On 9 Aug 2009, at 21:07, Bernard Hurley wrote:
The default rule is really that one should have a minimum number of
time
signatures if the repeat construct is expanded*).
That seems like a reasonable rule. But a quick look through my pile
of full scores (Mostly Dover reprints of 18th and 19th century
scores) shows it is not always obeyed. I.e. there are plenty of
cases where the a time signature change is placed after the repeat
sign even though there is no time change inside the repeat.
If one of the alternatives of the repeat end in a meter different from
what it begins, then if the expansion rule should prevail, there must
be a time signature at the beginning of the repeat after the |:. And
it is then unnecessary to have one extra before.
But not even that was followed in the Bulgarian score. It looks like:
11/16 |: ... | 8/16 ... | 11/16 ... | 8/16 ... :|
As a composer/arranger I always put the time change after the repeat
sign! Rather than talk about the right or wrong way of doing this it
should just be a user's option.
I figure LilyPond might support what people want to typeset.
Hans
_______________________________________________
bug-lilypond mailing list
bug-lilypond@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-lilypond
