Walbaum-font
Haavard Rekstad
--
Haavard Rekstad
Department of Refrigeration and Air Conditioning
Norwegian University of Science and Technology
N-7034 Trondheim, NORWAY
DickWeltz
Fonts of the Walbaum family are produced by Linotype (Germany) you can buy
them from any Linotype font reseller. If they are posted anyplace for
downloading free of charge without Linotype's permission, that would be
illegal.
-- Dick Weltz, Spectrum Multilanguage Communications, NYC
America's leading translators & foreign language
typesetters
===================================================
Visit our Language News & Notes on the Web at
http://ourworld.compuserve.com/homepages/SpectrumLang
Bill Troop
having a Walbaum cut. The three versions I know of
are (1) Berthold Walbaum, Lange's ultra modern redesign,
released by Adobe three or so years ago; (2) the most
beautiful of all, Berthold Walbaum Standard, a
*sensitive* digitization of the true Walbaum types,
and (3) the oldest of them all and, I think, the
only one with an expert set, Monotype Walbaum, which is
based on a model similar to Walbaum Standard. The
only problem with Monotype Walbaum is that it is
spaced too wide, for some reason. Metal Monotype
Walbaum did not have this problem. It is relatively
easy to correct.
It is very much worth noting that a properly
spaced font of Walbaum is one of the very few designs
which does not require f-ligatures to look at its
best.
Berthold Walbaum is available from Adobe and
all its licensees; Berthold Walbaum Standard is
available from FontShops; Monotype Walbaum is
available from Monotype but has not yet been
cross-licensed into the Adobe library.
Bill Troop
Bill Troop
barry & Susan Stock
Berthold. The appearance of the page is markedly different from one
t'other. The Monotype digitization seems to be from a different source,
maybe brass pattern master. The Monotype Spectrum also is quite a bit
lighter in PS than any film or metal sample you'll see. This can (has)
lead to unpleasant surprises, because laser printer output of these faces
are less contrasty, meatier because of hinting.
Barry Stock
Bill Troop
adaptation of MT's metal Walbaum which in turn is an exceedingly
faithful adaptation of Berthold's foundry Walbaum, which
in turn appears to be a slightly touched-up version of
Walbaum's Walbaum) for being too generously spaced, but
looking at some of the old Berthold (metal) specimens for
sizes around 12-14 pt, noticed that the spacing in the
foundry type is quite wide, too. Berthold's digital Walbaum
standard, based on similar metal masters, is spaced
rigorously in accordance with Berthold's best photo-era
spacing practices -- which is to say, rather oddly (H kerns
to H!!!!), but exceedingly competently. Has anyone analyzed
Lange's unusual approach to spacing? Has anybody written
anything about it? Note that Adobe appears to be faithful
to this scheme in its Berthold-licensed fonts.
Tiro TypeWorks
> I had criticized MT Walbaum (which is an exceedingly faithful
>adaptation of MT's metal Walbaum which in turn is an exceedingly
>faithful adaptation of Berthold's foundry Walbaum, which
>in turn appears to be a slightly touched-up version of
>Walbaum's Walbaum) for being too generously spaced, but
>looking at some of the old Berthold (metal) specimens for
>sizes around 12-14 pt, noticed that the spacing in the
>foundry type is quite wide, too. Berthold's digital Walbaum
>standard, based on similar metal masters, is spaced
>rigorously in accordance with Berthold's best photo-era
>spacing practices -- which is to say, rather oddly (H kerns
>to H!!!!), but exceedingly competently. Has anyone analyzed
>Lange's unusual approach to spacing? Has anybody written
>anything about it? Note that Adobe appears to be faithful
>to this scheme in its Berthold-licensed fonts.
I had noticed, while examining the afm files for Poppl Pontifex, that
the kerning information was a little strange by contemporary
standards. This, obviously, follows the Berthold settings very
closely. In the Poppl Pontifex, what has been achieved is a certain
amount of letterspacing (ie. negative kerning) between the uppercase
letters. That is to say, the uppercase letters are fitted to the
lowercase, and are given additional space when setting all caps by
negatively kerning essentially all the uppercase combinations. Adobe's
practice with their own fonts seems to be a compromise: the uppercase
letters are set just wide enough to be legible (but not properly
letterspaced) and are then kerned to certain lowercase letters. I have
toyed with applying the Berthold system (which is, as you point out,
probably Lange's personal preference) to one of my own fonts, and may
do so when I have a chance to experiment. It does suggest a slightly
larger number of kerning pairs, but also permits the type designer and
manufacturer much greater influence over final typography -- one
doesn't have to rely on the 'desktop publisher's' ability to correctly
letterspace all cap settings... or letterspace them at all!
John Hudson, Type Director
Tiro TypeWorks
Vancouver, BC
ti...@portal.ca
http://www.portal.ca/~tiro
tra...@inforamp.net
>
> I had noticed, while examining the afm files for Poppl Pontifex, that
> the kerning information was a little strange by contemporary
> standards. This, obviously, follows the Berthold settings very
> closely. In the Poppl Pontifex, what has been achieved is a certain
> amount of letterspacing (ie. negative kerning) between the uppercase
> letters. That is to say, the uppercase letters are fitted to the
> lowercase, and are given additional space when setting all caps by
> negatively kerning essentially all the uppercase combinations. Adobe's
> practice with their own fonts seems to be a compromise: the uppercase
> letters are set just wide enough to be legible (but not properly
> letterspaced) and are then kerned to certain lowercase letters. I have
> toyed with applying the Berthold system (which is, as you point out,
> probably Lange's personal preference) to one of my own fonts, and may
> do so when I have a chance to experiment. It does suggest a slightly
> larger number of kerning pairs, but also permits the type designer and
> manufacturer much greater influence over final typography -- one
> doesn't have to rely on the 'desktop publisher's' ability to correctly
> letterspace all cap settings... or letterspace them at all!
>
> John Hudson, Type Director
>
If you get a hold of Berthold's type catalogues from about 10 years ago
when they had just about made up their minds about digitizing all of their
fonts, you'll perhaps see their philosophy regarding type and ICS (Inter
Character Spacing). Practically all of their fonts were set with 0 (zero)
ICS and 12 unit word space, this is the way Berthold suggested.
Each page of their catalogue shows sample settings with different ICS and
wd # values (please don't bring up the # thing again). It actually works
very well.
I used their system when it was glass fonts and with digitized fonts.
There was absolutely nothing to compare with Berthold's glass fonts -
nothing. I used to get first-time customers calling me up saying they were
flabergasted, they'd never seen anything so sharp and beautiful (their
words, not mine).
I still have their catalogues if you need any info from them.
Peter Stanbridge
Oakville, Ontario
tra...@inforamp.net
Bill Troop
Thanks to John and Peter for, it appears, correctly unveiling the mystery
of the intent behind Berthold's spacing schemes which, of course, did
not work in every *single* case, such as their version of Monotype Bell,
which must be one of the most hideous and misguided revivals of a beautiful
historical face. Strangely, it would appear that no other designer or
foundry has followed this style of kerning and spacing, possibly because they
could not figure it out. There must of course have been some considerable
degree of algorithmic automation involved, and one wonders if, perhaps,
they had a customized version of Kernus? Which leads to another question,
namely, is there some way in which Kernus can be better customized than merely
by editing the kerning tables, which is not much help?
For that matter, has anybody ever been able to figure out precisely how
Kernus works, and why? I had assumed at first that it only kerned, but it
appears also, under some circumstances, to alter spacing as well. Is this
based on the font's original spacing, when it happens?
All efforts to learn more about this from URW have been fruitless.
This leads to another question, which is, what is the theory behind Fog 4.1's
spacing and kerning? I have been not yet been able to discover the underlying
philosophy, or find a way of using it that I like. I suspect that certain
compromises may have been made to accomodate a broad range of different
types, ie, italics, scripts, sanses, etc. It would appear that a more
specific version might be desirable. One of the basic problems is that
fog does not even attempt to find out where the vertical strokes are in
characters, but bases its decisions on extremeties of characters, such
as serifs, which may be irrelevant to correct spacing.
One might have hoped that Kernus had some sort of intelligence built in to
analyze actual character shapes, including overhangs, but it appears to
be somewhat arbitrary. For instance, if you happen to put your ff ligature in
the place of the ascii circumflex character, and tell Kernus to kern ff (or ascii tilde)
to every combination that exists for f, Kernus will indeed provide you with
kerning pairs, but they will be unrelated to what it has done for the f and hence
are entirely incorrect (always assuming that its kerning for the f was
correct).
All in all, it seems to me that there is a need for a really intelligent kerning
and spacing program which has not yet been met.
Early on, I assumed that major foundries and design studios must have attempted
to automate these processes in some sort of effective, proprietary fashion,
but the numer of obvious errors in every foundries' fonts appears to make
this assumption untenable.
Bill Troop
Bill Troop
tra...@inforamp.net
<bilt...@ix.netcom.com> wrote:
Bill, this way too much for me to read on-line and not have my wife nag
me. I'll read it later and e-mail you.
I know exactly how Berthold's kerning tables worked, I used them for too
long. They were good but far from perfect, they could be used to kern or
letterspace between individual characters.
Peter Stanbridge
rwak...@thoughtport.com
>tra...@inforamp.net wrote:
All in all, it seems to me that there is a need for a really
intelligent kerning and spacing program which has not yet been met.
Early on, I assumed that major foundries and design studios must have
attempted to automate these processes in some sort of effective,
proprietary fashion, but the numer of obvious errors in every
foundries' fonts appears to make this assumption untenable.
Typographers who skillfully use software programs to produce a good
quality kerned font, usually follow these steps:
1. Open the font in Fontographer or FontStudio (which is still a
favorite of many). Ikraus is not recommend for this, because when it
opens a font already created, it replaces the outline points with
Ikarus points, and this can cause a slight variation in the letterform
sometimes. Ikraus is fine for creating new fonts.
2. Remove all kerning pairs.
3. Run your favorite 200 or 300 words that have most of the
combinations of the english language. (Most typographers have such
lists.) If you need to start a list, you can check such programs as,
FontHopper (Windows), KernEdit (Mac and Windows versions of Pagemaker
have it as a separate app), and see what their lists look like. You
can add you own words to these lists.
4. You then adjust as much of the spacing between character
combinations as you can do universally by adjusting the side-bearings
or side-walls. Fonts that have a lot of kerning pairs (over 1,000),
usually have bad or sloppy side-bearings. FontStudio is used used by
professionals for this purpose.
5. After doing the above, then, you can start adding kerning pairs. A
font with good side-bearings should look good with about 300 kerning
pairs.
Most of the answers to your questions about the various kerning
options in Fontographer are in their manual.
Robert Wakeman
Type Director
Ogilvy & Mather, NY