Univers, Linotype Univers and Univers Next

I am having some trouble discerning the better digitization of these three typefaces.

There do appear to be some differences in them, not withstanding the face that Linotype Univers and Univers Next offer a far greater number of weights than Univers, as well as the more desirable ampersand; the cut (so to speak) appears different. Both Linotype Univers and Univers Next seem to have more modulation in the strokes, and Linotype Univers 430, which I assumed to be an equivalent to Univers 55, seems, well, heavier!

Also, any idea which is closer to Frutiger's original design?

Any help appreciated.

Celeste's picture

— Hrant
I’ll try to make myself clear in spite of my struggling English writing — so please be clement.

Technical constraints are not arbitrary when they’re part of a design brief — from a design point of view (not limited to typeface design — think furniture, architecture, etc.), a solution which meets all the requirements of the brief can hardly be bettered (although it can be equaled) within the same parameters. If you remove the technical constraints, you also change the brief, and you end up with a completely new design – which I think cannot be compared fairly with the “old” one, because it exists on a whole different plane.

Now, about Sabon : Tschichold himself didn’t think that Sabon had been handicapped by the technical constraints imposed by the Linotype and Monotype typecasting systems — on the contrary, he was proud (and rightly so, in my opinion) of the way he had met these particular requirements of the original design brief. Respecting this singular design solution’s integrity is not “misplaced reverence” — you’re absolutely entitled to design another sixteenth-century-inspired text typeface within contemporary technical parameters (and that’s precisely what Jean-François Porchez did with Sabon Next), it will simply not be Sabon anymore.

hrant's picture

We can argue about the meaning of "arbitrary" (and I can point out that "fair" does nothing for me) but it's probably better if I instead say it this way: although there is no design without constraints, some constraints are stupider than others, often resulting in unacceptably inferior results. For example if I ask a right-handed person to write with his left hand the results will be Bad. And it's not the fault of the person writing - just like Sabon's flaws are not Tschichold's fault (in fact he's rightly commended for balancing the compromises beautifully).

Most of all, and practically speaking, when a constraint is removed sticking to it is very rarely beneficial to the end-result.

it will simply not be Sabon anymore.

Maybe (although I will note that Sabon is not Sabon much more often than that). BUT: it will be Better.


quadibloc's picture

It is quite correct that Sabon was subject to technical constraints.

It was constrained to meet the lowest common denominator of Linotype and Monotype. So the italics were spaced exactly the same as the Roman (as Linotype required) and the letters were all constrained to have widths belonging to the Monotype 18-unit system.

These constraints set limits to the design, and doubtless they did impair the aesthetic quality of the result - at least slightly.

But when you ask why a digital revival of Sabon should have the same metrics... then, despite agreeing with you up to this point in principle, I part company.

Given that Jan Tschichold did a good job of preventing Sabon's aesthetic qualities from being too gravely impaired by these constraints, do you really think that a net improvement is the most likely result of some newbie font designer at Adobe being given a free hand to change the metrics and the design as he sees fit?

It isn't necessarily reverence for the technical constraints in Sabon that leads people to consider, based on the principle that a good typeface is a beautiful set of letters, and not a set of beautiful letters, that Jan Tschichold's work of making all the letters in the typeface into a harmonious whole ought not to be tossed out the window in hopes that a different type designer, now working under fewer constraints, could produce a better result than he did.

It isn't the constraints we want to keep: it's Jan Tschichold's artistic touch and vision in the final phase of integrating the typeface we want to keep, and we despair of preserving that while throwing away the constraints.

hrant's picture

Who said anything about having a newbie do it?

And I already agreed it would be Different. But so is 10 point Sabon versus 18 point Sabon...


Celeste's picture

This question of “different doesn’t equal better” is precisely why I’m always deeply suspicious of recurrent marketing claims about “new and improved” versions of classic 20th-century typefaces — be it Sabon or Univers (the original topic, remember ?).

Not to mention the fact (that I already stressed) that the “flawed” versions have more often than not become integral parts of the graphic design canon (i. e. the culture of the end-users of typographical tools). Which makes the repeated marketing attempts to wipe them out with “new and improved” versions akin to utter contempt towards these very users : “we know better, and you should trust us because you don’t have a clue, and the reason we know that you don’t have a clue is that for years we’ve been selling you flawed versions and you’ve used them no matter what, you morons !”.

hrant's picture

Your suspicion is well-founded. Especially considering how frequently a better version seems to be required... However culture must always more forward, even over bumpy roads.


eliason's picture

I agree that your suspiciousness is warranted. But one can "improve" something in ways besides fixing "flaws."

dberlow's picture

The basis of this thread is that you have choices to make. No one is attempting to wipe out Alternative Versions or force anyone to do anything. Calm down.

Celeste's picture

— David
When a big, ancient, well-respected, reputable, dignified type manufacturer rolls out its marketing steamroller (as Linotype did with the launching of every typeface included in their Platinum Collection), it’s very hard for the average customer to resist the suggestion that he should discard the “old” version of a typeface to adopt the “new and improved” one. Especially when the argument comes loaded with endorsements from the original designer (if still available) and/or various luminaries from the type and graphic design communities.

Could the fact that you call these original versions “Alternative” be interpreted as a symptom of an interesting perceptual shift ? The “new and improved” versions are the (sometimes welcome) “alternatives”, not the other way round…

By the way, the only reason these “old” versions don’t usually disappear from the typefoundries’ catalogues is that they can still contribute to make these very catalogues look thicker and richer (also a marketing ploy, by the way).

Iain Farnsworth's picture

"New and improved" is a wonderful term for the marketeer, because it offers the promise of a better product, without necessarily providing criteria for how the claim may be proved (well, with regard to "improved" anyway). To be fair to Linotype, they did provide such a criteria in their rationale for a new Univers. (Here... http://www.linotype.com/1813-15545/whyanewunivers.html ) Which, on the face of it, seems to be to provide more weights, with careful attention paid to how the weights relate to each other, and also to redraw all the weights for "...improved legibility and on-screen application." Also implicit in their documentation is to make Linotype Univers follow Frutiger's original designs more closely.

By these criteria, they seem to have had some success. There certainly are more weights (!), and there are features of Frutiger's original designs which made a comeback, such as the numerals, the ampersand and the increased angle of the oblique.

But could you call it 'better'? More to the point, would Linotype? It seems they wouldn't go that far. The interview between Frutiger and Friedrich Friedl here... http://www.linotype.com/1813-15552/podiumdiscussionuniversisaclassicalty... has Friedl calling the 'old' (he means the original, 1957 version) Univers 'perfect'. With Linotype Univers, they "took it further."

Yet, it is remarkably easy, when dealing with twentieth century typefaces like Univers or , to ask which is 'better', as I did in my original question. I'm not sure we would measure the various cuts of Garamond against each other in quite the same way. I see that, like dealing with typefaces such as Garamond or Bodoni, it's not quite so simple as that.

quadibloc's picture

Who said anything about having a newbie do it?

Fair enough. But I thought it was clear what my point was: it wasn't what I was accusing you of wanting, but what I was warning you that you were likely to get.

I agreed with your statement

How could the removal of an arbitrary constraint not open the way to a better result?

but I wanted to emphasize that "open the way" does not equal "guarantee".

So while I don't think that designers of new typefaces inspired by Sabon should adopt its technical constraints, as that would be gratuitous, a digital version of Sabon itself should be faithful to it, and provide its users with Jan Tschichold's work, faithfully copied over, not someone else's.

New type designers will in some cases have something worthwhile to say - sacrificing old typefaces to get new ones, though, would be a losing gamble. Of new typefaces, only a few survive.

Celeste's picture

— Iain

But could you call it 'better'? More to the point, would Linotype? It seems they wouldn't go that far. The interview between Frutiger and Friedrich Friedl here... http://www.linotype.com/1813-15552/podiumdiscussionuniversisaclassicalty... has Friedl calling the 'old' (he means the original, 1957 version) Univers 'perfect'. With Linotype Univers, they "took it further."

You’re right, they’re not going as far as pretending that the “new” digital Univers is better that the original metal one. What they’re doing, however, is criticize the original digital Univers (the one graphic designers had been using with such success since the beginning of DTP) — which they had marketed and sold as a faithful version of Adrian Frutiger’s original design for more than 25 years !

Nick Shinn's picture

Of new typefaces, only a few survive.

Not so.
In digital type culture, there is no difference between the warehouse, the musem*, and the supermarket.
As long as there is a part of the long tail market that still wants the “old and unimproved” (i.e. “Classic”), product/artefact, the vast majority of foundries will not remove it from their shelves/displays.

*On display, not in the vaults.

Iain Farnsworth's picture

As long as there is a part of the long tail market that still wants the “old and unimproved” (i.e. “Classic”), product/artefact, the vast majority of foundries will not remove it from their shelves/displays.

And of course, for a large font family like Univers, there will be, as owners of a few weights are more likely to want to build on the family then start afresh with a newer version.

Celeste's picture

And of course, for a large font family like Univers, there will be, as owners of a few weights are more likely to want to build on the family then start afresh with a newer version.

Especially considering the example set by all the practitioners who did a tremendous job using the old version. Yes, I know, I’m stubborn that way.

dberlow's picture

Celeste: "...it’s very hard for the average customer to resist the suggestion that he should discard the “old” version of a typeface to adopt the “new and improved” one."

Nevertheless, it's your choice, and few if any ancient or modern false claims, are breaking your arm. If resistance is futile, I suggest it's your opportunity to vote with your pocketbook on design and not Linotype's rapacious greed.

"...the only reason “old” versions don’t usually disappear from the typefoundries’ catalogues is that they can still contribute to make these very catalogues look thicker and richer (also a marketing ploy, by..."

While I agree that catalog thickening occurs with more versions of families, digital distribution thins this to a great extent to the thickness of your screen, and I'm not entirely certain the marketing department likes more versions, in print or on screen, as much as they like more families, and versions addressing specific typographic rather than generational uses, not to denigrate the latter in any way. But the bottom line for any sane marketeer, says the average user most likely prefers one version.

I might also suggest, though I am not at liberty to detail, that there are other more important marketing issues involving the profit margins, that keep multi-versioning going in many foundries, and though I know nothing of Linotype's details, this is an important aspect of some font marketing.

You, may have the last word, and I thank you for your opinions.

Nick Shinn's picture

I suspect that Frutiger and Zapf didn’t/aren’t receiving much remuneration for the originals of their classics, relatively speaking, and that they are much better rewarded in royalties for the updates. Props to Linotype for that.

quadibloc's picture

Incidentally, the Selectric Composer provides an illustrative example of how a variety of typeface designs fare when subjected to a similar (but more severe) set of constraints than those to which Sabon was subject.

1) Type elements contained only 88 characters, and so the same unit system applied to Roman, Italic, and Bold weights of every typeface.

2) The device employed a 9-unit system. However, this was not twice as coarse as the Monotype 18-unit system; instead, most letters were assigned widths comparable to those used for Times Roman, scaled down from 18 units to the em to 11 units to the em.

3) Hence, the letters m, M, and W were narrower than they should have been to be in the correct proportion, being 9 units wide rather than 11, in order to fit on the type element (it being desired to support type sizes up to 11 points, and to use the same basic type element as used in normal Selectric typewriters).

I just thought it would be worth citing this example in this thread, as comparing Press Roman to Times Roman, and so on, might give a benchmark for the potential impact of technical constraints on Sabon, and how well Jan Tschichold did in overcoming them.

EDIT: Oh, yes - here's the link


Otmar Hoefer's picture

You are asking for the real Univers.
let me give you a few insights on this.
Univers was a face that was originally made for hotmetal at Deberny & Peignot. It had this 16 degrees slanted italic.
This was at Linotype for the matrices of the hotmetal machines not possible. Therefore the type needed a lesser slant angle 11 degrees. You know the regular (55) and the italic (56) were needed on one matrice, meant they had to share the same width. The same was necessary for the bold (75) which was on the same width like the regular (55).
There was also another thing that I did not understand in the beginning of my career at Stempel and Linotype. Why does the hotmetal version has such a nice et-sign and why is this dump &-sign in the fonts.
Clear answer, that was based on customer requests from the US, they are used to the & sign form and could not understand what the et-ligature from Frutigers Univers mean.
And therefore the fonts have this sign included. In the new versions we have the new sign, but again some US customers complained.....

These fonts than needed to be compatible with our phototypesetting machines, as our former customers did not wanted to let their customers identify in print whether it was hotmetal or phototype. It needed to be width compatible.
This we migrated through all our technologies until the Linotype Laserfonts. The only difference we made was that the italics were really drawn as italics. You know the differences between a slanted font and an italic form.
When Linotype entered into the production with Adobe to make the Postscript library Univers was one of the first fonts and Linotype at that time did not yet had the Adobe tools to make our own digitization. So we supplied Adobe with digital data and they made the conversion into PS fonts. For italic it was at that early stage of the development the habit to just make set a slant angle into the fonts and you have an italic. which was than called oblique. Therefore Univers 55 oblique and not 56 Italic.

Whan Adrian saw the digitization he was not pleased, but it was nearly similar to the Linotype version which was this mechanic restricted version. so this was transitioned into this technology.
In 1996 Gerards Unger was approaching Bruno Steinert (our former MD) and me at the AtypI and explained ushis philosophy that we should extend our classic faces and do a similar work like with neue Helvetica and revise the Univers family to a more consistant family.
We approached Frutiger with that idea and he at that time just recovered from a heart surgery with three bypasses and he was pleased that Linotype took up this idea to make a new version of the Univers family. At that time he was completely clear in his head and he was working hard on this project with our former type director Reinhard Haus. The beginning was a trial to look if we could use the exisiting data and refine just a few errors in the outllines, but we realized that this was not leading to the right result. So we scraped all this work and restarted by looking at the original prints form the hotmetal Univers. But there as you can imagine were also resulting inconsistancies in the family and weights that Adrian used the prints and used his scissors to cut and shape the letters into a form the was for his eyes pleasing.
So the Linotype Univers family was developed. It was the beginning of our Platinum Collection to which Gerard Unger was the originator.
More you can read at our website
There was one thing left over for many years. (the family should get small caps) and this gap was filled a few years ago and we detected that the naming of the font was not as good like all the other products form the Platinum Collection used nova or Next. And therefore we renamed the complete family into Univers Next and added the small caps to the fonts. Nothing else had been changed.
If you want more to know please contact me at ohoefer@linotype.com

hrant's picture

Otmar, great to see you here! And thanks for the "inside info".

but again some US customers complained

Sadly you have to please your biggest customer. That's why BMWs aren't what they used to be. :-/ I remember arguing with a former boss about the ampersand in Poppl Laudatio. But he was right.


gilinski's picture

Fantastic. Thank you for the clearing up the mysteries of Univers, Otmar!

quadibloc's picture

On the subject of an italic for Univers with a slope of 16 degrees being too sloped, it may be noted that the original italic for Times 327 had a slope of 16 degrees as well; this was reduced to 12 degrees for the italic of Times 569, the one used for their 4-line mathematics system.

Nick Shinn's picture

It would be interesting to see a Univers revival, based on printed specimens of the Deberny & Peignot cut.

And another, based on IBM Selectrix typed specimens.


I would like to see some samples of the 16˚ slanted italic, sounds very cool.
I inclined the Figgins Sans italic at 20˚, to match the slope of the Scotch Modern italic.

quadibloc's picture

@Nick Shinn:
And another, based on IBM Selectrix typed specimens.

I realize that the Univers for the Selectric Composer was an official adaptation.

Even so, because the spacing used for the machine was close to ideal for Times Roman, thus making their imitation, Press Roman, successful, the sans-serif faces such as Univers and Theme left too much space around the lowercase i and l, for example.

Still, that doesn't mean the face doesn't contain useful information about how Fruitger intended Univers to look - and how to modify it to respond to constraints. So I approve of studying it, but I can't consider it worthy of revival.

Nick Shinn's picture

I did some work with Selectric typesetting in the 1970s; sure, it was a bit rough compared to real typesetting, but it was incredibly empowering.

The typesetting of The Last Whole Earth Catalogue (the bible of the Counterculture) was done with Selectrix. A revival of Selectric Univers, in particular, if it captured that slightly off-kilter quality of the original would, to my mind, evoke the optimistic spirit of the technology in that era.

Then again, others might just think it was bad typesetting—which might be a useful thing if you want to avoid a slick corporate effect.

Jens Kutilek's picture

There is a scanned version of the Selectric Univers: http://ospublish.constantvzw.org/foundry/univers-else/

I wouldn’t call it a proper revival though.

Nick Shinn's picture

Some form of pseudo-randomization would be good, to translate the yaw and bounce of the original “strike-0n” technology.

quadibloc's picture

Univers Else is interesting; I remember being given here a link to an IBM Executive font; looking for it, I found:


but I don't think it's the same one. (I think "P45" was in the font name.)

Ah: F25 Executive, by Volcker Busse.

Iain Farnsworth's picture

Thought this thread had died, only to find it resurrected!

Otmar, thanks so much for your inside knowledge on Univers. It certainly does help clear up some of the mysteries of the 'old' digital Univers, such as the ampersand. How could anybody complain about Frutiger's lovely 'Et'? It's one of my favourite features of Univers, and a reasons why I regret having purchased a few weights of the old version.

As an employee of Linotype, I'm not sure you'll feel inclined to answer the following, but I'll ask it in case you do (indeed, if anyone has any insights on this)...

What was the deal with Berthold Univers' disappearance from the marketplace? I believe it was a favourite of some Univers officianados (such I Willi Kunz, I think). I know there was some kind of legal dispute, but I wondered why Bitstream still sell Univers (albeit by another name - Zurich), yet Berthold has withdrawn it's offering. Anyway, thought I'd put that out there.

Chris Dean's picture

Did you know that Univers would have been todays Helvetica?

I had a few conversations with Mike Parker about his time with Linotype, and of course, Times & Starling, and he wanted Univers to be their typeface of choice, but the italics were to italic to work on their machines properly.

Cool eh?

Syndicate content Syndicate content