On Sunday, 17 January 1999, John Sankey writes:
> I believe that the following are sound algorithms for barring,
> note-stemming and correcting enharmonics for 95% of baroque
> harpsichord music:
>
> 1. given a pulse (e.g. 4 in 4/4, 4. in 3/8),
> put [] around every pulse group excepting
> a) groups that contain a rest/skip,
> b) groups that only contain one note, or
> c) groups containing [ or * (i.e. irregular constructs)
Although you make a good point (below) to do this in mi2mu,
it would be a shame not to do this in Lily, because specifying
all the beams with [] can be a pain.
But now that you've made me think of it, perhaps it's not too
difficult to make an Auto_beam_engraver at all.
Until we have autobeaming in Lily, you might want to use a trick
like this:
melody=\type Voice=one\notes\relative c'{
c8 d e f
c d e f
}
beams=\type Voice=one\notes{
[s2] [s2]
}
\score{
\type Staff <
\melody
\beams
>
\paper{
linewidth=-1.;
}
}
> 2. stemup the notes of the higher-pitched voice/track,
> stemdown those of the lower-pitched voice/track,
> ignore cases where one voice/track straddles another within a
> pulse group.
Is this too much work to do by hand? I'm not sure that i understand
what you want.
> 3. a) in all keys with flats in the key signature, change
> ais to bes, dis to ees, fis to ges, gis to aes
Does mi2mu always chooses the sharp variant? Sounds almost like a
bug, it would always be better to choose flats for flat keys, i suppose.
I'll have a look.
> b) in all flat keys except d-minor, change
> cis to des
Why is this?
> Of course, there are exceptions to every rule in music! So, it seems
> to me that the best place to do this sort of thing is in mi2mu, so
> that the few exceptions can be corrected easily using standard
> Mudela.
>
> 3 is simple enough that I've been able to write a sequential Basic
> program to do it, but 1 and 2 require a map of everything in a bar,
> which is not so simple. But, it has to be done somehow, with 550
> sonatas to go (not to mention maybe J.S.Bach and Francois Couperin
> after that).
>
> Would you have time to consider adding any of them to mi2mu?
Although I hope you'll succeed with mi2mu, i consider all time
put into mi2mu as a loss. Because you'll always have to edit
the generated .ly, i think that typing mudela is a lot faster.
And i haven't seen any real positive feedback on mi2mu.
Jan.
Jan Nieuwenhuizen <[EMAIL PROTECTED]> | GNU LilyPond - The music typesetter
http://www.xs4all.nl/~jantien | http://www.xs4all.nl/~jantien/lilypond
