Re: relative mode occasionally gets forgotten?

[Tim McNamara]

On May 15, 2009, at 9:54 AM, Chip wrote:

> Tim McNamara wrote:
> > I'm not sure that the relative mode "gets forgotten" but that  
> > LilyPond follows its own internal rules.  I find that LilyPond  
> > behaves the way that the manual says it does: it picks the closest  
> > pitch.  If I write { c2 a2 } it picks the A below that C rather  
> > than the A above that C.  If I want the A above the C, I write  
> > { c2 a'2 } and all is well.  Sometimes when I first compile the  
> > song there are pitches wildly out of their intended octave; then  
> > it's a matter of finding the proper places to change the pitch  
> > octaves with ' or , and the problem is usually solved with two or  
> > three of these.  Pretty quick and simple without needing to resort  
> > to sticking \relative commands into the  middle of the music  
> > information- something I hate doing because it makes it hard to  
> > "debug" the song.
> Here is an example of the issue I came across -
> r4 r8 ef~ ef d c bf
> a4. g8~ g2~
> g2-- r
> In any mode the tied notes are the same octave, of course. But for  
> whatever reason, with the \transpose and the \relative split as  
> originally described, they were in different octaves, which  
> obviously, broke the ties. Putting the \transpose ef f \relative c'  
> together fixed the problem. With the \transpose and \relative split  
> there were quite a few random instances of such odd octave issues,  
> some would be just a single note or two, others 3 or 4 measures of  
> music where almost every other note jumped octaves up and down.  
> It's really a weird thing to see.
>
> In my template I typically use \transpose x x \relative x but for  
> whatever reason I don't even recall now I moved the \transpose up  
> to the notes section of the code. I think what I did was entered  
> the notes of the tenor sax part in their actual tenor sax key (and  
> all other instruments in concert key), then \transposed them, the  
> tenor sax notes, down to concert key for the purpose viewing the  
> score entirely in concert key. Thus the tenor sax part had the  
> really wicked octave issue.

I recall reading in the documentation for the \transpose command that  
transposition should be done to or from concert to the target key  
using C as one of the arguments for \transpose.  So, if I write a  
song that's nominally in Bb for the concert instruments, then for  
transposing for the Bb horns/Eb horns, etc. is done this way,  
respectively:

\transpose c d  (putting the tenor sax chart in C)

\transpose c a  (putting the also sax chart in G)

So my source code in the .ly file is all in concert, modified by the  
\transpose command.  There is a separate .ly file for each  
transposition, the way I do it, and I print a separate lead sheet for  
each transposition.  There is probably an elegant way to do this from  
only one .ly file and automagically print separate lead sheets for  
each instrument in the right key, but I don't know what it is.

Now, for orchestral music and such where you have a conductor who  
needs to see all the parts at once, my approach might not work very  
well, but for jazz it gets the job done because I don't need all the  
parts on one sheet.  I would think that the

It may be that \tranpose ef f is what's creating your problem by  
confusing LilyPond?