Re: MIDI dynamics parsing error

[David Kastrup]

Thomas Morley <address@hidden> writes:

> 2014-05-18 12:28 GMT+02:00 David Kastrup <address@hidden>:
>> Thomas Morley <address@hidden> writes:
>>
>>> 2014-05-18 7:20 GMT+02:00 David Kastrup <address@hidden>:
>>>> Knute Snortum <address@hidden> writes:
>>>>
>>>>> My situation is I need to transcribe a crescendo starting without a 
>>>>> dynamic
>>>>> marking.  If I put in a \mf then I don't get warning.  If I can hide the
>>>>> dynamic mark, all will be fine.  Is there a \hide syntax that will work?
>>>
>>> Omitting the DynamicText.stencil will not cause the Hairpin to start as 
>>> usual.
>>> Resetting the parent of the Hairpin my work:
>>>
>>> \version "2.18.0"
>>>
>>> \score
>>> {
>>>   {
>>>     \once \override Hairpin.after-line-breaking =
>>>       #(lambda (grob)
>>>          (let ((dyn-txt (ly:spanner-bound grob LEFT)))
>>>            (ly:grob-set-property! dyn-txt 'stencil #f)
>>>            (ly:spanner-set-bound! grob LEFT (ly:grob-parent dyn-txt X))))
>>>     c'1\mf\> c' c'\!
>>>   }
>>>   \layout {}
>>>   \midi {}
>>> }
>>
>> If \omit is too strong and \hide leaves too much space, maybe overriding
>> the stencil with point-stencil will work?
>
> Using point-stencil doesn't work either.
>
> \version "2.18.0"
>
> \score
> {
>   {
>    \override DynamicText.stencil = #point-stencil
>     c'1\mf\> c' c'\!
>   }
>   \layout {}
>   \midi {}
> }

Ok, I'll bite: what does "doesn't work" mean?  I see the same results
for overriding with point-stencil and omitting, and those results are

a) without error message
b) looking as expected when compared to the version with visible \mf

Either way, I find the stopping position of the hairpin somewhat
strange/early, but it is exactly the same as it would be without
override.

-- 
David Kastrup