suggest notating it correctly - with a real breve - and then using a fermata with a "time stretch" value of less than 1.0 to force the playback to move faster.

The bottom line, it is almost always better to enter the music correctly (eg, with a real breve, since that is what this is), then use workarounds as necessary to customize the playback, than to enter the music incorrectly (eg, as a quarter note, since it most certainly is not a quarter note) and then try to fool the layout system.

It won't solve this particular problem but it will solve other problems you may or may not have encountered yet.

Reply

• Nov 16, 2018 - 19:20

I can't say that your suggestion really makes much sense to me. Given that a "stretchy-note" as a note does not exist in music, if one is going to be used, and if most choose a breve to do it, then it assumes its own new meaning as a note. There is no "right" or "wrong" in this situation because the liturgical community that commonly uses this kind of note is trying to create a new note. Therefore there isn't a "right" way or a "wrong" way to use it or to play it back. Furthermore, seeing as a "stretch" note can potentially have a stretch time of only a single quarter note in length, it isn't even wrong on a technical level. Are you objecting to the concept of a stretchy note, or the use of a breve as a notehead? If the notehead, what would you suggest instead?

The point of playback isn't about how many beats we're hearing: it's about assuming that the "stretch" is only a quarter note long in some circumstances and therefore merely about hearing the note and then moving on with the rest of the song. I don't need to hear 8 beats of the same stretchy note to know what the chord sounds like. It's not playing it back wrong - the playback, the design, and the concept are all doing exactly what they are supposed to do; conventional scoring just hasn't recognized yet that this is part of the reality of how chant gets scored.

As for the fermata idea - that's quite a lot more work to have an effect that's already more than complete and which has been done for a long time even before I came around. I'm not the first to do this - I'm perhaps just the first to comment on it thoroughly.

Reply

• Nov 16, 2018 - 20:37

I'm not "objecting" to the concept of a stretchy note and more than I am "objecting" to the concept of Santa Claus. Neither exists, whether I object or not. The liturgical community is not inventing a new type of "stretchy note", they are simply using a breve as a generic placeholder for a long note of indeterminate value, much as other styles of non-metric music might use quarter notes as generic placeholders for shorter notes of indeterminate value. But a long note of indeterminate value is still a long note - there's no way to sing all those lyrics using only a quarter note, which is the whole reason a breve is used in this context. So "correct" playback should be long. And more importantly, whether the duration of a given breve is intended to be indeterminate or not, it should still have the same effect as other long notes on music spacing, on alignment with other voices, on MusicXML export, on playback, etc.

And that is what is objectively wrong with your method. Using a quarter note will get all of those details wrong. You can partially workaround a subset of those errors by changing the notehead as you have done, but the other details will remain wrong, even if in some cases it won't be readily apparent.

So, from a notation standpoint, there is no question that using breve is correct and using a quarter is not here. It also produces more correct playback. I do get that you seem to want the incorrect playback that a quarter note gives you - I guess because you aren't wanting the playback to reflect performance, but just to help you debug your work. That's fine. But given two choices:

1) enter music correctly, get correct results, then use a workaround (eg fermatas) to make the playback incorrect to suit your whims, while also making it easy to revert to the correct playback for the benefit of anyone who might want to hear that

versus

2) enter music incorrectly, get incorrect results, then use a workaround (eg, changing notehead and hiding stems) to only partially fix the layout errors but live with the rest, and also live with the fact that there will be no easy way to get correct playback either

I'd pick #2 every time. Adding a fermata is no more work than your current method, but it produces objectively better results, and then if you ever do decide you want to hear the proper playback, you can get that in a few clicks by removing or resetting the fermatas.

Reply

• Nov 17, 2018 - 16:30

I don't know what you mean by "objectively," Marc - there's nothing obviously objective about this. A note of indeterminate value is precisely that - indeterminate; moreover, if I wanted the music to be metric music, (e.g. if the lyrics were poetically designed to fit the melody syllable by syllable for the most part), then I would choose a quarter note for the indeterminate value - that is the way it's supposed to sound, and it is playing back correctly. The possibility that it can be longer than that is indicated by a notehead (which could be some other notehead, but happens to be the Breve one), and ease of debugging is certainly a priority, because I don't want to be sitting around for 30 seconds per score listening to 8-beat breves. As for #2 - the layout error is fully fixed, not partially (with the -1.00), and the playback is quite correct. As for the fermata and the "correct" play-back - there's never a time when the stretchy note lasts for 8 beats anyways (4 would be a "stretch" - haha), hence the note-head's being symbolic and not literal.

Reply

• Nov 17, 2018 - 17:29

What I mean by "objectively" is that there are objective rules for how notes should be spaced according to their notated duration, and by having a quarter note with a breve notehead, these rules will not be followed, leading to incorrect spacing. And similarly for the other factors I mentioned. It doesn't matter if the notes are expected to actually last for the exact notated duration or not, that is the not as relevant as the notated duration for these purposes. Your workaround of applying an offset to the lyric fixes one of the layout errors, it does nothing about the others. Like I said, maybe they aren't as readily apparent to the causal observer, but that doesn't mean they don't matter if you care about getting the best results.

It's your music, you can continue to notate it however you want. I am just trying to help you get the better results with less effort.

Reply

Do you still have an unanswered question? Please log in first to post your question.