\vsize108pt
\article{How well do you know your CM fonts?}
I am sometimes surprised at the uncritical way in which people use
Computer Modern  fonts, especially on 300\,dpi laser printers.
There are two problems:
{\vadjust{\vfill\eject}} 
\bi CM is a rather lightweight face, and
it seems to break up quite quickly at low resolution; 
\bi most
people use a small set of fonts, magnifying them on demand, rather than 
having a correct font for every point size needed. 

\vfill\eject
\noindent The reason often
quoted for the latter is the disc space that fonts take up, but  we
should recall that only extra {\tt tfm} files are needed, as the {\tt pk} (or
{\tt pxl} or {\tt gf} font files will differ little in size between cmr10
magnified 3 times, and cmr17. 

In an effort to show the differences between the `correct' font
sizes and the magnified fonts, I have
generated a set of tables designed to show a \LaTeX\ user what he or
she can expect to see at all the supported sizes and font styles; a
comparison between the first table (page 33), generated with the default
fonts (frequent magnifications), and the second (%
this page),  using fonts generated at the right design  sizes, should demonstrate
even to the casual eye that the difference is considerable. 

It is instructive to compare these Computer Modern
tables with the third table (page 35),  generated using \PS\ fonts -- the
reader is left to decide which of the three he/she finds preferable.


The `true-sized' \LaTeX\ fonts can be accessed by building a new
\LaTeX\ with a revised `lfonts.tex' 
based on fonts using John Sauter's scripts a time-consuming, but
interesting, exercise. The Aston archive contains\nl 
{\tt[tex-archive.fonts.sauter]}\nl
to enable the `true-size' fonts to be built,
and\nl
{\tt [tex-archive.latex.contrib]lfonts.moretrue}
\vfill\eject\noindent
to allow \LaTeX\ to be rebuilt to take advantage of them.

It had been intended to print these tables with a comparative set at
1270\,dpi printed on a phototypesetter; unfortunately, experiments so
far have simply produced VM errors on the \PS\ interpreter on the Linotron
300. If the charts ever appear, they will be made available to \TeXline.
\author{\sl Sebastian Rahtz}
\vfill\eject
%Computer Science
%University
%Southampton S09 5NH

%spqr@uk.ac.soton.ecs