Custom key signatures with grand staff problem

• Apr 16, 2021 - 19:48

Create a grand staff, piano for example. Create a custom key signature, eg 1 flat on the second space.
Place the key signature on the upper staff of the grand staff. It appears as A flat.
In the bass clef, it appears as C flat.
Is the solution – not to use a grand staff and to create two different key signatures, one with a flat on the second space, and the other with a flat on the top line, and assign them independently?
The problem is alluded to in the documentation in the context of clef change and [non]-transposition.


You indeed need one coustom key signature per clef. The treble clef shown on the dialog to create a custom key signature is basically fake.

In reply to by Marc Sabatella

When I place a flat on the second space, depending on the clef, it will be a different note. In the example I gave, the second space flat was Ab in the treble clef and C flat in the bass clef. As was pointed out, this dialog box is a 'graphic'.
I consider the score to be a visual representation of an underlying 'intelligence'. This allows for such things as intervals and keys. I view that there is one major scale, 0, 2, 4, 5, 7, 9, 11. This reflects the structure of the major scale. From this, the function of scale degrees can be described, [using the ^]. It shows that, for example, in C major, F# is not Gb, they are ^#4 and ^b5.
In this representation of the underlying intelligence, 'A flat' I added to the staff, being in 0#/0b, is ^b6. In what I did, the Ab ic ^b6 in the treble clef, and ^b1 in the bass clef. I find this idiosyncratic because I am not attached to the 'note' -- or the image of the note, as being 'the intelligence'.
This form of notation reminds me of having a capo on a guitar, where the TAB notation does not represent the notes, but the position of the fingers on the neck.

There was a [futile] discussion on the Finale list probably 15-20 years ago over whether or not to include the 'doubly diminished unison ascending' in a list of possible transpositions. By theory, transposing the note [PC] C, up a doubly diminished unison results in Cbb. This pointed out a division in how people represented and understood notes [pitch classes, pitch relationships]. One group argued that it is not possible because the result is this ascending interval is lower in pitch than the original.
In analytic terms, it is the confusion of the object, for the function. In tonality, in C major, Cb is not B. Even Chopin knew that. Cb is ^b1 and therefore has the melodic function that indicates descent. This is used in the so-called 'Polish sixth', the german sixth with a flattened tonic. See the C minor Prelude for the consequences of the Polish Sixth.

Best wishes


In reply to by kevin.austin@v…

Either I'm missing something, or maybe you are. If I create a custom key signature of "flat sign on second to bottom line" and apply it to treble clef, that line is the G line, the key signature correctly applies to all G's in any octave. If I then then create another custom key of "flat sign on bottom line" and apply it to bass clef, that line is also the G line, and that key signature also correctly to all G's in any octave.

Are you saying you don't want this - that you'd rather accidentals in a custom custom key signature apply only to the octave specified? That would be unusual, even within the already-unusual world of custom key signatures. But no unheard of, and this was discussed recently as a potential enhancement to support someday.

In reply to by Marc Sabatella

My first post described what to do. I am applying the k/s to the measure stack, not to only one staff.
Create a grand staff. Create a custom k/s of a flat on the second space. Drag the k/s to the staff. It is put on the second space of each staff. In the treble clef this is Ab. In the bass clef it is Cb.
In the treble clef, the A is Ab and C is C.
In the bass clef, C is Cb and A is A.
When I do the same thing with a standard signature, eg 2#, both staffs have the same notes.
As noted in the documentation, when the clef changes, the key signature doesn't transpose. I would call this 'idiosyncratic', partly because I have never encountered this in music notation.

In reply to by kevin.austin@v…

Well, right, this is exactly why we say you can't do it that way, you need to add to the staves separately. I thought you were saying you were having some sort of octave problem with doing it the correct way.

Yes, it might be nice in theory is adding a custom key signature to one staff could magically figure out the correct way to rewrite it for different clefs and different transpositions. The reality is, that's AI. After all, a second space key signature on treble clef is Ab, OK, fine, but then which line or space should be chosen for bass clef? For alto clef? How about when it is transposed for English horn? So while we could try to apply some heuristics and guess which lines/spaces you might want, in practice, chances are excellent you'd end up wanting to override our guesses anyhow.

In reply to by Marc Sabatella

I have been teaching basic music theory for more than 55 years. I teach these things in the 'naming of the notes' part of the first class. I represent the 12 pitch classes as a table. I introduce accidentals as naturals, flat [one semitone below a natural], sharp, double sharp etc etc. Which introduces the concept of enharmonics. I introduce intervals as having quantity and quality. Transposition is simply the rotation of the table.
Bartok for a period of time placed F# on the first space of the treble clef. It didn't catch on. The idea was based on the idea that the first accidental in the k/s appears on the line or space closest to the middle line. But people have exceptions to rules, and, as in spoken language, idiomatic forms. He probably borrowed the design from the tenor clef.
I would imagine that many professional composers would view MuseScore's idiosyncratic response to this simple formality as an error, or a design feature by someone who does not understand basic music 'theory' principles.

In reply to by kevin.austin@v…

I agree it's not ideal, but I'm not sure how you propose resolving the AI dilemma I described. Sure, it would be great if MuseScore could read our minds about which lines/psace we wanted the Ab to appear on in bass clef. But I am betting whichever way it guessed, half the world would say it was wrong.

As for basic theory, I've been teaching it almost as long *(only 40+ years, so my hat is off to you there), but of course, there is nothing "basic" about key signatures like this!

FWIW, I was just looking at an early edition of Bach WTC today and the G major prelude & fugue had the F# on the bottom space treble clef.

In reply to by kevin.austin@v…

That may well be. But also, traditional Western / Renaissance "modal" key signatures are only a small fraction of the possible uses for custom key signatures. Consider also the various non-Western scales, including Turkish scales, differently-tempered scales involving microtonal accidentals, etc. In order to really do a useful job of this is a pretty big undertaking. That said, I do think it would be worth doing someday.

In reply to by Jojo-Schmitz

fwiw, my personal opinion is that the treble clef shown while constructing a custom k/s shouldn't be a mere fake - a sharp on the top line for example would correspond to the absolute pitch F#, and the same custom signature could then be applied to a bass/alto/whatever clef and the sharp would appear on the appropriate place setting for the given clef, removing the need for multiple custom signatures for different clefs. Just an opinion, of course.

In reply to by worldwideweary

I do not see the accidental as a 'fake'. It functions. It is redefined by the clef. I believe I read that it also doesn't function for transposing instruments. Even Finale 1.0.1 got this right. I remember basic notation programs on 8-bit machines that didn't have a full encoding of octave equivalence, but they ran on 64k of memory so some things couldn't be easily implemented.

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