Thanks a lot David and all,
It's quite clear now,
the fact is by now I write everything in relative mode, so I take notice of
your last remark, and will look further if this habit could/should be amended.
Cheers,
Paul
Le sam. 24 10 2026 à 17:33, David Wright a écrit :
> At the risk of repeating some of what has already been said, the
> \relative { }, \fixed { } and \absolute { } constructions are input
> methods/shorthands that have the effect of "baking in" the octavation
> of the enclosed notes. A naked { } leaves them floating, uncommitted
> until they are used.
>
> So in your second example, the enclosing \relative c'' has actually
> been made impotent because the inner \relative's have already baked
> those notes into absolute pitches. It could just as well have been
> written as \relative c,, instead.
>
> Here's an example of uncommitted notes, where \jjj is set differently
> according to where it is placed:
>
> %%
>
> jjj = { c g c }
>
> \relative c'' {
> <<
> { c e c g } \\
> { c, \jjj }
> >>
> }
>
> \relative c'' {
> <<
> { c e c g' } \\
> { c, \jjj }
> >>
> }
>
> %%
>
> By way of contrast, \kkk below is always set at the same octave:
>
> %%
>
> kkk = \absolute { c g c }
>
> \relative c'' {
> <<
> { c e c g } \\
> { c, \kkk }
> >>
> }
>
> \relative c'' {
> <<
> { c e c g' } \\
> { c, \kkk }
> >>
> }
>
> %%
>
> The last example shows a few other points:
> you can specify the relative octave with the first enclosed note,
> IOW as if it was written \relative f { };
> baked-in sequences of music included within a \relative { } sequence
> do not affect the \relative's own octavation outside the inclusions;
> and enclosed sequences do not pick up the notelengths preceding them.
>
> %%
>
> unbaked = { g e e c }
> baked = \absolute { g e e c }
>
> \relative {
> c' e g2 c \unbaked e g c \baked c e c
> }
>
> %%
>
> BTW I would recommend using \relative { } et al only as input methods
> for variables like \voice's and simple lines of music, and not
> writing any \staff or larger structures within them.
>
> Cheers,
> David.