Re: On the meaning of ``single-space'' inter-line spacing

[Michael Piotrowski]

On 2006-06-28 address@hidden (Ludovic Courtès) wrote:

>> Engineers shouldn't write things like "6-1/2 inches".
>
> Admittedly, this looks weird, but I thought there were good reasons for
> writing it this way.  You seem to imply that it's gratuitous weirdness.
> ;-)

Well, maybe there are good reasons, but I don't know any.  I think
it's very easy to misread 6-1/2 (6.5) as 6 - 1/2 (5.5).  Some of the
measurements in centimeters are also quite inaccurate.  Ok, it doesn't
really matter in this case (and it's off-topic, too), but when you
give two decimal places (i.e., tenths of millimeters) you're
suggesting a certain precision, but in fact it's sometimes more than
1 mm off.

[...]

>> As an aside, due to the continued use of the body height (i.e., the
>> height of the--nowadays purely hypothetical--lead block on which the
>> letter sits) for measuring type instead of measurements which relate
>> to the *actual* appearance of the typeface (such as the caps height),
>> leading specifications are only valid for *one* specific typeface.
>> Different typefaces may have very different caps heights at the same
>> nominal type height: It's well known that Helvetica is much larger
>> than Times at the same nominal size.  This also applies to leading:
>> What may be fine for one typeface, may be too tight or too loose for
>> another.
>
> Does that mean that when one selects, say, an 11pt typeface, the highest
> glyph in that typeface will be _at most_ 11pt high?

Well, kinda.  See the attached example.  The specified type size is 30
pt; you can see that you get 30 pt + a bit if you measure from the
descender of the p to the ascender of the k, which is equivalent to
the body size in metal type.  You could thus use 30 pt as the minimal
line type.  However, you can also see that this won't necessarily work
with accents.

And of course, an application could give you something else.  The
problem is that it's all not well-defined.

> OTOH, AFM [0] defines `CapHeight' as the height of the letter `H' in the
> font, `XHeight' as the height of the letter `x', plus the ascender and
> descender size.  That would imply that AFM-using programs know only the
> actual font height, not that of a hypothetical lead block.  Is this
> correct?

No, CapHeight and XHeight are additional (optional) information.  The
body height isn't needed in the AFM, because that's the normal way of
selecting the font size in PostScript: If you say

  /Helvetica 12 selectfont

you get a body height of 12 pt.  A program using the additional
information from the AFM file could, for example, allow the user to
specify the caps size or it could automatically scale typefaces so
that their x-heights match or for special effects, like drawing boxes
around characters.

Greetings

-- 
Michael Piotrowski, M.A.                               <address@hidden>
Public key at <http://www.dynalabs.de/mxp/pubkey.txt>

typesize.ps
Description: Example