Re: Gis major key signature; Lily's key signature algorithm

[Hans Åberg]

> On 8 Feb 2018, at 01:59, Andrew Bernard <address@hidden> wrote:
> 
> Good question, and lots of good answers. Modern practice is to follow the 
> order of the circle of fifths. But that was not always the case. So, indeed, 
> depends on your historical context. Bach was writing in G sharp for the WTC, 
> and it was most certainly not intended to be enharmonic with A flat. But if 
> you look at18C manuscripts and JS Bach in particular there is wide variance 
> in how key signatures were done, and Bach often repeated the notes in the 
> signature, say having two c sharps, for reasons of his own (quite interesting 
> to see). Obviously his music teacher would fail him today. This was before 
> the age of standardization of everything.

The key signature was a way to simplify notation, not to actually indicate the 
key of the musical piece-there examples where they do not agree. There is a 
similar problem with Balkan music: it may be useful to write a nonstandard key 
to simplify the ornaments in LilyPond which cannot handle them as intervals, 
but if the ornaments are very chromatic, an appropriate key signature will not 
help.

> So I would discard advice about rewriting in A flat. G sharp is perfectly 
> good, even though the textbooks call it a 'theoretical key' - what they mean 
> is that it is hard to read when an alternative is available in an equal 
> tempered context. As Urs has said, there are plenty of valid musical contexts 
> for a key such as G sharp.

Blatter suggests to do it for a harp, because otherwise the harpist will have 
to do it, which adds to the performance cost.

> Since lilypond gives you the ability to change the ordering in the key 
> signature, you ahve complete freedom in what you do.
> 
> Nothing new here, but I just wanted to chime in on this interesting topic.

There is an additional complexity with microtonal accents, say like in Persian 
dastgah. It is indeed interesting what a good default might be.