lilypond-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Feature request (?): Horizontal spacing


From: Erik Sandberg
Subject: Re: Feature request (?): Horizontal spacing
Date: Sun, 13 Mar 2005 14:01:56 +0100
User-agent: Internet Messaging Program (IMP) 3.2.7

Citerar "Anthony W. Youngman" <address@hidden>:

> >I suppose a skilled human engraver would have a mental hierarchy of
> >which parameters to change first (from among staff spacing, note
> >spacing, staff sizes, etc), in order to "make it fit". The most lovely
> >thing would be if Lilypond was able to mimic that way of thinking, (i.e.
> >the first thought is "I want to make it fit on X pages", and everything
> >follows from there), rather than trial and error by the user.
> >
> Another item in the mental hierarchy ... I don't think lily does this, 
> but if someone does visit this and try to do it, then when you're 
> printing a part you want to try and put a page break by a multi-rest 
> bar. I've not actually met that many but it's really infuriating when 
> you have say four bars rest, two bars to play and then a tricky page 
> turn (or a turn, a couple of bars and a long rest).

There are a few more things a human would consider around line breaks, e.g.:
1. There should preferably be a MM rest at the break (as you said).
2. If this is not possible, the music should be easy to play around the break
3. It is bad to split a repeat over a page turn, especially if there is no rest
in the beginning or end of the repeat.
4. If it's not possible to find a good page turn, it's sometimes OK to just
ignore the problem, and let the musician use tape & have 3 pages visible.
5. When you have two facing pages, you can (and should) completely ignore page
turning issues between them. Considering (4), it is difficult to know if this
happens after odd or even pages.
6. Iff a piece is played Attaca, you want to avoid page turning before it.

I think it would be far too complex for a machine typesetter to take all these
things into account. Point (5) would have to be taken into account, which
basically would require that the user would state for each page whether it's a
left or right page. This would be about as much work as if he just would decide
manual page breaks manually (which is already possible). To me that feels like a
simpler and cleaner solution.

Erik





reply via email to

[Prev in Thread] Current Thread [Next in Thread]