[comp.text] Quality of TeX fonts

abrueggemann@watdaisy.UUCP (03/04/87)

In this newsgroup and in publications on document processing, there
have been several comments on the quality of TeX's built-in fonts.
The old am-fonts are described as horrible, and the new cm-fonts
are described as acceptable or even good.

Frankly, if I compare documents set with the old or the new fonts,
I can hardly see a difference. I am aware of the fact that very
subtle properties that can only be detected by specialists
can decide on the quality of a font family. I'd like to see a
list of improvements of the new cm-fonts compared
with the old am-fonts.

Any pointers or comments?

-- 
  UUCP:	  ...!{clyde,utzoo,decvax,ihnp4,allegra}!watmath!watdaisy!abrueggemann
  ARPA:	  abrueggemann%watdaisy%waterloo.csnet@csnet-relay.arpa
  CSNET:  abrueggemann%watdaisy@waterloo.CSNET

maslak@sri-unix.ARPA (Valerie Maslak) (03/05/87)

 I think Computer Modern is UGLYYYYYYYYY.

Just an opinion.

Valerie Maslak

langdon@lll-lcc.aRpA (Bruce Langdon) (03/06/87)

In article <8176@watdaisy.UUCP>, abrueggemann@watdaisy.UUCP writes:
> In this newsgroup and in publications on document processing, there
> have been several comments on the quality of TeX's built-in fonts.
> The old am-fonts are described as horrible, and the new cm-fonts
> are described as acceptable or even good.
> 
> Frankly, if I compare documents set with the old or the new fonts,
> I can hardly see a difference. ...

The first 'cm' output I saw looked thin and anemic (note the craftman's
jargon here). Then I saw 'cm' on another device, and it looked fine.
Independent of their appearance on perfect output devices, there are
VERY important judgements to be made when generating the font files
for a printer. Metafont, the program that does this from the language
that describes characters and their properties, has some parmeters
for this. 
Also, there are two kinds of laser printers. Some write black, others
erase white. For real-world printers, the font descriptions must differ.

To agree with your comment, and go further: to my eye, the results
from typical TeX users are so much better than from typical troff,
MacWrite, etc users that the difference between cm and am is minor.
(--but expert troffers, such as those who set my book in 1985, do
just as well, for my purposes.)
Anyway, it seems that output quality from 300dpi laser printers
depends on some device-dependent tuning.
----------------------------------------------------------------------
	Bruce Langdon  L-472                langdon@lll-lcc.ARPA
	Physics Department                  "langdon#bruce%d@lll-mfe.ARPA"
	Lawrence Livermore National Laboratory       
	Livermore, CA 94550                 (415) 422-5444
UUCP: ..{ihnp4,qantel,ucdavis,pyramid,styx,topaz}!lll-lcc!langdon
                  ..{gymble,ll-xn,seismo}!lll-crg!lll-lcc!langdon

ken@rochester.ARPA (SKY) (03/06/87)

The TeXbook, METAFONT book and other books in the series are set in
Computer Roman. I don't find them displeasing.

Here, on our Raven print engine, the native fonts look better, possible
because the Raven is a write-white engine (hope I got that right). On
the other hand, I have seen really anemic troff glyphs from an Imagen
8/300.

So it would seem local printing conditions have a lot of effect on the
appearance.

	Ken

elwell@osu-eddie.UUCP (Clayton Elwell) (03/07/87)

The major improvements I've noticed in going from the "am" fonts to the
"cm" fonts are:

1.  The x-height (the distance from the baseline to the top of a lower
    case "x") was increased.  The main effects of this are to make a
    paragraph of text a more even texture, and to make it more legible
    on low resolution output devices (like 300 dpi laser printers).

2.  The proportions of several characters were twiddled around to make
    the letterforms more consistent with each other.  This makes the
    text more attractive and legible.

Even with these improvements, I still find Computer Modern to be cramped and
inelegant.  This is a matter of personal taste, however.


-=-
====================
Total Nuclear Annihilation:				Clayton Elwell
The Ultimate Error Message.			Elwell@Ohio-State.ARPA
					   ...!cbosgd!osu-eddie!elwell
====================

gary@darth.UUCP (03/07/87)

I have been using TeX for several years on a diverse set of output devices
ranging from Epson printers through APS-5 typesetters and I tend to agree
with the general perception that cmr / amr fonts are not the best.  These
sentiments have been echoed by many TUGboat (TeX User's Group) contribu-
tors and others who have a critical eye for such things.

If you happen to have a LaserWriter, TeX, and dvips from Textset in Ann
Arbor, try printing out the Textset dvips manual and turn to pages 14
and 15.  You can compare amr fonts at 300dpi with Adobe's Times Roman in
the LaserWriter.  Amr looks somewhat conservative---the serifs are stocky
and thick; the proportions not quite right.  On the other hand, the Times
Roman characters are open and slender, with fine complimentary serifs.

In the METAFONT book, Knuth says: "The characters you are reading ...
were developed hastily by the author of the system, who is a rank amateur
at such things."  Considering that Knuth designed a complete family of
compatible fonts, and didn't do too badly, perhaps he's being a bit hard
on himself.  Still, even he admits that the value of METAFONT is its
utility to top-notch typeface designers.

Unfortunately, METAFONT is a much more suitable companion for a
mathematician than for an artist of type design.  METAFONT provides no
interactive way to explore the subtle nuances of shape and relationship
and allow typefaces to evolve visually.  Instead, the type designer must
be constantly aware of the relationship of visual character forms to the
abstract mathematical declarations which generated them.

I believe the complexity of type design with METAFONT is the single largest
reason why we are stuck with amr / cmr.  Given METAFONT's capacity to
generalize font descriptions, so that entire families can be designed more
easily, it would be interesting to see what gems might be produced by
professional typeface designers using tools like METAFONT.

I still feel that TeX is the way to go, however.  I will soon be writing
and typesetting a book which will accompany a new product.  I'll probably
use TeX with Adobe fonts and have the camera-ready made from the PostScript
output.  Few tools provide the flexibility and control TeX does, providing
you have the patience.

Gary Wisniewski
uucp: {allegra, bellcore, cadre}!pitt!darth!gary