[smufl-discuss] Re: Glyph Registration and Graphical Metadata

[David Webber]

From: Mark Johnson

> David: Two objections.
>> It doesn't actually *need* to be adjustable.

> 1. You're talking about taking away an option that millions of Finale
& Sibelius users have had for a quarter century. Depending on the
design of a clef, it might look better at 3/4 or some other size.<

Absolutely not - they can still go on doing it the way they do it (as I said
explicitly) even is a 2/3 size clef is also present in the font.

For a standard font, the size should be 2/3 - Gould is quite specific about
that.  But in a different font (ie one with a different clef design) it
might be included at a slightly different size.  ('Depending on the design
of a clef' translates into 'depending on the design of the font' here.)

> 2. The problem with having 2/3-size clefs as separate codepoints is
that "Unicode does not encode glyphs"; it encodes characters.

Well this is all a bit semantic, and depends on your precise definition of
'character'.  Unicode includes all sorts of symbols, including a telephone,
a cup of coffee, a football, a petrol pump,...    A 'glyph' is just an
'element of writing' a suitably generic term, and includes diacritical marks
(for which there are unicode code points) which I would normally think of as
forming only part of a 'character'.

> If it functions as a G-clef, it's a G-clef regardless of the size or
> design.
Daniel is right that scaled-down clefs need to be character
alternates.<

Just consider that the symbol at the left hand side of the line is a CLEF
and the symbol in the middle of a line is a CLEF CHANGE - they have
different purposes.   They become different symbols, both of which are
required in principle, to draw a single line of music.

Unicode has lots of precedent for this.   Consider the letter 's' in middle
English.   In the middle of a word it was tall (mistaken by some for an f,
but it didn't have the cross bar) but at the end of a word it was drawn like
a modern s.   They're different styles of 's' but in Unicode these are not
put at the same code point:  they're  U+017F and U+0073 respectively.  [They
were also in different boxes in the type setter's font.]

It means you can write 'classics' (as used to be done) with long s in the
middle and a normal one on the end just by outputting a string with the
appropriate characters: you don't have to write 'cla', then switch to a
different font (or font style), then write 'ss' and the switch back and
write 'ics'.    In fact this idea of having all the characters at different
code points in the same font, so that you can do exactly this, lies at the
heart of the raison d'etre of Unicode.

Just as the two forms of 's' have well-defined different usages in the same
text, the two forms of a clef have well defined usages in the same piece of
music.   And it is in exactly the same spirit of Unicode that they should
both be there, at different code points.

> For this reason it now occurs to me to object to having codepoints for
grace notes in the PUA range of SMuFL. They are already in Perry
Roland's spec at U+1D194~5, and they're only there because of the
legacy Sonata character set. (Notice there are no characters for
stem-down grace notes, etc.)

I had.  And, unless I've missed them,  there are no characters for
accidentals on grace notes.   I believe they should be included.  [I'd be
very happy to *exclude* a whole host of things like upside-down and mirrored
clefs, but I guess they're there because somebody wants them, so I'm not
shouting about it.]

All of this depends on whether you are just collecting pictures of musical
symbols or defining a *font*  (originally a collection of wooden trays
holding the movable type required to write everything in the given style)
whose purpose is to facilitate writing music.   I thought the objective was
the latter.   (And I don't regard 'Sibelius doesn't do it like that' as a
definitive argument.)

Dave

David Webber
Mozart Music Software
http://www.mozart.co.uk/ 


#############################################################
This message is sent to you because you are subscribed to
  the mailing list <[hidden email]>.
To unsubscribe, E-mail to: <[hidden email]>
To switch to the DIGEST mode, E-mail to <[hidden email]>
To switch to the INDEX mode, E-mail to <[hidden email]>
Send administrative queries to  <[hidden email]>