Note spacing - voice collision in Dorico et al.?

[John Ruggero]

As a side note, you might get away with the best of both worlds with a little bit of fudge-factor by shifting the second 16th slightly right and the fourth 16th slightly left to even out those 16th runs:

Fudge-factoring was an important part of the plate engravers arsenal because of some built-in illogic in our notational system, as has been pointed out on this and other occasions. I wonder if the creators of Dorico have considered that and included approximation in their calculations?

[Knut]

I agree with MJCube.

I don't understand why Dorico would have its default for seconds set so wide because it exaggerates the unevenness of the 1/8th notes, when it should instead be hiding it. With this setting corrected, the Dorico layout would be the winner for me. Maybe Knut might want to show what that would look like.

Thing is, this has more to do with the font (i.e., the angle of the noteheads) than anything else. Here's the black noteheads from three different fonts (my own to the left, Bravura in the middle, and Maestro to the right) at the interval of a second, and with no horizontal space between them:

Skjermbilde 2016-10-26 kl. 01.08.48.png (24.04 KiB) Viewed 11019 times

Bravura's default setting for these noteheads is 1/16 spaces, which is next to nothing. After reviewing a couple of scores from major publishers, a real horizontal overlap between noteheads in these situations does not seem to be customary, or at least they are very marginal. Notehead angles do vary, however, which greatly influences the perceived gap.

After discussing this with Daniel, it appears that Dorico unfortunately does not allow negative values to be set for this option, at least not currently. Then again, Finale doesn't have a setting for this particular case at all, and will always set intervals of a second in different layers to 0.

[John Ruggero]

Knut:

Thing is, this has more to do with the font (i.e., the angle of the noteheads) than anything else.

This points up an interesting issue about seconds with fonts that have slightish note heads like Bravura and Maestro: they often need their stems aligned.

Here are aligned stems (and even slightly inset) from the Paderewski Edition of Chopin (New Etude no 2). All of the many seconds in this piece are engraved like this, The note heads in this edition are on the slight side.

Chopin stems.jpg (10.93 KiB) Viewed 11001 times

I looked through hand engraved literature and see a wide range of approach, from completely aligned stems like those in Finale to stems that touch on the outside edges like Dorico and some with even more gap than this, as well as variations within individual pieces depending on the situation. Generally there is no issue with gaps when the note heads are heavier.

Obviously, software aimed at experienced engravers needs to give them the tools to handle these different fonts and situations easily and flexibly.

[OCTO]

FWIW, LilyPond follows the same logic as Dorico. MuseScore seems to follow Finale's approach by spacing the lower staff notes more evenly and pushing the first pair of 16th notes together. I'm not a computer with Sibelius or SCORE to see what they do. Any takers?

This is the default output in Sibelius. It is definitely heavily malformed. Maybe that is one of the reasons I never liked Sibelius output.

shot 5.png (47.67 KiB) Viewed 10996 times

As a side note, you might get away with the best of both worlds with a little bit of fudge-factor by shifting the second 16th slightly right and the fourth 16th slightly left to even out those 16th runs:
MuseScore 2_ note-spacing-test.png

Oh, that is very interesting! This "fudge-factor" (I never heard it before) could be good to have implemented in all notation applications. It seems that there is really a lot left to implement in our current softwares' state!

Keep in mind that Finale, Sibelius and MuseScore don't justify that uneven 16-notes even without the third voice in 8-notes. For me it doesn't look good.

[John Ruggero]

Keep in mind that Finale, Sibelius and MuseScore don't justify that uneven 16-notes even without the third voice in 8-notes. For me it doesn't look good.

The fact that there are no accidentals involved in the example makes the unevenness stand out even more. Accidentals can distract and could be fudge-factors. In a case like this one without accidentals, the fastest notes look best perfectly even with the others fit against them using a little fudging to mitigate their unevenness.

[tisimst]

I stumbled something curious today as I was dabbling in Dorico. It turns out that the horizontal spacing that Dorico shows at its (A) "resting" state may be coincidentally so, because if you expand the measure's length a bit (B), you see that it begins to show a different picture not unlike Finale, Sibelius, and MuseScore, spacing the lower staff 8th-notes more evenly instead of the upper staff's 16th-note runs:

Dorico-true-note-spacing.png (39.06 KiB) Viewed 10953 times

I admit that I was quite surprised to see this so I double checked LilyPond again with different measure lengths to see if it did the same thing as Dorico, but sure enough it stayed consistent through each one, spacing according to the upper staff 16th notes:

LilyPond-note-spacing.png (76.37 KiB) Viewed 10953 times

[OCTO]

Whoh - that is very interesting!
I noticed the same with Finale and Sibelius. So, basically, Dorico doesn't differ here?
Here is Finale, the compressed spacing:

shot 1.jpg (25.31 KiB) Viewed 10952 times

[Knut]

Thanks for reporting this, tisimst. Dorico 0 – Lilypond 1. I've forwarded the report to the Dorico forum.

[John Ruggero]

I think that the LilyPond LH spacing would look even better if the second 1/8th in each pair were moved slightly to the left.

There is an optical illusion created between the 16ths and the 8th because the stems are on different sides of the note heads. This pulls the eye in a diagonal as if the lower note were too far to the right, even though the note heads are completely centered. See A.

This illusion doesn't exist when the stems are on the same side of the note. See B.

The illusion would allow the second eighth to be pulled left to help even out the LH spacing when the third voice is introduced. See C

Spacing.jpg (69.09 KiB) Viewed 10932 times

[Knut]

Interesting aspect, John.

I've seen this done from time to time in manual engraving (although, not as large a compensation as you propose), but have never bothered to spend time worrying about it myself. I'm wondering, though, if the effect of this isn't also largely dependent on the distance between the staves? Seems to me, there's less of a need for this compensation if the notes are closer together than farther apart.

I suspect I'll find the compensation in your screenshot to be too much, but it's har to tell with the vertical guidelines going through the staves. Could you add a screenshot without the guidelines to your post?