\centerline{\bf About Face!}
\medskip
\noindent
Sub-titled ``Design, Choice and use of `Type' in
Electronic Publishing and Page Design'', this
was {\it another} {\sc bcs-ep} meeting, this time
held on February 11th, at the usual venue, the
School of African and Oriental Studies in Bloomsbury.
This was a very well attended meeting, with
many people who were not {\sc bcs-ep} regulars
(as well as a number of familiar faces).

The first talk was from David Saunders of Monotype.
He presented a historical account of the rise
of typefaces, from their early, handwritten origins,
via Trajan's column through the rise of moveable type
to the present day. Naturally covering such a sweep
meant that he was selective, but he was also fascinating,
interesting, educational and thought provoking.

One of his themes was the interaction between design
and technology. This is inevitable, but often the
constraints are not obvious after the event. 
Early typefaces were derived from calligraphy (especially
lower case, for which there were no stone cut originals).
But gradually the calligraphic characteristics disappear,
to be replaced by conventions more suited to the medium.


Saunders also noted that new letterforms arose in the
early industrial revolution, notably sans serifs, which
were then regarded as expressions of the new age.

In comparing quality we were shown comparisons between
various laser printed and `conventionally' printed
letterforms. The key issue here was that with good quality
paper, laser printed material could be of good quality,
and that all conventional printing was not automatically
good. There are (or can be) problems with all types of
printing. A conclusion to be drawn
was that higher resolution does not ensure higher quality.


From looking at the way in which typefaces have evolved,
we looked next at some of the physiological and psychological
bases for recognising the marks on the paper.
John Richardson of Loughborough University described
some of the results of studies which has been done
by psychologists in this area. Most of this work is rather
old now. From the
way in which the eye handles text, it would seem that
speed of reading for type of 9--12 points is about
the same, but that it reduces for both smaller and larger
sizes. Not surprisingly, there is a certain amount
of visual interference in type set solid. The eye
has difficulty skipping to the beginning of the next line.
In the end we would seem to have to quote Eric Gill,
who, in 1931 wrote ``Legibility, in practise, amounts
simply to what one is accustomed to''.

John Durham, assisted by the keyboard skills of Jenny
Hearne,  described the font editing
software which has been developed by Archetype to
permit modification of the fonts used by the large
high volume laser printers (Xerox 9700). 

Essentially the system permits bitmap editing. Durham
argued that hand tuning of algorithmic fonts would still
be needed, and that simply scaling of a font to another
size could not be expected to give good results.
In order to achieve some degree of fidelity between
screen and page, they used a high resolution monitor
--- 100\,dpi, and grey scale fonts. In other words, the
monitor's pixels were not simply black and white, but
black, white and a range of grey.
He also touched briefly on the problem of font licensing,
since of course some of the fonts on these laser
printers are proprietry.

This was an interesting example of a small company
siezing on a niche which the original vendor had
failed to fill (for whatever reason), and exploiting it
rather thoroughly. 

After lunch Kathy Lang 
described some font practicalities. In most ways she
seemed most interested in  larger scale typographic
issues, but she described the sorts of things you expect to
get in the better known dtp packages.
I was surprised to see letterspacing included as
something one expected, since I had just
finished reading in Kleper's book that
``letterspacing should be avoided if possible''. Perhaps this
is a feature incorporated because it is easy to do,
regardless of its desirability (like `strike through' and
underlining --- easy to implement, but useless). 
Lang went on to discuss the relative merits of
`write white' and `write black' laser printers.
One of the audience was able to reveal that the new
Apple LaserWriter II is capable of 400\,dpi
(although it only does 300\,dpi just now).

John Miles, of Banks \& Miles
described the employment of dtp in his design
studio.  Even if you don't realise, you
will be familiar with some of the output of this studio.
They designed the logos for the Post Office and British
Telecom (among other well known corporate identifiers). The studio
uses a couple of Macs with PageMaker and RSG!.
But before describing what they did with the Macs,
he stood back a little to explain what it was they
were trying to achieve in the first place. He identified
a necessary focussing on function. What was the intention
of the material, how would it be used? One of the distinctions he made
was between legibility (the property of a typeface)
and readability (a function of the arrangement of typefaces).

Most of this talk was illustrated with reference to designs
on bits of paper, rather than a screen, but tempered by
the knowledge that they would end up as grids on a
dtp system. The starting point was a sheet of paper,
not the screen.
He noted that it was the job of the designer to
fulfill the function expected by the client, not
to demonstrate the versatility of the equipment.
He also attempted to dispell the myth that there was
anything valuable to be gained by changing typeface
or leading etc.\ `at a stroke'. Any designer worthy of
his hire would know these characteristics before he started
to look at the computer.

The last talk was by the type designer, Richard Southall, who, as
many will know bears some responsibility for some bits of
Computer Modern, and who has a considerable ability
with \MF. But even if you didn't know that,
you would be familiar with Richard's work. He designed
the sub-titling font which is used by the {\sc bbc}
for its films (also seen on Channel 4). 

Early in his talk Richard made the distinction between
\item{$\bullet$}a character image --- the marks which
a marking engine makes for readers to see. This includes
the letters made by laser printers, dot matrix printers, or
even those made on a vdu screen.
\item{$\bullet$}a font --- a set of instructions to a marking
engine to make a character image. This might be a bit map
or an algorithm, or even a piece of type.
\item{$\bullet$}typeface --- the sets of visual attributes for the
characters of a script. As such a typeface has no concrete
existence, it is an abstract object. The nearest concrete
entity is probably the designer's original drawings.
\smallskip
Richard expressed the view that the changes which were
taking place now, since the introduction of dtp in 1984
closely paralleled those which took place between 1965
(with the introduction of photo composition) and the virtual
extinction of hot metal typesetting in 1971.
 He also pointed
out the short period which had elapsed since the introduction
of digital photocomposition (in  1965). To an extent the old
constraints had been removed, such that almost anyone
can now design a typeface.

He pointed out that the commercial pressures to use
outline fonts were overwhelming, but that the
outline was a long way distanced from the typeface it
purported to represent. He commented too on the use of
intelligent outlines, citing \PS, but also the more
recently announced Bitstream, Compugraphics and Imagen
fonts which add information to the character outlines.
But he also said that he could not ``see how to make the
outline intelligent enough''. 

However, Richard has slightly retracted from his earlier
view that \MF\ was unusable, and he now expresses the
view that, subject to some pretty massive constraints,
`You can use \MF'

A good set of talks which from the extent
of interest shown and the questions asked
met some sort of need. The next {\sc bsc-ep} meeting
is on April 28th. Later this year there will be a meeting
on technical typesetting.

\rightline{\sl Malcolm W. Clark}
\bigskip