lilypond-user
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Q regarding best practice with \relative


From: David Kastrup
Subject: Re: Q regarding best practice with \relative
Date: Sun, 01 Nov 2015 12:59:59 +0100
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/25.0.50 (gnu/linux)

Michael Gerdau <address@hidden> writes:

> Hello lilyponders,
>
> in (very?) recent versions of the manuals I've seen notations like
> \relative { c' d e f g a b c }
> while older editions and e.g. all the templates in frescobaldi use
> \relative c' { c d e f g a b c }
>
> Obviously both produce exactly the same result.
>
> Is there something like a "best practice" w/r to the use of \relative?
> Or why does this difference exist ?

Historically, \relative { ... } was equivalent to \relative c' { ... }.
This historic use (at one time the only possible one) was deprecated
since c' seems rather arbitrary and it seemed to make better sense to
make the reference point explicit.  Particularly so since
\relative { f } resolved to f' while \relative { g } resolved to g.

The reference point itself was the issue of several conflicting
conventions.  The documentation suggested using an octave of c (but did
not really follow this convention entirely consistently).  Others
preferred using the first pitch in the music, just with the correct
octave added.

After it was pointed out that one could instead just add the correct
octave to the first pitch itself when using \relative f { ... } which
makes the first pitch inside be interpreted just like absolute pitch.
So this started sort of another convention that some people were fond
of.  In contrast to other conventions, it had a somewhat less than
arbitrary reason for its choice of the reference pitch (namely being
able to just write the first pitch absolute).  However, the reason for
this being a somewhat non-arbitrary choice of reference pitch is a
calculated reason (namely f being exactly the middle of the normal
scale) rather than a reason _associated_ with the pitch f in any
discernible manner.

So \relative f { ... } just looked weird because the "f" was just an
artifact of the way the scale is organized while still having some
reasonable justification _not_ really associated with f.

So there was a (not particularly overwhelming) agreement on making this
the default behavior of \relative without explicit reference pitch (it
actually does not use "f" but rather the middle of the current scale, so
when you use something other than the standard diatonic scale, the
salient point of the first pitch being specified absolutely is still
retained).  And after a while, a change to mostly use this convention in
our documentation.

So it was sort of an evolutionarily achieved agreement to go with this
change as it benefits those who like leaving the reference pitch as the
first pitch inside of \relative itself since people with other
preferences can still express them reasonally naturally with an explicit
reference pitch while the reference pitch of "f" detracts from the
_meaning_ of this construct.

There was also weak(er) agreement over using this convention in
preference to others in our documentation examples as it's basically the
only one without an arbitrary element.

So there's not really a "best practice" but rather one that requires
least choice-making for those not interested in making more choices than
necessary.

-- 
David Kastrup



reply via email to

[Prev in Thread] Current Thread [Next in Thread]