Re: What to do wanting a 4th order Bézier?

[Urs Liska]

Am 19.09.2016 um 18:41 schrieb Urs Liska:
>
> Am 18.09.2016 um 22:16 schrieb Thomas Morley:
>> 2016-09-18 20:26 GMT+02:00 Urs Liska <address@hidden>:
>>> Am 18.09.2016 um 18:28 schrieb Simon Albrecht:
>>>> May I incorporate that into the LSR version?
>>>>
>>>>
>>> I suggest not to do anything right now. I assume that once I'll come up
>>> with the next iteration there will be more things to be discussed. I
>>> suggest to bring that to an end before deciding what to do with it.
>>> Maybe it would be good to include to LilyPond itself instead of the LSR.
>>>
>>> Urs
>> Hi all,
>>
>> I have my doubts we will ever find a possibility to do it reasonable 
>> automatic.
>> What to do at linebreaks or if more than two splines are needed?
>>
>> You probably already know the most extreme case, from Sorabji's Opus
>> Clavicembalisticum (1930):
>> http://homes.soic.indiana.edu/donbyrd/InterestingMusicNotation_files/Sorabji_Slur+Ctxt.png
>>
>> The paper where the image is taken from lists some more extreme notations:
>> http://homes.soic.indiana.edu/donbyrd/InterestingMusicNotation.html
>>
>> Cheers,
>>   Harm
> OK, I have prepared a proof-of-concept score (see attached) where we'll
> be able to really test the multiple-segment slurs ;-)
>
> As it is I'm once more astonished about how perfectly readable this
> score is by default (the only manual "tweaks" are some forced \voiceXXX
> in the sixtuplet voice).
>
> I'll upload that to the openLilyLib directory where I'll integrate the
> code, then start with one two-segment slur and take it as an incentive
> to support multi-segment slurs and later even broken slurs.
>
> Urs
>
>
First substantial limitation of my approach: The "slope" of the
inflection can't go backwards but can only approximate a vertical line.
What would be a better calculation for the inflection's angle?

compound-slur-example-sorabij.png
Description: PNG image