Re: Automatic vertical spacing when *some* values are fixed

[Urs Liska]

Hi Torsten,

thank you for these experiments and thoughts which will probably bring 
me forward. Although they essentially miss the point - probably I should 
have copied more information from the earlier thread.

Am 13.02.2018 um 20:45 schrieb Torsten Hämmerle:
> Hi Urs,
>
> With the top and bottom fixed, the only flexibility left lies in varying the
> distances between systems and staves.
> Unfortunately, the system-system-spacing has a default minimum distance of 8
> (between the systems) and squeezing three systems into one page seems to be
> better than stretching them too far apart.
>
> I'd play around with system-system-spacing.minimum-distance.
>
> 1st attempt: When setting minimum-distance to 11, thus forcing a wider
> distance between systems, LilyPond still complains about compressing (and
> it's even worse).
> Now, there is much more compression within the stave groups. No, that's not
> good, either...
>
> Even pushing up system-system-spacing.minimum-distance to 12 still squeezes
> three systems on page 1, but only two on page 2 (because of the p standing
> out in the last line).
> That's just (!) because of the p in Vc bar 14., one might say.
> So a tiny difference in hairpins, stems, ... makes a huge difference in the
> resulting layout (three systems or two systems per page).
> This looks awful and very unbalanced. But there's nothing in between: it's
> either two or three systems per page.
>
> Setting system-system-spacing.minimum-distance to 13, finally pushes the
> systems so far apart that only two systems (but consistently at last) fit on
> one page.
>
> Doesn't look too good, either... No, none of the solutions really works.

The example file I gave is not the challenge I need to solve, it's just 
an example where the problem becomes very visible. My actual problem is 
that when I specify the margins the way I do (by setting stretchability 
of a number of variables to zero) LilyPond makes different decisions 
regarding page breaking than with the usual top-margin/bottom-margin layout.


> *The actual problem*
> The space available on the page does not at all go well with the system size
> for a string quartet.
> Just reduce the stave size and everything will fit neatly...
>
> Sometimes, I just use one of the system-count, systems-per-page,
> min-systems-per-page oder max-systems-per-page paper variables to achieve a
> uniform number of systems on all pages.

I know that the music can't be properly printed with the given 
combination of paper size, margins and staffsize. But I want LilyPond to 
prefer the solution with two systems per page (in this example) because 
the compressed variant is *really* inacceptable.

Using further paper variables is of no help either. Of course, if I'd 
want to engrave that certain piece I'd experiment along these lines too, 
but that's not what I'm after.
I'm working on a LaTeX package that includes scores in text documents 
and matches the layout as good as possible.
While it is trivial to set the margins to exactly match the LaTeX ones I 
don't really like that solution because the pages will look uneven when 
the top margin does not match the stafflines but the top-most element. 
What I would like to achieve is something like this: 
https://user-images.githubusercontent.com/1812148/35498550-1780de08-04cf-11e8-9553-2bb1aabbe33e.png

And actually it was possible to achieve it by fixing the margins of the 
staff, but it turns out that this will often result in absolutely 
inacceptable global spacing of the score.

Does all that finally mean I'll have to experiment with all the elements 
of the paper variables until I find something that promises to work 
generically (i.e. when I don't know the actual music and page layout 
beforehand)?

Best
Urs

> HTH
> Torsten
>
>
>
>
>
> --
> Sent from: http://lilypond.1069038.n5.nabble.com/User-f3.html
>
> _______________________________________________
> lilypond-user mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/lilypond-user