Page Set-up

• May 27, 2022 - 16:25

Moving staves/systems is incredibly convoluted and usually does not work, especially on a last page where there are only say 2 staves. The distance between them cannot be adjusted to match the other pages. A system should be able to be moved in total with a click and hold. There should not be further menus/math formulas/spacers, etc. That's PC computer type of logic. Use MAC logic when creating items. I don't know the coding difference, etc., but the final product will be TONS better as far as ease. While this program is very good generally, I cannot send out music when a feature doesn't work. I've searched the NET and Musescore info and the only thing I see is either the spacers (doesn't work for large distances between systems), or page items (convoluted and not easily understandable, etc). Use a clear MENU item that makes sense.

I'm always dumbfounded with these notation programs that don't respond in terms of ease. Please see what you can do with this issue while I'm off looking for a different program.

I've included a screenshot that shows the system difference between pages three and four.

Attachment Size
Screen Shot 2022-05-27 at 8.25.01 AM.png 66.63 KB

Comments

Distribute your staves more evenly accros pages, by using page breaks for example
Or use a spacer at the last system
Or use Format > Style > Page > Disable vertical alignment of staves

Moving a system by manually dragging is definitely something that might become possible in a future release.

But, consider: while this might seem to make sense if you only have one or two systems to adjust, it's incredibly impractical for an entire score that might have dozens or hundreds of systems. Especially when one considers, what happens if you then add measures, change staff size, or otherwise modify something that would then invaldiate that work.

That's why having simple numeric control that guide the behavior globally is actually superior in the long run.

The issue here is that there is a bug preventing "max system distance" from working correctly with vertical justification enabled. While one could try to workaround it by disabling vertical justification, that produces much worse results everything else in your score. So ut's better to avoid that except in the specific case of single-staff scores (ewg, lead sheets or instrument parts), where it is indeed safe and effective.

But even so, the best solution really is to add breaks so you don't have this problem in the first place. It's generally pretty unprofessional to leave a whole bunch of empty space on any page, whether it's spread between systems or bunched up at the bottom.

In cases where having extra space on the last page is unavoidable, then the workaround for the max system distance bug is to add a spacer below the last staff. It actually ends up being less effort than dragging each system individually even if that were possible., and unlikely disabling vertical justification, it doesn't have negative side effects elsewhere.

You say it doesn't work, but my guess is you aren't adding the spacer below the last system and are instead trying to do something else. The spacer below the last system absolutely works. If you continue to have difficulty, just attach your score (as opposed to just a picture) and we can assist further.

In reply to by Jojo-Schmitz

As far as I can recall, the bug was fixed for master before 3.6.2 was even released, but as part of a larger / riskier change that wouldn't have been suitable for 3.6.2 or even 3.6.3. I don't think there was ever an issue for this problem specifically since the fix was already in place before the bug was ever reported - I think originally on Telegram, and then of course over and over and over and over on the forum. And I also don't know if a 3.x version of the fix was ever produced. Presumably you could test using your "special" 3.x build - just create a piano score with two systems on the second page and justification enabled, then turn the max system distance down gradually and see if the score responds or not.

See for instance this thread, although this was already a few months later - https://musescore.org/en/node/323865

Part of the issue in tracking this is, I don't think anyone ever pinpointed exactly which PR fixed this for master. But I can definitely verify it is fixed there. I think it was a small part of a larger PR, though.

BTW, as far as I am concerned there are two bugs here in 3.6.2. The main one is that the max system distance setting is ignored with justification on. The other smaller bug is that the default value of 32 is way too large for cases like this. This is why the hack workaround of disabling justification "works" - it allows the max system distance to be honored, and it changes the default to a much more reasonable 15sp. More reasonable for single and double staff scores, anyhow - 32sp may well make sense for scores of 3-5 staves per system and two systems per page where you actually do want to spread them to fill the page. But it's way too big for one or two staff scores - lead sheets, instrumental parts, piano scores, SATB closed score, etc. So ideally, the solution would be to set that 15 or whatever. But there's where the main bug prevents that from working. The hack of disabling justification "fixes" that, but at the expense of, well, disabling justification. Meaning staff spacing in the rest of your score goes back to pre-3.6 jankiness.

In reply to by Marc Sabatella

Seems indeed fixed in 3.x, except for changed defaults for "Max. system distance" (currently 32sp) or "Max. staff distance" (currently 20sp). Which settings would be recommended there? 15sp for both or just for "Max. system distance" and a smaller value for "Max. staff distance"? Would most probably be needed for master too, which uses the same defaults still.

In reply to by Jojo-Schmitz

I’m not sure what makes sense, but the way I’d come up with a figure is create scores with systems of 1 staff, 2 staves, 3 staves, 4 staves, and 5 staves, put two systems on a page, and eyeball what seems to look best in each case, then see which combination of max system and page fill covers most of these best. It’s going to be more about five sing the best compromise than finding the One True Value.

Realistically, I expect that’s the sort of thing @oktophonie would wanting to make the call on.

In reply to by Jojo-Schmitz

Making these values too small will cause problems on some "normal" pages since the algorithm can not add enough space to make sure the last system is touching the bottom margin. This is the reason why these values are pretty large.
The basic issue is, what too do on the last page (or before a page break) and there is too much space left. Nicest would be to use a similar spread as the previous page(s) but what if there is no previous page? This is still an open question.

In reply to by njvdberg

There probably is no one size fits all right answer, not unless the algorithm gains a new parameter or two anyhow.

The fact that MU4 fixes the bug that causes max system distance to be ignored is great, as it means the workaround in cases where the default value is either too big or too small is to simply change that value.

Still, it would be good for that default value to be reasonable in the largest number of cases. Right now, I can pretty easily characterize the case where it is too big even in MU4 where the setting works correctly: it's scores with one staff per system (eg, leads sheets, or instrument parts generated from an ensemble score), or two systems on the last page. This produces spacing on that last page that is noticeably wider than on the preceding pages, and yet, it doesn't reach the bottom margin (nor should it), so the extra space doesn't really accomplish anything except make the last page look inconsistent:

Screenshot 2022-05-30 9.18.08 PM.png

My math tells me that if you normally fit ten systems on a page in your instrumental parts, then exactly 10% of these parts will have two systems on the last page and thus be "off" in this way. It's nowhere near as bad as MU3 where a) the spacing is much worse in this case, and b) it also often looks bad with 3 or 4 systems. But it still makes me wonder about the frequency of the situations where 32 sp is needed.

On the other hand, the distance isn't actually 32 sp here, because the page fill is kicking in and preventing the page from filling. I actually think the default may be too low, as many of my experiments ended up with ragged bottom margins where I might have expected the page to be filled. But, if we turn this value up, then the 32 sp for max system distance becomes increasingly problematic.

I don't know, I certainly don't claim to have the answers. And I'm very happy MU4 doesn't have the bugging that's been plaguing MU3, as even the "bad" case here isn't that bad. But it's still noticeably less good than with justification disabled.

A really simple optimization that might help is to automatically disable justification for scores with only a single staff. It wasn't going to have any particular beneficial effect in that case, and by disabling it, we get back to the old 15 sp max system distance that worked well for these cases. This automatically solves the case of lead sheets and instrument parts without affecting anything else.

In reply to by Marc Sabatella

"The fact that MU4 fixes the bug that causes max system distance to be ignored is great, as it means the workaround in cases where the default value is either too big or too small is to simply change that value."
I wasn't aware this issue exists in MU4 but will have a look.

In the meantime another solution came into my mind. The spreading algorithm does two things. First it spreads the space between systems and staves so this space becomes as evenly as possible. In a next stage, when there is still space left on the page, it adds extra space so the lowest staff will hit the bottom margin.
When this last step is disable for the last page (or a page having a page break), this might solve the issue of the strange/bad spread on the last page.

In reply to by njvdberg

To be clear, the actual bug is fixed for MU4 - there is no issue with max system distance being ignored. it's honored, at least to the extent that the max page fill distance setting allows.

When I talk about a workaround, I mean, for the cases where 32 sp is simply too large a default setting, which as I said, is all lead sheets and all instrumental parts. Because the actual spread is limited by page fill distance, you never actually get 32 sp of space between systems - the most you get is around 20 sp or so. But that's still too much in these cases - the value of 15 sp that is used in the non-justification case is more appropriate for these scores.

So my question remains, what sort of scores do require a max system distance of 32 sp in order to justify correctly? Seeing an actual example would help in trying to find a way to tweak the algorithm to handle both cases well.

I'm not fully understanding your proposed solution, but it sounds like it has potential indeed!

In reply to by Marc Sabatella

Every score that has more systems on all but the last page (and at least 2 systems there). So a leadsheet with 10 systems on the 2nd to last page and just 2 - 5 on the last, or a piano score with 5 on the 2nd to last and 2 on the last, same for closed score SATB, or a voice + piano with 4 systems on the 2nd to last and 2 on the last, etc.

In reply to by Jojo-Schmitz

Not sure if you're responding to something else, but to be clear: yes, the cases where large max system distances are not good are already very clear. What I am still not understanding are which cases exist where the large values are good. Presumably, some situation where you have only two systems of 3-4 staves and you do want them to fill the page, might require a system distance of 32 sp to accomplish. But the cases where I tried, don't actually fill the page, because the max page fill limits it.

In reply to by Marc Sabatella

The max page fill won't limit the fill when the staves within a system are spread first, e.g. when in one system there is quite some space between two staves to prevent a collision between the high elements on a staff and the low elements on the staff above. In this case the staves will be spread first and the max page fill won't kick in. When the staves are distributed evenly in this first phase and there is still space left this remaining space is distributed over the gaps between the staves, however limited to max page fill.
Especially in cases where there is an initial large gap between two staves, a larger max system distance is required to prevent the space between the staves within the system becomes much larger than the space between the systems. That's the reason why the default of the max system space is pretty large.

I agree the algorithm doesn't behave nicely on the last page (or before a page break) but I'm not sure decreasing the max system distance is the right way to go. Disabling the last step of the algorithm (trying to fill the remaining space on the page) in such cases might give a better result by default. I will experiment with this.

In reply to by njvdberg

Could well be; I definitely don't claim to have all of the answers - just maybe some of the questions.

I still think it would be very useful to see an actual real world score for which values of max system distance this large actually needed. It would help the rest of us understand how things work better, and either shut us up, or allow us to make more relevant suggestions :-)

In reply to by Marc Sabatella

I found an example in which, for me, a Max system distance of 32sp is better than a smaller value (like 16).
This is of page 2 from Linkin Park Medley and shows 3 systems. The right picture is with a Max system distance of 16sp, the right picture uses the default of 32sp.
q.png
For me the right image is easier to read since the systems are better separated from each other.

In reply to by njvdberg

I agree the image on the right looks better. Although, I'm confused, because as far as I can tell from the picture, the actual system distance is nowhere near 32sp - seems to be more like around 20sp (as measured by a sheet of notebook paper held up to my monitor).

Having the actual score would be useful too in discussing further. But, already I can see why I was having trouble coming up with good examples - I was thinking in terms of systems of a more consistent height. In other words, systems of consistently 2 staves, or consistent 3, or 4. Also, I was using more default staff space values; this appears to be significantly smaller. And I was thinking more about whether the last system would reach the bottom margin, not about how the intra-system would be affected.

So what you have is a case of three systems on a page with and total of twelve staves between them but actually distributed as 2+5+5, and a staff size of probably around 75% normal. In both cases, the last system reaches the bottom margin, but allowing a larger max system distance lets the space be distributed better - more space between systems, less within.

Makes me wonder if somehow the "factor for distance between systems" here isn't a factor also.

In reply to by Marc Sabatella

the actual system distance is nowhere near 32sp
This could be a big but also it is possible we are running into limit set by other parameters. it's difficult to experiment since MU4 seems to have some performance issues with "full" scores, certainly compared with MU3. I have to find another way to easily test these kind of scores.

Having the actual score would be useful too in discussing further.
I got this score for testing and am not sure whether I can freely distribute it. I'm looking at a good example at musescore.com.

I was using more default staff space values; this appears to be significantly smaller
Yes, the values are smaller when using vertical staff spreading. Initial spacing is smaller to give the algorithm more space to work. Also empty staves don't give a good result because everything is placed as closed as possible and the algorithm is very happy, there is not much to do. By adding some elements which causes a different initial spacing of staves gives an idea what the algorithm is doing and gives a decent result.

In reply to by Marc Sabatella

For me, the problem with the last page (please not, with last page I also mean before page break) is the last stage of the staff spreading algorithm where it tries to fill the remaining space on a page. This doesn't make much sense in cases as shown (like leadsheets).
Therefor I've been playing with an idea to optimize the algorithm which tries to decide whether it makes sense to fill the last page or not. The first trial looks like this:
Screenshot_20220607_185022.png
Now the the staves on the second page are kept closer to each other.
Here I used a Last page fill threshold, similar to the Last system fill threshold but I need more experimenting to see this give good result is all (or at least the majority of) cases. Also I'm not sure a plain page fill threshold is enough, it might depend on number of staves in a system. Will be continued.

In reply to by njvdberg

It's better, although now too tight on the second page. You don't want no spread on these underfull pages, you just want to limit the spread to not look out of proportion with other pages. After all, the first page got to enjoy some spread, and would enjoy even more if it had only 8 or 9 systems as is common.

So naively I might want two different max settings - regular "max system distance" versus "max system distance for pages that weren't going to reach the bottom margin anyhow", and just keep 15sp as the default for the latter. I think that looking it that way avoids some of the complication of needing threshold settings - it's black and white, either the page was going to reach the margin or it wasn't. But, if you do stick to the threshold idea, still, there does need to be some spread, and it makes sense this should be a configurable setting.

Results might not be perfect with this approach, but it seems like it would handle most of the cases I can think of. And the goal isn't perfectly, it's just better default behavior more often.

In reply to by Marc Sabatella

you just want to limit the spread to not look out of proportion with other pages
This is a tricky one since we have no idea what spacing previous page(s) will have. Also I'm not sure this what we want, the previous page might have a lot of extra spacing because some staves already had a large initial spacing. And what to do if this and the previous page contains systems with some hidden staves.

I might want two different max settings -...r
This might be something to explore. But, after seen that many different scores, my gut feeling tell me this can cause strange effect in some scores. But it might be worthwhile to investigate.

In reply to by njvdberg

Right, I wouldn't suggest literally trying using the previous page's spread in an algorithm (besides, this might be the only page in the score - a short lead sheet). I'm pointing out that this is the visual cue I am using in this instance to tell me, page two is too crowded as shown. So I'm fully accepting that there needs to be some spread on these underfill pages. There just needs to be a way to limit it in cases like this.

Another possibility I thought of is to somehow scale the max system distance with the size of the system. The taller the system, the more space you can put around it and not look bad. But, I'm not sure there is a simple way to quantify that. The amount of space that I'd tolerate for a grand staff isn't that much more than the amount I'd tolerate for a single staff, for instance. As with the idea of basing spacing on previous pages, it might be more useful as a gut feel guiding principle than as an algorithm.

Looking at all the comments in this thread I wonder if someone could publish the actual algorithm that is used when enable vertical justification is chosen.
Also I find the various options in the Style -> page dialogue box could be better arranged. Attached a screen shot trying to show the relationship between the dialogue box and the page elements. Happy to share the graphic so others can improve it.

Attachment Size
musescore_style_page_properties.jpg 1.36 MB

Do you still have an unanswered question? Please log in first to post your question.