Page 3 of 7

Re: Dorico

Posted: 21 Oct 2016, 01:26
by John Ruggero
Alexander Plötz wrote:
While having more finished features (especially from the "basic" category) would certainly be preferable, I think that a simple note-offset available right now would probably just lead to fewer users unlocking the true power of Dorico, which comes much more from a deeply ingrained understanding of notation than on any approach of just moving graphics around.
I think that human intervention will always be required to produce the best result in music engraving no matter how excellent the computer program used because engraving rules alone are not sufficient to deal with the complexity of musical notation—and for that reason, human intervention should be facilitated, not made more difficult.

For example, in the last few measures from the Bach 6-voice Ricercare shown in your article, Dorico made many errors that you corrected. Your final result was much improved, aside from a few quarter rests that are very close to the note heads in bars 3-5 and the two side-by-side half-notes in measure 3 that might better be a single double-stemmed one, as in the composer's MS and the New Bach Edition. The composer's MS also has superior distribution of the notes between the staves (and hands) in these measures, and the whole piece is in a subdivided 4/2 rather than 2/2. http://petrucci.mus.auth.gr/imglnks/usi ... ikel_1.pdf

Re: Dorico

Posted: 21 Oct 2016, 02:14
by John Ruggero
Nice work, Knut. Dorico is doing a much better job than Finale in the initial positioning of the rests and the lengths of the stems and beam angles. I am particularly impressed with the latter. Here is the first measure done using Finale's default file, with a corrected version that required 6 edits.
Finale Example.jpg
Finale Example.jpg (29.45 KiB) Viewed 9661 times
But I am very surprised that Dorico does not find the correct position for the rests given all the care that has gone into this kind of thing. For example, why is the first rest in measure one of your example wrong and the first rest in measure two correct? The rests don't seem to know when to stay put and when to adjust!

Re: Dorico

Posted: 21 Oct 2016, 07:10
by Knut
John Ruggero wrote: Here is the first measure done using Finale's default file, with a corrected version that required 6 edits.
6 edits in a single measure. That's quite the difference!

Finale Example.jpg
But I am very surprised that Dorico does not find the correct position for the rests given all the care that has gone into this kind of thing. For example, why is the first rest in measure one of your example wrong and the first rest in measure two correct? The rests don't seem to know when to stay put and when to adjust!
So am I, actually. But then again, in large part this may come down to differing opinions or individual taste. I'll have to check with Gould later, but off the top of my head there are (at least) three (often conflicting) considerations to keep in mind when determining the ideal rest placement in the context of 3-voice polyphony: (1.) Rests should be horizontally aligned, at least within the same measure; (2.) they should follow the musical contour of their respective voice as closely as possible, and (3) should deviate from it's default vertical position whenever possible. For practical purposes, rule number 3 is given less weight for the inner voice in a 3-voice context, but in principle, rules 1 and 2 apply equally to all voices.

These rules no doubt will challenge the simplicity of any computer algorithm for vertical rest placement, and as you can see, Dorico seems to emphasize the second rule for the middle voice, interpreting the eight rests as the first 'note' in an ascending 3-note figure in most cases (which indeed might be too much of an assumption in a fragmented case such as this). At the same time the program seems to strive to uphold the thirds rule even for the middle voice when possible.

With this in mind I'm not entirely sure that the results produced in measure 1 can be considered 'wrong'; too me there are several much clearer cases of misplacement than that one, at least. But I certainly agree with you that less vertical movement of rests than what Dorico seems to produce in this case is preferable.

I think this merits a deeper discussion on the matter, after which we should share our consensus with the developers of Dorico. At the moment, I think they have enough on their plate, but it would be great if this could be improved in a version down the line.

Re: Dorico

Posted: 21 Oct 2016, 12:14
by RMK
John Ruggero wrote:
But was it possible to move notes independently of the others in the earlier versions of Sibelius by typing numbers?
Yes, as far as I can remember. Sibelius 1 appeared (in the US) in the mid-90's.

Re: Dorico

Posted: 21 Oct 2016, 12:31
by OCTO
Wow, thanks Knut!

Actually, I think I will buy Dorico once it can meet my needs. (an example from my score: https://issuu.com/octoechos/docs/on_the ... _the_heart ).

I think that Dorico team should win better if they introduced it with the price of 99€ instead of 600€. The price is comparable with other notation software, yet it lacks many many important features. Maturing to v2 and more, the price would rise of course.

MuseScore 2 and soon 3 is appealing more to me than Dorico, for now.

Re: Dorico

Posted: 21 Oct 2016, 12:33
by OCTO
....And I would definitely love to see more screenshots and engraving examples and experiences done in Dorico.

Re: Dorico

Posted: 21 Oct 2016, 12:59
by Knut
OCTO wrote:I think that Dorico team should win better if they introduced it with the price of 99€ instead of 600€. The price is comparable with other notation software, yet it lacks many many important features. Maturing to v2 and more, the price would rise of course.
You're certainly not the only one who thinks so. To me, however, it has become clear that the music notation market is too small to expect any degree of revolution without users investing both money and trust in a company for it to deliver. As Finale has demonstrated time and time again, especially in recent years, doing so does not guarantee that you will avoid functioning as a beta tester and having to wait several months before a buggy first release is updated to a useable version, even when the application has been in development for 30 years. Many who have been part of the computer-based music notation field from the beginning can attest to the initial releases being quite horrible to use, and even if Dorico can't be said to deliver the same amount of innovation, there are similarities, both good and bad.

Re: Dorico

Posted: 21 Oct 2016, 13:50
by John Ruggero
Knut wrote:
But then again, in large part this may come down to differing opinions or individual taste.
I think that there are rules that govern the position of rests with a little wiggle-room allowed for taste. At least let's hope so or it will be impossible for a computer to position rests correctly.
(1.) Rests should be horizontally aligned, at least within the same measure; (2.) they should follow the musical contour of their respective voice as closely as possible, and (3) should deviate from it's default vertical position whenever possible.
I think your rule 1 is nice when it happens, but isn't a determining factor. I am not quite clear about your rule 2, and disagree completely with rule 3.

The rule that I have always used, and assumed was the standard one, is that a rest should only be displaced from its normal position if another voice impinges on it and should be shifted by the least amount necessary.

What constitutes "impingement" and what constitutes over-shifting is what computer analysts are paid for; but I know it when I see it and see incorrect positioning in many measures of your Dorico example: 1, 4, 5, 10, 11, 12, 13 etc.

Gould hasn't formulated exact rules regarding rests in polyphonic situations other than saying if one voice in two-voice writing lies outside the staff, rests for the other should remain in their normal positions. (p. 36) A similar situation occurs in measure 1 and several others of your example, since two of three voices lie outside the staff. For this reason, I would think that these are incorrect by Gould's rule.

Re: Dorico

Posted: 21 Oct 2016, 15:21
by Knut
John Ruggero wrote:I think that there are rules that govern the position of rests with a little wiggle-room allowed for taste. At least let's hope so or it will be impossible for a computer to position rests correctly.
In truth, the fewer and more rigid the rules, the easier it is for a computer to follow them, but with regard to rest placement I agree that some flexibility is appropriate.
John Ruggero wrote:I think your rule 1 is nice when it happens, but isn't a determining factor. I am not quite clear about your rule 2, and disagree completely with rule 3.
Rule 1 prevents the rests from jumping around needlessly, which can be pretty annoying and harder to read. Gould does indeed consider this rule, but only for rests within a beat (p. 37). In Dorico's engraving options, however, the rule is taken into account for outer voices in larger contexts, giving you the option to align strive for alignment of rests on either a measure by measure basis, or for longer stretches.

Rule 2 just means that you treat rests as if they had stemlets, placing them so that they take the direction of the surrounding lines into account. You can clearly see this principle bing followed in the inner voice of the Dorico example.

Even Gould seems to me to clearly consider rule 3 with her first two statements on p. 36, saying that for clarity, rests are usually placed of center in double stemmed writing, except when one part lies outside the staff. To me, this is also sensible to more clearly distinguish consolidated rests from split rests.
John Ruggero wrote:The rule that I have always used, and assumed was the standard one, is that a rest should only be displaced from its normal position if another voice impinges on it and should be shifted by the least amount necessary.
It's very interesting that we have such different takes on this. To me, this rule seems way to simplistic to account for all the variations in a musical score. I think proximity to the voice it applies to is a much more important consideration than to have the rest stay in the center of a staff at all times as long as there's room.

Re: Dorico

Posted: 21 Oct 2016, 17:15
by John Ruggero
Knut, my rule is a rule-of-thumb that is not intended to account for all situations.

Just so that I understand your style more clearly:

In measure 1 of your Dorico example, would you override the standard position of the first rest so that it can be closer to the note G that follows, even though both of the other notes are off the staff and there is no other pressing reason to move the rest?

My preference is shown in the example in my post above showing the Finale default with my corrections. This looks normal to me. While I certainly take the position of surrounding notes and rests into account when adjusting rests that NEED adjustment (i. e. your rule 1), I wouldn't adjust a rest simply to position it near surrounding notes.

And if Dorico is attempting to position all the interior rests near surrounding notes (your Rule 2), it is not doing this consistently. See rests in measures 4, 10, 14, 16 etc. At other times it seems to be creating conflicts by following such a rule as in 16 etc.