[Menvi-discuss] text-based music systems

[Dale Lieser]

Mark,

 

Several times you have referenced a window you look at to verify, etc. Among
the best ways to help us, in my opinion, would be for you to choose a screen
reader and use it in these programs. The only way you'll gain familiarity
with a screen reader and a given music program is to use them paired. A
limiting factor with Sibelius has always been that the developers have never
tried their program via a screen reader. Neither have the programmers at
Amazon, Schwan's, ad infinitum. It's one thing to ask the blind and visually
impaired what would be desirable. It's quite another to approach the
situation without relying on ones vision.

 

I am excited about your eagerness to provide solutions.

 

Dale

 

 

From: Menvi-discuss [mailto:menvi-discuss-bounces at menvi.org] On Behalf Of
Marc Sabatella
Sent: Tuesday, August 13, 2013 2:38 PM
To: This is for discussing music and braille literacy
Subject: [Menvi-discuss] text-based music systems

 

I'd like to pick up on a number of threads from these discussions, as well
as parallel discussions I've been having on the music21 mailing list
(music21 being the programming environment I mentioned having used to
develop some tool prototypes recently) and a couple of other offline
discussions.

The idea of using text-based notation systems has a lot of merit, I think,
and I glad to see others expressing interest in it.  In separate post, I'll
talk about tool support, but here I'd like to focus on the notation systems
themselves.

I have described some of the reasons I settled on ABC for my work with
Samantha.  I like the relatively simple syntax and the large variety of
tools that can work with it.

I mentioned octave specification as something that I don't particularly love
about ABC.  As with many other systems, each octave in ABC goes from C to B.
Some systems use numbers - eg, middle C is C4, C5 is an octave above (other
systems assign those same numbers differently).  ABC does it by making
middle C be C (capital), the C above that being c (lower case).  Octaves
above the c-b octave get one or more apostrophes (c' c'' etc), octaves below
the C-B octave get one or more commas (C, C,, etc.).  There are ways of
resetting things so that C becomes an octave or two below middle so you can
write bass clef parts without tons of commas.  I don't love the apostrophes
and commas, but it is what it is.

The real problem is that since octaves are always C-based, it's easy to
forget to change octaves in the middle of a line.  For example, "Twinkle
Twinkle Little Star" in G major in ABC is

G G d d | e e d2 | c c B B | A A G2

and you had better remember to start making those letters lower case
starting at c.  Standard notation doesn't have that issue.  I found myself
making octave errors in ABC all the time, although looking at the standard
notation that was displayed in another window made these easy to catch.  Sam
found it harder to find and correct her own errors.

I guess Braille music has a more clever way of solving the octave issue, by
making octave specification relative to context.  For that matter, notation
programs like MuseScore and Sibelius use relative octave specification way
too.  If you type "C" into either of these programs, you get the C that is
closest to the last note you entered.  If I were designing a text-based
language from scratch, I'd consider trying to use a relative octave
specification rather than an absolute one.  And the idea of designing out
our text-based notation language is worth considering.  But developing the
full set of tools necessary to work with it makes this a more long term
project.  I'd still like to concentrate on what can be done right now with
existing tools and maybe just a little bit of extra work on integrating
them.

Someone offline asked about how ABC handles multiple voices.  ABC is
actually pretty elegant here, to a point.  The main mechanism is to have
declarations up front listing all the voices and giving them names (perhaps
just V1, V2, etc).  Then you can switch between them by referencing their
names.  For example, here is a simple two-voice arrangement of "Twinkle
Twinkle Little Star" (and note the octave indications): 

% first declare your voices 
V:upper 
V:lower 
% then reference them by name 
[V:upper] G G d d | e e d2 | c c B B | A A G2 || 
[V:lower] G,2 B,2 | C2 B,2 | A,2 G,2 | D,2 G,2 || 

During the declaration you can give details like which way stems should go,
etc.  The elegant thing here is that this is the *exact* same mechanism used
to define a score with multiple staves.  As far as the music itself is
concerned, it really doesn't matter if these two voices are represented on
one staff or two, and that's how it is with ABC as well.  I gather Braille
might be similar?

If you wish to convert ABC to standard notation and have them display on one
staff, you do that with another declaration that lists how you want the
voices combined.  By default, each voice is its own staff, but the following
declares them to share a staff, by listing the voices you want combined in
parentheses:

I:score (upper lower) 

This system works well if each voice is used consistently through the piece.
For piano music where a voice might just sort of show up for a beat or two
here, it's overkill, but ABC provides an alternative that often works well:
just use an ampersand at the end of a measure followed by the contents of
the temporary voice for that measure. So the above example could have been
written: 

G G d d & G,2 B,2 | e e d2 & C2 B,2 | c c B B & A,2 G,2 | A A G2 & D,2 G,2
|| 

Things get uglier in more complex cases, but it's doable.  There's
practically nothing that can't be represented in ABC.  But realistically, as
a sighted user, I rely on the window showing me the standard notation as I
type my ABC to let me know if I've made a mistake in my syntax.  The more
complex the score, the more likely it is that I might type something wrong,
rendering the whole thing garbage to a program trying to play it back or
convert it to standard notation.  So while it's *possible* to do a whole
orchestra score in ABC, I wouldn't recommend it.  But as a way of notating
short exercises, even moderately complex pieces with some practice, it is
something that can work *today*.

Anyhow, my goal here isn't to convert people to my religion here - if others
are more comfortable with Mup or something else, that's great.  And I'd
still love to hear about what you like or don't like about any of these
systems - or why you feel none of them could work for you.

Again, in a separate post, I will talk more about some interesting tools
that already exist or that could exist.

Marc

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.menvi.org/pipermail/menvi-discuss_menvi.org/attachments/20130813/706fb50b/attachment.htm>