The point size refers to the overall height including the ascenders and
descenders (tails up on letters like bdfhl and down like gjpqy) and any
other protrusions like the height and depth of brackets. It originally
referred to the size of the body of metal on which type used to be cast.
It rarely has much to do with the actual height of capitals or the body
of lowercase letters. So 12pt Times will look quite different from the
12pt of a different typeface, quite apart from differences in shape).
But 12pt TNR from one typefounder should be very close to 12pt TNR from
another. Can you create a reliably-sized image of the two side by side
at 100% magnification so we can see?
///Peter
The name of an optic size of a font.
> Is it my imagination, or is MS Word's Times
> New Roman in 12 point font much smaller than 12 point font when using
> LaTeX?
Times for Word is not the same Times for LaTeX. It could be some
differences. In theorie, a 12 point font is a font which provides a
parenthesis which is 12 point tall. Well, it's the theory. In practice,
the font designer does what he wants!
The <n> point size for a font is not precise at all. Look closely the
result below: all the fonts are 12 point ones.
\documentclass[12pt]{article}
\usepackage[T1]{fontenc}
\newcommand*\showexample[3]{%
\begingroup
\usefont{T1}{#1}{#2}{#3}%
(abp)%
\endgroup
}
\begin{document}
% avant garde
\showexample{pag}{m}{n}
% bookman
\showexample{pbk}{m}{n}
% charter
\showexample{bch}{m}{n}
% courier
\showexample{pcr}{m}{n}
% helvetica
\showexample{phv}{m}{n}
% new century
\showexample{pnc}{m}{n}
% palatino
\showexample{ppl}{m}{n}
% times
\showexample{ptm}{m}{n}
% Utopia (via fourier)
\showexample{futs}{m}{n}
% Zapf chancery
\showexample{pzc}{m}{it}
\end{document}
Jean-Côme Charpentier
Smaller when printed out, or just smaller on the screen?
It's rather the baselineskip (which is not the size of the font).
Isn't it?
>
> It rarely has much to do with the actual height of capitals or the body
> of lowercase letters. So 12pt Times will look quite different from the
> 12pt of a different typeface, quite apart from differences in shape).
>
> But 12pt TNR from one typefounder should be very close to 12pt TNR from
> another.
Do you mean the same typeface from different founders? Elsewhere,
it's wrong:
\documentclass[12pt]{article}
\usepackage[T1]{fontenc}
\newcommand*\showexample[3]{%
\begingroup
\usefont{T1}{#1}{#2}{#3}%
T%
\endgroup
}
\begin{document}
% avant garde
\showexample{pag}{m}{n}%
% bookman
\showexample{pbk}{m}{n}%
% charter
\showexample{bch}{m}{n}%
% courier
\showexample{pcr}{m}{n}%
% helvetica
\showexample{phv}{m}{n}%
% new century
\showexample{pnc}{m}{n}%
% palatino
\showexample{ppl}{m}{n}%
% times
\showexample{ptm}{m}{n}%
% Utopia (via fourier)
\showexample{futs}{m}{n}%
% Zapf chancery
\showexample{pzc}{m}{it}%
\end{document}
In this example, all the fonts are Adobe ones.
Jean-Côme Charpentier
> What is a 12 point font?
There are two ways of looking at this.
Once upon a time, when a fount[1] of type was a collection of type
metal, fount designers designed a fount with the letters and whatnot to
look good at a particular size.
Generally speaking, you'd find that the larger sizes had letters that
look tall and thin compared to the smaller sizes.
When Knuth created TeX, he decided to carry on doing things this high
quality way - so the Computer Modern family of founts has (for example)
cmr5 and cmr10 - Computer Modern Roman designed to be printed at 5pt,
and a different version of Computer Modern Roman designed to be printed
at 10pt. Either one can be scaled to any size you like - but they're
meant to look right at their design size.
When almost everyone else looked at computer typesetting, they decided
to create a single fount in each design (say, Times), and that fount was
meant to be a good compromise so that it'd look okay at whatever point
size you scaled it to. I recall reading that the target size is roughly
what you'd expect for an 18pt fount designed the old-fashioned way.
So in the TeX world, a 12pt fount is a fount designed for use at 12pt,
while everyone else thinks that a 12pt fount is a fount scaled to 12pt.
> Is it my imagination, or is MS Word's Times
> New Roman in 12 point font much smaller than 12 point font when using
> LaTeX?
Are you comparing MS Word's Times New Roman with the `standard' founts
used with LaTeX - that is, the Computer Modern family?
Founts from different families often look very different sizes even when
scaled to the same nominal point size. It's all a bit complicated -
but, for example, one of the major factors behind how big a fount looks
is the ex-height (the nominal height of the lower case `x'), while the
point size is a different measure which has more to do with the overall
height of the maximum extent of the letters than anything else.
Rowland.
[1] Everyone else these days uses the spelling `font'. I don't like
it.
--
Remove the animal for email address: rowland....@dog.physics.org
Sorry - the spam got to me
http://www.mag-uk.org http://www.bmf.co.uk
UK biker? Join MAG and the BMF and stop the Eurocrats banning biking
No, because I could be using a 12pt type on a 14pt baseline.
It's still 12pt type though. The baselineskip setting has no
effect on type size, only on the spacing.
>> It rarely has much to do with the actual height of capitals or the body
>> of lowercase letters. So 12pt Times will look quite different from the
>> 12pt of a different typeface, quite apart from differences in shape).
>>
>> But 12pt TNR from one typefounder should be very close to 12pt TNR from
>> another.
>
> Do you mean the same typeface from different founders?
Yes.
> Elsewhere, it's wrong:
No, see what I wrote. Different typefaces look to be different sizes
even if their nominal (design) size is 12pt.
///Peter
One way to understand this is to go to a print shop and see some metal
type for yourself.
If you look at some 12pt sorts (which is what individual pieces are
called) and measure them, you'll see that the pieces of metal (or wood,
in some larger fonts) are 12pt tall.
If you now look at the letters on the sorts, you will see that they take
up only part of the available height.
Most typefaces include extra padding space both above and below. This
is included when you measure the font height. Think of small caps
fonts, for example, in which none of the letters takes up the full
height. (A common exception is large wooden type used for printing
posters.)
The next complication is differencing between type founders. They
modify the letter heights and proportions to suit themselves, sometimes
for improvement, other times for less savoury reasons. So you SHOULD
expect different fonts with the same name to look slightly different,
though in many cases the differences can be hard to spot. You could
compare Arial with Helvetica and Swiss, or Book Antiqua with Palermo and
Palatino. I remember when Arial first appeared in the 80s. It was so
clumsily proportioned that it looked conspicuously like a cheap rip-off.
It also had a name that didn't seem to relate to what it mimicked.
But it was cheap, and the mass market didn't know any better, so it took
off strongly in the low-budget Windows world while Adobe Helvetica and
Bitstream Swiss went more to Macs and graphics designers. Arial has
been improved since then.
Are you using the same Times New Roman for LaTEX as you are for Word?
Or are you using a Microsoft version for one, and an Adobe or Metafont
or other version for the other?
You might also compare titling (or display) fonts with body fonts of the
same face and size.
How to measure the font size:
If you know the specifications of a font, you can deduce its type size
from the x-height, which is the height of the lower-case x. The
required knowledge includes the place where and time when the face was
designed, because units of measure, including the point, have a history
all of their own.
You cannot deduce the point size from inter-baseline spacing on a
printed page, because you can't tell how much leading is in there.
Leading is flat strips of metal that lie between the lines of type.
The point size matters for spacing.
A quad is a square space, as wide as it is high. So in a 12-pt font,
one quad is 12 points wide. That distance is called an 'em'. It has
less to do with the width of the lower case m than is popularly claimed.
This is especially obvious if you consider expanded and condensed
faces where the quads do not change even though the letter m does.
Standard spaces are one third of an em (which can be selectively
exchanged for quarters and fifths of ems for justification and other
adjustments).
If you look at what happens when you turn a space on its side (e.g. a
60pt en becomes a 30pt double quad) you'll start to get some inkling of
why the type in a manual print shop comes in only certain sizes.
Computer typesetting has moved on from much of this, and there are many
typefaces that deliberately break the 'rules'. But there are also a
great many computer typesetters and graphic designers who insist that
the best way to get good layout is to learn the conventions of manual
typesetting, which had several centuries to rid itself of bad design.
You could have a look at lulu.com if you want to see some examples of
books typeset by people who aren't all that interested in this kind of
thing, but who have had immense power put into their hands by the new
technology. In the old days, you had to go through an apprenticeship at
the very least before you could freely produce hardcopy output, and a
printing press with type cost much more than a computer and laser
printer do now. Early Mac users did the same sort of thing by including
a dozen fonts per page.
Systems like Metafont can try to draw upon the accumulated knowledge of
ages, but when you come across Knuth's claim that point sizes are
arbitrary, one has to wonder what kinds of sources he was learning his
typography from at the time. It strikes me as odd to get the niceties
of re-proportioning the opticals, and yet to miss the absolute basics of
type size. Then again, his Computer Modern strikes me as one of the
most peculiar typefaces in common use. Like Courier, it is clear, but
also so grossly mannered as to draw so much attention to itself that it
detracts from the text that it's meant to be conveying. But this is
just opinion. Each to his own. And if you're exposed to enough of it,
you will get used to it, as with most aesthetic things.
I was talking about your "size of the body of metal...". It seems to
me that this size is not that TeX calls "font size". Instead, it is the
"normal" \baselineskip. Maybe I'm wrong, I don't know very well the
story of the old typographic technics. If I'm right (not sure at all), a
parenthesis in the body of metal don't extend on the whole piece: there
is a gap below and above the parenthesis. If I'm wrong, it was another
metal pieces which had to be inserted between lines of text.
Jean-Côme Charpentier
Both. I should have specified I was comparing Word's default font
(Times New Roman 12 point) with LaTeX's default font (Computer Modern
something) in 12 point. Word's standard 12 point font seems to me
closer in size to LaTeX's standard 10 point font. I'm sometimes asked
to submit papers in 12 point font, double-spaced, with 1-inch margins,
but I'm not sure how these requirements should be fulfilled using LaTeX.
You certainly should have. There's no point comparing apples with
oranges. Different fonts look different sizes, for a number of reasons.
> Word's standard 12 point font seems to me
> closer in size to LaTeX's standard 10 point font. I'm sometimes asked
> to submit papers in 12 point font, double-spaced, with 1-inch margins,
> but I'm not sure how these requirements should be fulfilled using LaTeX.
Easy. Assuming they mean Times New Roman: (if they don't specify, a
trick to make your document look longer is to use Palatino -- for
example -- at 12pt which takes up more space than Times)
\documentclass[12pt]{article}
\usepackage{mathptmx}% times font, including maths
\usepackage{setspace}
\doublespacing
\usepackage[margin=1in]{geometry}
\begin{document}
There is no just ground, therefore, for the charge brought against me
by
certain ignoramuses---that I have never written a moral tale, or, in
more
precise words, a tale with a moral. They are not the critics
predestined
to bring me out, and \emph{develop} my morals:---that is the secret. By
and by
the ``North American Quarterly Humdrum'' will make them ashamed of
their
stupidity. In the meantime, by way of staying execution---by way of
mitigating the accusations against me---I offer the sad history
appended,---
a history about whose obvious moral there can be no question whatever,
since he who runs may read it in the large capitals which form the
title
of the tale. I should have credit for this arrangement---a far wiser
one
than that of La Fontaine and others, who reserve the impression to be
conveyed until the last moment, and thus sneak it in at the fag end of
their fables.
\end{document}
> I'm sometimes asked
> to submit papers in 12 point font, double-spaced, with 1-inch margins,
> but I'm not sure how these requirements should be fulfilled using LaTeX.
When people ask for this, 12 point font (in a technical sense) is not
what they really mean. They mean that the letters (probably best gauged
by the x-height) should be large and legible.
In most cases, you also want to avoid picking a typeface that the reader
will find distracting for its unusualness.
So there's reason to pick Times or Palatino in twelve points.
The margins might seem a bit narrow, but double-spacing restores the
legibility, in most cases at the price of elegance.
In 'real' typesetting, there's no such thing as double spacing, we think
of leading instead, though even that's a misnomer if you use wooden
friskets. Double spacing belongs to typewriter culture.
In manual lead typesetting, to get a 12-on-14pt typesetting, one
placed a 2pt thick peace of lead between the lines. This is called
"leading" in English ("Durchschuß" in German). Actually, I always
assumed that this term has "lead" as an etymologic [sp?] root, but
that's hard to tell without further research.
Cheers,
Joachim
PS: AFAIR, Peter has experience with manual lead typesetting. You
can trust him when he explains something like that that he's right.
(For the math inclined: probability is near 1 that he is. :-)
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Joachim Schrod Email: jsc...@acm.org
Roedermark, Germany
Thanks to this answer. I don't know it.
> This is called "leading" in
> English ("Durchschuß" in German). Actually, I always assumed that this
> term has "lead" as an etymologic [sp?] root, but that's hard to tell
> without further research.
>
> Cheers,
> Joachim
>
> PS: AFAIR, Peter has experience with manual lead typesetting. You can
> trust him when he explains something like that that he's right. (For the
> math inclined: probability is near 1 that he is. :-)
Since the probability that I'm right was not defined, I choose the
probability near 1 without hesitation :-)
Jean-Côme Charpentier
fwiw, i went looking for the etymology of this meaning of lead, and
it's grouped in the oxford english dictionary under the metal (8th
meaning of that classification of the word). the etymology given is from
old english l\'ead, together with a bewildering array of (mostly
n. european) languages that had similar words (e.g., old frisian
l\^ad).
i particularly liked the following citation of the word, which seems
to categorise extended leading with hysteria. tell that to these
specifiers of thesis styles, eh? ;-)
1889 Harper's Mag. Apr. 819/1 A newspaper which..avoids double
leads..and all forms of typographical hysteria.
--
Robin Fairbairns, Cambridge
I think that the short answer is: In these days of electronic typesetting
it's whatever the font designer says it is (although font designers tend
not to deviate too far from the norm).
--Paul Vojta, vo...@math.berkeley.edu
Indeed, asking for a 12pt font dates back to the old days, when there
were typewriters (mechanical or otherwise) where you could change the
size of the typeface. A 12pt font meant something like Courier New
12pt: a large typeface with serifs, easy to read and annotate.
FWIW, my font of choice is Concrete Modern Roman 12pt: simply add
\usepackage{ccfonts} to your preamble. The typeface is far more
regular than Times or Palatino, and its edges imitate something of the
old Courier typewriters, yet it's still a proportional font: you keep
the look and feel of typewriters of lore, with all the elegance of a
proportional serifed font. With that and the fullpage and spacing
packages, you can imitate papers of old times.
Good Luck,
Luis.
Even in the days of metal the apparent size of a font could vary quite a
lot.
For example Times was designed to be used on newspapers on fixed leading, so
the glyphs were cut small so that the lines could be set without using any
extra lead, so saving money.
Helvetica is cut big. Put that against Times and you will see the
difference.
--
Eric
Leading comes in multiple thicknesses and we also use wooden strips
called 'friskets' but yes, this is indeed the process.
> PS: AFAIR, Peter has experience with manual lead typesetting.
I also still do it the old way at times. It has its advantages...
% Indeed, asking for a 12pt font dates back to the old days, when there
% were typewriters (mechanical or otherwise) where you could change the
% size of the typeface. A 12pt font meant something like Courier New
% 12pt: a large typeface with serifs, easy to read and annotate.
In the "old days" of typwriters with interchangable type balls, sizes
were typically specified in terms of the number of characters per inch.
The equivalent of "12pt" was probably "10 pitch". Asking for a 12pt
font is certainly tied to word processor culture, and they probably
mean the equivalent of "ptmr8r at 12bp" in TeX speak (unless they
ask for APA style, which I think expects Courier).
--
Patrick TJ McPhee
North York Canada
pt...@interlog.com
You're right; that's what I meant. Sorry for the confusion...
Luis.