Soft flat?

[MRoth1910]

I'm working on a Kyriale. In the gabc score of Agnus IX found at Gregobase, the flat of "miserere" and "dona" is only on the first note, as per the source (mi(gxfgh)se(hvGF)) However, in my case, it appears that I'm stuck, for now, with a line break which would require inserting an additional flat. But I'd rather not insert a flat into the gabc, as that would make the file less portable (although I will if it means that the flat is printed correctly).

I mean "soft" just like in "soft hyphen", only printed if the line needs to be broken at that point; a soft flat would be printed in a case like mine, where the first note is already flat but where the second note to be sung with a flat could unfortunately wind up on another line.

This has some similarities to the problem mentioned in #157, but I think the problem is only at the ends of lines in a word with several flattened notes, due to the less predictable nature of line breaks, which cannot be as easily adjusted, unlike when setting the original scores by hand. I'm far from an expert, but awkward breaks like in Agnus IX are fairly rare…

Since a word division, a natural, or a bar cancels the flat, the first flat of a word or the first one after a bar, can be coded as normal without trouble. Nothing needs to be coded one way or the other to return the note to natural after a bar either even if theory one could add the natural just to be safe (cf. the Gradual of Ss Peter and Paul) If I understood the other discussion correctly, a line break is known already by Gregorio, and these two things together mean that only the line breaks need to be considered for printing some of the flats, not the other rules.

[xmarteo]

Is there a rule that a flat gets canceled after a line break? If not, why print it a second time?

[MRoth1910]

Yes.

Laus in ecclesia (pp. 26 and 27 of the French edition for those who have it) gives the Solesmes rules: Si is flat until a bar line (this is an issue in the gradual of Ss. Peter and Paul), until a natural, or until a line break. For that, the book gives the tract of Palm Sunday; "me" (at the verse beginning "Libera me") is split over two lines, before the quarter bar which would otherwise cancel the flat.

I propose this soft flat because I think having the flat twice on one line where it's not needed looks silly, which is the case for Square Notes (which does its own thing, I know, but the principle os the same) on my iPhone, but the flat is essential if the line might break and never does in the Solesmes books or even in the Vaticana.

After looking at hundreds (thousands maybe, though not as much as you) of chants and other typeset parts of the office, I have noticed that they were strict with conventions and knew how to make sure that they had as few exceptions as possible. So Agnus IX would never be typeset so as to cause a line break requiring a second flat.

[xmarteo]

OK.
This is clearly harder to do than "like soft hyphens in lyrics, but with flats", because this one relies on LaTeX, who simply /always/ inserts a hyphen when a word is spread over two lines, whereas the soft flat requires inspection of the presence of a Si before the next bar or word when a word is spread over two lines.

[MRoth1910]

Signum ♭ molle ante notam si positum minuit acuitatem notæ uno semitono. Effectus autem ejusmodi signi se extendit usque ad finem lineæ, nisi prius ♮ quadro evictus fuerit.

This is the Dominican rule. So, bar lines don't affect things, and it's not clear if you can have a flat for the entire chant as in the Roman books. Fr Innocent Smith noted that the simplicity of Domnican notation made for a hard time with Gregorio (back in 2016, at least).

We had an issue in the pre-1955 Holy Week book (as Matthias knows, the chant scores were made with Gregorio via Richard Chonak's website); most of the chants were OK, but there is a tract where Si flat occurs with the flat before the note, then there is a second Si in the same syllable on the same line, without an intervening quarter bar, making it flat in the Roman system. But the limitations of the page size forced this second note to the next line, where it was printed without a flat, because the score as coded only has one flat. This is exactly the circumstance which would require a flat to be printed sometimes, once Gregorio/LaTeX knows where to break lines — and really, I now basically think that people should always run latexmk even for simple projects, which would make this sort of adjustment happen behind-the-scenes with a score taken from Gregobase, if such a feature could be implemented.

But I do understand that, since it depends on line breaks, it's not easy! I'm just providing ammo for my case.

[xmarteo]

That is solid ammo. I can imagine having "flatrule" as an option, with "roman", "dominican", "ignorelines"...

My humble opinion is that relying on line breaks for musical semantics is a terrible idea, and that the editors of the Vatican edition and subsequent editions were wrong for adopting this system. But I also understand that the proverbial cat is out of the bag now.

I make a note to not follow that rule in the NR: Nocturnale-Romanum/nocturnale-romanum#32

[MRoth1910]

(Thanks for the hint. I'll add a note for my own stuff.)

My best guess is that the VE partially followed modern convention in a way that, owing to the non-corresponding nature of Gregorian bars and line breaks, you get this problem when using the computer.

So, I'll leave that there.

Hopefully we can get this in a future edition. I like the idea as proposed so far.

[rpspringuel]

Basically for something like this we need a flag that indicates whether or not alterations are being printed. This flag starts out true and the is switched to false after printing an alteration. We’d then need to define what causes it to flip back to true. For bar lines or ends of words, this is fairly easy as Gregorio knows where those are on its own. Line breaks, however, are a different matter. Gregorio knows where manual line breaks are (a z or Z in the gabc) but not the automatic ones. Those are determined by LuaTeX and we’ve yet to identify a mechanism for communicating that information back to Gregorio. This has come up before: for instance in multi-line initials. You’re required to manually specify line breaks for those because we can’t track the automatic ones in order to adjust the length and placement of the staff lines. If someone can figure out how to track the automatic line breaks, then several things which are currently not happening become possible. ✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝✝Fr. Samuel, OSB(R. Padraic Springuel)St. Anselm’s Abbey4501 South Dakota Ave, NEWashington, DC, 20017202-269-2300(c) 202-853-7036PAX ☧ ΧΡΙΣΤΟΣOn Apr 15, 2024, at 3:09 PM, MRoth1910 ***@***.***> wrote: (Thanks for the hint. I'll add a note for my own stuff.) My best guess is that the VE partially followed modern convention in a way that, owing to the non-corresponding nature of Gregorian bars and line breaks, you get this problem when using the computer. So, I'll leave that there. Hopefully we can get this in a future edition. I like the idea as proposed so far. —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[xmarteo]

@MRoth1910 upon closer inspection, I don't see how Laus in Ecclesia justify the line break rule. In the preface to the Vatican edition (found in Latin in the 1908 AR, in English in the LU) only a natural, a new word, or a bar line (not a line break) cancel the flat.

[MRoth1910]

I know. I have always taught that, and have been taught that.

But the actual printing tells a different story — at the very least, it becomes ambiguous when the first flat note is on one line with the sign and the next note on the following line without any sign whatsoever.

There is a line break in the tract of Palm Sunday in the 1908 Gradual and at the same place in the Solesmes edition of 1961. Both give a second flat for clarity; the 1961 English Liber Usualis has no break, so the flat isn't repeated. So this suggests that it isn't a Desclée-Solesmes rule adopted by Laus in ecclesia; it's a shared convention that the Dominicans formalized.