ATypI's (old) stance on cloning vs. yours

John Hudson

Yes, Hermann Zapf is not with is anymore, but his wife, Gudrun Zapf von Hesse is still alive.

Alas, no longer. She passed away 13 December.

John Hudson

There are other organisations that have fairly extensive and practical ethical guidelines and documentation, and which manage to do so without contravening antitrust legislation. See, for example, the AIGA Business and Ethic series, which includes a pretty good one on Use of fonts.

Thomas Phinney

Albert_Jan_Pool said:

Yes, Hermann Zapf is not with is anymore, but his wife, Gudrun Zapf von Hesse is still alive.

Sadly, no. She passed away a month ago, December 13, 2019. Just a couple of weeks short of her 102nd birthday!

Daniel Benjamin Miller

Albert_Jan_Pool said:

In the case of Sabon, you write that copying font data is unacceptable. In the case of Bitstream that is probably exactly what has happened. Virtually no one could digitize such a large number of typefaces in such high quality and thereby perfectly matching the sources in such a short time. Not in those days, not even Matthew Carter, no matter how nice he is. And he is. I can tell you.

In the case of Palladio you halfway missed a point I think. Yes, Zapf approved of Palladio, but he never approved of URWs PostScript clone of Palatino, no matter under what name URW ever sold these. Compare the two typefaces, they are not the same.

• As I was never involved in the bitstream fonts, I cannot tell you how they were made. According to Adam Twardoch, I believe they were digitized from catalogues --- presumably the sample catalogues for Linotype machines in this instance. And I'm not entirely sure that there was even another digital Sabon at that point. http://luc.devroye.org/fonts-45464.html
• Examine URW Palladio and URW Palladio L and you will see that they are almost exactly identical. They are certainly not different typefaces. And the quote above also says Zapf approved of Bitstream's copy of Palatino as well.
  

Daniel Benjamin Miller

John Hudson said:

There are other organisations that have fairly extensive and practical ethical guidelines and documentation, and which manage to do so without contravening antitrust legislation. See, for example, the AIGA Business and Ethic series, which includes a pretty good one on Use of fonts.

Interesting --- but do you know of any guidelines for font digitizers?

As for my personal view, I think it is usually fine to digitize a font from the pre-digital era. I don't think it's fine to just copy somebody's new font today. The expectations were different then, and the culture of the early digital era established that those pre-digital fonts are public domain, in my view. In many cases the same typefaces were available from different foundries, so there isn't a particular foundry from which it is stolen. In any case, if there are many digitizations out there, then it seems that they have grown acceptable, and I think it is OK to produce a new digitization. And doing so can even bring something new to the table, like ET Bembo did, for example. And I think one should keep in mind that the font designers of that era had the understanding --- I'd believe --- that their fonts would be adapted to new machines. And if your design is 60, 70 years old, I think that is also a significant consideration in comparison with a font from the 1990s or 2000s.

As for digital fonts I think it is perfectly fine to take loose inspiration --- which we all have been doing since forever --- but not to directly copy without permission. (In this case, direct copying includes using font data as a basis as well as tracing.)

But these are just my two cents.

Adam Twardoch

I wasn't there but from what I got from conversations with older designers was that Hermann Zapf was one of the first prominent post-WWII designers who negotiated _licensing_ of his designs to foundries rather than “sold” them. And he negotiated hard, so much of his licensing was not on exclusive basis. 

This way, he was able to license the same designs to Stempel (as Palatino), and later to Bitstream (as Zapf Calligraphic) and to URW (as URW Palladio). Zapf was a leader on many fronts, was extremely interested in digital font technology, worked closely with URW on hz and the font projects, and with Knuth on TeX. 

What I was told was the reason he left ATypI was that Monotype produced the Palatino ripoff Book Antiqua without entering a contract with him. So Book Antiqua was an unauthorized ripoff while Zapf Calligraphic and Palladio were authorized clones. I imagine this was possible because he was an assertive person with star qualities, so either Stempel (later Linotype) did not have exclusivity clauses in the contracts with Zapf, or he later persuaded them to let him do the multiple licensing. 

Adam Twardoch

(I mean, Zapf is probably the only person whose “private” drawings, i.e. his Dingbats that he himself conceived and selected, ended up in Unicode, so there you go.)

Adam Twardoch

Never in my life I met a second person who would combine the actual highest-possible skill with a very strong (yet still healthy) confidence. Hermann Zapf was a true force majeure of the type design in the 20th century, and I think using him as a model for anything else is a bit pointless. It's like debating the practice of music composition citing Mozart as an example. 

Adam Twardoch

Bitstream was founded by a group of ex-employees of Mergenthaler Linotype, led by Mike Parker who used to run Linotype.

The Bitstream library was created on the basis of many Linotype designs, though I'm not sure what exactly were the models. But it must be noted that this was a very modern, high-quality digitization effort, while at the same time or a bit later, Linotype was subcontracting the digital data creation to Adobe, URW, Scangraphic and other companies. So on the digital level, much of Bitstream’s fonts were advertisment better than the ones that were later marketed as Linotype. 

When I was talking to the Bitstream people, they ensured me that they have made every effort to sign alternative contracts with designers or companies who received royalties from Linotype. For example, they did get contracts from Zapf, and from Neufville for the Futura trademark. They did “rip off” the designs that Linotype “owned”, i.e. had acquired decades earlier on a work-for-hire basis. Of course Linotype owned the trademarks so Bitstream used the funny names.

I was told that pretty much the only living designer they didn't manage to convince to sign a contract with them was Adrian Frutiger. Frutiger was loyal to Linotype — and Bitstream decided to create Zurich, Humanist 777 etc. anyway.

This was, I believe, Bistream’s only “major fault” (it was major indeed), and was the source of the general disdain about Bitstream. 

Albert_Jan_Pool

John Hudson said:

ATypI simply doesn't have any teeth, so it doesn't make sense to give it a mission that involves biting things.

In several European countries ATypI can have its teeth fixed. Health Care would take care of that and patients do not need to fill out forms for which they need the assistance of a lawyer. Now how does that sound? 

Albert_Jan_Pool

Thomas Phinney said:

Albert_Jan_Pool said:

Yes, Hermann Zapf is not with is anymore, but his wife, Gudrun Zapf von Hesse is still alive.

Sadly, no. She passed away a month ago, December 13, 2019. Just a couple of weeks short of her 102nd birthday!

Thank you Thomas, sad to hear this …

For some people it may be interpreted differently though. For them it would be perfectly OK to ‘digitize’ Diotima and engineer it to the widths of Palatino. What a bright and shiny future. And what about
Gerard Unger and Wim Crouwel? They are dead too! So why not take Crouwel’s New Alphabet and make it width-compatible width Unger’s Swift? How Dutch would that be?

Albert_Jan_Pool

Dave Crossland

Albert, I'm curious, what do think about Bahnschrift?

Albert_Jan_Pool

To me, Bahnschrift is a typical “me too” product.

Albert_Jan_Pool

Adam Twardoch said:

Bitstream was founded by a group of ex-employees of Mergenthaler Linotype, led by Mike Parker who used to run Linotype.

The Bitstream library was created on the basis of many Linotype designs, though I'm not sure what exactly were the models. But it must be noted that this was a very modern, high-quality digitization effort, 

Adam, your attempts to white-wash Bitstream are touching …

Unfortunately I happen to have some other, first hand information. A colleague once visited Bitstream in the 1980s. He saw that they had quite an impressive collection of what he referred to as “original Ikarus data carriers (computer tapes if I recall rightly)” on their shelves. Note that in those days, Linotype had digitised hundreds of their typefaces using URWs Ikarus. It was the de-facto standard for mastering digital outlines at that time. Several sources confirmed that Bitstream’s proprietary outline format for digitising, storage and editing of typefaces was quite similar to URWs Ikarus format. 

I do not think that they had those Ikarus data carriers for pure decorative purposes on their shelves ;-) Digitizing the Linotype library from specimen catalogues would have taken them several years … obviously they had managed to find a short cut.

George Thomas

@Adam Twardoch

Mike Parker did not run Linotype; he was the Director of Typographic Development. He did have a huge influence on their success in later years.

Albert_Jan_Pool

Adam Twardoch said:

Linotype was subcontracting the digital data creation to Adobe, URW, Scangraphic and other companies. 

Yes, Linotype subcontracted URW to digitize their typefaces, they also bought digital data at URW, notably the ITC typefaces that URW digitizer, interpolated and mastered for ITC as from 1985 to the early 1990s. Not quite sure wether Adobe was ever subcontracted by Linotype. Adobe originally bought Ikarus data at URW for a few hundred fonts. Some of the first Adobe fonts were converted from those outlines. But as soon as Linotype bought Adobe’s software Build Font to create PostScrip Type1 fonts and both parties agreed on creating what they called the Adobe/Linotype font library by joining forces, most of that URW data was not used anymore at Adobe. Linotype’s idea behind that library was that they wanted the metrics to be backwards compatible with their digital fonts as they had been selling along with the Linotronic and Linotron typesetting systems. Many URW fonts did not fit that strategy because they were digitized from other sources than Linotype artwork, so they would not fit on Linotypes metrics without extra work. Converting the Linotype library to PostScript was a huge task and they wanted to be fast. Probably because Bitstream was out there on the PS clone market using that very similar data.

Where did you get the story that Scangraphic was subcontracted by Linotype? If at all, this must have been after 1991, otherwise I would have known it. To me it seems more likely that Linotype bought Ikarus data at URW for some typefaces that were in the Linotype Library. Scangraphic had subcontracted URW to proivide them with Ikarus outlines for the Scangraphic Digital Type Collection. Amongst these were many Linotype fonts, most of them were already in the URW library, but some of these had to be digitized. It occured that URW had digitized Linotype typefaces that Linotype had not digitized themselves yet. So what I recall is that they bought these at URW.

Dave Crossland

Albert_Jan_Pool said:

To me, Bahnschrift is a typical “me too” product.

So no moral concerns about it?

Mark Simonson

@Albert_Jan_Pool said:
When Bitstream was founded, they somehow magically released hundreds of fonts which quite exactly matched Stempel/Linotype’s digitisations.

Are you sure about this? I licensed fonts from both libraries back in the nineties. In comparing Bitstream vs Linotype digitizations of the same faces, the data does not match, not even the metrics. They were different enough that I had opinions about which digitization I preferred.

John Hudson

Re. Bahnschrift, I've lost count of how many different digital versions of DIN 1451 have been made, including pre-DTP versions made by Linotype, and before that there were dry transfer versions from Letraset and photocomp versions from Berthold. The original design is out of copyright and, so long as new versions are not illicitly derived from other maker's data, there's no legal or, I would say, ethical issue with new versions. There are, of course, issues of boringness and lack of originality, and the success of FF DIN surely has been an impetus for the creation of some of the other versions, just as the success of some geometric grotesques in recent years led to loads of foundries adding that style to their catalogues to try to cash in.

Bahnschrift came about because a team at Microsoft wanted a version of DIN 1451 for some purpose, so it is an example of something being created by MS for an internal client, and later released (same thing happened for Sylfaen).

Albert_Jan_Pool

Dave Crossland said:

Albert_Jan_Pool said:

To me, Bahnschrift is a typical “me too” product.

So no moral concerns about it?

What makes you think that I would have no moral concerns? I am just not going to discuss that with you. I only chose to answer, to not leave this question open. Now no one can come up with something like: “Oh Albert is such a nice guy, he did not even bother to answer that question, so I think that he simply looooooves the idea that I also made a version of DIN.”

Albert_Jan_Pool

Mark Simonson said:

@Albert_Jan_Pool said:
When Bitstream was founded, they somehow magically released hundreds of fonts which quite exactly matched Stempel/Linotype’s digitisations.

Are you sure about this? I licensed fonts from both libraries back in the nineties. In comparing Bitstream vs Linotype digitizations of the same faces, the data does not match, not even the metrics. They were different enough that I had opinions about which digitization I preferred.

From catalogues I have I got the impression that many of these were remarkably similar. On the other hand, some of the Bitstream versions were better because in the course of converting the Linotype library to PostScript, many faces lost their subtleties. Egyptienne F. was a good example of a Bitstream Version being much better. Also, many sans serif italics were merely electronically slanted versions of the upright version. Notably Helvetica, Avant Garde and Univers and some others too. Some of these “mistakes” were corrected in Linotype’s LT versions. As I said before, in some cases, the Ikarus data for some of the Linotype/Adobe PS fonts came from URW, not from Lino. This may also explain some differences because in many cases URW had digitized from Berthold Staromat versions of Linotype faces. Berthold fonts had a unified cap height and so was URW data in Ikarus. Fitting these on a unified body size could have been an issue as well, as there are no standardised and/or accurate rules for doing this.

Adam Twardoch

I can well-imagine that Bitstream may have used URW’s Ikarus tech. They had developed their own format and some tools over time, but they also did convert. 

But I’m reasonably sure that Bitstream’s fonts were original digitizations. I don't have proof of that but I did get my large Bitstream collection bundled with CorelDraw 4.0, the first software I bought, in 1993. Only much later I also bought two URW font collections (that had CE accents), and I actually got a Linotype Gold CD as a present, as my father bought some remnants of a small German design studio that went out of business, and that (plus some other stuff like Adobe PageMaker, books about PostScript and other things) was in their inventory.

I then spent quite a lot of time comparing these digital versions, and they really were very different.

The Bitstream data was remarkably “clean” technically. The proportions, the details, they were all different. This was very different from, say, variants of URW-digitized stuff that was then released in different formats by different foundries.

So while I don't have any way to confirm that the original data of the BT fonts was legit, I never heard accusations that it was not the case, not even in private conversations. The problem was with trademarks, and with royalty contracts for some designers (like AF), and with the older designs with the Lino/Mono libraries that were not royalty-based, and with Bitstream’s pricing and the CorelDraw bundling. But I don't think anyone ever suggested to me that the Bitstream fonts were some “conversions”. 

(My thing about Scangraphic and Linotype is most likely wrong. What I should have said is that there was a lot of digital data interchange between URW, Adobe, Linotype, plus Scangraphic and others, going on in different directions, but I am not really qualified to provide any details.)

LeMo aka PatternMan aka Frank E Blokland

In principle it is a bit off-topic to work out the following statements. However, I want to make a few comments.

If you want to digitize types once cut by Garamond, I think there is no one out here that would complain for reasons other than this being superfluous as this has been done often enough.

I disagree a bit with the ‘superfluous’ part. A type revival is by definition an interpretation that shows as much of the style period in which the original foundry type was made, as from the time (Zeitgeist) in which the revival was produced. New research can lead to new insights. My students in Antwerp and The Hague examine standardization in the work of Garamont, Granjon, Van den Keere and their peers. For example, they distill the fitting/spacing from the matrices and investigate what the relation is between fitting and details, such as the length of the serifs. This small video shows how a digital interpretation, including the spacing, is distilled directly from matrices that are justified for fixed mould registers of Granjon’s Ascendonica Romain.

Another difference could be that stems in revivals of Renaissance type, such as Adobe Garamond, are often straightened, while the original models, such as Garamont’s Gros Canon Romain above, show evidence of curvature.

Why bother about Bembo? It is a poor rendering of the original typeface anyway. No matter which version you take.

As I mentioned, a revival will ultimately reveal how the source material was viewed at the time when it was made. Our way of interpreting what we see is, after all, largely shaped by the way in which we are conditioned. As such, Monotype Bembo is interesting, if only because of the obvious deviations from the original model. Especially when you realize how successful it has become.

Too much technical restrictions. 18 unit system, keyboard arrangement, poor insight in typeface design […]

The standardization of character widths for the Monotype ‘hot-metal’ composing machines was a clear deviation from nineteenth-century foundry practice, but clearly had no noticeable negative effect on the quality of the designs. That this adjustment did not lead to distortions of the proportions of interpretations of historical models such as Monotype Bembo, can be explained by the fact that comparable standardization was an intrinsic part of font production in the early Renaissance.

Daniel Benjamin Miller

LeMo aka PatternMan aka Frank E Blokland said:

In principle it is a bit off-topic to work out the following statements. However, I want to make a few comments.

If you want to digitize types once cut by Garamond, I think there is no one out here that would complain for reasons other than this being superfluous as this has been done often enough.

I disagree a bit with the ‘superfluous’ part. A type revival is by definition an interpretation that shows as much of the style period in which the original foundry type was made, as from the time (Zeitgeist) in which the revival was produced. New research can lead to new insights. My students in Antwerp and The Hague examine standardization in the work of Garamont, Granjon, Van den Keere and their peers. For example, they distill the fitting/spacing from the matrices and investigate what the relation is between fitting and details, such as the length of the serifs. This small video shows how a digital interpretation, including the spacing, is distilled directly from matrices that are justified for fixed mould registers of Granjon’s Ascendonica Romain.

Another difference could be that stems in revivals of Renaissance type, such as Adobe Garamond, are often straightened, while the original models, such as Garamont’s Gros Canon Romain above, show evidence of curvature.

Why bother about Bembo? It is a poor rendering of the original typeface anyway. No matter which version you take.

As I mentioned, a revival will ultimately reveal how the source material was viewed at the time when it was made. Our way of interpreting what we see is, after all, largely shaped by the way in which we are conditioned. As such, Monotype Bembo is interesting, if only because of the obvious deviations from the original model. Especially when you realize how successful it has become.

Too much technical restrictions. 18 unit system, keyboard arrangement, poor insight in typeface design […]

The standardization of character widths for the Monotype ‘hot-metal’ composing machines was a clear deviation from nineteenth-century foundry practice, but clearly had no noticeable negative effect on the quality of the designs. That this adjustment did not lead to distortions of the proportions of interpretations of historical models such as Monotype Bembo, can be explained by the fact that comparable standardization was an intrinsic part of font production in the early Renaissance.

For me, actually, the reason I am interested in these revivals is because I want open-source digital fonts which are reminiscent of the results of Monotype and Linotype machines — since I spent my time in the library reading books set with those machines, not with legitimate renaissance type.

Stephen Coles

I concur with Mark. There are many Bitstream revivals that are unique. One of the most obvious may be their Futura, which is very different from Linotype’s and any other version. It’s more of a redesign more than a revival.

Christopher Slye

For what it's worth, here are a couple pages from Bitstream’s “Version 2 Catalog” (1991).

Mark Simonson

The irony is that Monotype now owns Linotype and Bitstream and distributes it all, including Swiss 721. Monotype Imaging was called AgfaType before it acquired Monotype in 1999, and some of AgfaType's clones, such as Triumvirate (a clone of Helvetica), are distributed by Monotype as well.

Christopher Slye

“... the typeface popularly known as Helvetica” is an amusing way to put it.

James Puckett

I always wondered if Zapf got royalties for Book Antiqua after Monotype bought Linotype.