Colin Hall | 11 Apr 02:36 2012
Picon

Re: lilypond bug 2368


On Sat, Apr 07, 2012 at 05:32:28PM +0200, Karol Majewski wrote:
> 
> Is it possible to mark 2368 bug as critical? I think this kind of
> regression is a serious problem. Just take a look at my last
> comment:
> 
> http://code.google.com/p/lilypond/issues/detail?id=2368#c6

I agree with you that the notation output is not correct, especially
in the sample you provide in comment 6.

I'm going to do my best to explain why this issue should not be marked
critical. I hope that others will correct my statements as necessary.

You might not be aware that there are several open issues for Lilypond
that relate to the typesetting of ties and slurs. For example see
issue 298, referenced from issue 2368.

http://code.google.com/p/lilypond/issues/detail?id=298

There are others.

As I understand it the development team have identified that a better
implementation of ties is possible but requires aesthetic design work
to establish how this feature should work before technical
implementation work can begin. The technical work may be substantial.

Janek has a study of Lilypond's typesetting of ties and slurs which is
work in progress and which may provide the aesthetic design we
(Continue reading)

Janek Warchoł | 11 Apr 07:59 2012
Picon

Re: lilypond bug 2368

On Wed, Apr 11, 2012 at 2:36 AM, Colin Hall <colinghall <at> gmail.com> wrote:
>
> On Sat, Apr 07, 2012 at 05:32:28PM +0200, Karol Majewski wrote:
>>
>> Is it possible to mark 2368 bug as critical? I think this kind of
>> regression is a serious problem. Just take a look at my last
>> comment:
>>
>> http://code.google.com/p/lilypond/issues/detail?id=2368#c6
>
> I agree with you that the notation output is not correct, especially
> in the sample you provide in comment 6.
>
> I'm going to do my best to explain why this issue should not be marked
> critical. I hope that others will correct my statements as necessary.
>
> You might not be aware that there are several open issues for Lilypond
> that relate to the typesetting of ties and slurs. For example see
> issue 298, referenced from issue 2368.
>
> http://code.google.com/p/lilypond/issues/detail?id=298
>
> There are others.
>
> As I understand it the development team have identified that a better
> implementation of ties is possible but requires aesthetic design work
> to establish how this feature should work before technical
> implementation work can begin. The technical work may be substantial.

Yes, i do think that it will make more sense to rewrite tie code
(Continue reading)

Karol Majewski | 11 Apr 20:33 2012
Picon

Re: lilypond bug 2368

OK, thanks for explanation! Actually, I've menaged to solve my problem by
entering hidden notes with unHidden slur in separate voice. Now the tie
refers to the right notes and appeals to my sense of aesthetic - despite
colliding with an accidental (take a look at the attachment).

Janek, I'm not a student of Politechnika Warszawska - neither former nor
present. I'm a former student of Akademia Muzyczna w Gdańsku :)

Best Wishes
Karol Majewski

2012/4/11 Janek Warchoł <janek.lilypond <at> gmail.com>

> On Wed, Apr 11, 2012 at 2:36 AM, Colin Hall <colinghall <at> gmail.com> wrote:
> >
> > On Sat, Apr 07, 2012 at 05:32:28PM +0200, Karol Majewski wrote:
> >>
> >> Is it possible to mark 2368 bug as critical? I think this kind of
> >> regression is a serious problem. Just take a look at my last
> >> comment:
> >>
> >> http://code.google.com/p/lilypond/issues/detail?id=2368#c6
> >
> > I agree with you that the notation output is not correct, especially
> > in the sample you provide in comment 6.
> >
> > I'm going to do my best to explain why this issue should not be marked
> > critical. I hope that others will correct my statements as necessary.
> >
> > You might not be aware that there are several open issues for Lilypond
(Continue reading)

Janek Warchoł | 11 Apr 20:41 2012
Picon

Re: lilypond bug 2368

On Wed, Apr 11, 2012 at 8:33 PM, Karol Majewski
<karol.majewsky <at> gmail.com> wrote:
> OK, thanks for explanation! Actually, I've menaged to solve my problem by
> entering hidden notes with unHidden slur in separate voice. Now the tie
> refers to the right notes and appeals to my sense of aesthetic - despite
> colliding with an accidental (take a look at the attachment).

I see additional dots, you may want to hide them as well.
As for the tie, a perfect solution would involve not only avoiding
accidental, but also moving the augmentation dot a bit... there's a
lot of work to be done on ties.

> Janek, I'm not a student of Politechnika Warszawska - neither former nor
> present. I'm a former student of Akademia Muzyczna w Gdańsku :)

Ah :)  I found a Karol-Majewski-Elka-student on the net and thought it
may be you.
Nice to see people from Poland!
Are you interested in becoming a contributor? ;)
BTW, i will take the opportunity for an obvious self-promotion: see
how i represent our country in LilyPond newsletter
http://news.lilynet.net/?The-LilyPond-Report-25 !

best,
Janek

_______________________________________________
bug-lilypond mailing list
bug-lilypond <at> gnu.org
https://lists.gnu.org/mailman/listinfo/bug-lilypond
(Continue reading)

Phil Holmes | 13 Apr 16:10 2012
Picon

Re: lilypond bug 2368


  ----- Original Message ----- 
  From: Karol Majewski 
  To: Janek Warchoł 
  Cc: LilyPond Developmet Team ; Lilypond Bugs 
  Sent: Wednesday, April 11, 2012 7:33 PM
  Subject: Re: lilypond bug 2368

  OK, thanks for explanation! Actually, I've menaged to solve my problem by entering hidden notes with
unHidden slur in separate voice. Now the tie refers to the right notes and appeals to my sense of aesthetic -
despite colliding with an accidental (take a look at the attachment).

  Janek, I'm not a student of Politechnika Warszawska - neither former nor present. I'm a former student of
Akademia Muzyczna w Gdańsku :)

  Best Wishes
  Karol Majewski
Check my suggestion at http://code.google.com/p/lilypond/issues/detail?id=2368 - it works better
and is designed for this.

--
Phil Holmes
Karol Majewski | 24 Apr 22:48 2012
Picon

Re: lilypond bug 2368

Thank you Phil.

Dear LilyPond friends: Janek, Phil, Colin,

I have another big problem now. I'm trying to find a global setting that would allow to ignore collisions between ties and accidentals. For me, a tie-accidental collision looks a way better than a shifted tie - especially in tied chords in which one of the ties is unnaturaly shifted horizontaly. I'm almost sure that it is possible to achieve this, but I don't know how.

Best wishes

PS Janek, I saw that report. I really appreciate your contribution :)




W dniu 13 kwietnia 2012 16:10 użytkownik Phil Holmes <mail <at> philholmes.net> napisał:
 
----- Original Message -----
Sent: Wednesday, April 11, 2012 7:33 PM
Subject: Re: lilypond bug 2368

OK, thanks for explanation! Actually, I've menaged to solve my problem by entering hidden notes with unHidden slur in separate voice. Now the tie refers to the right notes and appeals to my sense of aesthetic - despite colliding with an accidental (take a look at the attachment).

Janek, I'm not a student of Politechnika Warszawska - neither former nor present. I'm a former student of Akademia Muzyczna w Gdańsku :)

Best Wishes
Karol Majewski

Check my suggestion at http://code.google.com/p/lilypond/issues/detail?id=2368 - it works better and is designed for this.

--
Phil Holmes
 


_______________________________________________
lilypond-devel mailing list
lilypond-devel <at> gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel
Mark Mathias | 28 Apr 19:19 2012
Picon

Re: lilypond bug 2368

Karol,
I hope someone can help you do what you wish. It's definitely beyond my level of expertise. From my point of view on the bug squad, this doesn't look like something that needs to be handled by us, so I'm going to suggest that future correspondence on this issue not be sent to the bug squad.
Thanks,
Mark

2012/4/24 Karol Majewski <karol.majewsky <at> gmail.com>
Thank you Phil.

Dear LilyPond friends: Janek, Phil, Colin,

I have another big problem now. I'm trying to find a global setting that
would allow to ignore collisions between ties and accidentals. For me, a
tie-accidental collision looks a way better than a shifted tie - especially
in tied chords in which one of the ties is unnaturaly shifted horizontaly.
I'm almost sure that it is possible to achieve this, but I don't know how.

Best wishes

PS Janek, I saw that report. I really appreciate your contribution :)




W dniu 13 kwietnia 2012 16:10 użytkownik Phil Holmes
<mail <at> philholmes.net>napisał:

> **
>
>
> ----- Original Message -----
> *From:* Karol Majewski <karol.majewsky <at> gmail.com>
> *To:* Janek Warchoł <janek.lilypond <at> gmail.com>
> *Cc:* LilyPond Developmet Team <lilypond-devel <at> gnu.org> ; Lilypond Bugs<bug-lilypond <at> gnu.org>
> *Sent:* Wednesday, April 11, 2012 7:33 PM
> *Subject:* Re: lilypond bug 2368
>
> OK, thanks for explanation! Actually, I've menaged to solve my problem by
> entering hidden notes with unHidden slur in separate voice. Now the tie
> refers to the right notes and appeals to my sense of aesthetic - despite
> colliding with an accidental (take a look at the attachment).
>
> Janek, I'm not a student of Politechnika Warszawska - neither former nor
> present. I'm a former student of Akademia Muzyczna w Gdańsku :)
>
> Best Wishes
> Karol Majewski
>
> Check my suggestion at
> http://code.google.com/p/lilypond/issues/detail?id=2368 - it works better
> and is designed for this.
>
>  --
> Phil Holmes
>
>
>
_______________________________________________
bug-lilypond mailing list
bug-lilypond <at> gnu.org
https://lists.gnu.org/mailman/listinfo/bug-lilypond

_______________________________________________
lilypond-devel mailing list
lilypond-devel <at> gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel
mike@apollinemike.com | 28 Apr 19:49 2012

Re: lilypond bug 2368

On 28 avr. 2012, at 19:19, Mark Mathias wrote:

Karol,
I hope someone can help you do what you wish. It's definitely beyond my level of expertise. From my point of view on the bug squad, this doesn't look like something that needs to be handled by us, so I'm going to suggest that future correspondence on this issue not be sent to the bug squad.
Thanks,
Mark

2012/4/24 Karol Majewski <karol.majewsky <at> gmail.com>
Thank you Phil.

Dear LilyPond friends: Janek, Phil, Colin,

I have another big problem now. I'm trying to find a global setting that
would allow to ignore collisions between ties and accidentals. For me, a
tie-accidental collision looks a way better than a shifted tie - especially
in tied chords in which one of the ties is unnaturaly shifted horizontaly.
I'm almost sure that it is possible to achieve this, but I don't know how.

Best wishes

PS Janek, I saw that report. I really appreciate your contribution :)






Hey Karol,

One difficult-to-document feature of LilyPond ties is the details list.  For example, if you do:

\override Tie #'details #'vertical-distance-penalty-factor = #1000

it'll make certain vertical shifts very unlikely.

I've copied the details from define-grobs.scm below.  There are a couple other details that for the moment are just used internally.

There's been a lot of back and forth on a few LilyPond lists about improving ties.  I'm of the opinion that the system in place is sound (it is a similar system as that of slurs and beams), but what needs improvement is an understanding of how much factors are worth and where they make the difference.  We've made some headway in slurs in 2.15 by removing unnecessary factors and decorolating them but we haven't been able to do the same yet for ties.  So, my advice to everyone would be to play around with this details list.

The more we know how the details list functions in real music, the better we can make it.  User input matters: if you exhaust all combinations of details and it still doesn't work, then it's much easier for a developer to know how to modify the infrastructure.  Furthermore, if you find a cocktail of details setting that leads to a result that, in your opinion, should be out-of-the-box, then it's helpful to know that as well.

Cheers,
MS

        (details . (
                    ;; for a full list, see tie-details.cc
                    (ratio . 0.333)
                    (center-staff-line-clearance . 0.6)
                    (tip-staff-line-clearance . 0.45)
                    (note-head-gap . 0.2)
                    (stem-gap . 0.35)
                    (height-limit . 1.0)
                    (horizontal-distance-penalty-factor . 10)
                    (same-dir-as-stem-penalty . 8)
                    (min-length-penalty-factor . 26)
                    (tie-tie-collision-distance . 0.45)
                    (tie-tie-collision-penalty . 25.0)
                    (intra-space-threshold . 1.25)
                    (outer-tie-vertical-distance-symmetry-penalty-factor . 10)
                    (outer-tie-length-symmetry-penalty-factor . 10)
                    (vertical-distance-penalty-factor . 7)
                    (outer-tie-vertical-gap . 0.25)
                    (multi-tie-region-size . 3)
                    (single-tie-region-size . 4)
                    (between-length-limit . 1.0)))
_______________________________________________
lilypond-user mailing list
lilypond-user <at> gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-user
Janek Warchoł | 29 Apr 19:42 2012
Picon

Re: lilypond bug 2368

On Sat, Apr 28, 2012 at 7:49 PM, mike <at> apollinemike.com
<mike <at> apollinemike.com> wrote:
> There's been a lot of back and forth on a few LilyPond lists about improving
> ties.  I'm of the opinion that the system in place is sound (it is a similar
> system as that of slurs and beams), but what needs improvement is an
> understanding of how much factors are worth and where they make the
> difference.  We've made some headway in slurs in 2.15 by removing
> unnecessary factors and decorolating them but we haven't been able to do the
> same yet for ties.  So, my advice to everyone would be to play around with
> this details list.

Surely i will experiment when i'll be finishing my tie report :)
However, given how often current ties fail, i'm pretty sure that the
current system is not good.  I have a strong feeling that it's an
overkill; from what i've already seen i could probably write a
deterministic algorithm that would perform better in many cases and
use half as much parameters.
I may be wrong; however, i think that the core of the problem is that
while slurs are really complicated and need advanced heuristics and
guesswork, ties are significantly more straightforward.

cheers,
Janek

_______________________________________________
lilypond-user mailing list
lilypond-user <at> gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-user

Gmane