Re: Choice of pitch input mode

[David Wright]

On Fri 29 Apr 2016 at 17:52:16 (-0300), Caio Giovaneti de Barros wrote:
> 
>  On 4/29/16 7:27 AM, "Carl Sorensen" <address@hidden> wrote:
> >>About a year ago, Kieren indicated that he has decided to go completely
> >>absolute mode, even to the point
> >>of redoing his historic code:
> >>
> >>https://lists.gnu.org/archive/html/lilypond-user/2015-04/msg00846.html

One hopes that he is aware of the ly script that can make this a lot
easier than it otherwise might be (though I'd couple it with "find"
rather than involve Frescobaldi).

> I was not aware of Kieren's opinion, but after my last work on
> Lilypond I'm arriving at almost the same conclusion. At least for
> music that works outside the tonal (and in a way also modal)
> paradigm of pitch ranges and intervals it is preferable to use
> absolute mode almost always. What a pain it is to write, say, a line
> by Webern in relative mode. And if you make a mistake along the way
> (and you will make it) you have to be prepared to hunt the right
> note to correct the octave next. At least with Frescobaldi you can
> click on the pdf viewer and it puts your cursor directly where you
> want, which is awesome.
> 
> >>
> >>I find his arguments interesting.  And if I were engraving things as
> >>complex as his, I might also move to absolute.
> >>
> >>But I am working on simple, short, pieces.  The longest I have done to
> >>this point is about 8 pages.
> >>For my use case, the convenience of less typing (and to be fair, not
> >>having to worry about the octave
> >>I'm using) overrides the inconvenience of the occasional octave mistake.
> >>
> >>So I use relative.
> 
> I do believe that at the end this is a choice the engraver has to
> make based on the type of music she or he is working with. Many
> decisions in code organization comes down to convenience, your
> habits and comfort, but just like music engraving in general there
> are good practices to make your work more productive, and if you
> need to be read and understood by others, this is even more
> critical.

I certainly think that absolute mode suits LP archives because it
lacks any ambiguity, so Kieren's converting all his historic code is a
sensible move. But the Subject line says "input mode" and then it's a
matter of horses for courses. Most of my music input consists of vocal
lines written out conventionally, so I find relative far easier and
faster. (Thank goodness octavation ignores accidentals.)

However, I find absolute easier if a voice is fragmentary; for example,
I write vocal performing scores where the accompaniment is sketched in
only when noone is singing (so they don't get lost). Similarly, when
writing keyboard parts where you sometimes have to juggle the voices
to get LP to render it correctly. (I don't enjoy that.)

But it's no surprise that composing directly into LP is only really
possible in absolute mode. The nearest I ever get to that is
transcribing from a recording, and anyway I do this voice by voice in
relative mode (making many more mistakes).

I haven't yet used the new duration-only mode. I've never investigated
what ly is able to do with its reformat option, but it would be great
if it could convert into a canonical style, where canonical could be
defined in ways such as: every note with pitch&duration; duration (or
even pitch) on only the first note of each line (omitted elsewhere);
single or zero space between each pitch&duration pair; and so on.

Cheers,
David.