Re: Problems with \inversion

[Trevor Daniels]

David Kastrup wrote Tuesday, June 19, 2012 10:57 AM

> "Trevor Daniels" <address@hidden> writes:
> 
>> Colin Hall wrote Tuesday, June 19, 2012 1:12 AM
>>  
>>> It looks like the behaviour has been explained by David Kastrup here:
>>> 
>>> http://lists.gnu.org/archive/html/lilypond-user/2012-06/msg00380.html
>>> 
>>> So, no bug in Lilypond to report then.
>>
>> Correct.
>>  
>>> If either of you, Alexandre or Trevor, would care to create a
>>> documentation suggestion we'll create a tracker for it.
>>
>> OK.  There are two points here.  The first is that it is not
>> immediately obvious that the inversion takes place around
>> from-pitch, and the result is then transposed by the interval
>> between from-pitch and to-pitch.  So simply change the
>> second sentence to "... first inverted interval by interval about
>> from-pitch and then ..."
>>
>> The second point is to insert a warning that the inversion of a music
>> expression in absolute notation should not later be relativised.
>> Maybe say, "The source and inverted music expressions should
>> both be in absolute notation or both in relative notation."  This
>> is not entirely the whole story, but anything more accurate gets
>> very convoluted - unless anyone has a better suggestion.
> 
> "Both in relative notation" is asking for trouble.  I'd strongly suggest
> only inverting absolute music, and the result of \relative ... _is_
> absolute music.

OK, how about this: "If the pitches in a motif are expressed in 
relative form be sure to apply any inversions to the motif outside 
a \relative block, not within it."

Or maybe, "Motifs to be inverted should be expressed in
absolute form or be first converted to absolute form by
enclosing them in a \relative block."

I prefer the second wording.

Trevor