Horizontal alignment of chords with offset notes

Discuss the rules of notation, standard notation practices, efficient notation practices and graphic design.
Knut
Posts: 864
Joined: 05 Oct 2015, 18:07
Location: Oslo, Norway

Horizontal alignment of chords with offset notes

Post by Knut » 06 May 2017, 17:02

Proper horizontal spacing of chords with offset notes across different staves has previously been discussed in these threads:
viewtopic.php?f=2&t=224&p=3667&hilit=Chopin+Etude#p3667
viewtopic.php?f=2&t=307

As mentioned, current notation software is unable to take multiple staves into account when distributing notes in different columns. Even within a single staff, the rules used to determine the vertical alignment automatically are often too simplistic, requiring the engraver to manually adjust the spacing and make alternate decisions about column distribution.

Elaine Gould does not seem to address this problem at all, although an example on p. 309 in her book does seem to indicate that she follows the default alignment determined by the notation software rather than any contextual practice established by manual engraving.

Based on quality engravings, a proposed general rule to determine the ideal alignment would be that:
Any notes of movement and/or similar value should be vertically aligned across staves.

This seems to be a good guideline in most cases, and some editions does seem to adhere rather strictly to it. In other editions, however, the application of this rule seems, at best, rather inconsistent.

For starters, here are a few examples from Debussy's Preludes, Book II which seems to me to be inconsistent:

A typical case following the above rule would be:
Skjermbilde 2017-05-06 kl. 12.34.15.png
Skjermbilde 2017-05-06 kl. 12.34.15.png (261.41 KiB) Viewed 1671 times
Here, the l.h. voice aligned with the moving voice in the r.h., and the half note is pushed to the left rather than the right (1.) to avoid collision of the slur, (2.) to adhere to most common practice (facing stems) and (3.) for consistency with the following measure.

The following examples, however, follows a more arbitrary, software-esque rule:
Skjermbilde 2017-05-06 kl. 12.38.08.png
Skjermbilde 2017-05-06 kl. 12.38.08.png (300.25 KiB) Viewed 1671 times
Skjermbilde 2017-05-06 kl. 12.50.49.png
Skjermbilde 2017-05-06 kl. 12.50.49.png (211.39 KiB) Viewed 1671 times
A more complex set of circumstances can be seen in the following:
Skjermbilde 2017-05-06 kl. 12.36.25.png
Skjermbilde 2017-05-06 kl. 12.36.25.png (333.81 KiB) Viewed 1671 times
Here, not only does the column distribution seem inconsistent upon first glance, but the eight notes does not lign up across hands. Even so, this seems like the best choice for the situation, given that there are seconds involved in both hands, and the idea is to limit the disruption as much as possible.

There even seems to be some doubt as to which column of chord notes on the same stem another voice should align with:
Skjermbilde 2017-05-06 kl. 14.38.37.png
Skjermbilde 2017-05-06 kl. 14.38.37.png (54.67 KiB) Viewed 1671 times
Normally the r.h. notes would align with the left column in the l.h.
Last edited by Knut on 07 May 2017, 21:18, edited 1 time in total.

User avatar
John Ruggero
Posts: 1229
Joined: 05 Oct 2015, 14:25
Location: Raleigh, NC USA

Re: Vertical alignment of chords with offset notes

Post by John Ruggero » 06 May 2017, 21:19

You have brought up so many interesting points that need careful consideration that I am going to comment in several different posts.
Knut wrote:
06 May 2017, 17:02
Based on quality engravings, a proposed general rule to determine the ideal alignment would be that:
Any notes of movement and/or similar value should be vertically aligned across staves.
I think that this rule is correct as far as it goes, but doesn't handle all cases, such as when there none of the voices have the same rhythm. I would go with the rule that I mentioned in the previous thread:

"… (in) cases where a note (or notes) is offset before or after a chord, the parts that are similar in some way, for example, in rhythm or (if the rhythm is the same) similar in the number of notes or because they are the most important voices, are always directly over each other so that the parts that properly belong together, are together."

In this example from Chopin's Scherzo no. 4, none of the voices have the same rhythm. The most important voices are therefore positioned over each other and the filler voice is offset:
Offset Spacing 1.jpg
Offset Spacing 1.jpg (24.26 KiB) Viewed 1658 times
Mac mini (OS 10.8.5) with dual monitors, Kurzweil Mark 5 with M-Audio Midisport 2 x 2,
Finale 2014d with GPO 4, JW Plug-ins, SmartScore X Pro, Adobe InDesign CS4,
Inkscape .48.5 and .91, FontForge 20150526
http://www.cantilenapress.com

User avatar
John Ruggero
Posts: 1229
Joined: 05 Oct 2015, 14:25
Location: Raleigh, NC USA

Re: Vertical alignment of chords with offset notes

Post by John Ruggero » 06 May 2017, 21:49

The abnormal case from the Chopin Scherzo no. 4 in the Schirmer (ed. Mikuli) edition that Knut presented in the previous thread:
Chopin Scherzo 4 ms.367-8 Schirmer Mikuli.jpg
Chopin Scherzo 4 ms.367-8 Schirmer Mikuli.jpg (12.44 KiB) Viewed 1601 times
This results from an attempt to keep the notes distributed on the staves as in the first French edition (and the autograph), yet follow more modern conventions regarding three voices on one staff. But the original edition and autograph have no offset at all, which keeps the similar parts over each other and for me is the best choice:
Chopin Scherzo 4 ms.367-8 1st Fr ed.jpg
Chopin Scherzo 4 ms.367-8 1st Fr ed.jpg (13.04 KiB) Viewed 1601 times
Or a very slight offset of the LH E would be fine too, but the large offset in the Schirmer (Mikuli) is undesirable.

Most modern editions rearrange the notes as seen in the Schirmer (ed. Joseffy) edition and avoid the issue completely:
Chopin Scherzo 4 ms.367-8 Schirmer Joseffy.jpg
Chopin Scherzo 4 ms.367-8 Schirmer Joseffy.jpg (18.21 KiB) Viewed 1601 times
Last edited by John Ruggero on 07 May 2017, 16:02, edited 1 time in total.
Mac mini (OS 10.8.5) with dual monitors, Kurzweil Mark 5 with M-Audio Midisport 2 x 2,
Finale 2014d with GPO 4, JW Plug-ins, SmartScore X Pro, Adobe InDesign CS4,
Inkscape .48.5 and .91, FontForge 20150526
http://www.cantilenapress.com

Knut
Posts: 864
Joined: 05 Oct 2015, 18:07
Location: Oslo, Norway

Re: Vertical alignment of chords with offset notes

Post by Knut » 07 May 2017, 10:10

John Ruggero wrote:
06 May 2017, 21:49
This results from an attempt to keep the notes distributed on the staves as in the first French edition (and the autograph), yet follow more modern conventions regarding three voices on one staff. But the original edition and autograph have no offset at all, which keeps the similar parts over each other and for me is the best choice:
I'm not sure if I agree with that analysis. There's nothing in that example which, to me, reveals an attempt to meet modern guidelines. Nor does it seem to have been brought on by a desire to represent the MS as closely as possible without shortening the stems of the middle l.h. voice. In fact, I'd say that following the most common modern convention, which Dorico does automatically, would give you a much more desirable compromise:
Skjermbilde 2017-05-07 kl. 11.49.45.png
Skjermbilde 2017-05-07 kl. 11.49.45.png (36.22 KiB) Viewed 1635 times
Pushing the lower voice to the right still keeps it within the borders of the upstem second, while the most important, moving voices are kept aligned.

User avatar
John Ruggero
Posts: 1229
Joined: 05 Oct 2015, 14:25
Location: Raleigh, NC USA

Re: Vertical alignment of chords with offset notes

Post by John Ruggero » 07 May 2017, 16:19

I just reattached the examples to my second post since they were not displayed when I just brought up the post.

Knut, all I meant to say is that the situation itself is the result of the Mikuli edition attempting to preserve the look of the autograph and first edition. Modern editions would distribute the notes as in the Joseffy to avoid the problem. Along with this more modern convention of placing the music for each hand on a different staff in a situation like this, I was also referring to the offsetting one of the voices when three occur on one staff rather than shortening the stems.

I would prefer the following distribution as an alternative to the original edition (if one is desired), rather than the Dorico default:
(I just noticed an error in stem direction in the example and corrected it.)
Note Positions revised.jpg
Note Positions revised.jpg (15.15 KiB) Viewed 1587 times
Last edited by John Ruggero on 07 May 2017, 19:55, edited 2 times in total.
Mac mini (OS 10.8.5) with dual monitors, Kurzweil Mark 5 with M-Audio Midisport 2 x 2,
Finale 2014d with GPO 4, JW Plug-ins, SmartScore X Pro, Adobe InDesign CS4,
Inkscape .48.5 and .91, FontForge 20150526
http://www.cantilenapress.com

Knut
Posts: 864
Joined: 05 Oct 2015, 18:07
Location: Oslo, Norway

Re: Vertical alignment of chords with offset notes

Post by Knut » 07 May 2017, 16:57

John Ruggero wrote:
07 May 2017, 16:19
I just reattached the examples to my second post since they were not displayed when I just brought up the post.

Knut, all I meant to say is that the situation itself is the result of the Mikuli edition attempting to preserve the look of the autograph and first edition. Modern editions would distribute the notes as in the Josephy to avoid the problem. Along with this more modern convention of placing the music for each hand on a different staff in a situation like this, I was also referring to the offsetting one of the voices when three occur on one staff rather than shortening the stems.

I would prefer the following distribution as an alternative to the original edition (if one is desired), rather than the Dorico default:

Note postions.jpg
Ah, I see.

I'm not opposed to decreasing the offset somewhat, but I think your preferred alternate version is a bit too tight. My quarrel with Schirmer's solution is that the moving voices aren't lined up, and therefore not in accordance with the principles outlined above.

User avatar
John Ruggero
Posts: 1229
Joined: 05 Oct 2015, 14:25
Location: Raleigh, NC USA

Re: Vertical alignment of chords with offset notes

Post by John Ruggero » 07 May 2017, 20:05

You have a good point, Knut. My slight offset was an attempt to have my cake and eat it. The following version keeps the corresponding voices completely over each other:
Note Postions 2.jpg
Note Postions 2.jpg (8.42 KiB) Viewed 1586 times
I am not really sure which I prefer.
Last edited by John Ruggero on 07 May 2017, 21:08, edited 1 time in total.
Mac mini (OS 10.8.5) with dual monitors, Kurzweil Mark 5 with M-Audio Midisport 2 x 2,
Finale 2014d with GPO 4, JW Plug-ins, SmartScore X Pro, Adobe InDesign CS4,
Inkscape .48.5 and .91, FontForge 20150526
http://www.cantilenapress.com

User avatar
John Ruggero
Posts: 1229
Joined: 05 Oct 2015, 14:25
Location: Raleigh, NC USA

Re: Vertical alignment of chords with offset notes

Post by John Ruggero » 07 May 2017, 21:07

Knut's first example from Debussy Preludes Bk 1 no. 3: Along with the reasons given in Knut's excellent analysis, the placing of the half notes may also have to do with the composer's concern that the player understand the importance of these notes structurally (note the tenuto markings in the first edition*) against an important RH line that could easily cover them. Note the same distribution in the composer's autograph as in the first edition given above:
Debussy Preludes Bk 2 no 3 .jpg
Debussy Preludes Bk 2 no 3 .jpg (31.69 KiB) Viewed 1583 times
In this case, I don't think that placing the half notes directly over the bass tone, as Finale does by default, would be incorrect—just more orthodox and in accordance with the idea that the main parts be over each other when the voices are in different rhythms. Of course, the first edition was right in engraving the passage exactly as in the autograph.

* It is worth noting that the F with its auxiliary E dominates the music immediately before this passage and several measures after it.
Mac mini (OS 10.8.5) with dual monitors, Kurzweil Mark 5 with M-Audio Midisport 2 x 2,
Finale 2014d with GPO 4, JW Plug-ins, SmartScore X Pro, Adobe InDesign CS4,
Inkscape .48.5 and .91, FontForge 20150526
http://www.cantilenapress.com

User avatar
OCTO
Posts: 1003
Joined: 05 Oct 2015, 06:52
Location: Sweden

Re: Horizontal alignment of chords with offset notes

Post by OCTO » 08 May 2017, 09:56

If I may jump on...

For my personal taste, in all these cases I would consider two principles:
1. That voices of the same rhythmical structure should be aligned equally,
2. That voices that consist of shorter rhythmical values should have note-spacing priority.
Freelance Composer. Self-Publisher.
Finale 25 • Sibelius 8 • MuseScore 2 • Logic Pro X • Ableton Live 9 • Digital Performer 9 /// OS X El Capitan, (side system: Debian 9, Windows 7)

Knut
Posts: 864
Joined: 05 Oct 2015, 18:07
Location: Oslo, Norway

Re: Horizontal alignment of chords with offset notes

Post by Knut » 08 May 2017, 10:30

OCTO wrote:
08 May 2017, 09:56
If I may jump on...

For my personal taste, in all these cases I would consider two principles:
1. That voices of the same rhythmical structure should be aligned equally,
2. That voices that consist of shorter rhythmical values should have note-spacing priority.
I agree, OCTO.

Your first point is pretty much incorporated into the principles already discussed. Your second point is also very important in the greater context of these adjustments, and I agree that in most cases it is a sound principle. There are exception where you might want to reduce the space compensation somewhat to avoid contrasts between similar rhythmic structures form being too large, but I would consider those reasonably rare.

Post Reply