Re: Hairpin.minimum-length gives wrong output

[Thomas Morley-2]

ptoye wrote
> What I want is the first quaver sf and the hairpin to finish before the
> next quaver, so that the 3rd quaver is at the same dynamic  as the 2nd
> one.
> 
> Here's a better MWV
> 
> \version "2.19.52"
> \language "english"
> {
>   <<
>     \new Staff {
>       \time 2/4
>       \clef "treble"
>       c''8 8 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4\break
>       c''8 8 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4\break
>       c''8 8 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4\break
>       c''8 8 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4
>     }
>     \new Dynamics {
>       s8\sf\> s8\! s4 | s2 2 2 2 2 2 2 2 2 2 2 2
>       s8\sf\> s8\p s4 | s2 2 2 2 2 2 2 2 2 2 2 2
>       \override Hairpin.minimum-length = #10
>       s8\sf\> s8\! s4 | s2 2 2 2 2 2 2 2 2 2 2 2
>       s8\sf\> s8\p s4 | s2 2 2 2 2 2 2 2 2 2
>     }
>   >>
> }
> 
> Line 1 has the hairpin too late at both the start and endpoints. Line 2
> has a warning that the hairpin is too small. Line 3 has the end of the
> hairpin too late. Line 4 is fine.

Hi Peter,

I'm having a hard time to shorten your mail to the part I want to reply. 
I'd recommend to use plain text.
Right now I use the nabble-interface, because I failed to do it otherwise.
That's unneeded tedious.

Anyway, single Hairpins (not surrounded by other dynamic signs) should start
at the left of the starting note and end at the right of the target note. If
dynamic signs are present they are centered below their notes and the
Hairpin is printed in between. 
For subsequent Hairpins, end/new-start are printed a little left and right
from the corresponding notehead-center.
This is the typesetting rule as far as I can tell.

See LilyPond doing it:
\paper { ragged-right = ##f }
{ b4\> b\! }
{ b4\> b\> b\! }


Thus I think you're wrong saying
"Line 3 has the end of the hairpin too late."
And even your initial quoted statement is not at the lines of usual
typesetting rules.
The Hairpin _is_positioned correctly, although admittedly it looks terrible.
But this is due to the uneven spacing, not a wrong Hairpin.

Here: "Line 1 has the hairpin too late at both the start and endpoints" I'm
not sure what you want instead for the Hairpin-start, wherelse than after
the sforzato should it start? 
The endpoint is correct either (see above).
Alas, this line looks terrible as well.
By all means, the Hairpin is too short.

Similiar for line 2 of your example.

So, all these three examples look weird because the Hairpin is not long
enough.
This may be a weakness of LilyPond: A spanner doesn't push it's bounds,
unless forced. P.e. with setting minimum-length to a higher value. This is
not done automagically.

Thus line 4 looks best: The Hairpin is lengthened and positioned between the
two dynamic signs, each centerd below their notes. Although the uneven
spacing is still ugly.


So if you really want the Hairpin to end before the target note, you have to
ensure/set two things
(1) Help LilyPond, make the Hairpin long enough, i.e. set minimum-length to
a suitable value.
(2) For further tweaking use shortern-pair 

\paper { ragged-right = ##t }
{ 
        \override Hairpin.shorten-pair = #'(0 . 1.5)
        \override Hairpin.minimum-length = #10
        b4\> b\! 
}


Cheers,
  Harm








--
Sent from: http://lilypond.1069038.n5.nabble.com/User-f3.html