Double stems

• Sep 10, 2014 - 02:06

Since starting this test suite on double stems months ago, which is based on pages 49-50/52-54 of 'Behind Bars', Marc has made tremendous improvements to the layout code.

However, there are a few instances in which things aren't quite right (bar 6), or perhaps a different decision was taken (bars 2 and 3).

I just wanted to ask Marc for his take on this.

Attachment Size
Double stems.mscz 3.38 KB


Thanks for the comments!

Actually, it turns out there is a good explanation for most of this.

In bar 6, it appears you manually mirrored some of the note heads when you created the file in 1.3 and that is why they remain incorrect when imported into 2.0. I can see that even in 1.3 just looking at Note Properties. If I recreate that same example from scratch,either in 2.0 or in 1.3, and it looks fine when I load the score into 2.0 (either beta or current development build) it looks right.

Bar 2 is a very straightforward application of Gould's basic advice on p. 49: align two adjacent-note chords so the notes on the correct side align vertically. She gives a special case exception for adjacent-note chords that are also themselves separated by a second, and that is what this example demonstrates. Obviously we don't handle that exception automatically. Looking at the code, it *might* turn out to be a very simple change, or it might turn out to be next to impossible without a major rewrite. Feel free to file this as a minor bug, and if it's not too hard, I will take a look.

Bar 3 is the same story, except for the fourth chord. Here it took a few minutes to figure out what is happening. It's because you created this whole passage as one measure, so there are multiple voices, meaning that chord is actually upstem (which you can't see because there is no stem). That's the proper arrangement for an upstem chord of that configuration. Had you created the first part of this as a separate measure requiring only a single voice - or if you flip the stem down manually - it would have come out as in Gould.

Measure 21: because horizontal spacers exist for a variety of purposes, it's hard to say definitively if there should be courtesy time signatures before them. My gut feel is that it seems better create them and then be able to turn them off than to not create them and then leave the user with no choice.

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