Not how I’d spell it, but the F# seems logical enough to avoid G and Gb. Of course, if I input the Tpt 2 line alone, I get this, which is horizontally perfect:
It’s only when I input into multiple staves at once that the Tpt 2 line has a C#. If it’s not coming from a vertical structure or horizontal, then where is that C# coming from? I’m just trying to understand what possible rationale is being used for the C# so I know to look out for it.
I think we are at the point encountered in the situation of the “pig that can walk on its hind legs.”
It’s not that it does it well that is notable, but that it can do it at all.
Needless to say, when Dorico does not choose the desired accidental, it is disappointing; but so many users will have so many different expectations, that asking a computer to decide instantaneously on matters that many humans might have to think about for a moment before proceeding is expecting a lot.
No doubt, if the programmers find a way to have the computer guess more accurately in the future, they will add it.
I generally get along with the default harmonic choices quite well. Sometimes I have to change a note, which is simple enough. The more I work with it the more I understand how it “thinks” and what enharmonics to expect. I certainly have fewer wrong ones to correct than I had with Sibelius or Finale.
That was really my issue with the C# in the multi-staff input gif above. Perhaps there’s a bug with multi-staff input , but I couldn’t come up with any possible rationale for a C#, especially when it doesn’t occur with single-staff chordal or single line input. If I could understand how Dorico was “thinking” here, I could anticipate it, but that enharmonic simply seems wrong by any sort of thought process.
I think this is actually the flipside of the original problem. If you input just the chord G-Db-Gb onto a piano staff in a G flat major key signature then it will spell it as G-C#-F#, just as in your extended caret case. So I think the problem with the extended caret is not so much that the initial spelling is worse than normal, but that it can’t go on to use the local context to correct the spelling after the fact.
(As to why it starts off as a C#, I’m not sure of that, though it doesn’t happen in the equivalent case in C major. At a guess it’s because it wants to avoid augmented/diminished intervals but isn’t able to use double-flats in the first instance, which prevents it starting with Abb-Db-Gb. That is just a guess though.)
I have had a similar kind of problem. I input a chord using a MIDI keyboard, and Dorico chooses a spelling that doesn’t match the context, so I use the respell command [shift-I G#min]. Dorico gets most of the notes right, but weirdly spells one of the notes as Gb! In the key of G# minor, F# is a scale tone but there is no such thing as Gb!
While I am on this topic, why is it that the respell command requires me to type G#min, whereas the key signature tool requires typing only g#? Shouldn’t these both accept g#?
