Gerald,

Calligraphers don't change to a fatter brush or pen when they are working with different sizes of letterforms.

Very enlightened observation.

> Calligraphers don't change to a fatter brush or pen
> when they are working with different sizes of letterforms.

Which makes the results inferior to typography, of course.

--

GG, still waiting on those answers...

hhp

> Nonsense.

Think about it. It's actually pretty obvious.

hhp

:->

hhp

Gerald,

Do you like poetry? :-)

Addison.

Tell me this, Mr. Giampa. Frederic Goudy cut his own type, right? Did he anticipate the finer execution of printing that you expect? In other words, do you feel your digital versions of his types live up to his (Goudy's) expectations of his metal types? And please, this is no attack--I just want to know your opinion.

(1.) The patterns were very lightly inked, proofs were pulled on a Vandercook #4.

http://lanstontype.com/GiampaProofPress.html

(2.) The inked proofs were hand digitized on a very high resolution tablet. The software is Ikarus, 15,000 units vs Fontographer 1000 units.

(3.) The digitized outlines were laser output on onion skin paper at "exact" same size as the inked proofs.

http://lanstontype.com/PatternIkarus.html

(4.) Laser output positioned over the inked proofs would illustrate clearly the minutest deviance allowing for easy perfection.

If Goudy was happy with his own lead master patterns certainly he would be equally happy with the digital mastering.

As objects he may prefer lead to polymer, I know I do.

As far as his opinion of desktop publishing goes I can not speak to that matter.

Gerald Lange has spoken of Goudy's welcoming Photocomp with open arms. I wonder how happy he really was "without" drawn small caps, ligatures and tied characters.

I suspect he was greatly disappointed.

Would Goudy be happy with digital type? Well, if he liked Photocomp I see no reason he would not like digital type. I don't, but that's just me.

I believe Goudy would have continued drawing letterforms with pen and paper. He would prefer, I predict, Ikarus over Fontographer or FontLab for outlining.

I know I do, I know Jim does. Ikarus is the finest method for digitizing paper artwork. Not so for originating typefaces. Ikarus is not a drawing tool.

> I am sure you mean "different" rather than "inferior".

Nope - sorry.

And your elaboration makes it clear that you're thinking like an artist, focusing on the act of creation instead of the necessities of usage. I never said calligraphy is easier than type design - that's moot. I'm looking at it from the viewpoint of the user/reader, who could care less what tools you enjoy fondling and what your childhood was like.

In calligraphy, the lack of variance in stroke thickness with respect to size is a problem when it comes to optimal letterform quality. Because type design is non-instantaneous, it allows for the considered, conscious control of things like weight (and its ideal distribution). This should be obvious.

--

Thanks for the answers, in the end.
I'll wait for news on the optical scaling software, etc. Thanks.

> You may need better glasses.

I have 20/15 vision.

> not until you and I determine the best way to go with optical scaling.

I thought I didn't know anything?

But lacking a fancy method to streamline the selection of optically-correct fonts from a typeface family, providing separate fonts isn't so bad - remember that optical scaling is a specialist thing, and specialists are usually willing to make the extra effort.

> Lanston's Californian is the most authentic vesion of Goudy's Californian.

At one point size.

> photopolymer as a type form does not have the je ne sais quoi of metal

I found metal type a lot of fun, but:
1) Mostly just to fondle; actually setting metal type is pretty pissy.
2) Only in larger sizes.

When I was doing my ATF Garamond cascade the smaller sizes were nothing but displeasure to work with (not to mention all the spacing matter you have to struggle with - it feels so pointless). You spend half your time fighting the technology - I'd rather spend my time conceptualizing. But yeah, rubbing some 48 point Palatino in your palm is fun.

hhp

Gerald L

Yes, photopolymer as a type form does not have the je ne sais quoi of metal, but my concern has always been the printed page rather than idle homage to technology.

I would like your work even if you were a litho printer.

> At all point sizes.

How so? The original metal Californian had multiple optical masters, right? If you're reviving such a face, one optical size cannot represent what that typeface really was.

> I think it is time to look for funding.

I have funding for the trapping, but not the optical scaling yet. Once I put some of my ideas in a formal written piece, it will be easier to fish. But I need the ATF stuff first: they were the sultans of scaling.

hhp

Valdonega does indeed use optical scaling. They have a line of "classical" fonts with 4 optical masters each. Here's a size-normalized cascade of their Garamond (Jannon), scanned from a specimen booklet:


Although their stuff has "only" 4 stops on the scale, the non-linearity of the scaling is great. A bigger issue though is that the fonts are font internal use (for typesetting clients), not sale.

hhp

Hrant,

The original metal Californian had multiple optical masters, right?

Can I say yes, and no?

Very few mats were ever struck for Californian. Mackenzie-Harris had the matrices for setting books and other work for the University.

Mackenzie & Harris are trade setters and type founders. They set type for printers, for that matter, they still do. And they are a good source for lead type.

http://lanstontype.com/SanFranTruck.html
http://www.arionpress.com/mandh/index.htm

I am not sure how many sizes were cut. Goudy cut (1) set of patterns in lead. Some lead patterns were substituted by brass in fitting the face for composition by the Lanston Monotype Machine Company.

Mackenzie & Harris can probably report how many sizes were made available.

The point: Californian by and large was a private typeface. Therefor University economics prevailed over multiple optical masters.

However two sets of patterns were made. One for display, the other for composition.

Gerald G

Californian was cut from 8 point to 36. Lanston Monotype produced the comp sizes for the University's use, and Goudy cut the 14 to 36 point display at his workshop.

You are aware since you own the patterns that the fce when it was done ws named Univerist of California Old Style. Goudy had wanted to name it simply University, but the directors of the school felt the name to be too generic.

In the mid fifties Lanston revised the design for its own purposes of production and sold the design under the current name, Californian.

If you dig out your original "Typologia" (mine went astray years ago on a loan to an unprincipled-book-non-returner- bugger!) you can search out just what Goudy cut. I believe that at the time all the sizes made for the University were done by Goudy at his workshop, and the type cast by him. I seem to recall that he cut 14, 18 and 24, with few sorts for titling cut in 36.

You can tell from Typologia just how much he did. I do know that all the early sizes and cast types were done by Goudy.

Incidentally, and for what it's worth, I have been involved in letterpress since I was a kid, and I will always like it. It feels good and it suits me. I Have no negative opinions about printing with polymer plates. If that feels good to the person doing it, then that's a good enough reason for them to do it; and they will make a buck much easier into the bargain.

I don't give a rat's butt about all of letterpress's perceived drawbacks mentioned by others on this thread. Most of the problems noted were the result of poor craftsmanship and are hardly the fault of the craft. About the "bad" fitting of Linotype: it was one of the best ways of getting type set quickly and economically in acommercial world, and refinement mats were made to overcome all of the problem pairs. The refinement matrices had to be placed in the assembler by the operator by hand, and if was not done it was simply for lack of caring or for lack of the special refinement mats. Much work done on the Linotype was good by any standard. If there were a few inherent weaknesses in the technology, there seems little to be gained in lamenting it now. The Linotype, like other composing machines that preceded was the result of a gradual progression toward excellence and production.

I am certain that 30 years from now the same kind of stuff will be said about the very best of what is available to us today in typography.

Without Buddy Bolden there would have been no popular music as we know it now.

Jim Rimmer

And William Morris

http://web.uflib.ufl.edu/spec/rarebook/catalog/morris/initials.htm

At one point MacKenzie-Harris made the Californian type available for sale. The promo was that they had been given permission by UC to do so. I think that was in the early 1980s. I bought near a cabinet's worth of all sizes available (I can't remember what the range of sizes were though). It was a joy to work with and we did a couple of books and various other printed pieces with it before I sold it off. Quite nice were the small caps that ran all the way up to 36-point. I remember that at some point they stopped providing the larger small cap sizes but I don't know why.

The company eventually changed hands and the name was changed to M&H Type Foundry.

G. Lange wrote: "The 5% I was talking about was not during the letterpress printing process, but a gain from neg to plate during the platemaking process. I read a technical report where it was suggested as 5%. "

Gerald--I was wondering where you might have read this? While I have noticed that different types of plate material seem to gain differently during platemaking (case in point, a Printight plate I processed recently seemed to gain a lot more than a Jet plate I made from the same neg), it would seem to me hard to hold fine line screens on material that gains 5% in platemaking. 5% seems like the "Mac Truck" that G. Giampa was talking about! At 200 lpi I can't imagine these plates holding a reverse dot.

And if this indeed happens, I wonder if offset plates have the same sort of gain during platemaking. I've never read anything about this phenomenon before, but since it's also a film-based platemaking process I have to wonder.

But I have to admit I'm a skeptic that the platemaking is the root of letterpress's gain. Pulling a plate with minimal ink on a coated piece of paper shows minimal gain. Placing the neg overtop this proof shows very little difference between the neg and the plate. Print that same plate with enough ink to transfer well to an uncoated book paper, and a lot of gain occurs--enough to be visible, anyway. While gain might occur during platemaking, I still think that more occurs from the printing process.

Another reason that thin typefaces might appear more substantial when printed by letterpress from photopolymer is the thicker ink film that transfers to the paper on press. I believe I've read that the ink film is four times as thick as the ink film transfered during offset printing. This extra density gives weight and, when properly printed, a sparkle to the type.

Okay, sorry to be talkin' shop here. This has been an interesting read so far and I hope I'm not deadening it.

Harold

> a sceptic like Hrant might like to note that it is very much in the
> economic interests of digital foundries to push optical variants

Sure - but I'm just happy it's an improvement! Most efforts to save/make money are the opposite.

Same with OpenType: a company who axes MM is not about to suddenly start caring about quality in the craft; but OT is still a good thing for the user, so I like it.

> running the different point sizes as separate pressruns

This is exactly what Gerald L had me do in his class when I was printing my ATF Garamond cascade.

> three dimensional printing will always look different

But if you advocate a very light-pressure impression (at least I do, something else I've learned from GL), the 3D effect is probably too slight to matter. On the other hand, letterpress type does have a slightly "soft" effect (compared to offset), and I think this might in fact be very significant in explaining letterpress's greater readability.

> there is nothing to stop that person from messing with the tracking.

But in the hands of a non-idiot that's a good thing, simply because type designers make mistakes too. FF Eureka for example is too loose, and some negative tracking makes all the difference. (To be fair to Bilak, his later designs are much improved in this respect - although he's always been great with the black forms.)

hhp

BTW, two less-mentioned reasons for the first digital fonts/revivals being too light:
1) Lasers printers make fonts darker (especially back then at 300dpi).
2) Culturally (again, especially back then) lightness was fashionable.

> hopefully that is a matter of time.

I fear the opposite. Under capitalism, cultural refinement is all downhill from here, baby.

> I do not believe that we can realistically demand that a digital version of a metal type

If you do it right, and you use photopolymer letterpress, virtually nobody would have a hope of telling them apart.

hhp

I agree with Paul's viewpoint about the mytholigization (phew!) of letterpress, and the accepted wisdom that digital is more efficient, yes, but the main task is to recover what has been lost. This is absurd. As Joni said, "somethings lost and somethings gained by living every day".

Let's face it, letterpress does a really awkward job of what it was invented for -- mechanizing handwriting.

At the time of its invention, only a small % of the population was literate.

During the letterpress era, only a small % of the population set type.

Now we have a new technology, and the majority can read and write, and set type.

The present great interest in script fonts is, I would hazard, an expression of this reality.

Let's start over, relegate the print era and its typography to the important, yet minor role of cultural legacy, and build a digital typography that can express individual humanity far better than print typography did -- this, I suggest, is the "potential" that Hrant had mentioned earlier in the thread.

Let's start over, relegate the print era and its typography to the important, yet minor role of cultural legacy, and build a digital typography that can express individual humanity far better than print typography did....

I'm struck recently by how beholden to the metal type paradigm digital typography still is. Despite the absence of any physical 'body', we are still essentially treating type as sequences of rectangles, rather than as sequences of letterforms. This paradigm is particularly obvious in our discussions about spacing: we talk about sidebearings and kerning as if this were the normal way to space letterforms, when any scribe or stonecutter would look at you funny if you suggested that he imagine a rectangle behind every letter and shift these rectangles and overlap or space them out. We talk solemnly about how the design of white space between and within letters is as important as the design of the letters themselves, but we continue to work with a paradigm that massively limits the ability to actually design the space between letters without the perverse explosion of thousands of kern pair adjustments to overcome the limitation of the rectangle.

A few years ago, on the comp.fonts newsgroup I believe, Laurence Penney described a possible spacing model for digital type that involved designing the invisible, non-rectangular envelope that every letter should occupy regardless of its neighbours. Sequences of letters would be spaced by butting the envelopes of the letters against one another. Intrigued by this notion, I made some tests, first with unitised strips of spacing material because it was easiest, and then trying to actually design the ideal envelopes for each letter. It is difficult, but probably only because one isn't used to thinking in this way, but it works and is by far the best spacing model I've seen (and that includes the legendary Kindersley Optical Character Spacing system). However, I think the likelihood of the software industry en masse dumping the existing paradigm of digital type is very unlikely.


While I tend to agree with Gerald that our humanity needs a physical connection to the past, I think we should also be aware that the manufacturing of metal type also compromised that humanity because of the nature of the material: we're not made of metal, and much as I love good letterpress printing and admire the craftsmanship that it represents, I'm not sure it represents a sufficiently deep past in which to locate our humanity. My work on Arabic and, to a lesser extent, Hebrew has made me much more senstitive to pre-typographic ideas of spacing and layout, and I think there is a particular humanity in the freedom of arrangement that precedes the rectangle of the 'body'.

>our only link with "humanity"

Gerald: For "humanity", I'm thinking of the medium being able to express the individuality of its users. The physicality and craft of letterpress does this to an extent, expressing the qualities of the professionals who work with it, and by extension, the culture at large.

What digital media can do, customized and data-base driven, is allow the democratized tasks such as typesetting, to have an individual quality. In theory, but the counter-force to this is standardization and uniformity.

Gerald, you missed my point and are responding to the afterthought. I am saying that there is no logical reason why digital typographic technology should be based on a metal type paradigm (whether handset or mechanical); the fact that today's digital font technology is based on such a paradigm is due to historical exigence: it is easier to model an existing technology than to invent new ways of doing things in a new technology, especially when existing companies have a massive collection of design and metrics data that they want to leverage. But when the paradigm of the existing technology is based on the properties of a particular material -- in this case, metal -- does it really make sense to carry that paradigm over to a new technology from which that material is absent? Historical exigence is important in understanding why things are the way they are, but it should not be accepted as the way things must be, let alone as the best way things can be.

Why does digital type have sidebearings? Why is spacing based on a combination of glyph metrics and kerning? What other paradigms can we find in earlier technologies (including handwriting)? What new paradigms can we create that take into account the specific material (bits and bytes) of digital typography?

FontLab's half-implemented anchor based kerning is a step in the right direction, it seems to me, but the bigger step would be to abolish the whole notion of kerning, which is nothing more than the exceptions to the rule of the rectangle, and have anchor or envelope based spacing per se.

And perhaps abandon point sizes? Or leading?

Worked for Aldus Manutius.

Don't confuse particular technologies (the point size system, leading) with fundamentals (type size, interline-spacing). I made no argument against mathematics: I questioned whether the system of metrics worked out for type as little bits of metal made sense as a paradigm for the arrangement of digital type. I'm not saying it is a bad system per se -- although it has clear problems for many non-Latin scripts --, only questioning whether there might be a better system. I've just spent the better part of three weeks kerning a family of what I consider medium-size fonts (c. 1000 glyphs per font). This is kerning applied after 'basic' spacing, i.e. setting sidebearings. I'm wondering why these two processes need to be separate, since what I'm really interested in is the space between letterforms, not the arrangement of rectangles.

John,

This is kerning applied after 'basic' spacing, i.e. setting sidebearings.

I am not so sure how separate these proccesses are. With Fontographer this is accomplished in one and the same window. But I may suggest one logic, side bearings are good starting points for kerning. What I am saying is this. You would not wish to kern first and build side bearings afterwards? And If you were going to simply kern you will have to work hard to avoid creep. I have never been happy with auto kern, or auto side bearings. Both have been toyed with. Ikarus offered such features.

Can I sell you guys my self tracking ligatures?

John

Real good sense. Brilliant actually. Let's do it!!!

Naive question though. While still locked into the rectangle concept, didn't PS3 allow for over extensions of the rectangle? All those "your own handwriting" font offers and the randomized fonts were/are possible in PS3, correct?

Maybe the attempt to standardize, PS1, OT, are movements away from where we would ideally like to be?

Gerald

John,

"I think the rectangle is an anachronism: a virtual hold-over from a previous technology.

It strikes me curiously ironic that you see rectangles, where I, with a long hot metal background do not. For that matter never did. I view the product on the paper. When grids irritated I worked to escape them and found it possible even with single unit quad units. See: http://lanstontype.com/

Or this:http://lanstontype.com/Granjon.html this example has evidence of grid only on the outer corners. Those would be elimated in a further step and repeat performance.

Linotype is not a rectangle but a long slug, or for those not familiar with the term, a bar of metal.

But as far as grids go, or retangles as you put them, they never to be limitations for hot metal types, at least not often, when they were, such as "script types", particularly the extremes from German and Dutch foundries you can view some very peculiar looking type bodies. Other faces, not even scripts, but Caslon Oldstyle Italic Swash was cast in a most unusual fashion. But wait, think about the obvious, why should we search for such extreme examples? Lower case italic y, j, overhanging f's, ligatures. Solutions, not limitiations. That's what I see, not rectangles and prisons.

Also, no matter how you look at it, the objectives are the same even when the solutions are not