[smufl-discuss] Re: Glyph registration proposal

[Emil B. Wojtacki]

Daniel wrote:

>>>> TIME SIGNATURES
>>>>
>>> > >I would add, that they should vertically centred on the baseline.
>> >Then, the scoring application draws them on the 4th (numerator) and
>> >2nd (denominator) staff line (this behaviour should be possible to
>> >override by the user, e.g. for "handwritten" styles).
>>> > >It is because for very small staves, the staff lines are
>> >relatively thick, so the height (i.e. the size) of the digits should
>> >be adjusted so that the staff lines do not overshot the digits too much.
>> >
>> >I agree with your point -- I think this was an oversight in my
>> >proposal. You are correct, time signature digits conceptually fit
>> >within a vertical extent, or are centered around some baseline.
> How should fonts that require or prefer larger digits work? For example,
> several handwritten fonts feature time signature digits that protrude
> above and below the staff, but still abut at the middle staff line. If
> these digits were drawn centered on the 2nd and 4th staff lines per Emil's
> proposal, this wouldn't be possible. Unless I'm missing something (which
> is by no means unlikely).

Let me refer to the "classical" fonts first. In some circumstances, the
font size for time signatures may differ from the staff size. For
instance, when staves are small, staff line width gets relatively thick
(like 0.16 or 0.19 spaces; some raterizers make thin lines thicker, so
that no line is thinner than 0.5 PS point). With this staffline
thickness, time signatures digits are overlapping the staff lines too
much, so an experienced engraver makes them smaller (and changes the
font to a wider/bolder one, if possible). The correct behaviour of a
scoring application should be the following:

- numerators are still centred on the 4th line
- denominators are still centred on the 2nd line

I think that the easiest way to achieve this behaviour is that the time
signature is drawn within a single text object which has the following
properties:
- the digits in the first row are centred on the 4th line
- the leading is fixed to the half of the staff size (not depending on
the font size nor on font metrics stored in OS/2 table).

For this purpose, it is the most convenient to have the digits centred
explicitly on the baseline (like in Opus and Petrucci). Then, if font
size is changed, no additional corrections of positioning time
signatures are necessary.

For the handwritten fonts with protrusions you mention, I would suggest
that the digits are supposed to be primarily  numerators (so they are
positioned still on the 4th line), and the additional metadata file
stores the value for the proper leading (expressed in staff spaces).

I am tempted to suggest, that perhaps C and ¢ should be positioned so,
that they are placed correctly if font baseline is aligned to the centre
of 4th line.


[FLAGS]
>> >I don't favor any registration rules that "bake in" a standard stem
>> >length because there are so very many exceptions and adjustments to
>> >stem sizing and placement relative to note heads, many more than the
>> >multivoice case you mention. Stem length doesn't feel to me as
>> >though it should be part of the standard.
> I agree with Joe here. The approach Joe proposes is basically what most
> music fonts (including the venerable Sonata) seem to do anyway.

Yes, but currently there are notes with octuple flags encoded in SMuFL.
Their height can vary from font to font (it does not make a visible
difference for semiquavers, but I can imagine, that even with quadruple
flags, the stem height may vary in a range of 0.5 space depending on the
design). The metadata file must store the additional stem height for the
different rhythmic values anyway. My strong feeling is that the
additional length of stem should be easy to figure out  from the glyph
registration (i.e. by comparing the Y position of the upper extreme of
quaver flag and 1024-note flag)
The other adjustments that were mentioned are what they are - exceptions
and adjustments, that must be handled additively, anyway.



[RESTS]
> I'm not sure about this. The positioning of e.g. a bar rest is different
> when on a one-line staff than on a five-line staff, since a bar rest hangs
> from a one-line staff, but a half rest sits on top of a one-line staff,
> which is the exact opposite of their normal positions relative to each
> other in the more common five-line staff case. For what it's worth, Joe's
> proposal also appears to be the convention followed by Opus, Sonata and
> Petrucci.
>
> My vote would be to stick to Joe's original proposal in this area.
It is only a semibreve (and on even) rest that must be placed in a
special way. Other rests are actually placed relative to the middle line
(which is also a case of 5-lines staff). And this seems to apply to any
staff: think about a 2-lines staff for percussion (with widely spaced
lines, like 3 spaces between lines): minim (and breve) rest sits on the
bottom line and semibreve rest hangs down from the top line, other rest
are somehow centred in the staff. Things get more complicated, if you
consider polyphonic notation with rests above/below staff (on ledger
lines) in 4/2 time. Minim rest following a semibreve rest may be placed
on the ledger line that has the same Y position, or one space lower or
it can be a black rectangle positioned exactly like the semibreve rest,
just with the ledger line placed on the opposite side.

In fact, Petrucci and Opus have rests centred on the font baseline (so I
did not propose anything new) with exception for semibreve rest, which
is placed one space higher. But for the easy of handling rests on
non-standard staves (which are standard for percussion) and for more
flexibility with polyphonic notation in the old-style alla breve time, I
suggest this one change concerning the semibreve rest.

I would like to add also the following:

[FONT METRICS]
It could be a good idea to have the font baseline in the middle of the
em, so that F-clef and other symbols do not jump below the em too much.
A standard five-lines staff would be then centred in the em, with the
5th line on font ascent and bottom line on font descent.


#############################################################
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]>