Page 4 of 5

Re: [puzzle] Slur solutions 2

Posted: 09 Mar 2016, 14:55
by DatOrganistTho
Knut wrote:John is right.

As I see it, there are two main problems when trying to simulate plate engraved slurs in a scoring application:

1. Generally, plate engraved slurs vary in thickness (shorter slurs are thinner than longer slurs), and this variable is impossible to recreate in most scoring apps without some kind of post production tool.

2. Plate engraved slurs have round tips, which allow the ends to be somewhat thicker without being perceived as 'sausagy'.

I'm currently experimenting with slur tips between 0.06 and 0.08 spaces (1.5 – 2.0 EVPU). 0.08 spaces is closer to the relative thickness used in plate engraved scores, as it is the same as the thickness of my stems. However, the resulting abrupt end points becomes very visible on screen, and it therefore seems too thick to me. I'll have to do some print outs and see how it looks on paper.
Lilypond's default output is such. Still don't understand how default outputs are undesired and sausage like. Please see attached and help me understand.

Re: [puzzle] Slur solutions 2

Posted: 09 Mar 2016, 14:57
by DatOrganistTho
John Ruggero wrote:DatOrganistTho wrote:
So old fashioned scores and their engravings are inferior because they look like sausages (the natural result of plate engraving)?
No, well-done hand engraved slurs do not look like sausages because good balance between thick and thin is maintained. I think that the slurs in Henle hand-engraved scores, for example, maintain beautiful thick and thin balance.

Here is an example from the past that I think has many beautifully hand-engraved slurs that are on the thicker side but never appear inartistic. Of course, there are also imperfections because of the nature of hand work, but the overall impression is warm and delightful to the eye, which is so typical of the Brahms Complete Edition:

http://javanese.imslp.info/files/imglnk ... 4_scan.pdf
I understand what you are saying, and yet when I look at the default output on Lilypond versus the example you gave, I cannot see a difference in what you are saying. I'd love to understand better.

Re: [puzzle] Slur solutions 2

Posted: 09 Mar 2016, 18:21
by John Ruggero
DatOrganistTho wrote:
Lilypond's default output is such. Still don't understand how default outputs are undesired and sausage like
My goodness, when did I ever say that LilyPond's slurs were sausage-like? I said there was a danger that slurs in general could become sausage-like IF slur tips got too fat:

"If someone prefers thicker slurs, then thicker endings are in order. However, the slur should have a nice thin-thick-thin balance or it looks more like a sausage rather than an something artfully drawn on the page."

As far as I am concerned, your slur in the example above has a very nice balance of thick and thin elements and is fine. My slur settings produce a more tapered ending because the tip is smaller. Knut is rethinking his tip settings, and I am, as well. I think that my slurs could be more pleasing with slightly larger tips.

Re: [puzzle] Slur solutions 2

Posted: 09 Mar 2016, 18:59
by Knut
DatOrganistTho wrote:
Knut wrote:Lilypond's default output is such. Still don't understand how default outputs are undesired and sausage like. Please see attached and help me understand.
Since this was a reply to one of my comments, something I've said earlier in this thread must have given you the impression that I find Lilypond's default slurs to resemble sausages.

The only such critical comment I've raised was towards one of tisimst's examples, which in my opinion had too little contrast between the tips and middle of the slur. Otherwise I've just supported John's general opinions on what distinguishes a well formed slur shape from a bad one, regardless of software. In fact, I specifically said that Lilypond's slur shapes otherwise seems to be superior to those of Finale.

Once again I second Johns opinion in that your slur above seems very well balanced.

Re: [puzzle] Slur solutions 2

Posted: 09 Mar 2016, 19:23
by tisimst
In all fairness, the original image I posted should be the same settings as that shown by DatOrganistTho. I think the confusion is being caused by the resolution of my original image that was captured from the PDF viewer, which I've noticed can cause the contrast to appear smaller than it actually is.

Either way, who's to say that LilyPond's default settings can't be improved? I find stuff all the time that could use improvement. It also depends on the intended appearance of a given house-style, which may require a curve with less-contrast than another house-style. In general, however, I agree that an obvious contrast along the curve is more pleasing to the eye and functionally discernible among the other elements on the page.

Re: [puzzle] Slur solutions 2

Posted: 10 Mar 2016, 13:03
by John Ruggero
Dear DatOrganistTho,

I am so sorry that I didn't understand that your comments were alluding to tisimst's original solution to the puzzle with LilyPond defaults, which appeared a little thick to Knut and to me—so tisimst modified them.

Yes, I do prefer more taper than the LilyPond defaults. That would be produced by having a fuller middle, smaller tips and more taper in between as shown by tisimst in his modification. But the LilyPond defaults are fine. It is just a fuller look, such as in the Brahms example.

Re: [puzzle] Slur solutions 2

Posted: 10 Mar 2016, 23:26
by DatOrganistTho
John Ruggero wrote:DatOrganistTho wrote:
Lilypond's default output is such. Still don't understand how default outputs are undesired and sausage like
My goodness, when did I ever say that LilyPond's slurs were sausage-like? I said there was a danger that slurs in general could become sausage-like IF slur tips got too fat:

"If someone prefers thicker slurs, then thicker endings are in order. However, the slur should have a nice thin-thick-thin balance or it looks more like a sausage rather than an something artfully drawn on the page."

As far as I am concerned, your slur in the example above has a very nice balance of thick and thin elements and is fine. My slur settings produce a more tapered ending because the tip is smaller. Knut is rethinking his tip settings, and I am, as well. I think that my slurs could be more pleasing with slightly larger tips.
Ah! I see where I missed it (also having read your comment later). I really appreciate your patience in helping me see what you are saying! I apologize for the communication breakdown in my questions: I could have certainly been clearer. Thank you... :)

Re: [puzzle] Slur solutions 2

Posted: 10 Mar 2016, 23:28
by DatOrganistTho
Knut wrote:
DatOrganistTho wrote:
Knut wrote:Lilypond's default output is such. Still don't understand how default outputs are undesired and sausage like. Please see attached and help me understand.
Since this was a reply to one of my comments, something I've said earlier in this thread must have given you the impression that I find Lilypond's default slurs to resemble sausages.

The only such critical comment I've raised was towards one of tisimst's examples, which in my opinion had too little contrast between the tips and middle of the slur. Otherwise I've just supported John's general opinions on what distinguishes a well formed slur shape from a bad one, regardless of software. In fact, I specifically said that Lilypond's slur shapes otherwise seems to be superior to those of Finale.

Once again I second Johns opinion in that your slur above seems very well balanced.
I understand now. Thank you for bearing with my questions and lack of understanding. I didn't intend to sound defensive; I apologize! :)

Re: [puzzle] Slur solutions 2

Posted: 10 Mar 2016, 23:29
by DatOrganistTho
tisimst wrote:In all fairness, the original image I posted should be the same settings as that shown by DatOrganistTho. I think the confusion is being caused by the resolution of my original image that was captured from the PDF viewer, which I've noticed can cause the contrast to appear smaller than it actually is.

Either way, who's to say that LilyPond's default settings can't be improved? I find stuff all the time that could use improvement. It also depends on the intended appearance of a given house-style, which may require a curve with less-contrast than another house-style. In general, however, I agree that an obvious contrast along the curve is more pleasing to the eye and functionally discernible among the other elements on the page.
That would make sense, and it is all coming together.

I absolutely agree that LilyPond can be improved, but perhaps not nearly to the degree that Finale and other culprits of "bad defaults" might need... ;)

Re: [puzzle] Slur solutions 2

Posted: 11 Mar 2016, 10:51
by Peter West
Knut wrote:Unfortunately, I don't think there is any way to improve the default settings in Finale to accommodate these situations, without compromising others at the same time.

Here's my effort:
Skjermbilde 2016-03-03 kl. 12.18.00.png
My solution for manually adjusting these situations is:

1. Not to overstate the angle of the slur when moving the left end point down the stem.
2. Pull the left inset point up alongside the tangent, until it's parallel with the centre point (or slightly below it).
3. Move the right inset point slightly to further to the right to compensate.
4. Move the centre point up or down to clear the staff line slightly. Be sure not to make the slur too flat. These slurs usually need to be a bit higher than the default.

Edit: My idea here is to come up with something along the lines of OCTO's results for challenge B, but to make the slur more reflective of the ascending motion.
I like this solution, though I would keep more curve in the right half, at the moment it looks almost like a straight line. Apart from that this has my vote.

And I agree, any automatic setting that correctly places these will adversely affect others. Least damage, I think is generally to have these wrongly placed then manually adjust as needed.