Hi Don,

thanks for your exhaustive and as always helpful explanations. My experience is also that with some experimentation one gets finally a decent solution. Obviously I managed to create a vshrink situation. Maybe one could put a warning into the next version documentation, that under certain conditions AI does not work as expected.

Regards,
Dieter


Am 22.08.2016 um 09:27 schrieb Don Simons:

I’m afraid I still cannot confidently provide a checklist of everything that affects PMX’s internal calculation of \interstaff. It’s very complicated, I worked it out a very long time ago, and I didn’t document or comment it very well in the code. If anyone cares to look in the FORTRAN, be my guest. There are some intriguing comments at line 19170 of pmxab.for ver 2.74, but the operations are scattered throughout the code.

I will however offer some comments about Dieter’s issue and about vertical spacing in general.

First, I’m virtually certain that using npages=0 had nothing to do with it. In all my tests on Dieter’s 2-system example, I’ve gotten exactly the same results with (npages,nsystems) = (0,4) as with (1,2).

When PMX sets up a score, it computes a value of \interstaff based on how much vertical space is used by individual staves, titles, and some other items. It tries to set \interstaff so that when TeX spreads things out over the full page height (after \eject) it looks OK. If there is lots of blank vertical space, as there would be with just one or two systems in a normal page, then the computed \interstaff would come out unpleasantly big, and PMX goes into “vshrink” mode, setting \interstaff to 10 and using \vfill before \eject, thereby grouping everything at the top of the page. (I think the cutoff \interstaff is 20 but there’s some uncertainty there). In vshrink mode, AI does not seem to have any effect. One way to check if PMX is using vshrink mode is to look in the tex file and see if \interstaff is 10. If this happens and you still want to adjust vertical spacing, you have to get out of vshrink mode. There are two ways to do that. One is to issue the Av command. Then you will get things spread out more evenly over the full page height. \interstaff may come out bigger than you like, but now you can change it with AI. But if you only have two systems, you might also want less vertical space between them. I don’t know how to make PMX do that directly; you’d have to resort to inline tex, possibly inserting a \vfill at the end of the page before \eject. A second way to avoid vshrink is to make the page height smaller. There will always be some smaller page height that will coerce PMX not to go into “vshrink” mode. Then you can use AI, and you’ll also have reduced the vertical space between staves.

PMX’s vertical spacing decisions should work fine, with no tinkering, for just about any ordinary music without lyrics, and you shouldn’t have to worry about anything in the previous paragraph. PMX’s automatic choice to go into vshrink mode will make sparse pages look just fine, with a few systems grouped at the top. The problem Dieter came upon was caused by the need to make room for lyrics in the space between the two staves in a system, on a page with just two systems of two staves each. And even that problem went away when he added a title, instrument, and author, because those used enough extra vertical space to avoid going into “vshrink” mode.

--Don Simons

*From:*Don Simons [mailto:dsim...@roadrunner.com]
*Sent:* Saturday, August 20, 2016 1:58 PM
*To:* 'Werner Icking Music Archive' <tex-music@tug.org>
*Subject:* RE: [Tex-music] Vertical spacing: AI-Option in PMX

Dieter and Andre—

This issue is more complicated than it seems, and I’m afraid it’s been so long since I did the programming that it going to take me some more time to sort it out. The good news is that if you’re making a score, it’s very easy to experiment and you can almost always find a set of inputs that give you the vertical spacing you want.

One thing I’ve noticed so far is this. Andre’s suggestion about changing (npages,nsystems) from (0,4) to (1,4) does lead to different vertical spacing, but it also leads to 4 systems on the page rather than 2, and that in itself completely changes the game. I don’t think npages=0 by itself changes the way AI operates, but it certainly can (and does in the example) influence the number of systems, and if it does change the number of systems, then of course it will change the vertical spacing.

Another thing I’ve noticed is that if you remove the instrument, title, and composer from Dieter’s latest file, and don’t change anything else, you get a major change in the vertical spacing.

I’ll report back when I’ve had a chance to review this more thoroughly.

--Don Simons

*From:*TeX-Music [mailto:tex-music-boun...@tug.org] *On Behalf Of *Andre Van Ryckeghem
*Sent:* Friday, August 19, 2016 8:48 PM
*To:* Werner Icking Music Archive <tex-music@tug.org <mailto:tex-music@tug.org>>
*Subject:* Re: [Tex-music] Vertical spacing: AI-Option in PMX

I believe the

----

% npages nsystems musicsize fracindent

0  <----

4

20

0.05

-----

does not obey the rules anymore.

I tried

----

% npages nsystems musicsize fracindent

1

4

20

0.05

-----

with

AbepI.6 and AbepI1.0 and all seems to be normal

Andre

*From:*Dieter <mailto:d.gloet...@web.de>

*Sent:*Friday, August 19, 2016 3:08 PM

*To:*Werner Icking Music Archive <mailto:tex-music@tug.org>

*Subject:*[Tex-music] Vertical spacing: AI-Option in PMX

Hi,

I do have trouble in working with the AI Option of PMX for interstaff spacing.

According to Noack:

========================================================================================
I understand that \interstaff will be multiplied by x, when I use this option. But what I find is for a certain configuration, that the staff distance in a piano piece becomes bigger, when I set x to 0.5 insted of 1.0.

x = 0.5


x = 1.0

I am quite happy with the x =0.5 solution, but I would be happy to gain some understanding of what is happening.

Regards, Dieter

------------------------------------------------------------------------

-------------------------------
TeX-music@tug.org <mailto:TeX-music@tug.org> mailing list
If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music



-------------------------------
TeX-music@tug.org mailing list
If you want to unsubscribe or look at the archives, go to 
http://tug.org/mailman/listinfo/tex-music


-------------------------------
TeX-music@tug.org mailing list
If you want to unsubscribe or look at the archives, go to 
http://tug.org/mailman/listinfo/tex-music

Reply via email to