Re: modern accidentals rule

[Hans Aberg]

On 10 Nov 2009, at 07:25, David Kastrup wrote:

> > > BTW, I have found a strange behaviour of "modern" accidentals rule:
> > > LilyPond considers volta alternatives as "previous" measure. Is this
> > > correct? (Note the natural sign at the 2nd and 3rd alternatives, not
> > > related to the previous _played_ measure.)
> >
> > I do not know.  I am not an expert on the different accidental rules,
> > nor on the standard engraving practices for accidentals in
> > alternatives.  But I would expect that the last measure of the repeat
> > should serve as the previous measure for each of the alternatives.
>
> And it should serve as the previous measure for starting ties and  
> slurs
> and key and meter changes.  And the first measure of the repeat should
> serve as the following measure for unfinished ties and slurs for all  
> of
> the alternatives except the last one (which means that tie and slur
> directions should match and the tie/slur should extend to the bar line
> rather than trigger an error).

One way to express it, is that it should be fit together when  
expanded: repeats is just a shorthand for the expanded version.

When writing down some Balkan music, I have found it might be suitable  
with alternatives in other positions, like the beginning and the  
middle. One use is to indicate variations in the performance.

  Hans